A Long COVID Definition: A Chronic, Systemic Disease State with Profound Consequences (2024)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

2 Defining Long COVID A DEFINITION DESIGNED TO EVOLVE AS NEW EVIDENCE AND UNDERSTANDING EMERGES Evidence on the different presentations of Long COVID continues to emerge as the clinical and scientific understandings of Long COVID deepen. Because of this dynamic, the definition of Long COVID is expected to evolve, and the version described should be considered the “2024 NASEM Long COVID Definition.” At this stage of understanding Long COVID, no definition can ensure that every affected individual will be correctly classi- fied and that no others will be wrongly included. The 2024 NASEM Long COVID definition can only go as far as current science and evidence permit. With new evidence and better scientific understanding accumulating over time, the definition of Long COVID can and should be improved. Relying on findings reported in the literature as well as on the engage- ment process, the committee proposed the 2024 NASEM Long COVID Definition. This definition identifies Long COVID as an infection-associated chronic condition (IACC), specifies a minimum duration of 3 months, and expressly incorporates common symptoms and diagnosable conditions characteristic of Long COVID. The bolded “Core Definition” is designed to be accurate and inclusive; the “Important Features” provide context and highlight notable aspects of Long COVID. 31 PREPUBLICATION COPY—Uncorrected Proofs 

32 A LONG COVID DEFINITION The committee hopes that this definition will: • Aid clinicians in the consistent diagnosis, documentation, and treat- ment of Long COVID. • Encourage further research into the pathophysiology, diagnosis, prognosis, consequences, and treatment of Long COVID. • Enhance patient access to appropriate care, treatment, services, and benefits. • Harmonize research and surveillance efforts on Long COVID, while providing researchers flexibility in the design of studies on Long COVID. • Raise awareness and educate the public and policy makers about Long COVID. Balancing Risks of False Negative and False Positive Classification Throughout its deliberations, the committee grappled with two con- tending objectives: to ensure that patients who experience Long COVID will be included in the definition and to avoid wrongly including patients whose condition is not related to prior SARS-CoV-2 infection. This is a familiar dilemma in any diagnostic challenge, to balance the risks of false negative and false positive classification. The 2024 NASEM Long COVID Definition is intentionally inclusive, to satisfy the first objective. The com- mittee acknowledges the potential for false positives with the definition. The committee believes the patient’s treating clinician is best poised to strike the right balance between avoiding a false positive and a false negative classification. All the symptoms and concomitant diagnosable conditions that are characteristic of Long COVID existed before the COVID-19 pandemic, and none are specific to Long COVID. What makes the case for associa- tion between symptoms and Long COVID is the temporal relation to the SARS-CoV-2 pandemic that may be documented as clinical findings or through patient self-report. The lack of biomarkers with high sensitivity and specificity for Long COVID is a major knowledge gap. As discussed later in this report, studies are underway to examine possible biomarkers for Long COVID, though none is yet dispositive for the disease state. Biomarkers may reinforce, or help exclude, a diagnosis of Long COVID, reducing the risk of false positives. Also discussed later is evidence for risk factors for develop- ing Long COVID. As with any diagnosis, risk factors can be helpful in the clinical assessment of the probability of disease in any individual patient. The definition does not require documentation of prior infection. Based on antibody surveys, more than three quarters of adult Americans had evi- dence of SARS-CoV-2 infection by the end of 2022, and it is reasonable to PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 33 estimate that 80–90 percent of all adult Americans were infected at least once between 2020 and the end of 2023 (CDC, 2024b). In assessing a patient’s condition and reaching a diagnosis, the clinician must take account of all aspects of a patient’s history, symptoms and signs, and test findings against the background of different disease probabilities. A clinician may need to overcome language, cultural, or educational barriers, and it may be difficult to elicit a complete and reliable history in patients who have experienced medical trauma or have cognitive impairment. It is incumbent on the clinician to consider alternative explanations of a patient’s presentation and to balance the risks of false inclusion and of false exclusion in reaching any possible diagnosis. A later section elaborates on applying the definition in clinical care, along with applications for research and for public health surveillance, and Appendix B defines key terms in the measurement of diagnostic performance. A Single Definition The committee chose to put forth one broad definition for Long COVID that could apply to adults and children. Findings from the com- mittee’s engagement process indicate a desire for consistency between how Long COVID is defined in adults and children. An inclusive definition that sets appropriate boundary conditions (such as the minimum duration of symptoms) is designed to promote coordination among different audiences and purposes. For example, the results of research are more relevant when the study participants involved are like those identified in clinical practice. Public surveillance data can better inform disability policy when subjects of the former are defined similarly to those attempting to access benefits. Such a definition also helps ensure that compensation programs for loss of work or disability reflect the entirety of those affected by this chronic medical condition. The committee sought to find balance between a single definition and the need to operationalize it for different purposes. Findings from the committee’s engagement emphasize that the definition should encompass most, if not all, patients who are affected. More broadly, the committee sought to craft a definition usable in a variety of ways and for a range of purposes. In the final section of this report, the committee discusses how the definition can be applied to specific use cases. For example, as research and surveillance strive to answer many of the unknowns surrounding Long COVID, specific inclusion or exclusion criteria may be necessary to achieve the purposes of the particular study or surveillance. PREPUBLICATION COPY—Uncorrected Proofs 

34 A LONG COVID DEFINITION The Value Proposition for this Definition The committee’s central charge was to put forth a refined definition for Long COVID that lays the foundation for increased scientific understanding of the disease state and for better diagnosis and treatment of patients. The definition presented here includes a few notable features and specifically introduces a few new features that existing definitions lack. The new definition provides explicit examples of common symptoms and conditions that are characteristic of Long COVID. Educating clinicians about common symptoms can improve the accuracy and speed of Long COVID diagnosis. Recognizing diagnosable conditions allows for the treat- ment of the specific conditions and communicates that these disorders are part of Long COVID. The definition reinforces divergent course patterns by describing the ongoing disease state as continuous, relapsing and remitting, or progressive. The 2024 NASEM Long COVID Definition requires symptoms or conditions to be present for a duration of 3 months or longer. Notably, while symptoms need to be present for at least 3 months, the timing of those 3 months is unspecified. In particular, it is not necessary for symp- toms to be experienced continuously from the time of the acute infection. The committee notes that this 3-month period allows for the spontaneous resolution of symptoms, evaluation for other diagnoses that might explain symptoms, and possible therapeutic trials to clarify the diagnosis. Although the definition specifies a minimum duration of 3 months to qualify as Long COVID, a clinician should recognize, acknowledge, and monitor concerning symptoms before the 3-month mark. These symptoms should be assessed and treated appropriately, and the ICD-10 code U09.9 (post COVID-19 condition, unspecified) may be used even before establishing a Long COVID diagnosis. The committee developed this definition with equity in mind to recog- nize that social determinants and structural inequalities intersect to create health disparities (HHS, 2020) and to discourage stereotypical assump- tions and biases that could prevent patients, clinicians, public health prac- titioners, researchers, and policy makers from recognizing all those who experience Long COVID. For example, a stereotypical assumption about conditions like Long COVID, which often present with more symptoms than definitive clinical signs, is that they are more common among affluent, highly educated white women. Such assumptions might deflect clinicians from recognizing that persons of color or young men might potentially have Long COVID. Considering this, one of the definition’s important features is “Long COVID can affect children and adults, regardless of health, dis- ability, or socioeconomic status, age, sex, gender, sexual orientation, race, ethnicity, or geographic location.” Risk factors, per se, do not play a direct PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 35 part in the 2024 NASEM Long COVID Definition, though they are relevant to differential diagnosis, and a later section of this report summarizes cur- rent evidence about risk factors. The definition does not require laboratory confirmation or other proof of initial infection. The initial infection may or may not have been recog- nized, in part due to the lack of availability of and limited access to tests early in the pandemic, limited sensitivity of some SARS-CoV-2 tests and the potential for false negatives, and an overall decline in testing rates later in the pandemic. However, the symptoms and diagnosable conditions char- acteristic of Long COVID can have alternative origins. If and as the back- ground frequency of acute SARS-CoV-2 infection diminishes, as discussed earlier, other sources will likely become responsible for a growing propor- tion of cases. The probability that a particular clinical case is attributable to SARS-CoV-2 will be increased when there is evidence of prior, acute infection, or multiple infections, with that virus. The definition firmly acknowledges the profound impact of Long COVID on function, and this has serious implications for the provision of services, accommodations, and benefits to patients. EVIDENCE SUPPORTING KEY ELEMENTS OF THE DEFINITION The committee found no published, standardized guidelines for the development of disease definitions (Doust et al., 2017). The committee gleaned lessons from the process of developing and modifying definitions for other multi-symptom conditions such as ME/CFS and Gulf War Syn- drome (IOM, 2014, 2015). As a first step in the committee’s process, the committee articulated several possible key elements of a disease definition (Table 1). These key elements served as a framework for the committee as it developed and refined the definition of Long COVID. In the follow- ing section, the committee describes how the 2024 NASEM Long COVID Definition approaches each key element and provides a summary of the evidence from the scoping review and primary literature and also provides a summary of the findings from the engagement of multiple groups in sup- port of its decisions. In addition to the above key elements, the committee considered other foundational criteria in its deliberations. These foundational criteria are multi-factorial, layered, and interacting, and they were informed by many different frameworks and checklists (Doust et al., 2017; Moberg et al., 2018; NASEM, 2020). Those foundational criteria were: PREPUBLICATION COPY—Uncorrected Proofs 

36 A LONG COVID DEFINITION TABLE 1  Possible Key Elements of a Disease Definition Element Description Attribution Source responsible for causing the disease Time Onset of disease Clinical Features Symptoms, symptom course and duration, and symptom severity of the disease to be defined Equity Identify persons affected and consider equity implications Functional Impairment Effect of the disease on daily activities Exclusions/Alternative Diagnoses Consideration of alternative diagnoses Biomarkers and Laboratory Criteria Objective tests (e.g., blood tests, neuroimaging, cognitive batteries) that help identify the disease Risk Factors Characteristics associated with a higher probability of disease or adverse outcome • Precision: A precise definition should be repeatable (agree in identi- cal conditions), reproducible (agree across comparable conditions), and accurate (specific and sensitive). • Feasibility: The definition must be reliably translated into opera- tional terms for multiple different purposes and should be adapt- able to a range of possible circumstances. The definition should consider the impact on resource usage. • Acceptability: The definition should take account of stakeholders’ values and preferences. • Accessibility: The definition must be easily and equally well under- stood and applied by diverse stakeholders. • Balancing benefits and harms: The definition should be guided by a balanced assessment of the anticipated benefits and harms, using the best available evidence, and considering both the individual and societal level. • Potential impact on health equity: The definition should be equi- table and should not perpetuate discrimination or inequities. The definition should also be perceived as equitable by socioeconomi- cally, racially, culturally, and educationally minoritized groups and by those who have distinct historical experiences with the health system. • Unintended consequences: The definition should consider potential for misuse, effect on measured incidence, changes to the apparent natural history of disease, suboptimal treatment of patients, psy- chological and financial consequences, and other adverse effects on individuals and society. PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 37 How Does the Definition Address Attribution to Infection? Definition occurs after SARS-CoV-2 infection Important Features LC can follow asymptomatic, mild, or severe SARS-CoV-2 infection. Previous infections may have been recognized or unrecognized. Linking symptoms to a confirmed or suspected SARS-CoV-2 infection may be desirable in a Long COVID definition to reduce the risk of misdiag- nosis of other conditions as Long COVID. However, the reality is that some, if not many, SARS-CoV-2 infections have gone unrecognized throughout the COVID-19 pandemic. Therefore, the 2024 NASEM Long COVID Defini- tion states that Long COVID occurs after acute SARS-CoV-2 infection but does not require laboratory confirmation or other proof of initial infection. The definition emphasizes that Long COVID can follow infections of any severity (including asymptomatic infections), whether they were initially recognized or not. Findings from both the engagement process and the evidence review highlighted the large number of SARS-CoV-2 infections, both throughout the pandemic and currently, that are not captured by testing and/or are not recorded in patients’ medical records. Combined with the lack of a specific and sensitive test that can detect a past SARS-CoV-2 infection (e.g. antibody testing), this supports the committee’s decision not to require laboratory evidence or formal diagnosis of an initial SARS-CoV-2 infection as part of the definition of Long COVID. Requiring such evidence would likely lead to underdiagnosis and raises equity issues. This aspect of the NASEM 2024 Long COVID definition is in accord with prior definitions the committee reviewed, none of which require laboratory evidence or formal diagnosis of infection. The evidence review also highlighted the challenges of attributing research study participants’ symptoms and manifestations to SARS-CoV-2 infection; this task is likely to become more difficult as the population of never-infected individuals available for control groups decreases. Findings from the Evidence Review Because no test for SARS-CoV-2 infection has perfect sensitivity and because the rates of false negatives on antigen and polymerase chain reac- tion (PCR) tests vary with time and other factors, some infected individu- als will receive negative test results. For example, median real-time PCR PREPUBLICATION COPY—Uncorrected Proofs 

38 A LONG COVID DEFINITION (RT-PCR) false negative rates vary from 38 percent on the first day of symptoms to 20 percent at day 3 after symptom onset and to 66 percent on day 21 (Davis et al., 2021; Dinnes et al., 2022; Kucirka et al., 2020). In addition to the possibility of false negatives on testing, some individuals were not tested or could not access testing during a suspected acute SARS- CoV-2 infection. Individuals experiencing an asymptomatic infection may not be tested for SARS-CoV-2, yet a variety of sequelae can occur after asymptomatic infections (Ma et al., 2023) or mild infections (Malkova et al., 2021). Individuals with an unrecognized or unconfirmed initial infec- tion (including, potentially, those with a false negative test) can still develop Long COVID. A team at Northwestern Medicine investigated 61 patients with neurological symptoms linked to a suspected post-viral condition. They found no substantial difference in the average number of symptoms or the subjective perception of recovery between those who had a positive test for SARS-CoV-2 (n=32) and those without a positive SARS-CoV-2 test but who were found to have evidence of humoral or cellular SARS-CoV-2-spe- cific immune responses during investigation (n=12) (Orban et al., 2023). Because of these limitations, some clinicians and researchers have sought other ways to link ongoing symptoms to a possible past infection. Antibody testing can sometimes indicate a past SARS-CoV-2 infection, but antibody levels can fluctuate or wane over time. Vaccination against COVID-19 complicates antibody testing as it can cause positive results on some antibody tests (Fogh et al., 2022). Furthermore, the trajectories of antibody concentrations over time appear to differ between women and men after SARS-CoV-2 infection, and some research suggests that the sen- sitivity of antibody testing may be lower in women and may differ among age groups (Korte et al., 2021; Vashisht et al., 2021). Attributing symptoms to a previous SARS-CoV-2 infection in a research study is a separate issue from attributing a particular patient’s symptoms to a previous infection in clinical care. However, over time, the former may inform the latter. Sensitivity may be prioritized in diagnosis, while specific- ity may be prioritized in research. The inclusion of an uninfected control group in studies can be helpful in distinguishing manifestations specifically related to COVID-19 from those resulting from other causes. For example, one systematic review and meta-analysis with seven observational studies, including studies in adult and pediatric populations, found an increased risk of newly diagnosed diabetes mellitus in individuals with a past SARS-CoV-2 infection compared with uninfected individuals and compared with those with a past severity-matched influenza infection (Banerjee et al., 2022). Another systematic review and meta-analysis with over 20 million patients found an increased risk of acute myocarditis in individuals in their first year after confirmed SARS-CoV-2 infection compared with control subjects (Zuin et al., 2022). PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 39 A 2022 systematic review and meta-analysis of studies on Long COVID in children and adolescents (<19 years of age) by Behnood and colleagues highlights the need for high-quality data. Based on their analysis of five controlled studies, the only symptoms significantly more prevalent in cases than in controls were loss of smell (8 percent more), headaches (5 percent), cognitive difficulties (3 percent), sore throat (2 percent), and sore eyes (2 percent). Meanwhile, reported symptom prevalence was much higher in the 17 uncontrolled studies, and higher study quality was associated with lower prevalence of symptoms, except for smell loss and cognitive problems. The authors point out that including control or comparison groups in studies can help researchers distinguish symptoms resulting from SARS-CoV-2 infection from those resulting from other factors, such as background symp- toms and concurrent social changes during the pandemic (Behnood et al., 2022). But control group comparisons also have limitations in knowing for certain that those in control groups do not have unrecognized prior SARS- CoV-2 infections. An evidence-mapping study based on a search conducted in November 2021 found that only 15 percent of 565 included studies on Long COVID had a control group; this may be a limitation of the current evidence base (Franco et al., 2022). Researchers may face rising challenges in including control or comparison groups in studies as the percentage of the U.S. and global populations with no history of SARS-CoV-2 infection continues to shrink. Findings from the Engagement Process Participants suggested that requiring laboratory confirmation of SARS- CoV-2 infection would be too exclusive and observed that numerous indi- viduals never received a laboratory confirmation of SARS-CoV-2 from a medical professional. Therefore, requiring a laboratory confirmation of a SARS-CoV-2 infection may result in barriers to health coverage or benefits. Regarding access to testing, a focus group participant said, “A lot of mar- ginalized people didn’t have access to testing, and a lot of people in city centers got infected very early in the pandemic when testing was not avail- able. That includes some of our poorest citizens.” Phrases such as “probable COVID-19” or “suspected COVID-19” were supported in the definition to foster inclusivity. Some suggested that a patient-centered definition that attributes infection according to the patient’s lived experience would be more appropriate. Lessons from Existing Definitions The existing Long COVID definitions considered in this report do not mention whether a patient ever had a positive COVID-19 test. The WHO PREPUBLICATION COPY—Uncorrected Proofs 

40 A LONG COVID DEFINITION Adult and WHO Pediatric definitions state that Long COVID may follow “probable” or “confirmed” SARS-CoV-2 infection, and the NICE definition uses the phrase “during or after an infection consistent with COVID-19.” The CDC and NIH definitions acknowledge the possibility of Long COVID developing after an asymptomatic or unrecognized acute infection. How Does the Definition Address Onset and Duration? Definition is present for at least 3 months Important Features LC can be continuous from the time of acute SARS-CoV-2 infection or can be delayed in onset for weeks or months following what had appeared to be full recovery from acute infection. One of the most important elements that the committee wanted to clar- ify is the timeframe used in a definition for Long COVID. A Long COVID definition may address the minimum or maximum time after the onset of initial COVID symptoms that a medical condition can be designated as Long COVID as well as the minimum duration of symptoms needed to qualify for a Long COVID diagnosis. Precision with time course and dura- tion may be an opportunity to create a more meaningful, specific definition of Long COVID that can prevent the misdiagnosis of patients who have continuing symptoms from a pre-existing illness or symptoms from a new- onset, unrelated illness. Decisions regarding duration and latency time can also affect eligibility for care, reimbursement, or benefits. From the engagement process, the committee determined that 4 weeks of symptoms was likely too short to be defined as Long COVID, because in many cases, SARS-CoV-2-related symptoms that persist for 4 weeks will resolve shortly thereafter. The current evidence base regarding symptom resolution trajectories has limitations, but the literature review generally suggests many people have symptoms that resolve in a matter of weeks, while the smaller group of individuals with symptoms that last at least 3 months have a higher chance of persistent symptoms to at least 1 year. Both the evidence review and the engagement process highlighted that some individuals have onset of symptoms that is delayed by weeks or months after an apparent recovery from initial infection, although the frequency of delayed onset is unclear. Because there is still ambiguity regarding the relationship between the timing of SARS-CoV-2 infection relative to Long COVID onset, the committee chose not to include a maximum latency PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 41 period. Although this action may lead to an increase in the number of people diagnosed with Long COVID, any maximum latency chosen would be speculation without the backing of scientific evidence. Furthermore, such a move might exclude people who develop delayed onset Long COVID, did not recognize they might be affected by Long COVID until later in its course, or were not able to access care due to the availability or restrictive criteria of some Long COVID clinics. Additionally, research into delayed onset Long COVID or the long-term consequences of Long COVID might be discouraged were an artificial cut-off time established. The committee anticipates that as more knowledge is gained about Long COVID, methods (e.g. biomarkers for Long COVID) will emerge to help distinguish Long COVID from other conditions regardless of latency time. The 2024 NASEM Long COVID Definition specifies 3 months as the minimum duration of symptoms, which means that 3 months after infec- tion, whether consistent or relapsing and remitting, is the earliest that symptoms can be designated as Long COVID. In cases with delayed onset, the minimum 3-month duration would still apply but could begin at any time. Some individuals may present with persistent symptoms linked to COVID-19 before 3 months have elapsed; the ICD general code U09.9 (post COVID-19 condition, unspecified) is available for these patients. The committee would like to emphasize that individual symptoms or diagnos- able conditions should be recognized and treated regardless of whether 3 months have elapsed. Findings from the Evidence Review Most patients with acute SARS-CoV-2 infection recover after a period of days to weeks. A 3-month cutoff will likely provide enough time for most patients to recover from acute manifestations of COVID-19. The Global Burden of Disease Long COVID Collaborators conducted an observational analysis of 1.2 million people in 22 countries, using data from 54 studies plus two medical record databases. Considering three Long COVID symp- tom clusters (persistent fatigue with bodily pain or mood swings, ongoing respiratory problems, cognitive problems), the authors estimated that 6.2 percent of individuals with a history of symptomatic SARS-CoV-2 infec- tion had one or more of these symptom clusters at 3 months post-infection (Wulf Hanson et al., 2022). Among 10 studies analyzed within a systematic review on Long COVID prevalence and manifestations, the prevalence of failure to recover full health and fitness by 12 weeks after infection ranged from 8 percent to 70 percent, with a pooled estimate of 34.5 percent (prediction interval, 4.3–85.9 percent) (Woodrow et al., 2023). The choice of a 3-month minimum duration may allow for the resolution of tempo- rary symptoms that are due to non-medical circumstances (e.g., overwork, PREPUBLICATION COPY—Uncorrected Proofs 

42 A LONG COVID DEFINITION stressful situations, medication side effects) or due to medical conditions other than COVID-19. A 3-month cutoff may also allow for evaluation and treatment for alternative conditions with similar initial presentations. Another factor that influenced the committee’s decision to choose a 3-month duration is that several primary studies suggest that people who have symptoms that persist for several months may have a high chance of still having symptoms at 1 year after infection. Among a cohort of unvaccinated adults randomly sampled from all confirmed wild-type SARS- CoV-2 infections in Zurich, Switzerland (which had mandatory reporting) (n=1,106), 72.9 percent of the infected reported that they had recovered fully within 3 months, including 55.3 percent within 1 month and 17.6 percent between 1 and 3 months post-infection. Meanwhile, 27.1 percent, 22.9 percent, and 18.5 percent of the infected individuals said they had not fully recovered at 3, 6, and 12 months, respectively, meaning that 4.2 percent of patients experienced recovery between 3 and 6 months and 4.4 percent between 6 and 12 months (Ballouz et al., 2023). An analysis of the ComPaRe Long COVID cohort found that 85 percent of patients who were experiencing symptoms at 2 months were still experiencing symptoms at 12 months after infection. The limitations of this study include the lack of an uninfected comparison group and the possibility that patients with more symptoms were recruited (Tran et al., 2022). Delayed onset of Long COVID symptoms after apparent recovery from acute COVID-19 has been reported. In a survey-based, 7-month interna- tional study of individuals with confirmed or suspected Long COVID, some respondents reported recovery from symptoms within the first four weeks to two months of their initial illness followed by a relapse or the appearance of new symptoms in months 3, 4, or 5. However, this pattern was uncom- mon within the study population (Davis et al., 2021). Findings from the Engagement Process Participants indicated that the disease’s onset should be included in the definition; however, there were varying opinions on the specific approach to include such information. One participant said, “The time frame is critical. I think 4 weeks is too short. WHO uses 3 months after the initial SARS- CoV-2 infection, with these symptoms lasting for at least 2 months with no other explanation. I think this is much better.” Furthermore, participants raised the duration of Long COVID as an important but ambiguous topic that should be addressed in the definition. Many noted that Long COVID may manifest in different ways and last different lengths of times, and yet PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 43 it is unknown whether Long COVID symptoms will persist indefinitely. Participants emphasized the importance of including the relapsing and remitting nature of Long COVID symptoms. One participant said, “There are people for whom acute symptoms fade, and they have a period of wellness and then their long-term symptoms arise several months after the original acute infection. It’s not the most common pattern, but it is common enough that we keep seeing it, and I am not sure that the current timelines accommodate that.” Lessons from Existing Definitions The WHO Adult definition states that Long COVID symptoms usually appear 3 months after initial infection, and the WHO Pediatrics definition includes a maximum latency time, stating that Long COVID symptoms ini- tially occur “within 3 months of acute COVID-19.” Both the WHO Adult and the WHO Pediatric definitions state that symptoms must last “at least 2 months” to be designated Long COVID. By contrast, the 2024 NASEM Long COVID Definition requires symptoms or manifestations to persist for at least 3 months and does not give a maximum latency time, instead recognizing that Long COVID can first appear weeks or months after an apparent recovery. The U.S. definitions either reference 4 weeks (OASH and CDC) after infection as the earliest that Long COVID can be identified or do not give a specific earliest time point (NIH). Patients experiencing symptoms 3–6 weeks or 4–12 weeks after infec- tion may or may not develop illness that persists beyond that point. It may be desirable to accommodate patients seeking treatment before 3 months have elapsed while acknowledging the uncertainty over individual patients’ future course of symptoms. Of the existing definitions covered here, only the NICE definition separates patients into two temporal subsets: signs and symptoms occurring between 4 and 12 weeks are “ongoing symptomatic COVID-19,” and signs and symptoms that last beyond 12 weeks are “post- COVID-19 syndrome.” According to NICE, both subsets can be considered Long COVID. The NICE definition also states, “Post‑COVID‑19 syndrome may be considered before 12 weeks while the possibility of an alternative underlying disease is also being assessed.” PREPUBLICATION COPY—Uncorrected Proofs 

44 A LONG COVID DEFINITION How Does the Definition Address Symptoms, Temporal Pattern and Duration of Symptoms, and Symptom Severity? Definition continuous, relapsing and remitting, or progressive disease state that affects one or more organ systems.  LC manifests in multiple ways. A complete enumeration of possible signs, symptoms, and diagnosable conditions of LC would have hun- dreds of entries. Any organ system can be involved, and LC patients can present with • single or multiple symptoms, such as shortness of breath, cough, persistent fatigue, post-exertional malaise, difficulty concentrating, memory changes, recurring headache, light- headedness, fast heart rate, sleep disturbance, problems with taste or smell, bloating, constipation, and diarrhea. • single or multiple diagnosable conditions, such as intersti- tial lung disease and hypoxemia, cardiovascular disease and arrhythmias, cognitive impairment, mood disorders, anxiety, migraine, stroke, blood clots, chronic kidney disease, postural orthostatic tachycardia syndrome (POTS) and other forms of dysautonomia, myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS), mast cell activation syndrome (MCAS), fibromyalgia, connective tissue diseases, hyperlipidemia, dia- betes, and autoimmune disorders such as lupus, rheumatoid arthritis, and Sjogren’s syndrome. Important Features LC can range from mild to severe. It can resolve over a period of months or can persist for months or years. LC can exacerbate pre-existing health conditions or present as a new condition. The 2024 NASEM Long COVID Definition does not list any symptoms or conditions as required and does not list any symptoms or conditions as exclusionary; this may have the effect of lessening the specificity while increasing the sensitivity of the diagnosis. The symptoms and conditions listed in the definition are chosen to be representative of the hundreds of symptoms currently identified in the Long COVID population to date, but the intent was not to downgrade other symptoms. A complete enumeration PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 45 of signs, symptoms, and diagnosable conditions of Long COVID would have more than 200 entries. For example, one study estimated the preva- lence of 203 symptoms in multiple organ systems (systemic, neuropsy- chiatric, reproductive, cardiovascular, musculoskeletal, immunological, head–ear–eye–nose–throat, pulmonary, gastrointestinal, and dermatologic) (Davis et al., 2021). Furthermore, Long COVID may present differently in different individuals, and a single set of illustrative symptoms cannot pre- cisely capture all presentations of Long COVID. Not every delayed consequence of acute SARS-CoV-2 will satisfy the definition of Long COVID. For example, chronic impairment of cardiac function could satisfy the definition’s criterion of persistence for at least 3 months. However, an elevated incidence of acute myocardial infarction in the year following acute SARS-CoV-2 infection, as reported by Xie et al., 2022, would not meet the 3-month persistence criterion. Such delayed, acute events, although attributable to prior SARS-CoV-2 infection, would not in themselves qualify as Long COVID in accordance with the proposed definition. In reviewing the evidence for Long COVID symptoms, the committee reviewed a large and diverse evidence base describing and investigating new symptoms and manifestations of Long COVID in different organ systems. The committee also reviewed studies indicating exacerbation of pre-existing conditions. The process of narrowing down a list of hundreds of reported signs, symptoms, and diagnosable conditions identified in the evidence review was challenging. In selecting the illustrative symptoms and diagnosable conditions, the committee considered prevalence data from the evidence and the engagement process. In future iterations of the Long COVID definition, it may be desirable to enlist specific inclusion criteria for the illustrative symptoms, such as a threshold frequency of their occurrence among patients and specificity to Long COVID. Multiple studies reported variable severity and variable time courses of symptoms among individuals with Long COVID, including examples of fluctuating, worsening, new- onset, improving, resolving, and stable courses for individual symptoms. Participants in the engagement process favored inclusion of a non-exhaus- tive list of possible symptoms and conditions in the definition, and there was strong support for including language around the fluctuating time course of some individuals’ symptoms. Some of the previously developed Long COVID definitions do give examples of common symptoms, and some mention relapsing-remitting and other possible time courses of symptoms, while others do not discuss these aspects. The committee elected to include both aspects in the 2024 NASEM Long COVID definition. PREPUBLICATION COPY—Uncorrected Proofs 

46 A LONG COVID DEFINITION Findings from the Evidence Review Signs, Symptoms, and Conditions: Disease definitions may or may not include listings of specific signs, symptoms, or symptom clusters; some definitions include required symptoms or symptom clusters. These choices can affect the balance between specificity and inclusivity of a definition. The 2024 NASEM Long COVID Definition includes a non-exhaustive list of possible manifestations but does not require any specific manifestation. Several studies and reviews have examined the most common or most specific symptoms and manifestations of Long COVID. For example, a RECOVER Initiative prospective cohort study (Figure 4) with 9,764 U.S. adults identified symptoms that were more frequent in individuals at a study visit 6 months or more after a positive SARS-CoV-2 test result than in individuals with no known infection history. The authors developed a Long COVID (post-acute sequelae of COVID-19, or PASC) score based on 12 symptoms; among these symptoms, there was a 15 percent or greater abso- lute difference between infected and uninfected individuals in frequencies FIGURE 4  Frequencies of new onset symptoms. NOTE: Among participants with Long COVID, the most common symptoms were post-exertional malaise (87 percent), fatigue (85 percent), brain fog (64 percent), dizziness or lightheadedness (62 percent), GI (59 percent), and palpitations (57 percent). SOURCE: Thaweethai et al., 2023. PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 47 of post-exertional malaise, fatigue, brain fog, dizziness, and gastrointestinal (GI) symptoms. Several other symptoms, such as shortness of breath, were common but were not included in the score because they were correlated with included symptoms (Thaweethai et al., 2023). Woodrow and col- leagues conducted a systematic review of studies on Long COVID preva- lence and manifestations that were published in 2020 and 2021. They found that fatigue, breathing problems, sleep problems, tingling or itching, and joint or muscle pain were among the most commonly reported Long COVID symptoms (Woodrow, 2023). In a 2021 analysis, Deer and colleagues used the Human Phenotype Ontology (HPO) standardized vocabulary to integrate data from 81 cohorts of patients who were at least 3 weeks post SARS-CoV-2 infection. The use of HPO vocabulary allowed the authors to combine data from cohorts based on patient reports, clinical examinations, and electronic health record (EHR) data pulls, helping overcome the challenge of heterogeneity in how patients and clinicians describe the same symptoms. The authors identified 287 phenotypic abnormalities; 155 of these abnormalities were reported in two or more cohorts and 25 were reported in 12 or more cohorts (Deer et al., 2021). Numerous primary studies and systematic reviews have investigated sequelae of SARS-CoV-2 infection in specific body systems. A non-exhaus- tive sampling from this large literature base is presented here. Cardiovascular and Pulmonary Symptoms and Conditions Many research teams have documented cardiovascular and pulmonary sequelae of SARS-CoV-2, including interstitial lung disease and other lung abnormali- ties (Bazdar et al., 2023; Woodrow et al., 2023); a new or increased require- ment for supplemental oxygen (Admon et al., 2023); platelet pathology and thromboembolic disorders (Pretorius et al., 2022; Shah et al., 2023; Turner et al., 2023); and arrhythmias and tachycardia (Mohammad et al., 2022). Autonomic Symptoms and Conditions Multiple research teams have investigated autonomic symptoms and conditions following SARS-CoV-2 infection. A Stanford-based, international study that recruited mainly through social media and in COVID-19 and Long COVID support groups surveyed adults with a history of test-confirmed or test-unconfirmed SARS- CoV-2 infection. Of 2,314 surveyed adults (87.3% female) who had symp- toms persisting beyond 30 days, 66% had median scores of 20 or greater on the Composite Autonomic Symptom 31 (COMPASS-31) assessment, indicating moderate to severe autonomic dysfunction (Larsen et al., 2022). Among 42 patients with moderate to severe post-COVID-19 fatigue and exertion intolerance who were evaluated by a medical team in Berlin, 32 patients also had COMPASS-31 scores above 20 (Kedor et al., 2022). An PREPUBLICATION COPY—Uncorrected Proofs 

48 A LONG COVID DEFINITION Australian team evaluated 33 adults with PASC who either presented to a cardiology clinic or were recruited from a Long COVID support group. These participants had a median total COMPASS-31 score of 38, and 79 percent met the criteria for POTS (Seeley et al., 2023). A systematic review of the literature as of April 2022 examined cases of new-onset autonomic dysfunction affecting the cardiovascular system within 6 weeks of confirmed SARS-CoV-2 infection. In the acute phase, reflex syncope was the most common form of autonomic dysfunction affecting the cardiovascular system, while in the chronic phase (>4 weeks after infection), postural orthostatic tachycardia syndrome (POTS) was the most common form. Patients with autonomic dysfunction >4 weeks after infection were mostly female and younger; only 15 percent of these patients experienced a full recovery during the follow-up period (19 ± 16 weeks) (Reis Carneiro et al., 2023). Neurological and Psychiatric Symptoms and Conditions Multiple authors have investigated neurological and psychiatric complications and symptoms following SARS-CoV-2 infection, including stroke, migraine, smell and taste problems, cognitive impairment, difficulty concentrating, anxiety, and mood disorders (Crivelli et al., 2022; Nuzzo et al., 2021; Ong et al., 2023; Park et al., 2022; Premraj et al., 2022; Wingrove et al., 2023; Xu et al., 2022). A weighted analysis of a large Department of Veterans Affairs dataset, based on electronic health records, showed increased risk of epilepsy and seizures (hazard ratio 1.80), of cognition problems (haz- ard ratio 1.80), and of other neurologic sequelae in patients 12 months post SARS-CoV-2 infection compared with uninfected controls (Xu et al., 2022). To investigate sleep disturbances following COVID-19, Linh and colleagues conducted a systematic review and meta-analysis of worldwide studies focusing on adults at least 1 month following a SARS-CoV-2 infec- tion. In the subset of studies with an uninfected or pre-infection control group, the authors found an increased prevalence of sleep disturbances in post-COVID-19 patients, with an odds ratio of 2.00 (1.28, 3.14) (Linh et al., 2023). A large community-based study in the United Kingdom (112,964 participants aged 18 or older) used a computerized cognitive test battery to evaluate cognitive functioning in individuals who had recovered from COVID-19 within 4, 4–12, or >12 weeks; individuals who reported unre- solved persistent symptoms and were at least 12 weeks post-infection; and individuals who had no history of COVID-19. Results showed that participants whose COVID-19 symptoms had resolved had small deficits in cognitive performance (approximately -0.2 standard deviation on average) compared with those with no history of infection. Those with unresolved PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 49 symptoms beyond 12 weeks after infection had greater deficits in perfor- mance on cognitive tasks (approximately -0.4 SD on average), particularly in memory, reasoning, and planning tasks. Having had a more severe acute infection and having been infected during earlier periods in the pandemic were each associated with higher probabilities of cognitive impairment (Hampshire et al., 2024). It is important to note that some studies have not confirmed cognitive impairment in cohorts of Long COVID patients. In a narrative literature review by Garmoe and colleagues, most studies supported the occurrence of cognitive impairment in some individuals after SARS-CoV-2 infection, but a subset of three small studies (n=49- 189 patients) failed to demonstrate cognitive dysfunction attributable to COVID-19 up to 6 months after infection (Garmoe et al., 2024) Systemic, Musculoskeletal, Rheumatic, and Immune-Related Symptoms and Conditions  Multiple research teams have documented post-COVID-19 systemic, musculoskeletal, rheumatic, and immune-related complications, including new-onset lupus, rheumatoid arthritis, and Sjogren’s syndrome (Chang at al., 2023; Ciaffi et al., 2023; Kioi et al., 2023). Studies have also reported hyperlipidemia (Xu et al., 2023) and new-onset diabetes (Qeadan, et al., 2022). For example, a retrospective analysis of data from 27 million U.S. patients found that having a diagnosed SARS-CoV-2 infection was associated with significantly increased risk of new-onset type 1 diabetes mel- litus compared with individuals with no history of SARS-CoV-2 infection, and this risk was disproportionately higher among American Indian/Alaska Native, Asian/Pacific Islander, and Black participants (Qeadan et al., 2022). Additionally, other teams have investigated similarities and possible overlap between Long COVID mast cell activation syndrome (MCAS) (Szukiewicz et al., 2022), fibromyalgia (Clauw and Calabrese, 2024; Gavrilova et al., 2022; Savin et al., 2023), and connective tissue disorders like Ehlers-Danlos syndrome (EDS) (Lim et al., 2023; Pollack et al., 2023). The RECOVER Initiative and several studies have documented that up to 40 percent of patients with clusters of persistent symptoms meet the diagnostic criteria for ME/CFS following SARS-CoV-2 infection (Bonilla et al., 2023; Kedor et al., 2022; Mancini et al., 2021; Sherif et al., 2023). Persistent fatigue following SARS-CoV-2 infection is well-documented (Sherif et al., 2023; Zhao et al., 2023). Exacerbation of Pre-Existing Conditions Long COVID can present as a new condition that develops after SARS-CoV-2 infection. It can also worsen pre-existing health conditions, as shown in several studies. Among individuals with pre-existing type 1 diabetes, a history of SARS-CoV-2 infection was associated with an increased risk of experiencing diabetic PREPUBLICATION COPY—Uncorrected Proofs 

50 A LONG COVID DEFINITION ketoacidosis (Qeadan et al., 2022). The American Diabetes Association’s Standards of Care in Diabetes–2024 guideline provides information on the occurrence of new-onset diabetes after SARS-CoV-2 infection and includes a recommendation that individuals with underlying diabetes who have had COVID-19 should receive follow-up to assess possible complications and symptoms (American Diabetes Association Professional Practice Com- mittee, 2024). Among patients with POTS evaluated at a University of Toledo center, 68 percent (28/41) of those infected with SARS-CoV-2 had a worsening of their baseline POTS symptoms that persisted more than 1 month post-infection, and 29 percent had persistent, exacerbated symptoms despite escalation of therapy (Meenakshisundaram et al., 2024). Temporal Pattern and Duration of Symptoms: Another notable feature of Long COVID is the variable course of symptoms. An international sur- vey distributed primarily to participants in COVID-19 support and advo- cacy groups found that 85.9 percent of respondents who had experienced COVID-19-related illness for at least 28 days reported that they had experi- enced relapses of symptoms during the survey period (which spanned up to 7 months) (Davis et al., 2021). In the ComPaRe Long COVID cohort (968 patients), part of a prospective study in France, 33.3 percent of 150 par- ticipants who reported full symptom remission later experienced a relapse (Tran et al., 2022). Other studies provide evidence that neurocognitive symptoms can worsen over time and that new manifestations, such as cognitive symptoms, post-exertional malaise, paresthesia, and parosmia, commonly appear for the first time months after the initial infection (Apple et al., 2022; Davis et al., 2021; Tran et al., 2022). In a ComPaRe Long COVID cohort study, the prevalence of individual symptoms decreased (e.g., cough, change/loss of taste), increased (e.g. paresthesia, back/neck pain), or remained stable (e.g., dyspnea, word finding problems) between 2 months and 1 year after infection (Tran et al., 2022). Similarly, in a sys- tematic review and meta-analysis based on 63 articles published worldwide in September 2021 or earlier, Alkodaymi and colleagues found that certain symptoms (cough, loss of taste, loss of smell, headache) were most com- monly reported during the 6- to <9-months follow-up interval, while other symptoms (fatigue, myalgia, dyspnea, and sleep disorder) were most com- mon in the >12-month follow-up period (Alkodaymi et al., 2022). In an integrated model for overall symptom trajectory proposed by Fernández-de-las-Peñas and coauthors, post-COVID-19 symptoms are divided into “new-onset” (symptoms that first appeared after infection with COVID-19) and “exacerbated” (pre-existing symptoms that wors- ened after infection with COVID-19); symptoms can be further labeled as PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 51 “fluctuating,” “progressive,” and “continuous.” (Fernandez-de-las-Penas et al., 2021a). Several studies provide data on the overall durations of symptoms among patients. Among 1,106 randomly sampled adults with wild-type SARS-CoV-2 infections from Zurich, Switzerland (which had mandatory reporting), 27.1 percent, 22.9 percent, and 18.5 percent of infected indi- viduals said they had not fully recovered at 3, 6, and 12 months after infection, respectively (Ballouz et al., 2023). The ComPaRe Long COVID cohort study found that, of patients experiencing symptoms at 2 months post-infection, 85 percent still had symptoms at 12 months after infection (Tran et al., 2022). Symptoms and conditions associated with Long COVID can persist for multiple years. For example, a large Department of Veterans Affairs cohort study found that 2 years after infection, participants who had not been hospitalized during their acute infection still had an increased risk for 24 of 77 sequelae analyzed, including sequelae affecting the neurologic, musculoskeletal, and gastrointestinal systems. Meanwhile, participants who had been hospitalized during the acute phase had an increased risk of 50 of 77 sequelae at 2 years, including those affecting every organ system (Bowe et al., 2023). Severity of Symptoms: Long COVID symptoms can range from mild to severe. In the RECOVER Initiative study by Thaweethai and colleagues in which the authors developed a PASC score based on 12 symptoms, patients who met the authors’ criteria for PASC had responses on the Patient- Reported Outcomes Measurement Information System (PROMIS) Global 10 scale ranging from “not at all” to “completely” for ability to carry out every day physical activities, and responses ranging from “poor” to “excel- lent” on both quality of life and general physical health. Higher PASC score was associated with worse responses on PROMIS Global 10 (Thaweethai et al., 2023). As of 2023, estimates suggest that over 5,000 U.S. death cer- tificates have attributed Long COVID (or related terms) as an underlying or contributing cause of death (Ahmad et al., 2022; Rapaport, 2024).1 A large Department of Veterans Affairs cohort study found that veterans who had been hospitalized with COVID-19 had an elevated risk of death which persisted throughout the 2-year follow-up period, compared with a control group of veterans with no known SARS-CoV-2 infection. Compared with the control group, veterans who were not hospitalized during their acute COVID-19 illness were at increased risk of death during days 91–180 and at increased risk of hospitalization during days 361–540 following their acute infection (Bowe et al., 2023). 1  This sentence was modified after release of the prepublication version of the report to be more consistent with the cited references. PREPUBLICATION COPY—Uncorrected Proofs 

52 A LONG COVID DEFINITION Findings from the Engagement Process Participants noted that many people will understand Long COVID through symptoms. Including a list of possible symptoms and condi- tions associated with Long COVID could help participants understand the breadth and severity of Long COVID. Participants emphasized inclu- sion of the most common symptoms: “I do think that listing the three or four major symptoms that most studies are showing—brain fog, fatigue, shortness of breath—should be part of the definition”; another proposed, “Cognitive dysfunction is as important to include as physical symptoms, especially since cognitive dysfunction is highly stigmatized.” Participants also emphasized the relapsing and remitting nature of symptoms. Some mentioned that these recurring symptoms hindered access to proper care, as some physicians lacked an understanding or were skeptical of this pattern. For instance, one participant said, “Many of my patients, when we would talk about the definition, would find it validating that the definition itself from the WHO said that the symptoms were intermittent. I think that’s a key part of the definition, because the patients have often experienced medi- cal gaslighting. A lot of that is due to the inconsistency of the symptoms. I liked having that phrase in the definition that I can tell my patients, ‘Look, that’s part of this condition, and this is what you could show your employer and your family.’ I like that.” Lessons from Existing Definitions The OASH and CDC definitions acknowledge the heterogeneity of presentations by including wording like “multisystemic” and “wide range” without listing specific symptoms. The NICE definition also mentions “clus- ters of symptoms” but does not list specific symptoms and acknowledges that symptoms may be multisystemic and heterogeneous. By contrast, the WHO Adults and Pediatrics definitions give specific common symptoms, while stating that other symptoms may appear. It is useful to consider whether any symptom or cluster of symptoms could be considered a hallmark or cardinal symptom of Long COVID. However, considering the heterogene- ity of Long COVID manifestations, a definition that does not require any specific symptom or symptom cluster has the advantage of reducing false negative classifications. A definition that provides a non-exhaustive list of common symptoms may help balance the goals of specificity and inclusiv- ity. The OASH definition includes the phrase “with the possibility of severe and life-threatening events even months or years after infection.” The other existing definitions considered here do not directly mention severity, though both WHO definitions state that symptoms generally impact everyday func- tioning. A sound definition asserts what a term means rather than stating PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 53 what it does not mean; hence the 2024 NASEM Definition avoids stating, for example, that Long COVID is not one condition. How Does the Definition Address Equity? Important Features LC can affect children and adults, regardless of health, disability, or socioeconomic status, age, sex, gender, sexual orientation, race, ethnicity, or geographic location. The 2024 NASEM Long COVID Definition does not exclude patients based on demographic factors, preexisting conditions, vaccination status, or history of antiviral use. This definition applies to both adult and pediatric patients. Equity needs to be considered at multiple steps in a Long COVID patient’s journey to obtain care and services. Socioeconomic factors, inequality, discrimination (based on race and gender, among others), bias, and stigma affect whether patients can receive a diagnosis and benefit from Long COVID-targeted health care or services. These factors include but are not limited to access to COVID-19 testing during acute illness, access to evaluation for possible Long COVID, the willingness of physicians to diag- nose a particular patient, access to insurance benefits, and patients’ fears of stigmatization that could result from having a Long COVID diagnosis (Bergmans et al., 2023; HHS and OASH, 2022; Kim et al., 2023; Kromydas et al., 2023). Among participants in the engagement process, there was support for recognition of the impact of social determinants of health on the risk of Long COVID, the impact of cultural factors on decisions to seek healthcare, and the impact of Long COVID itself on financial status, especially among those with fewer economic resources. In the evidence review, the commit- tee encountered data gaps, highlighting the possibility that underdiagnosis of Long COVID in communities with less healthcare access and under- representation of or under-reporting on some non-white groups in Long COVID research is influencing the Long COVID research base. However, the committee found evidence suggesting that Long COVID more fre- quently affects women, and some research supporting differences in Long COVID prevalence or manifestations among different age groups, among different racial and ethnic groups, or based on vaccination status. While the committee recognizes such differences in risk factors, the 2024 NASEM Long COVID Definition emphasizes that a Long COVID diagnosis may be considered regardless of health status, vaccination history, or demographics. PREPUBLICATION COPY—Uncorrected Proofs 

54 A LONG COVID DEFINITION Findings from the Evidence Review Social, economic, and environmental factors may affect individuals’ risk for and burden of Long COVID as well as their access to health care for diagnosis and management of Long COVID. Among non-hospitalized individuals who experienced acute SARS-CoV-2 infection, one study found an 11 percent increased risk (HR 1.11, 95% CI 1.07–1.16) of COVID-19 symptom persistence among people in the most socioeconomically deprived versus the least deprived category (Subramanian et al., 2022). Researchers analyzing zip code and electronic health record data from two large clinical research networks found that among people with COVID-19, exposure to disadvantaged environmental characteristics (including certain air pollut- ants, limited food access, and overall neighborhood deprivation) was associ- ated with increased risk for Long COVID (Zhang et al., 2023). Numerous studies indicate that Long COVID is more frequently diag- nosed and reported among women (Government of Canada, 2023; M. M. Jacobs et al., 2023; Sylvester et al., 2022). The U.S. Census Bureau and the National Center for Health Statistics’ Household Pulse survey data from March 5 to April 1, 2024, which is based on self-reports, estimates that 21.1 percent of adult women and 13.9 percent of adult men in the United States have ever experienced Long COVID-19 symptoms that lasted 3 months or longer (CDC, 2024a). In the pediatric population as well, girls may be at greater risk for Long COVID (Zheng et al., 2023). The frequencies of specific Long COVID symptoms and manifestations may differ between women and men, and there is some evidence for sex-specific differences in immunological responses during Long COVID (Jiang et al., 2023; Silva et al., 2024; Sylvester et al., 2022). Additional evidence regarding sex- and gender-related risk factors for Long COVID is discussed below in the sec- tion “How Does the Definition Address Risk Factors?” Results of some analyses suggest that the risk of Long COVID after an acute infection may differ among racial and ethnic groups in the United States, with most studies indicating a higher risk for Hispanic and Black individuals and a lower risk for Asian individuals compared with non-His- panic White individuals (CDC, 2024a; Cohen and van der Meulen Rodgers, 2023; M. M. Jacobs et al., 2023; Louie and Wu, 2023). Members of other racial and ethnic groups, such as American Indian, Alaska Native, Native Hawaiian, Pacific Islander, and multiracial individuals, are not tracked separately in most studies. The risk of Long COVID among these groups is currently a data gap. In the U.S. Census Bureau and National Center for Health Statistics’ Household Pulse survey data from March 1 to April 1, 2024, 21.2 percent of adult Hispanic or Latino respondents, 17.7 percent of non-Hispanic White respondents, 12.9 percent of non-Hispanic Black respondents, 12.1 PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 55 percent of non-Hispanic Asian respondents, and 20.7 percent of non-His- panic members of other races and multiple races reported having ever expe- rienced Long COVID-19 symptoms that lasted 3 months or longer (CDC, 2024a). A two-stage modeling analysis based on Household Pulse data (four releases between June 1, 2022, and October 17, 2022) by Jacobs and colleagues (2023) that considered only participants who had COVID-19 found a comparatively higher risk of Long COVID among Hispanic par- ticipants and a slightly higher, statistically significant risk of Long COVID among Black participants (M. M. Jacobs et al., 2023). Several studies sug- gest that the frequency of specific symptoms and manifestations of Long COVID may differ among racial and ethnic groups (Khullar et al., 2023; Xie et al., 2021). In a series of interviews by Bergmans and colleagues, multiple Black participants with Long COVID described their experiences of being treated dismissively by doctors or being sent home with COVID-19 or Long COVID symptoms (Bergmans et al. 2022). In another report by Bergmans and col- leagues, based on a series of interviews with Black Americans who have Long COVID, participants described challenges in receiving care (including racial bias in medical treatment, health care costs, insurance coverage, and others) as well as impacts of social determinants of health on individuals’ ability to manage their symptoms (including housing quality, neighborhood socioeconomic status, and access to healthy food). Black Americans are also underrepresented in Long COVID research (Bergmans et al., 2023). Existing gaps in access to health care for the diagnosis of Long COVID could in turn affect research on Long COVID. As part of the RECOVER initiative, Hill and colleagues conducted a retrospective case–control study based on EHR data from 31 health systems; though the study was designed to identify risk factors for Long COVID, the authors also found that living in a county with a greater number of physicians per capita was associated with higher likelihood of Long COVID diagnosis or care at a Long COVID clinic. This suggests that underdiagnosis may be occurring among people with less access to health care. The authors point out that studies relying on EHR data may be influenced by existing biases and disparities in health care access (Hill et al., 2023). Ensuring the inclusion of all ethnic and racial groups in future research will be critical to improving the Long COVID research base and informing patient care (Bergmans et al., 2023; Khullar et al., 2023). Long COVID can occur in the pediatric population and in adults of any age (SeyedAlinaghi et al., 2023). As in adults, Long COVID in children and adolescents can manifest with symptoms and conditions affecting a wide range of body systems or as an exacerbation of underlying conditions, or both (Rao et al., 2024). PREPUBLICATION COPY—Uncorrected Proofs 

56 A LONG COVID DEFINITION Findings from the Engagement Process Participants agreed that the definition should apply equitably to all people with Long COVID and that it is important to recognize that dif- ferent people have different symptoms and experiences. Participants also emphasized the importance of a definition that supports access to services for people with Long COVID, thereby advancing equity. Many participants also deemed it significant to include the financial challenges posed by Long COVID to individuals of low socio-economic status; a majority indicated a desire to acknowledge the social determinants of health that can influ- ence the likelihood of developing Long COVID, such as poverty, race, and geographic location. Comments were also shared about explicitly linking the definition for Long COVID to equity considerations due to the impact that COVID-19 and Long COVID have had at an individual level and also within communities that have less access to care or to economic resources. A participant said, “It could be helpful to include a specific statement around health equity in a Long COVID definition. That would maybe be a little unusual to include in a definition, but it is important, if not in the definition, somewhere else.” Participants also suggested that the definition reflect cultural humility and sensitivity. For example, one participant com- mented, “Then there is also a cultural aspect in some communities where if you literally are not dropping [to the floor], you are fine to go to work. That is really problematic with a condition like Long COVID where your prognosis is significantly negatively affected when you try to do something like push through.” Lessons from Existing Definitions While existing definitions of Long COVID exhibit an overarching aim towards inclusivity, a specific mention of equity considerations is notably absent from all definitions. All definitions approach Long COVID primar- ily from a clinical standpoint, largely overlooking the socioeconomic and demographic factors that could markedly influence Long COVID’s distribu- tion, severity, and patient outcomes. In 2022, a group of stakeholders informed by data from the CLoCK Study of Long COVID in children and young people and by patients’ experiences developed a research definition for Long COVID in children and young people using a modified Delphi process to achieve consensus (Stephenson et al., 2022). In 2023, the WHO released a separate Pediatric definition for Long COVID. PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 57 How Does the Definition Address Functional Impairment? Important Features LC can impair individuals’ ability to work, attend school, take care of family, and care for themselves. It can have a profound emotional and physical impact on patients and their families and caregivers. A definition for Long COVID may or may not address its potential effects on functionality and daily living or its effects on overall well-being (e.g., financial, employment, quality-adjusted life-years [QALYs]). The 2024 NASEM Long COVID Definition emphasizes that some individuals with Long COVID are severely affected and can have a variety of activity limi- tations. This can profoundly affect patients’ and caregivers’ lives and is an important feature of Long COVID. In the evidence review, the committee found publications documenting a range of mild to severe functional impairments, activity limitations, and quality of life impacts in individuals with Long COVID. It is also impor- tant to note that Long COVID symptoms may affect functionality in dif- ferent domains. For example, individuals may be impaired in daily home functioning, working capacity, or both (Ford et al., 2023). Participants in the engagement process noted that impairments in Long COVID can be invisible to others and supported inclusion of functional impairment in the definition of Long COVID. Findings from the Evidence Review Many primary studies and systematic reviews have highlighted the potential impact of Long COVID on activities of daily living. The authors of a systematic review based on a literature search of articles published in July 2021 or earlier investigated the range of activity limitations, physical function limitations, and health-related quality of life (HRQoL) impacts in COVID-19 survivors, many of whom had been hospitalized (de Oliveira Almeida et al., 2023). Participants were evaluated during time spans rang- ing from the day of admission to rehabilitation after acute care to 7 months after discharge from the hospital. The analysis revealed impaired perfor- mance among the COVID-19 survivors on pulmonary function and muscle strength tests, physical tests such as the 6-minute walk test, and daily living activities scales such as the Lawton scale (de Oliveira Almeida et al., 2023; Piquet et al., 2021). A systematic review by Figuereido and col- leagues focused on HRQoL after discharge among patients who had been PREPUBLICATION COPY—Uncorrected Proofs 

58 A LONG COVID DEFINITION hospitalized with COVID-19 (Figueiredo et al., 2022). Among six included studies that compared people hospitalized for COVID-19 with either the general population or matched controls, all found lower scores on HRQoL instruments (SF-36, 15D, St. George’s Respiratory Questionnaire, or Short Form-12) for the post-COVID-19 individuals at time points ranging up to 3 months post-discharge. The authors write that although lower HRQoL can persist for months after hospitalization for COVID-19, partial improve- ments in HRQoL are often observed soon after hospitalization (Figueiredo et al., 2022). As reported in the U.S. Census Bureau and the National Center for Health Statistics’ Household Pulse Survey data (March 5 to April 1, 2024), 23.8 percent of U.S. adults currently experiencing Long COVID were esti- mated to have significant activity limitations and 78.7 percent were esti- mated to have any activity limitation from Long COVID (CDC, 2024a). In a UK study, patients receiving care at post-COVID-19 clinics were evaluated using the Work and Social Adjustment Scale (WSAS), a patient-reported measure of functional impairment. Of the participants, 53 percent scored over 20, indicating “moderately severe” impairment; impacts were observed across all five domains of the WSAS (Walker, 2023). That study also found a large impact on quality of life for many patients, with a median score of 0.60 (IQR 0.41 to 0.71) on the EQ-5d, a measure of HRQoL (Walker, 2023). In the LONG-COVID-EXP-CM study, a multicenter cohort study in Spain, 1,593 patients who were hospitalized during the first wave of the pandemic were later interviewed regarding their self-perceived functional status and limitations in comparison with their status before their COVID- 19 illness. Interviews took place at T1 (mean 8.4 months after discharge) and T2 (mean 13.2 months after discharge). The percentages of patients reporting limitations at T1 and T2, respectively, were 20.9 percent and 12.8 percent for occupational activities, 30.1 percent and 20.8 percent for leisure/social activities, 27.1 percent and 18.1 percent for instrumental activities, and 19.9 percent and 13.7 percent for basic activities (Fernández- de-Las-Peñas et al., 2022c). Long COVID can have serious impacts on employment. An analysis of the University of Southern California’s longitudinal survey, Understand- ing America Study, found that 25.9 percent of “long haulers” in mid-2021 reported that their symptoms affected their employment or work hours; these affected individuals worked 50 percent fewer hours compared to people who had never had COVID-19 (Federal Reserve Bank of Minne- apolis, 2022). Among patients in Long COVID support groups who were surveyed in the fall of 2020, 45.2 percent said they required reduced work hours, and an additional 22.3 percent said they were not working due to their symptoms (Bowe et al., 2023). Based on estimates of Long COVID prevalence rates, impairment rates, and labor impacts, economist David PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 59 Cutler estimated that Long COVID has caused 3.5 million people to leave the labor force as of 2022. With the assumption that Long COVID cases last an average of 5 years with no change in severity, Cutler estimated that the cost of Long COVID would amount to $3.7 trillion (including the costs of lost QALYs, lost earnings, and increased medical spending) (Cutler, 2022). An April 2024 analysis by Economist Impact, found that Long Covid symp- toms have prompted some individuals to leave the workforce (estimated 953.6 million hours lost), others take time off work (estimated 363.3 mil- lion hours lost) or reduce their work hours due to symptoms (estimated 177.5 million hours lost)—resulting in 1.5 billion work hours lost in 2024 and a potential cost of more than $152.6 billion (Economist Impact, 2024). Although there is little available published data regarding the impact of Long COVID on caregivers and families of patients, anecdotal reports indicate that caregivers are experiencing a wide range of impacts including financial and logistical challenges, impacts to employment, and impacts to their own health, among others (McGowan, 2023; Olsen, 2021). While caregiver burdens and impacts have been documented in other chronic ill- nesses (Strang et al., 2018), caregiver and family burden in the context of Long COVID appears to be a research gap that merits further attention. Findings from the Engagement Process Participants consistently called for the concept of impairment to be included in the definition of Long COVID, as many patients find that their symptoms interfere with daily functioning (such as socially, occupation- ally, mentally, and other aspects of daily life). The inclusion of a statement on impairment could affect degrees of access to disability and services accommodations. For example, one participant noted, “What we need to be able to do with the definition, or a subset of it, is determine a degree of impairment because so many people are disabled.” Another underlined the importance of including impairment in the definition, because many of the symptoms are invisible: “When I think about Long COVID, it’s about the symptoms that are causing functional impairment that you do not see.” Lessons from Existing Definitions Only one of the existing definitions, that of the USG, directly discusses Long COVID severity. The USG stands out for its acknowledgment of Long COVID’s potential severity, noting that Long COVID “may present with a relapsing-remitting pattern and progression or worsening over time, with the possibility of severe and life-threatening events even months or years after infection.” This characterization underscores the broad spectrum of severity that Long COVID can encompass, hinting at its substantial PREPUBLICATION COPY—Uncorrected Proofs 

60 A LONG COVID DEFINITION implications for patient health and quality of life. Although this descrip- tion hints at the variable severity and potential for significant health con- sequences, it stops short of explicitly discussing “functional impairment.” Additionally, a Long COVID definition may need to include people with different current levels of severity or impairment. On the other hand, the WHO Adult and WHO Pediatric definitions directly address the functional consequences of Long COVID. The WHO Adult definition mentions that symptoms “generally have an impact on everyday functioning,” including issues like fatigue, difficulty breathing, and cognitive challenges. The WHO Pediatric definition broadens this perspective, noting that symptoms “gen- erally have an impact on everyday functioning” and can influence various aspects of a child’s life, including eating habits, physical activity levels, behavior, academic achievements, and social interactions. These insights from the WHO underscore the critical need to consider functional impair- ment and the spectrum of severity when fully understanding Long COVID’s effects on individuals’ lives. How Does the Definition Address Alternative Diagnoses? The 2024 NASEM Long COVID Definition does not include men- tion of alternative diagnoses. Many diseases, like Long COVID, can share symptoms or character- istics with other conditions, leading to overlap and confusion in diagno- sis. A disease definition could list exclusionary conditions or indicate the consideration of exploring alternative diagnoses (Lim and Son, 2020). For example, the Ramsay definition of ME/CFS considers depression and anxi- ety to be exclusionary conditions. In contrast, the IOM ME/CFS definition does not list exclusionary conditions (Lim and Son, 2020). The committee elected not to include a statement regarding exclusions or alternative diagnoses in the 2024 NASEM Long COVID Definition. First, there is no scientific evidence that any medical condition prevents or cannot exist alongside Long COVID. Second, the history of a similarly debilitating medical condition, ME/CFS, illustrates that such requirements can lead to a denied or delayed diagnosis. Patients have reported that clinicians were reluctant to diagnose them with ME/CFS if any other co-morbidity—like reactive depression—was present that could partially explain their symp- toms even as aspects of their presentation—like a post-acute-infectious onset or post-exertional malaise—were not typical for mood disorders. Conversely, clinicians have voiced hesitation to diagnose ME/CFS because PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 61 they were not confident they had eliminated enough alternative conditions when the disease was framed as a diagnosis of exclusion The 2024 NASEM Long COVID Definition does include ME/CFS and POTS, among others, as examples of diagnosable conditions that can be part of the picture of Long COVID. These and other potentially overlap- ping conditions are compatible with a diagnosis of Long COVID. To the contrary, evidence exists these conditions may be more common among Long COVID patients than the general population, and that they might share some common physiological pathways (Komoraff and Lipkin 2023). For example, patients may be diagnosed with Long COVID and POTS or Long COVID and ME/CFS. In addition, patients may be diagnosed with Long COVID and new-onset diabetes, Long COVID and new-onset rheumatoid arthritis, etc. Long COVID is thus an umbrella term and can be diagnosed alongside associated conditions. This is in accord with the WHO-Pediatrics definition for Long COVID, which states that other diag- noses do not exclude a diagnosis of Long COVID but is in contrast to the WHO-Adults definition and the NICE definition, which specify that Long COVID symptoms must not be “explained by an alternative diagnosis.” Lack of language concerning alternative diagnoses does not absolve health care professionals of their clinical responsibilities, which include generating a reasonable differential diagnosis, testing for other causes of a patients’ symptoms, re-considering initial impressions, and monitoring patients with uncertain diagnoses regularly and carefully. Despite not including this ele- ment in the definition, the committee articulates relevant findings below. Findings from the Evidence Review When discussing the long-term ramifications of acute SARS-CoV-2 infection, it is useful to consider another well-described condition called the post-intensive care syndrome (PICS). PICS is an intense form of suffering that persists during the months and years following treatment in medical and surgical intensive care unit (ICUs), which of course includes survivors of acute SARS-CoV-2 infection (Nanwani-Nanwani et al., 2022; Weidman et al., 2022). There are hundreds of thousands of people around the world who—after prolonged hospital courses on mechanical ventilation and dialy- sis during acute SARS-CoV-2 infection—are now experiencing cognitive, mental health, and physical disabilities (Hägglöf et al., 2023; Heesakkers et al., 2022; Jackson et al., 2010; Neville et al., 2022; Roedl et al., 2022; Sevin and Ely, 2022). These hospital survivors are struggling to recover from a constellation of symptoms and conditions that often include both PICS and Long COVID. These two conditions have a tremendous amount of overlap, and it is important to acknowledge that some patients have both PICS and Long COVID. PREPUBLICATION COPY—Uncorrected Proofs 

62 A LONG COVID DEFINITION There are some relevant distinctions. Both PICS and Long COVID involve structural and functional organ dysfunction of the brain, heart, lungs, kidneys, muscles, and nerves. The mechanisms of PICS and Long COVID likely vary (e.g., immune dysregulation, mitochondrial disease). PICS, unlike Long COVID, can include non-infectious triggers (pancreatitis) as well as infectious triggers (including infections unrelated to SARS-CoV-2). Clinically, both often leave patients with troublesome cognitive impairment that meets the criteria for mild to moderate dementia. While there is early evidence to indicate that PICS patients may find improvements in memory and executive function after cognitive rehabilitation, Long COVID patients anecdotally report that “brain exercises” or even simply reading causes them days of mental worsening. Similarly, while fatigue and neuromyopathy are common in PICS, post-exertional malaise that Long COVID patients experience is often more unpredictable and refractory than that seen in PICS (Jackson et al., 2012). It can be challenging clinically to differentiate within an individual patient between PICS and Long COVID related symp- toms, thus clinicians and patients must work together to decipher what therapies help versus which exacerbate symptoms. In summary, PICS can follow treatment in intensive care, and Long COVID can follow acute SARS-CoV-2 infection. While different in etiol- ogy, the two conditions overlap in symptoms of cognitive impairment and profound fatigue. When patients with Long COVID have also been treated with intensive care, either the infection or the experience of treatment, or both, could be contributing factors to their condition. Further research will be needed to understand better the pathophysiology of both PICS and Long COVID and to determine whether similar or different treatment regimens will provide the best care for patients. Similarly, a notable proportion of individuals presenting with symp- toms consistent with Long COVID also meet the diagnostic criteria for or exhibit symptoms of other diagnosable conditions such as POTS or other forms of dysautonomia (Larsen et al., 2022; Kedor et al., 2022; Seeley et al., 2023), or ME/CFS (Goldenberg, 2023; Morrow et al., 2022), although most ME/CFS criteria require at least 6 months of symptoms in contrast to the 3 months of symptoms required by the 2024 NASEM Long COVID Definition. POTS is a clinical syndrome often triggered by infec- tion that can include symptoms associated with a dysfunctional vasomotor and sympathetic nervous system (e.g., lightheadedness, palpitations, and headache) (Diekman and Chung, 2023; Morrow et al., 2022). ME/CFS is characterized by persistent fatigue, post-exertional malaise, unrefreshing sleep, orthostatic intolerance, and cognitive impairment; many but not all cases are reported to occur following an infectious or infectious-like illness (Goldenberg, 2023; Vivaldi et al., 2023). Long COVID may be con- sidered alongside dysautonomia presentations other than POTS, mast cell PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 63 activation syndrome (MCAS), or a range of other diagnoses (Diekman and Chung, 2023; Larsen et al., 2022; Morrow et al., 2022). Long COVID can be diagnosed alongside these and other conditions. Findings from the Engagement Process Most participants did not feel that the definition required language about excluding alternative diagnoses or considering Long COVID a diag- nosis of exclusion. However, one participant noted that when symptoms are solely attributed to Long COVID, there have been instances of missed diag- noses, like lung cancer, when patients await care at Long COVID clinics. Lessons from Existing Definitions The WHO Pediatrics definition acknowledges the potential for addi- tional diagnoses coexisting with Long COVID in children and adolescents, allowing for a broad diagnostic scope: “Workup may reveal additional diag- noses, but this does not exclude the diagnosis of post COVID-19 condition.” The WHO Adults definition and the NICE definition specify that Long COVID symptoms must not be “explained by an alternative diagnosis.” How Does the Definition Address Biomarkers? Important Features LC can be diagnosed on clinical grounds. No biomarker currently available demonstrates conclusively the presence of LC. Although Long COVID is defined as a chronic condition that occurs following a SARS-CoV-2 infection, the current evidence base does not indi- cate a clear pathobiology that is universal among patients. Long COVID symptoms appear to arise from pathobiological changes that span many different organ systems and tissues (Deer et al., 2021), and no definitive biomarker for Long COVID has yet been identified. It may be that no single or small number of biomarkers will explain the vast complexity of Long COVID (Al-Aly and Topol, 2024). From the evidence review and engagement process, the committee determined that the research and medical communities have not yet identi- fied any diagnostic tests that are well documented to have high specificity and sensitivity for Long COVID. Therefore, the 2024 NASEM Long COVID Definition does not include any requirements for biomarker testing that must be performed before diagnosing Long COVID. PREPUBLICATION COPY—Uncorrected Proofs 

64 A LONG COVID DEFINITION However, numerous studies and reviews have identified pathobiological abnormalities (e.g., immune dysfunction and coagulation abnormalities) in patients experiencing Long COVID (Figure 5) (Davis et al., 2023; Mohan- das et al., 2023; Turner et al., 2023). These findings, elaborated on below, make it clear that Long COVID is a physical health condition. These find- ings also raise the possibility that further discoveries will enable biomarkers to be incorporated in a revised, future Long COVID definition. Findings from the Evidence Review Over 200 potential biomarkers for Long COVID are currently under investigation, often in connection with hypotheses regarding pathobiology. However, more research is needed to overcome the limitations of the data, and candidate biomarkers need validation in larger studies, including clini- cal studies (Espin et al., 2023; Lai et al., 2023). In a systematic review of biomarkers in Long COVID published in January 2023, Lai and colleagues discussed 113 biomarkers that have been significantly associated with Long COVID in 28 eligible studies (Lai et al., 2023). These biomarkers include 38 cytokine/chemokines, 24 biochemical markers, 20 vascular markers, 6 neurological markers, 5 acute phase proteins, and 20 others. Of these, the authors state that upregulated interleukin 6, C-reactive protein, and tumor necrosis factor alpha could potentially be used to support Long COVID diagnosis or clinical management. In addition, increased levels of neurofilament light chain and glial fibrillary acidic protein could potentially serve as biomarkers for Long COVID with neurological manifestations, while increased transforming growth factor beta could potentially serve as a biomarker for Long COVID with pulmonary manifestations (Lai et al., 2023). A scoping review by Espín and colleagues published in May 2023 identified a longer list of 239 candidate biomarkers from 23 cohort studies; these include cellular biomarkers (e.g. T regulatory cells), immunoglobulins, cytokines/chemokines, and others. Some of the reviewed studies aimed to develop biomarkers for Long COVID risk prediction during acute illness, while others investigated potential biomarkers for specific Long COVID symptoms or manifestations (Espin et al., 2023). Several research groups have investigated the persistence of SARS- CoV-2 RNA or proteins after a SARS-CoV-2 infection. For example, one small study found S1 protein in the plasma of 64 percent of 22 people with Long COVID and in 35 percent of a control group of 17 people who had had COVID-19 but did not develop Long COVID; none of these par- ticipants had been vaccinated against COVID-19 (Schultheiss et al., 2023). Another research team used the Simoa (Quanterix) single molecule array detection platform to analyze the presence of SARS-CoV-2 spike, S1, and nucleocapsid antigens in 660 plasma specimens collected from 171 adults PREPUBLICATION COPY—Uncorrected Proofs 

Immune dysregulation Microbiota dysbiosis Autoimmunity and Blood clotting and Dysfunctional immune priming endothelial abnormalities neurological signalling Persistent infection Viral replication Reactivation Immune dysregulation, with Impacts of SARS-CoV-2 on Autoimmunity and primed Microvascular blood clotting Dysfunctional signalling in the or without reactivation of the microbiota and virome immune cells from molecular with endothelial dysfunction brainstem and/or vagus nerve underlying pathogens, (including SARS-CoV-2 mimicry including herpesviruses such persistence) as EBV and HHV-6 FIGURE 5  Hypothesized mechanisms of Long COVID pathogenesis. PREPUBLICATION COPY—Uncorrected Proofs NOTES: There are several hypothesized mechanisms for long COVID pathogenesis, including but not limited to immune dysregula- tion, microbiota disruption, autoimmunity, clotting and endothelial abnormality, and dysfunctional neurological signaling. EBV = Epstein–Barr virus; HHV-6 = human herpesvirus 6; SARS-CoV-2 = severe acute respiratory syndrome coronavirus 2. SOURCE: Davis et al., 2023. 65 

66 A LONG COVID DEFINITION during the 14 months following their RNA-confirmed SARS-CoV-2 infec- tions. They found at least one detectable SARS-CoV-2 antigen in 9.2 per- cent of the specimens, including those from 25 percent of the participants, compared to a 2 percent positivity rate for plasma specimens collected from 250 adults before 2022 (Peluso et al. 2024b). Evidence of SARS-CoV-2 persistence, including partial viral fragments, complete virus and viral RNA, have been identified months to years after acute infection in tissue biopsy studies and tissue autopsy studies, including in lymph nodes, central nervous system, lung tissue and many other tissue types (Proal et al., 2023) Multiple research groups have investigated immune dysfunction in Long COVID. According to the authors of a review published as part of the NIH-funded Researching COVID to Enhance Recovery (RECOVER) initiative, both innate and adaptive immune dysregulation may play a role, but many research questions remain open in this area (Mohandas et al., 2023). Potential immune-related contributors to Long COVID that are under investigation include persistent activation of inflammatory pathways in macrophages; transcriptional changes in monocytes, lymphocytes, or dendritic cells; activation of mast cells; and involvement of neutrophils, T cells, or B cells (Altmann et al., 2023; Mohandas et al., 2023; Peluso et al., 2024a). A cross-sectional immune phenotyping study by Klein and col- leagues compared participants with a Long COVID diagnosis to matched control individuals who either had no history of SARS-CoV-2 infection or had fully recovered after COVID-19. The participants with Long COVID had significant alterations in circulating myeloid and lymphocyte popula- tions and showed exaggerated humoral responses against both SARS-CoV-2 and herpesviruses (Klein et al., 2023). In an analysis based on the DigiHero cohort study in Germany, both individuals with Long COVID (median of 8 months post infection) and individuals without Long COVID (median of 7.5 months post infection) had markedly elevated levels of monocyte/ macrophage‐related soluble factors, including pro-inflammatory and pro- fibrotic cytokines, compared with people with no COVID-19 history (Schul- theiss et al., 2023). In a pre-print study of unvaccinated individuals 8 months post-acute COVID-19, those experiencing Long COVID symptoms (n=27) showed mis-coordination between humoral and cellular responses, signatures of inflammation, and T-cell perturbations suggestive of an ongo- ing immune response; these changes were not seen in those without Long COVID (n=16) (Yin et al., 2023). Multiple research studies have suggested that SARS-CoV-2 infection, particularly severe infection, can induce auto- antibodies. This phenomenon has also been noted with influenza infection, though to a lesser extent (Altmann et al., 2023; Proal et al., 2023). Other researchers have proposed that reactivation of latent Epstein-Barr virus (EBV) or other herpesviruses during SARS-CoV-2 infection could lead to PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 67 the production of autoantibodies (Altmann et al., 2023; Mohandas et al., 2023). Autoimmunity induced by SARS-CoV-2 infection has also been inves- tigated in multisystem inflammatory syndrome in children (MIS-C) and severe COVID-19 disease (Altmann et al., 2023; Mohandas et al., 2023). Several research groups have investigated the possibility that interac- tions between the effects of SARS-CoV-2 infection and pre-existing infec- tions (e.g. HIV) or reactivated, latent viruses (e.g. EBV) could contribute to the development of Long COVID through immune-mediated or other mechanisms (Chen et al., 2023; Peluso et al., 2022). Pointing to future research directions, an international scientific con- ference on Long COVID, held by the Keystone Symposia in August 2023, highlighted research on potential pathogenesis and mechanisms of Long COVID, connections to other post-infection complications, diagnosis and biomarkers, disease management, and potential treatment directions (Durst- enfeld et al., 2024). Findings from the Engagement Process Although there were varying views, overall participants emphasized that adding biomarkers at this time will not improve the definition. One participant said, “We don’t have good biomarkers, and I anticipate in a few years’ time, we probably will have more in the way of biomarkers. I think it’s good to acknowledge that we anticipate the evolution of the definition.” Lessons from Existing Definitions No existing Long COVID definitions include explicit mention of spe- cific biomarkers, but the USG definition includes an overarching statement that states, “Long COVID represents many potentially overlapping entities, likely with different biological causes and different sets of risk factors and outcomes.” How Does the Definition Address Risk Factors? The 2024 NASEM Long COVID Definition does not explicitly in- clude risk factors. Various risk factors, such as underlying comorbid conditions, may influence the risk and presentation of COVID-19 and Long COVID in a particular individual and may be useful in assessing individual patients or PREPUBLICATION COPY—Uncorrected Proofs 

68 A LONG COVID DEFINITION populations at risk. However, as risk factors do not in themselves define a disease, the committee chose not to include risk factors in the 2024 NASEM Long COVID Definition. Some participants in the engagement process supported including risk factors in the Long COVID definition. In the evidence review, the commit- tee found studies and systematic reviews that support possible risk factors for Long COVID, including female sex, certain health conditions, greater severity of initial COVID-19 illness, reinfection, and others. However, much of the data on Long COVID risk factors has limitations (e.g. some relies on EHR data, which is influenced by existing healthcare disparities) and some risk factors are difficult to disentangle from one another based on current data (e.g. certain SARS-CoV-2 strains encountered an increasingly vac- cinated population; certain health conditions could increase Long COVID risk directly or through increased risk of severe acute illness). The commit- tee’s decision not to include risk factors in the 2024 NASEM Long COVID Definition is in accord with the previous Long COVID definitions, which do not give risk factors. Findings from the Evidence Review Several systematic reviews have examined risk factors for Long COVID. A review and meta-analysis by Tsampasian and colleagues included 41 studies and over 860,000 patients. The authors found evidence in support of several risk factors for Long COVID, including female sex, age above 40, obesity, smoking status, history of hospitalization or ICU admission during acute COVID-19, and several pre-existing conditions (anxiety and/ or depression, asthma, chronic obstructive pulmonary disease, diabetes, immunosuppression, ischemic heart disease) (Tsampasian et al., 2023). The authors note that it is unclear whether obesity and smoking increase the risk for Long COVID directly or through an increased risk for severe COVID-19. In addition, the association between hospitalization/ICU admis- sion and Long COVID may be influenced by overlap with post-intensive care syndrome (Tsampasian et al., 2023). Some evidence suggests that individuals in middle age groups may be at highest risk for Long COVID. For example, a retrospective case–control study based on data from 31 health systems found that individuals from 40 to 69 years were more likely than adults in other age groups to be diagnosed with Long COVID or to visit a Long COVID clinic (odds ratio [OR] ranging from 2.32 to 2.58). This study also identified female sex, greater severity of acute infection, obesity, chronic lung disease, depression, and metastatic cancer as risk fac- tors. The authors note that the reliance on EHR data is a limitation of this study because such data may be influenced by existing biases and dispari- ties in health care access (Hill et al., 2023). An analysis of data from the PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 69 LONG-COVID-EXP-CM study, a multicenter cohort study in Spain, found that female sex was associated with greater odds of reporting ≥3 post- COVID symptoms at a mean of 8.4 months post-discharge (adjusted OR 2.54, 95% CI 1.671–3.865, p < 0.001) among 1,969 individuals who had been hospitalized with COVID-19 (Fernandez-de-Las-Peñas et al., 2022b). Among children and adolescents, a review and meta-analysis with 40 studies identified possible risk factors for Long COVID including older age; female sex; and history of severe COVID-19, multisystemic inflammatory disease, or hospitalization for COVID-19 (Zheng et al., 2023). Multiple studies support an association between greater severity of initial COVID-19 illness and higher risk for Long COVID. For example, Xie and colleagues investigated the impact of the severity of acute illness among 181,384 veterans (average age, 67.13; 90.47 percent male) using data in the U.S. Department of Veterans Affairs database. After adjustment for age, race, sex, comorbidities, and other factors, and after subtraction of background symptoms using a comparison group of uninfected veterans, the authors calculated the burden of Long COVID (defined as the number of individuals per 1,000 who had at least one of 33 specific SARS-CoV-2-at- tributed sequelae). They found that the adjusted burden of Long COVID at 6 months was 44.51 for non-hospitalized patients, 217.08 for hospitalized patients, and 360.16 for patients who had been in intensive care (Xie et al., 2021). As part of the LONG-COVID-EXP-CM study, 1,969 patients who were hospitalized with COVID-19 between February and May 2020 were interviewed a mean of 8.4 months post discharge. The number of symptoms at hospital admission (OR 1.309, 95% CI 1.15–1.49) and the number of days at the hospital (OR 1.01, 95% CI 1.007–1.017) were each associated (p<0.001) with greater number of post-COVID-19 symptoms at the time of the interview (Fernandez-de-Las-Peñas et al., 2022b). In addition to differ- ences in risk for Long COVID, some analyses support differences between hospitalized and non-hospitalized patients in the prevalence of specific post-COVID-19 symptoms (Fernandez-de-Las-Peñas et al., 2021b, 2023). Despite these associations, because non-severe COVID-19 cases have been far more prevalent than severe cases, most Long COVID cases occur in indi- viduals who did not have severe initial disease (Al-Aly and Topol, 2024). In support of the suggestion that SARS-CoV-2 reinfections are a risk factor for Long COVID or for increased severity of Long COVID, or both, an analysis of the U.S. Department of Veterans Affairs national health care database found that individuals with one or more reinfections (n=40,947), compared with those with a single infection (n=443,588), had increased risks of hospitalization, all-cause mortality, and sequelae affecting the pul- monary and multiple extra-pulmonary organ systems, in both the acute (0–30 days) and post-acute (30–180 days) phases after the reinfection. Dur- ing the post-acute phase, these risks gradually decreased, but elevated risks PREPUBLICATION COPY—Uncorrected Proofs 

70 A LONG COVID DEFINITION of death, hospitalization, and having at least one sequela persisted until 6 months after reinfection. In this study, reinfections were recorded between June 2020 and June 2022, when pre-Delta, Delta, and Omicron were the dominant variants (Bowe et al., 2022). There is some evidence to suggest that vaccination against SARS-CoV-2 may lower the risk of Long COVID. In the meta-analysis by Tsampasian and colleagues described above, among four high-quality studies that investigated the effect of vaccination (over 249,000 patients), individuals who had received two vaccine doses had a lower risk of Long COVID (OR 0.57; 95% CI 0.43–0.76) (Tsampasian et al., 2023). Another analysis of the Veterans Affairs database found that people with breakthrough SARS-CoV-2 infection (infection that occurs after vaccination) had a higher risk of post-acute sequelae at 6 months compared with uninfected controls (HR = 1.50, 95% CI: 1.46, 1.54), but a lower risk of post-acute sequelae compared with unvaccinated people infected with SARS-CoV-2 (HR = 0.85, 95% CI 0.82, 0.89) (Al-Aly et al. 2022). A staggered cohort study based on health records from the United Kingdom, Spain, and Estonia investigated whether vaccination reduces the overall risk of developing Long COVID (defined as having at least one of 25 Long COVID-associated, new symptoms >90 days after a SARS-CoV-2 infec- tion), whether through effects on the risk of infection or through effects on the risk of developing Long COVID after an infection. The analysis included >10 million unvaccinated and >10 million vaccinated individuals. After weighting to reduce confounding, the authors found that vaccination consistently reduced the risk of Long COVID; hazard ratios for a first dose varied from 0.48 to 0.71 depending on the database (Català et al., 2024). Some reports suggests that infection with SARS-CoV-2 strains circulat- ing earlier in the pandemic may be associated with higher risks for Long COVID (Hedberg et al., 2024; Antonelli et al., 2022). However, it can be difficult to disentangle the effects of the SARS-CoV-2 strain from the effects of concurrent changes in vaccination status, reinfection status, social cir- cumstances, and other confounders (Fernandez-de-Las-Peñas et al., 2022a). The authors of an observational cohort study in Italy used a multivariable logistic regression model to investigate the risk of Long COVID (defined in this study as persistent symptoms more than 4 weeks after SARS-CoV-2 infection) according to patients’ vaccination status and timing of infection. All participants in this study were health care workers and received their first and second vaccine doses in January–February 2021 and a booster dose in November–December 2021. Among 739 non-hospitalized adults with SARS-CoV-2 infection, Long COVID prevalence was 48.1 percent of those infected during wave 1 (February–September 2020, wild-type vari- ant), 35.9 percent in wave 2 (October 2020–July 2021, Alpha variant) and 16.5 percent in wave 3 (August 2021–March 2022, Delta and Omicron variants). However, in the same study, Long COVID prevalence varied from PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 71 41.8 percent among unvaccinated participants to 16.0 percent among those with three vaccine doses before infection (Azzolini et al., 2022). In the sec- ond report from the Canadian COVID-19 Antibody and Health Surveys, a series of self-report surveys based on random sampling of Canadian adults, the percentage of those infected who said they had persistent symptoms varied over time (27.3 percent of those infected in 2020, 26.7 percent of those infected from January 2021 to June 2021, 14.5 percent of those infected between July 2021 and Nov 2021, 12.7 percent of those infected between December 2021 and May 2022). However, these results may be affected by changes in both the predominant virus variant and changes in the percentage of vaccinated adults over time, as 72.1 percent of Canadian adults received their first vaccine dose by June 2021, and 86.7 percent had completed the primary vaccination series by December 2021. In the same survey, 25.0 percent of participants who were unvaccinated at the time of their initial infection reported persistent symptoms, compared with 13.2 percent of those who received two vaccine doses and 12.2 percent of those who received three vaccine doses (Government of Canada, 2023). Findings from the Engagement Process Participants underscored the importance of including specific risk fac- tors in the definition, citing them as key considerations for underserved and minority populations. Lessons from Existing Definitions No existing Long COVID definitions include explicit mention of specific risk factors, but the USG definition includes an overarching statement that “Long COVID represents many potentially overlapping entities, likely with different biological causes and different sets of risk factors and outcomes.” CONSIDERATIONS FOR IMPLEMENTATION AND DIFFERENT USES Participants highlighted the need to develop a definition that is both operational for varied uses and widely acceptable to all users of the defini- tion, with one participant noting, “It is important to consider the different ‘needs’ of the definition. For example, researchers want reproducibility; clinicians and patients want to help people get treatment.” The committee intends its definition to be applied to many purposes. These may include clinical care and diagnosis; eligibility for health services, insurance coverage, disability benefits, and school or workplace accommodations; public health; social services; policy making; epidemiology and surveillance; private and PREPUBLICATION COPY—Uncorrected Proofs 

72 A LONG COVID DEFINITION public research; and public awareness and education, especially for patients and their families and caregivers (Pan and Pareek, 2023; Soriano et al., 2022). In all these situations, an ideal Long COVID definition will likely need to interface with existing practices and policies without worsening health disparities or other problems. This section describes ways in which the 2024 NASEM Long COVID Definition may be applied for different purposes (clinical, research, and public health surveillance) and provides considerations as well as illus- trative examples. The committee also recognizes the need to adopt the definition for policy and service uses and refers the reader to the National Academies report, Long-Term Health Effects Stemming from COVID-19 and Implications for Social Security Administration (NASEM, 2024), to the HHS Guidance on “Long COVID” as a Disability Under the ADA, Section 504, and Section 1557 (HHS, 2021), and to the Social Security Administration’s guidance Long COVID: A Guide for Health Profession- als on Providing Medical Evidence for Social Security Disability Claims (SSA, 2023) for key considerations and resources. Long COVID qualifies as a disability under Titles II (state and local government) and III (public accommodations) of the Americans with Disabilities Act (ADA),2 Section 504 of the Rehabilitation Act of 1973 (Section 504),3 and Section 1557 of the Patient Protection and Affordable Care Act (Section 1557).4 The 2024 NASEM Long COVID Definition includes a key feature on functional impairment, but does not address when Long COVID may meet the legal definition of disability. All stakeholders involved in social safety net pro- grams, including payers, workplaces and employers, academic institutions and educators, and support services and government officials, need to be aware of Long COVID to properly support patients, and their families and caregivers in need. How Can Clinicians Apply the Definition? By implementing the 2024 NASEM Long COVID Definition, clinicians can enhance their ability to identify, diagnose, and manage Long COVID effectively. As previously stated, the goal of the definition is to aid clinicians in the consistent diagnosis of Long COVID. The definition is meant to be inclusive and should lead the clinician to provide comprehensive care for individuals experiencing Long COVID. This should also improve patients’ understanding of Long COVID and enhance access to care and effective treatment. 2 42 U.S.C. §§ 12101-12103, 12131-12189. 3 29 U.S.C. § 794. 4 42 U.S.C. § 18116. PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 73 Because of issues concerning COVID-19 test sensitivity, availability, access, and reporting, the 2024 NASEM Long COVID definition does not require a positive COVID-19 test to qualify for a Long COVID diagnosis. For those patients without a positive test, health care professionals will need to use their clinical judgement to decide whether the patients’ clinical picture fits a Long COVID diagnosis. For example, a history of contact with someone with a confirmed SARS-CoV-2 infection, persistence of symptoms from the time of an acute infection, and/or presence of symptoms like anosmia or post-exertional malaise which are not common features of other medical conditions would point towards Long COVID. On the other hand, a long intervening period between acute SARS-CoV-2 infection and occurrence of new symptoms, recurrence of symptom similar to a patient’s pre-COVID co-morbidities, or a diagnosis that better explains the patient’s presentation would not favor a Long COVID diagnosis. The minimum 3-month duration criterion ensures that patients with persistent symptoms are not overlooked, allowing clinicians to distinguish Long COVID from acute viral effects. This 3-month period can occur any- time: the committee chose not to specify that duration be counted from the instigating acute SARS-CoV-2 infection because studies have shown Long COVID symptoms can appear after a period of seemingly normal health. In cases of unconfirmed, asymptomatic SARS-CoV-2 infections, 3 months can be counted from the initial appearance of the patients’ symptoms. Overdi- agnosis of Long COVID can be mitigated by assessing whether a patient’s set of symptoms are consistent with Long COVID and whether another diagnosis could better account for their symptoms. While conducting their differential diagnosis, clinicians should recog- nize any concerning symptoms before the 3-month mark to provide appro- priate clinical care and avoid missing other unrelated conditions. The broad spectrum of symptoms and associated diagnosable conditions outlined in (but not limited to) the definition provide clinicians with a framework to recognize the diverse manifestations of Long COVID, ranging from respira- tory issues to neurological and autoimmune conditions. It is also imperative to recognize that following an acute SARS-CoV-2 infection, an existing diagnosis may notably deteriorate or exacerbate, necessitating clinicians to incorporate this into patient care and align it with the definition of Long COVID. For instance, complications such as controlled asthma escalating to severe asthma with frequent exacerbations exemplify this phenomenon. Moreover, the definition promotes a holistic approach to patient assess- ment by encouraging the clinician to consider the various symptoms, organ system dysfunctions, and conditions that may be present in a person with Long COVID. The definition also encourages interdisciplinary collaboration among health care specialists. Clinicians should form collaborative teams of relevant experts to address the multi-organ impact of Long COVID. PREPUBLICATION COPY—Uncorrected Proofs 

74 A LONG COVID DEFINITION A multi-disciplinary approach that is team-based will facilitate a more nuanced understanding of Long COVID and allow for tailored diagnostic and treatment strategies based on the specific organ systems affected. This approach aligns with the evolving understanding of Long COVID as a com- plex and heterogenous disease state, requiring a nuanced and collaborative health care response. Clinicians should be strongly encouraged to conduct thorough evaluations when presented with any symptomatology suggestive of Long COVID, prioritizing testing for potentially serious sequalae. Clinicians are encouraged to use the ICD-10 code for Long COVID of U09.9 as well as to participate in research collaborations to contribute valuable insights to the growing body of evidence around Long COVID. Clinicians should stay actively engaged in learning about Long COVID to drive continuous improvement in the understanding and management of Long COVID. Clinician goals include (1) improving patients’ and caregivers’ under- standing of Long COVID, (2) effectively assessing and documenting Long COVID, and (3) considering this diagnosis in any patient (inclusive of chil- dren and adults of all ages) who presents with persistent symptoms after SARS-CoV-2 infection. Table 2 provides considerations for the clinician who is implementing the 2024 NASEM Long COVID Definition. How Can Researchers Apply the Definition? The 2024 NASEM Long COVID Definition is by design inclusive and overarching, with many possible areas of use and application; this allows researchers flexibility in designing studies that are consistent with the defini- tion while also sufficiently specific to meet the study objectives. The com- mittee recognizes that most research studies are designed to answer a small number of very focused questions and that this will require identifying a subset of patients who satisfy the overarching definition and also satisfy more restrictive eligibility criteria for the study. Researchers may consider additional criteria to select the subset of Long COVID patients suited to the research project’s aims and hypothesis. The study investigators should specify the following elements: • Documentation of SARS-CoV-2 infection: Whether and how the initial SARS-CoV-2 infection is documented (e.g., suspected, prob- able, or confirmed SARS-CoV-2 infection) or based on self-reported symptoms. Some studies may also include a group of participants with no history of COVID-19 and negative tests for SARS-CoV-2. • Time period of infection and/or study (e.g., month, year): May be associated with different dominant SARS-CoV-2 variants. PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 75 TABLE 2  2024 NASEM Long COVID Definition Implementation Checklist for Clinicians Steps in the Process Key Considerations Initial Assessment • Consider Long COVID as a possible diagnosis when patients with a 3 months or more prior history of SARS-CoV-2 infection have persistent symptoms affecting one or more organ systems. • Assess patients for a history of asymptomatic, mild, moderate, or severe SARS-CoV-2 infection (or suspected infection), including reinfections. • Assess the duration of symptoms. Comprehensive • Obtain and document a detailed clinical history with attention Medical Evaluation to the onset and course of symptoms, recognizing that they can be continuous, intermittent, or delayed in onset. • Consider the use of a comprehensive symptom inventory to capture the wide range of manifestations of Long COVID, including post-exertional malaise. • Recognize that Long COVID can present as single or multiple symptoms and that severity may vary among individuals and even fluctuate for an individual. • Document the presence and measure the severity of all symptoms reported. Common symptoms include shortness of breath, persistent fatigue, post-exertional malaise, difficulty concentrating, memory changes, recurring headache, dizziness, fast heart rate, sleep disturbance, and problems with taste or smell. • Document and measure reported limitations and assess function, to see if and how symptoms and conditions are affecting an individual’s life (i.e., activities of daily living, work, and quality of life). • Most people affected by Long COVID can be diagnosed clinically based on history, physical examination, and/or symptom-directed diagnostic testing; yet thorough evaluations must be done. Testing for potential serious sequelae of SARS-CoV-2 should be prioritized. While certain symptoms like palpitations or shortness of breath are prevalent among individuals with Long COVID, it is crucial to recognize that in a minority of cases, they may signify underlying conditions requiring more acute medical intervention. • Eliminate common and uncommon causes for each individual’s symptoms. If, after a diagnosis of Long COVID, new symptoms appear or pre-existing symptoms worsen, consider re-evaluating these symptoms instead of automatically attributing them to Long COVID. Diagnosis of Specific • Identify specific diagnosable conditions associated with Long Conditions COVID. • Recognize that some conditions may not be diagnosable at the initial visit. • Assess comorbidities. continued PREPUBLICATION COPY—Uncorrected Proofs 

76 A LONG COVID DEFINITION TABLE 2  Continued Steps in the Process Key Considerations Diagnostics • Order appropriate diagnostic tests and imaging studies based on the suspected organ system involvement. Clinicians must ensure comprehensive imaging, functional, and physiological testing is done to exclude overt organ pathology necessitating immediate medical attention. • Consider relevant investigations to confirm specific Long COVID–related conditions including pulmonary function tests, cardiovascular assessments, autonomic testing, neuropsychological testing, relevant imaging, and labs. • Use both standard vital signs and pulse oximetry for patients with fatigue or respiratory or cardiac symptoms; abnormal blood pressure and heart rate findings should be evaluated using a 10-minute active standing test for those with orthostatic intolerance or other autonomic symptoms. Beyond this, patients with ongoing orthostatic symptoms with a normal active stand test should be referred for further autonomic testing. • Understand that many routine tests will be normal, and there is currently no diagnostic biomarker for Long COVID. • Carefully monitor and follow up with patients who continue to be sick 4–12 weeks after SARS-CoV-2 infection; clinicians might choose to take a conservative diagnostic approach prior to the 3-month time frame following the SARS-CoV-2 infection, but appropriate diagnostics should not be delayed when there are signs and symptoms of urgent/life threatening clinical conditions. Treatment • While some diagnosable conditions that fall within the definition of Long COVID, such as an increased frequency of diabetes are treatable, other manifestations of Long COVID have no FDA-approved treatment. And no biomarker exists to clearly ascribe symptoms and diagnosable conditions to Long COVID rather than to other causes. Treat a diagnosable condition related to Long COVID (e.g., diabetes, hypertension, POTS, Sjogren’s syndrome), based on standard, evidence-based guidelines. • Symptom management for single or multiple symptoms should be holistic and should use a multidisciplinary, patient-centered approach. • Treatments for Long COVID or related conditions with minimal potential for harm can be trialed while the patient is being evaluated or awaiting a diagnosis of Long COVID. For example, educating patients about how to conserve/ manage their energy (also known as pacing) and suggesting compression stockings for lightheadedness are relatively benign measures. • Treatments for possible alternative diagnoses (e.g., sleep hygiene for chronic fatigue) can also be trialed. Complete or substantial resolution of a patients’ symptoms would argue against a Long COVID diagnosis. PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 77 TABLE 2  Continued Steps in the Process Key Considerations Monitoring and • Recognize that patients may require multiple visits over time to Follow-Up capture the course of Long COVID’s clinical fluctuations and to allow various diagnostic entities to manifest clearly. • Recognize that subspecialty clinic referrals are frequently beneficial, particularly for patients with multisystem disorders like Long COVID, although access to such follow-up care may be a privilege within the health care system. Clear communication between the primary Long COVID clinician and the subspecialist can avoid poor coordination of care, conflicting guidance to the patient, and drug–drug interaction Patient and Provider • Educate patients and caregivers about Long COVID and its Education systemic, long-term impacts. • Provide informational resources as well as resources for accessing social services and support groups. • Thoroughly document symptoms, function, diagnostic findings, and treatment plans in the medical records for both clinical care and appropriate reimbursement. Careful documentation also assists patients in obtaining work/ school accommodation and, if needed, disability benefits. • Use the ICD-10 code for Long COVID, U09.9, and codes for any diagnosable condition related to Long COVID that may be present, such as ME/CFS (G93.32), POTS (G90.A), disorder of the autonomic nervous system, unspecified (G90.9), or MCAS (D89.42). • Pay attention to new developments surrounding COVID-19 and LC. In particular, we anticipated the definition of LC will evolve in the future as we learn more about this medical condition. Patient-Centered • Use a patient-centered approach that works toward “structural Collaborative Care competency” by employing “cultural and epistemic humility” and active listening strategies. • Understand that certain marginalized populations have experienced a higher burden of COVID-19, which has led to higher numbers of persons from these groups being affected by Long COVID, including racial and ethnic minority populations, persons with disabilities, and persons without access to health care. • Recognize the complexity of Long COVID and facilitate a coordinated, multidisciplinary approach that may include primary care clinicians, subspecialty clinicians, nurses, social workers/case managers, pharmacists, physical and occupational therapists, behavioral health professionals, and others from relevant fields. • Be aware that many individuals with Long COVID face stigma and their symptoms and conditions have been misattributed to psychiatric causes alone. Though there are psychiatric manifestations of Long COVID, a comprehensive work up should be done. PREPUBLICATION COPY—Uncorrected Proofs 

78 A LONG COVID DEFINITION • Predominant Long COVID sub-phenotypes: Long COVID may present as a single symptom, symptom cluster (e.g., neurocogni- tive, sleep disturbance), or conditions (e.g., postural orthostatic tachycardia syndrome [POTS], myalgic encephalomyelitis/chronic fatigue syndrome [ME/CFS]). Where multiple symptoms are pres- ent, describe the predominant Long COVID sub-phenotypes that are the focus of study eligibility, if any. Temporal pattern of symp- toms: Fluctuating, increasing, new onset, persistent, relapsing/ remitting, improving, or another pattern. • Minimum duration of symptoms: e.g., 3 months, 6 months, 12 months. • Time of onset of symptoms after infection: e.g., 0, 3, 6 months after SARS-CoV-2 infection. • Severity of illness at the time of SARS-CoV-2 infection: e.g., WHO Clinical Progression Scale: uninfected, ambulatory mild disease, hospitalized moderate, hospitalized severe (Marshall et al., 2020). • Co-morbid diagnoses: if relevant to the study’s purpose, analysis, or application. Assessment of specific co-morbidities using standard- ized and validated instruments is recommended and is dependent on the research question. • Exclusionary criteria: particular conditions (e.g., diagnosis of venous thromboembolism or interstitial lung disease prior to COVID-19) • Equity in research: equity must be considered, with additional awareness of and attention to the stigma and under-reporting that often occur with Long COVID. Consider access to and use of tests for SARS-CoV-2 and access to research studies. Consider popu- lations that may be disproportionately affected by SARS-CoV-2 illness (e.g., rural and medically underserved populations; mar- ginalized or minority populations; populations experiencing more severe disease). • Choice of comparison group: For randomized clinical trials or observational studies, defining the comparison group is critical to interpreting results. The selection of the comparison group depends on the research question and should be justified scientifically. • Additional design considerations including: • Treatment goals (e.g., mitigation of symptoms, cure, prevention) • Issues relevant in pediatric research, including children’s ability to communicate their experience and parent stigma/ under-reporting • The utility of including comparison groups with other IACCs • The implications of re-infections and co-infections • The value of long-term follow-up to understand the impact on conditions that may be slow to develop (e.g., cancer, Alzheim- er’s disease) PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 79 Research studies can include observational studies and clinical trials. Observational studies are designed to gather data on the prevalence, risk factors, or natural history of Long COVID symptoms. Study investigators do not assign participants to different treatments in an observational study. Clinical trials are studies designed to test different treatments, with the goal of identifying treatments that are both safe and effective. Studies may need to deviate from the 2024 NASEM Long COVID Defi- nition based on the underlying data sources that will be used or the specific study objective(s), as illustrated below for three hypothetical studies: (1) an observational study of MIS-C, (2) an observational study of neurologic sequelae in adults, and (3) a clinical trial of therapies for older adults with sleep disturbances (Table 3). When the study inclusion criteria differ from the 2024 NASEM Long COVID Definition, the investigators are encouraged to (1) explain the rationale for adopting more restrictive or more expansive study inclusion criteria and (2) explain how the use of alternate study-specific inclusion criteria may affect the generalizability or interpretation of study findings. How Can Public Health Practitioners Apply the Definition? The surveillance of conditions of public health significance is important for consistent tracking of trends within and across jurisdictions and over time. Federal, state, tribal, local, and territorial public health agencies have a long history of performing surveillance in a variety of ways to better understand and address health issues. Surveillance definitions will differ depending on a project’s purpose, the availability of resources, and the need to make comparisons among different groups. Public health surveillance provides critical information to understand the epidemiology of a newly emerging pathogen and to assess for the impact of a novel infection on society. Public health surveillance goals include: • Understanding the natural history and impact, e.g., the spectrum of physical, social, and mental illness that results from infection. • Identifying the epidemiologic characteristics of affected persons, e.g., demographics, underlying health conditions, occupational risk, geography. • Conducting risk assessments, e.g., probability, consequence • Comparing incidence and prevalence estimates among jurisdictions and nationwide. • Monitoring trends, e.g., incidence, prevalence, impact, epidemiol- ogy of affected persons • Identifying opportunities for interventions, e.g., primary, secondary, tertiary prevention PREPUBLICATION COPY—Uncorrected Proofs 

80 A LONG COVID DEFINITION TABLE 3  Study Design Considerations for Application of the 2024 NASEM Long COVID Definition Three studies as examples Study Type 1. Observational 2. Observational 3. Randomized clinical study: Retrospective study: Prospective trial in older study of cohort study adults with sleep multisystem of neurological disturbances inflammatory sequelae in adults syndrome in children (MIS-C). Study Goal Identify risk factors Characterize natural Evaluate and compare history potential therapies Data Source Electronic health Participant Participant records (EHRs) questionnaires questionnaires and polysomnography Target Population Children with Patients at community Age at least 65 years laboratory-confirmed medical centers serving meeting eligibility SARS-CoV-2 infection racial/ethnic minorities criteria with and without diagnosis of MIS-C Analytic Strategy Logistic regression Mixed-effects Analysis of covariance longitudinal models Long COVID Definition Elements and Important Features Documentation Confirmed infection Self-report of illness Suspected, probable, of SARS-CoV-2 with positive NAAT or confirmed SARS- infection test in EHR CoV-2 infection Time period of March 1, 2020– Any Any infection February 28, 2021 Predominant Long MIS-C Neurocognitive Sleep disturbance COVID phenotype Temporal pattern Any Relapsing/remitting Persistent of symptoms Minimum Any 6 months 3 months duration of symptoms Minimum time of 0 months 3 months 3 months symptom onset after infection Severity of illness Ambulatory mild to Mild to moderate/ Hospitalized moderate at time of SARS- hospitalized severe never hospitalized to severe CoV-2 infection (WHO Clinical Progression Scale) NOTE: NAAT = nucleic acid amplification test. PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 81 It is important that uniform criteria, also known as a case definition, are agreed upon to enable the above activities. As with any other health condition, each jurisdiction must determine, in accordance with its laws and regulations, whether to require the reporting of Long COVID as a condition of public health significance. The authority for implementing disease report- ing from health care providers to public health authorities can rest with state, tribal, territorial, and some local governments (CDC, 2023a).5 While no jurisdictions have made Long COVID reportable on an individual basis to public health authorities, many health departments have or may consider in the future conducting special surveillance or epidemiologic assessments of Long COVID in their jurisdiction. The 2024 NASEM definition of Long COVID will allow for implementation of a standard definition to be used for these purposes. Early in the evolution of an epidemic or pandemic, especially if out- comes include severe illness, hospitalization, and death, it is important for public health surveillance to include all potential cases of a new disease to prevent further spread by enacting appropriate control measures. This inclusiveness, or high sensitivity, helps to ensure that the full range of possible presentations of a given condition is captured. However, like all conditions, this must be balanced with specificity to avoid diluting our understanding of a new condition by being too broad and including cases that do not truly belong. For public health purposes, a case definition should be easy to apply consistently. The 2024 NASEM Long COVID Definition as proposed does not currently include a requirement for laboratory test evidence. This is due in part to known limitations in access to early testing that could have impacted some populations ability to access diagnostics, especially those with pre-existing difficulties accessing care. However, laboratory testing criteria might be added to the 2024 NASEM Long COVID Definition in implementation by researchers as previously discussed. And while it could potentially serve to augment surveillance, it is not necessarily required for all forms of surveillance activities. Public health surveillance approaches to Long COVID should focus on monitoring trends, understanding the epidemiology of this condition (including identifying populations at risk), and assessing physical, mental, and social impact over time. Various methodologies, other than individual case reporting, have been used including cross sectional surveys, leverag- ing existing healthcare and administrative data sources (e.g., all payer 5 For example, the current COVID-19 case definition as passed by CSTE and accepted by CDC can be found at https://ndc.services.cdc.gov/case-definitions/coronavirus-disease- 2019-covid-19/ (accessed March 6, 2024). PREPUBLICATION COPY—Uncorrected Proofs 

82 A LONG COVID DEFINITION databases, health information exchanges), modeling, sentinel surveillance, syndromic surveillance, or a combination of these approaches. Jurisdictions may not have all the resources needed to completely implement all these surveillance activities. However, they may elect to still benefit from the consistency that the 2024 NASEM Long COVID Defini- tion offers and begin to implement some, if not all, of these options. Con- siderations for jurisdictions seeking to implement the 2024 NASEM Long COVID Definition include the following: • Legal landscape: Does the jurisdiction have the authority to require (or simply encourage/ask for) reporting? • Technical landscape: Does the jurisdiction have the technical infra- structure necessary to implement surveys, electronic case reporting (ECR), health information exchange (HIE) extraction or mes- sage sharing, syndromic surveillance, sentinel surveillance, and modeling/forecasting? • Educational landscape: What is the level of awareness and under- standing of Long COVID among key stakeholders (e.g., the health care sector, public/community members, and/or elected officials)? • Capacity landscape: Does the jurisdiction have funding support for the systems and personnel required to perform surveillance activi- ties? If not, consider the cost of new systems (e.g., the Electronic Surveillance System for the Early Notification of Community-Based Epidemics, HIE) or the maintenance of existing systems. Also con- sider necessary support for training personnel to monitor and man- age systems and results (e.g., control measures, action steps, etc.) Table 4 provides examples of approaches that governmental public health jurisdictions may choose to employ to perform surveillance for Long COVID. Each of the activities below can be considered at the national, state, tribal, local, or territorial level depending on resource availability and needs. A WORKING DEFINITION AND RESEARCH AGENDA The committee expects that the 2024 NASEM Long COVID Defini- tion will evolve as new evidence emerges and our understanding of Long COVID continues to mature. The committee sought to be clear about the current understanding of Long COVID, and as one engagement participant commented, “Use this as an opportunity to be transparent about what we know and what we don’t know.” This is in line with lessons from defining other diseases such as HIV/AIDS, which took years and multiple iterations to define. Given the current pace of research, it is possible that the definition PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 83 TABLE 4  Public Health Surveillance Approaches for Long COVID Method Description Key Stakeholders Survey Any population-level survey design Public health officials, community (e.g., cross-sectional, retrospective, etc.) members/residents/patients, delivered through the secure platform community organizations, health of a jurisdiction’s choosing (e.g., Red care payers/providers, elected Cap, Qualtrics) that uses the 2024 officials, media NASEM Long COVID Definition to better understand trends in a given population Syndromic Use of key symptoms and/or billing Public health officials, hospitals, Surveillance codes (e.g., ICD codes) in an health care payers/providers, outpatient or inpatient setting that can elected officials be evaluated with the 2024 NASEM Long COVID Definition and integrated into current state and national surveillance systems. Medical Review of medical records to extract Public health officials, hospitals, record review symptoms and/or billing codes health care payers/providers, (electronic evaluated for agreement with the 2024 elected officials case reporting) NASEM Long COVID Definition. Some jurisdictions may consider feasibility of automated processes for electronic case reporting depending on software capabilities and associated resources. Sentinel Jurisdictions unable to broadly Public health officials, hospitals, Surveillance implement ECR due to resource health care payers/providers, limitations may consider identifying elected officials representative geographic or demographic groups to focus surveillance efforts on (e.g., one county, one hospital, one month of patient visits, etc.) Prospective The 2024 NASEM Long COVID Public health officials, academic modeling/ Definition can be applied to the institutions, hospitals, health care forecasting emerging body of forecasting and payers/providers, elected officials, modeling efforts to provide potential media estimates of burden based on circulating SARS-CoV-2 infections for a given jurisdiction (e.g., assess potential future economic, health care, school, and work needs/impacts) continued PREPUBLICATION COPY—Uncorrected Proofs 

84 A LONG COVID DEFINITION TABLE 4 Continued Method Description Key Stakeholders Education The 2024 NASEM Long COVID Public health officials, community and Care Definition can be shared widely with members/residents/patients, Coordination individuals and groups to provide community organizations, health greater understanding and recognition care payers/providers, elected of Long COVID in our communities officials, media Patient Voluntary patient registries can be Public health officials, community Registry facilitated by governmental public members/residents/patients, health the local, state, or federal levels community organizations, together with healthcare providers academic institutions, health care and academic institutions to provide a payers/providers, method for monitoring and supporting persons with Long COVID over time. Such cohorts can also serve as key cohorts for new treatments and technology as it arises. SOURCE: Adapted from CSTE, 2023. may need to be updated in no more than 3 years, and the reconsideration may benefit from a multidisciplinary approach. Other triggers for updating the 2024 NASEM Long COVID Definition could include the emergence of new treatments with clear benefits for patients identified by a refined defini- tion of disease, the development of a new test, new evidence on prognosis, or the need to improve the clarity or precision of the definition (Doust et al., 2017). Findings from the committee’s engagement process indicate the need to pair any new definition with a dissemination and education campaign to inform the public and key stakeholders. Furthermore, going forward, it may be valuable to have mechanisms in place for gauging how the 2024 NASEM Long COVID Definition is understood, how it is being used, what other ele- ments need to be added, and whether it is being applied in a consistent and standardized way (e.g., assessment tools). A research agenda to improve the definition could focus on the key elements articulated earlier: attribution to infection, time, clinical features, equity, functional impairment, exclusions and alternative diagnoses, biomarkers and laboratory criteria, and risk fac- tors. New evidence of the following may affect decisions to reconsider the definition: • Improved testing to identify those who have been infected, even when tested weeks, months and years later. However, a large pro- portion of the population has been infected with COVID-19 at PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 85 this point, and, as a result, finding control groups will become an increasing challenge in conducting research. • Symptoms and organ damage that distinguish Long COVID from healthy people and other medical conditions. • Onset and duration, including delayed onset of Long COVID after an ostensible period of recovery from acute infection. • Recovery trajectory and natural history over longer periods of time. • Presence and prevalence of co-morbid conditions. • Biomarker(s) to diagnose Long COVID. • Risk factors for Long COVID. • Prevalence and outcomes of Long COVID by sex, gender, race, ethnicity, socioeconomic status, and other factors. • Patterns in Long COVID among special populations such as older adults; children and adolescents; pregnant, lactating, and postpar- tum persons; people with disabilities; people experiencing home- lessness; tribal communities; and imprisoned populations; among others. • Longitudinal consequences (e.g., risk and development of other diseases, disability, hospitalization, and death). • Effects on functionality and daily living, overall well-being, and caregivers and families. • Social sciences research aimed at understanding the social and economic consequences of Long COVID. • New treatment and management options that could potentially affect the sensitivity threshold and elements of the definition. Ongoing and new research should lead to a more complete understand- ing of the natural history, etiology, therapy, and clinical management of Long COVID, and this enriched knowledge may prompt reconsideration of the definition of Long COVID. SHORTCOMINGS IN THE AVAILABLE EVIDENCE Research on Long COVID has been complicated due to heterogeneous study methods and lack of common data elements (e.g., different terms to describe the same symptom or condition) (Deer et al., 2021). Currently, most studies are among people with documented evidence of SARS-CoV-2 infection, which omits some who have been infected but have gone untested or had a false negative result. Differences in test availability and access could have affected the populations selected for these investigations. Time factors in studies were quite heterogenous—for example, how long people had to be sick before being included in a study, when they were assessed during a study, and the total follow-up period—which made it difficult to PREPUBLICATION COPY—Uncorrected Proofs 

86 A LONG COVID DEFINITION synthesize data. Furthermore, when the vast majority of people have been infected with SARS-CoV-2, it is difficult to create a valid, uninfected control group. Additionally, it seems few studies were adjusted for duration of illness when analyzing or interpreting results. Some studies lasted years, some only followed up for months to 1 year, and it was not clear if those who may have “recovered” were still followed to see if they had relapsed. Scientific studies take time and often confirm patterns that patients and clinicians may already be aware of. On the other hand, patient and clinician reports are usually not systematically gathered, and what appear to be common or prominent findings may appear different after more data are collected. Because the very nature of Long COVID requires evaluation over time, additional information and knowledge will continue to emerge, as will opportunities to document and understand the natural history of Long COVID. LIMITATIONS AND UNINTENDED CONSEQUENCES OF THE DEFINITION The committee confronted many difficulties in its efforts to define Long COVID including: • Varying working definitions and terminology used in studies that attempt to characterize Long COVID. • The tradeoffs of a broad definition and effects of this broad defini- tion on the IACC field. • A lack of confirmatory tests for SARS-CoV-2 infection. • Tests with limited performance characteristics (i.e., sensitivity, specificity). • Changes in test types (e.g., shift from PCR to antigen tests, espe- cially to those self-administered at home and not typically reported to public health authorities). • Varying levels of access to diagnostic tests across the population at different times in the pandemic response. • The social and cultural factors that can impact individuals’ ability and decisions to seek health care and therefore ultimately impact who is diagnosed, managed and studied for acute infections as well as Long COVID (e.g., cost, economic ability to sustain a work absence, family care obligations, etc.). • The inescapable circularity of relying on symptoms to define Long COVID and using the definition to indicate what symptoms are attributable to Long COVID. PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 87 • Striking the right balance between recognizing wide susceptibility to Long COVID while also acknowledging evidence of differential host risk factors (such as higher risk for women, those with mul- tiple previous infections, or more severe prior infection). • Understanding that for certain purposes (such as many research projects or public health surveillance), selection criteria that are more restrictive than the overall definition will be needed to char- acterize the target population. • The reality of the unknown pathophysiologic mechanisms that are inherent to any new condition and that will ideally be further elucidated as more time passes. • Limitations in the evidence base, as described above. Disease definition is an essential tool in health care that aids diagnosis, treatment, and research; it also has implications for access to supportive services post diagnosis. Throughout its deliberations, the committee sought to define Long COVID while simultaneously recognizing that there is more to learn and understand about it, and this means acknowledging inher- ent limitations and anticipating that changes will be made as the science advances. Particularly, the absence of an independent diagnostic standard for Long COVID, such as a definitive biomarker, is a noteworthy limita- tion. In any current effort to define Long COVID, there is an unavoidable circularity in using symptoms to define Long COVID and then relying on the definition to recognize characteristic symptoms. If SARS-CoV-2 had infected just one-tenth of one percent of the population, there could be the same proportional increase in Long COVID, but it would likely never have been recognized as such, lost in the background noise of similar IACCs with uncertain etiology. The present exercise to develop a definition is feasible because SARS-CoV-2 infected such a large number of persons and was fol- lowed by an upsurge in multiple chronic symptoms among those who had experienced infection. This strong temporal association lends confidence to the pathobiological, clinical, personal, and social reality of Long COVID. However, in the absence of a highly sensitive and perfectly specific bio- marker, the edges of inclusion and exclusion remain fuzzy: patients with a newly developed condition, such as diabetes, may have it attributed, rightly or wrongly, to Long COVID, while other patients with a single, persisting symptom, such as headache, that is due to Long COVID may fail to be properly diagnosed because there are many other possible causes of that symptom. Over time, with growing knowledge of the epidemiology and etiology of Long COVID and especially with the advent of one or more definitive biomarkers, the probability of both erroneous exclusion and erroneous inclusion will diminish. PREPUBLICATION COPY—Uncorrected Proofs 

88 A LONG COVID DEFINITION CONCLUDING REMARKS AND RECOMMENDATIONS The committee hopes the 2024 NASEM Long COVID Definition will, first and foremost, benefit the Long COVID community by creating a shared understanding of what Long COVID is and that it will lend added recogni- tion to IACCs within the medical community and society at large. The 2024 NASEM Long COVID Definition reflects and promotes a more holistic and integrated approach to understanding disease. Such an approach reflects a broader recognition of the complexities inherent in human health, includ- ing health disparities, social determinants of health, and the importance of including patient experience and knowledge in decision making. Further- more, the 2024 NASEM Long COVID Definition recognizes the heterogene- ity and complexity of Long COVID presentations and how Long COVID can affect individuals in multiple and profound ways. Through a synthesis of current research, multidisciplinary clinical insights, and patient perspectives, the committee attempted to unravel the complexities of Long COVID and to define Long COVID in a way that can serve a range of needs. In support of its definition, the committee puts forth three recommendations about the adoption, implementation, and updating of the 2024 NASEM Long COVID Definition. Recommendations RECOMMENDATION 1. Adopt and Implement the 2024 NASEM Long COVID Definition. The federal government, state, tribal, local, and territorial health authorities; clinical societies and associations; public health practitioners; clinicians; payers; researchers; drug industry; employers; educators; international organizations; and patients should adopt the 2024 NASEM Long COVID Definition and should use the term Long COVID. The 2024 NASEM Long COVID Definition is intended to be applied to many purposes, but the com- mittee notes that there is flexibility within the broad definition, for example, to restrict research eligibility to a subset of Long COVID patients. RECOMMENDATION 2. Promulgate and Monitor the Implementation of the 2024 NASEM Long COVID Definition. The Office of the Assistant Secretary for Health’s Office of Long COVID Research and Practice and the Long COVID Coordination Council should lead the coordination and collaboration efforts across federal, state, tribal, local, and territorial agencies and other relevant entities, including inter- national organizations, in the wide dissemination and implementation of PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 89 the 2024 NASEM Long COVID Definition. Such implementation efforts should: • Give special attention to the definition’s equity implications to maximize appropriate inclusion. • Develop standardized communication for key stakeholders and the public about the revised definition and understanding of Long COVID. • Empirically test the 2024 NASEM Long COVID Definition; moni- tor, evaluate, and identify barriers to implementation and adoption of the definition in research and in practice (including supporting an individual’s ability to apply for and receive Social Security dis- ability benefits) that may be improved in future revisions. • Develop a standard protocol for screening and diagnosing patients with Long COVID in clinical settings and enhance clinical educa- tion and training on infection-associated chronic conditions. • Catalogue and summarize the application of the definition in research settings and identify sub-phenotypes of Long COVID that inform the need for further investigation across the translational research spectrum from discovery to delivery science. • Take advantage of a unique opportunity to learn from epidemiologic surveillance of an infection-associated chronic condition and support, for example, improved data infrastructure, technologic and legal support for more efficient cross-jurisdictional information- sharing, and improved test types and access to testing. • Continue to listen to and collaborate with the Long COVID com- munity to learn from lived experience. RECOMMENDATION 3. Update the 2024 NASEM Long COVID Definition. In no more than 3 years or when triggered by the emergence of relevant new knowledge, the Office of the Assistant Secretary for Health’s Office of Long COVID Research and Practice should convene a multi-disciplinary group, including individuals with lived experience, to reexamine and update the 2024 NASEM Long COVID Definition set forth in this report. The Office of the Assistant Secretary for Health’s Office of Long COVID Research and Practice should put into place the necessary infrastructure, policies, and mechanisms to support and prepare for future updates to the 2024 NASEM Long COVID Definition, including a process to track and assess new scientific knowledge that may inform the definition. PREPUBLICATION COPY—Uncorrected Proofs 

90 A LONG COVID DEFINITION REFERENCES Admon, A. J., T. J. Iwashyna, L. A. Kamphuis, S. J. Gundel, S. K. Sahetya, I. D. Peltan, S. Y. Chang, J. H. Han, K. C. Vranas, K. P. Mayer, A. A. Hope, S. E. Jolley, E. Caldwell, M. L. Monahan, K. Hauschildt, S. M. Brown, N. R. Aggarwal, B. T. Thompson, and C. L. Hough. 2023. Assessment of symptom, disability, and financial trajectories in patients hospitalized for COVID-19 at 6 months. JAMA Network Open 6(2):e2255795. Ahmad, F. B., R. N. Anderson, J. A. Cisewski, and P. D. Sutton. 2022. Identification of deaths with post-acute sequelae of COVID-19 from death certificate literal text: United States, January 1, 2020–June 30, 2022. NVSS Vital Statistics Rapid Release, no. 25. https:// stacks.cdc.gov/view/cdc/121968 (accessed March 30, 2024). Al-Aly, Z., and E. Topol. 2024. Solving the puzzle of long COVID. Science 383(6685):830–832. Al-Aly, Z., A. Agarwal, N. Alwan, and V. A. Luyckx. 2023. Long COVID: Long-term health out- comes and implications for policy and research. Nature Reviews Nephrology 19(1):1–2. Al-Aly, Z., B. Bowe, and Y. Xie. 2022. Long COVID After Breakthrough SARS-CoV-2 Infec- tion. Nat Med 28(7): 1461-1467. Alkodaymi, M. S., O. A. Omrani, N. A. Fawzy, B. A. Shaar, R. Almamlouk, M. Riaz, M. Obei- dat, Y. Obeidat, D. Gerberi, R. M. Taha, Z. Kashour, T. Kashour, E. F. Berbari, K. Alkat- tan, and I. M. Tleyjeh. 2022. Prevalence of post-acute COVID-19 syndrome symptoms at different follow-up periods: A systematic review and meta-analysis. Clinical Microbiology and Infection 28(5):657–666. Altmann, D. M., E. M. Whettlock, S. Liu, D. J. Arachchillage, and R. J. Boyton. 2023. The im- munology of long COVID. Nature Reviews Immunology 23(10):618–634. Alwan, N. A., and L. Johnson. 2021. Defining long COVID: Going back to the start. Med 2(5):501–504. American Diabetes Association Professional Practice Committee. 2024. Comprehensive medi- cal evaluation and assessment of comorbidities: Standards of care in diabetes–2024. Diabetes Care 47:S52–S76. Antonelli, M., J. C. Pujol, T. D. Spector, S. Ourselin, and C. J. Steves. 2022. Risk of Long CO- VID associated with Delta versus Omicron variants of SARS-CoV-2. Lancet (London, England) 399(10343), 2263–2264. Apple, A. C., A. Oddi, M. J. Peluso, B. M. Asken, T. J. Henrich, J. D. Kelly, S. J. Pleasure, S. G. Deeks, I. E. Allen, J. N. Martin, L. C. Ndhlovu, B. L. Miller, M. L. Stephens, and J. Hellmuth. 2022. Risk factors and abnormal cerebrospinal fluid associate with cogni- tive symptoms after mild COVID-19. Annals of Clinical and Translational Neurology 9(2):221–226. Au, L., C. Capotescu, G. Eyal, and G. Finestone. 2022. Long COVID and medical gaslight- ing: Dismissal, delayed diagnosis, and deferred treatment. SSM—Qualitative Research in Health 2:100167. Azzolini E., R. Levi, R. Sarti, C. Pozzi, M. Mollura, A. Mantovani, and M. Rescigno. 2022. As- sociation between BNT162b2 vaccination and long COVID after infections not requiring hospitalization in health care workers. JAMA 328(7):676–678. Baig, A. M. 2021. Chronic COVID syndrome: Need for an appropriate medical terminology for long-COVID and COVID long-haulers. Journal of Medical Virology 93(5):2555–2556. Ballouz, T., D. Menges, A. Anagnostopoulos, A. Domenghino, H. E. Aschmann, A. Frei, J. S. Fehr, and M. A. Puhan. 2023. Recovery and symptom trajectories up to two years after SARS-CoV-2 infection: Population-based, longitudinal cohort study. BMJ 381:e074425. Banerjee, M., R. Pal, and S. Dutta. 2022. Risk of incident diabetes post-COVID-19: A system- atic review and meta-analysis. Primary Care Diabetes 16(4):591–593. PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 91 Bazdar, S., A. K. A. L. Kwee, L. Houweling, Y. de Wit-van Wijck, F. A. A. Mohamed Hoesein, G. S. Downward, E. J. Nossent, and A. H. Maitland-van der Zee. 2023. A systematic review of chest imaging findings in long COVID patients. Journal of Personalized Medi- cine 13(2):282. Behnood, S. A., R. Shafran, S. D. Bennett, A. X. D. Zhang, L. L. O’Mahoney, T. J. Stephenson, S. N. Ladhani, B. L. De Stavola, R. M. Viner, and O. V. Swann. 2022. Persistent symptoms following SARS-CoV-2 infection amongst children and young people: A meta-analysis of controlled and uncontrolled studies. Journal of Infection 84(2):158–170. Bergmans, R. S., K. Chambers-Peeple, C. Yu, L. Z. Xiao, R. Wegryn-Jones, A. Martin, S. Dell’Imperio, D. Aboul-Hassan, D. A. Williams, D. J. Clauw, and M. DeJonckheere. 2023. “I’m still here, I’m alive and breathing”: The experience of Black Americans with long COVID. Journal of Clinical Nursing 33(1):162–177. Bonilla, H., M. J. Peluso, K. Rodgers, J. A. Aberg, T. F. Patterson, R. Tamburro, L. Baizer, J. D. Goldman, N. Rouphael, A. Deitchman, J. Fine, P. Fontelo, A. Y. Kim, G. Shaw, J. Stratford, P. Ceger, M. M. Costantine, L. Fisher, L. O’Brien, C. Maughan, J. G. Quigley, V. Gabbay, S. Mohandas, D. Williams, and G. A. McComsey. 2023. Therapeutic trials for long COVID-19: A call to action from the interventions taskforce of the recover initiative. Frontiers in Immunology 14:1129459. Bowe, B., Y. Xie, and Z. Al-Aly. 2022. Acute and postacute sequelae associated with SARS- CoV-2 reinfection. Nature Medicine 28:2398–2405. Bowe, B., Y. Xie, and Z. Al-Aly. 2023. Postacute sequelae of COVID-19 at 2 years. Nature Medicine 29(9):2347–2357. Callard, F., and E. Perego. 2021. How and why patients made long COVID. Social Science & Medicine 268:113426. Català, M., N. Mercadé-Besora, R. Kolde, N. T. H. Trinh, E. Roel, E. Burn, T. Rathod-Mistry, K. Kostka, W. Y. Man, A. Delmestri, H. M. E. Nordeng, A. Uusküla, T. Duarte-Salles, D. Prieto-Alhambra, and A. M. Jödicke. 2024. The effectiveness of COVID-19 vaccines to prevent long COVID symptoms: Staggered cohort study of data from the UK, Spain, and Estonia. The Lancet Respiratory Medicine 12(3):225–236. CDC (Centers for Disease Control and Prevention). 2023a. National Notifiable Disease Sur- veillance System (NNDS): What is case surveillance? https://www.cdc.gov/nndss/about/ index.html (accessed March 6, 2024). CDC. 2023b. Long COVID or Post-COVID Conditions. https://www.cdc.gov/coronavirus/ 2019-ncov/long-term-effects/index.html#:~:text=Some%20people%2C%20espe- cially%20those%20who,kidney%2C%20skin%2C%20and%20brain. (accessed April 23, 2024). CDC. 2024a. Long COVID: Household Pulse Survey. https://www.cdc.gov/nchs/covid19/ pulse/long-covid.htm (accessed March 30, 2024). CDC. 2024b. 2022-2023 Nationwide COVID-19 infection- and vaccination-induced antibody seroprevalence (blood donations). https://covid.cdc.gov/covid-data-tracker/#nationwide- blood-donor-seroprevalence-2022 (accessed April 23, 2024). CDC Foundation. 2024. Infection-associated chronic conditions: Opportunities for action. https://solvecfs.org/wp-content/uploads/2024/02/ICUEPublicReport.pdf (accessed March 30, 2024). Chaichana, U., K. K. C. Man, A. Chen, I. C. K. Wong, J. George, P. Wilson, and L. Wei. 2023. Definition of post–COVID-19 condition among published research studies. JAMA Net- work Open 6(4):e235856. Chang, R., T. Y.-T. Chen, S.-I. Wang, Y.-M. Hung, H.-Y. Chen, and C.-C. J. Wei. 2023. Risk of autoimmune diseases in patients with COVID-19: A retrospective cohort study. EClini- calMedicine 56:101783. PREPUBLICATION COPY—Uncorrected Proofs 

92 A LONG COVID DEFINITION Chen, B., B. Julg, S. Mohandas, S. B. Bradfute, and RECOVER Mechanistic Pathways Task Force. 2023. Viral persistence, reactivation, and mechanisms of Long COVID. eLife 12:e86015. Choutka, J., V. Jansari, M. Hornig, and A. Iwasaki. 2022. Unexplained post-acute infection syndromes. Nature Medicine 28(5):911–923. Ciaffi, J., E. Vanni, L. Mancarella, V. Brusi, L. Lisi, F. Pignatti, S. Naldi, E. Assirelli, S. Neri, M. Reta, C. Faldini, and F. Ursini. 2023. Post-acute COVID-19 joint pain and new onset of rheumatic musculoskeletal diseases: A systematic review. Diagnostics 13(1):1850. Clauw, D. J., and L. Calabrese. 2024. Rheumatology and Long COVID: Lessons from the study of fibromyalgia. Annals of the Rheumatic Diseases 2024;83:136-138. Cohen, J., and Y. van der Meulen Rodgers. 2023. An intersectional analysis of long COVID prevalence. International Journal for Equity in Health 22(1):261. Crivelli, L., K. Palmer, I. Calandri, A. Guekht, E. Beghi, W. Carroll, J. Frontera, D. García- Azorín, E. Westenberg, A. S. Winkler, F. Mangialasche, R. F. Allegri, and M. Kivipelto. 2022. Changes in cognitive functioning after COVID-19: A systematic review and meta- analysis. Alzheimer’s & Dementia 18(5):1047–1066. CSTE (Council of State and Territorial Epidemiologists). 2023. State, tribal, local, and ter- ritorial public health agency approaches to long COVID-19/post COVID-19 condition surveillance: Lessons learned, gaps, and needs. https://preparedness.cste.org/wp-content/ uploads/2023/09/CSTE-STLT-Long-COVID-Surveillance-August-2023.pdf (accessed April 1, 2024). Cutler, D. M. 2022. The economic cost of long COVID: An update. Harvard University. https:// scholar.harvard.edu/files/cutler/files/long_covid_update_7-22.pdf (accessed April 4, 2024) Davis, H. E., G. S. Assaf, L. McCorkell, H. Wei, R. J. Low, Y. Re’em, S. Redfield, J. P. Austin, and A. Akrami. 2021. Characterizing long COVID in an international cohort: 7 months of symptoms and their impact. EClinicalMedicine 38:101019. Davis, H. E., L. McCorkell, J. M. Vogel, and E. J. Topol. 2023. Long COVID: Major findings, mechanisms and recommendations. Nature Reviews Microbiology 21(3):133–146. de Oliveira Almeida, K., I. G. Nogueira Alves, R. S. de Queiroz, M. R. de Castro, V. A. Gomes, F. C. Santos Fontoura, C. Brites, and M. G. Neto. 2023. A systematic review on physical function, activities of daily living and health-related quality of life in COVID-19 survi- vors. Chronic Illness 19(2):279–303. Deer, R. R., M. A. Rock, N. Vasilevsky, L. Carmody, H. Rando, A. J. Anzalone, M. D. Basson, T. D. Bennett, T. Bergquist, E. A. Boudreau, C. T. Bramante, J. B. Byrd, T. J. Callahan, L. E. Chan, H. Chu, C. G. Chute, B. D. Coleman, H. E. Davis, J. Gagnier, C. S. Greene, W. B. Hillegass, R. Kavuluru, W. D. Kimble, F. M. Koraishy, S. Kohler, C. Liang, F. Liu, H. Liu, V. Madhira, C. R. Madlock-Brown, N. Matentzoglu, D. R. Mazzotti, J. A. McMurry, D. S. McNair, R. A. Moffitt, T. S. Monteith, A. M. Parker, M. A. Perry, E. Pfaff, J. T. Reese, J. Saltz, R. A. Schuff, A. E. Solomonides, J. Solway, H. Spratt, G. S. Stein, A. A. Sule, U. Topaloglu, G. D. Vavougios, L. Wang, M. A. Haendel, and P. N. Robinson. 2021. Charac- terizing long COVID: Deep phenotype of a complex condition. EBioMedicine 74:103722. Diekman, S., and T. Chung. 2023. Post-acute sequelae of SARS-CoV-2 syndrome presenting as postural orthostatic tachycardia syndrome. Clinical and Experimental Emergency Medicine 10(1):18–25. Dinnes, J., P. Sharma, S. Berhane, S. S. van Wyk, N. Nyaaba, J. Domen, M. Taylor, J. Cun- ningham, C. Davenport, S. Dittrich, D. Emperador, L. Hooft, M. M. G. Leeflang, M. D. F. McInnes, R. Spijker, J. Y. Verbakel, Y. Takwoingi, S. Taylor-Phillips, A. Van den Bruel, and J. J. Deeks. 2022. Rapid, point‐of‐care antigen tests for diagnosis of SARS‐CoV‐2 infection. Cochrane Database of Systematic Reviews 7(7):CD013705. PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 93 Doust, J., P. O. Vandvik, A. Qaseem, R. A. Mustafa, A. R. Horvath, A. Frances, L. Al-Ansary, P. Bossuyt, R. L. Ward, I. Kopp, L. Gollogly, H. Schunemann, P. Glasziou, and Guide- lines International Network Preventing Overdiagnosis Working Group. 2017. Guid- ance for modifying the definition of diseases: A checklist. JAMA Internal Medicine 177(7):1020–1025. Durstenfeld, M. S., S. Weiman, M. Holtzman, C. Blish, R. Pretorius, and S. G. Deeks. 2024. Long COVID and post-acute sequelae of SARS-CoV-2 pathogenesis and treatment: A keystone symposia report. Annals of the New York Academy of Sciences. https://doi. org/10.1111/nyas.15132. Economist Impact. 2024. An incomplete picture: understanding the burden of Long Covid. https://impact.economist.com/perspectives/sites/default/files/download/ei264_-_an_in- complete_picture_understanding_the_burden_of_long_covid_v8.pdf. (accessed May 3, 2024). EpiCore. 2023. Obtaining Long COVID definition through EpiCore. EpiCore Long COVID definitions. https://endingpandemics.org/wp-content/uploads/2023/03/EPICORE-Long- Covid-Definitions-NASEM-2023-4.pdf (accessed April 19, 2024). Espin, E., C. Yang, C. P. Shannon, S. Assadian, D. He, and S. J. Tebbutt. 2023. Cellular and molecular biomarkers of long COVID: A scoping review. EBioMedicine 91:104552. Federal Reserve Bank of Minneapolis. 2022. Long-haulers and labor market outcomes. Avail- able at https://www.minneapolisfed.org/research/institute-working-papers/long-haulers- and-labor-market-outcomes (accessed April 1, 2024). Fernández-de-Las-Peñas, C. 2022. Long COVID: Current definition. Infection 50(1):285–286. Fernández-de-Las-Peñas, C., D. Palacios-Cena, V. Gomez-Mayordomo, M. L. Cuadrado, and L. L. Florencio. 2021a. Defining post-COVID symptoms (post-acute COVID, long COVID, persistent post-COVID): An integrative classification. International Journal of Environ- mental Research and Public Health 18(5):1–9. Fernández-de-las-Peñas, C., L. L. Florencio, V. Gomez-Mayordomo, M. L. Cuadrado, D. Pa- lacios-Cena, and A. V. Raveendran. 2021a. Proposed integrative model for post-COVID symptoms. Diabetes & Metabolic Syndrome 15(4):102159. Fernández-de-las-Peñas, C., D. Palacios-Cena, V. Gomez-Mayordomo, L. L. Florencio, M. L. Cuadrado, G. Plaza-Manzano, and M. Navarro-Santana. 2021b. Prevalence of post- COVID-19 symptoms in hospitalized and non-hospitalized COVID-19 survivors: A sys- tematic review and meta-analysis. European Journal of Internal Medicine 92:55–70. Fernández-de-Las-Peñas, C., K. I. Notarte, P. J. Peligro, J. V. Velasco, M. J. Ocampo, B. M. Henry, L. Arendt-Nielsen, J. Torres-Macho, and G. Plaza-Manzano. 2022a. Long-COVID symptoms in individuals infected with different SARS-CoV-2 variants of concern: A sys- tematic review of the literature. Viruses 14(12):2629. Fernández-de-Las-Peñas, C., O. J. Pellicer-Valero, E. Navarro-Pardo, D. Palacios-Cena, L. L. Florencio, C. Guijarro, and J. D. Martin-Guerrero. 2022b. Symptoms experienced at the acute phase of SARS-CoV-2 infection as risk factor of long-term post-COVID symptoms: The long-COVID-exp-cm multicenter study. Int J Infect Dis 116:241–244. Fernández-de-Las-Peñas, C., J. D. Martín-Guerrero, I. Cancela-Cilleruelo, P. Moro-López- Menchero, J. Rodríguez-Jiménez, E. Navarro-Pardo, and O. J. Pellicer-Valero. 2022c. Exploring the recovery curves for long-term post-COVID functional limitations on daily living activities: The LONG-COVID-EXP-CM multicenter study. Journal of Infection 84(5):722–746. Fernández-de-Las-Peñas, C., J. D. Martin-Guerrero, L. L. Florencio, E. Navarro-Pardo, J. Rodriguez-Jimenez, J. Torres-Macho, and O. J. Pellicer-Valero. 2023. Clustering analysis reveals different profiles associating long-term post-COVID symptoms, COVID-19 symp- toms at hospital admission and previous medical co-morbidities in previously hospital- ized COVID-19 survivors. Infection 51(1):61–69. PREPUBLICATION COPY—Uncorrected Proofs 

94 A LONG COVID DEFINITION Figueiredo, E. A. B., W. T. Silva, S. P. Tsopanoglou, D. F. M. Vitorino, L. F. L. Oliveira, K. L. S. Silva, H. D. H. Luz, M. R. Ávila, L. F. F. Oliveira, A. C. R. Lacerda, V. A. Mendonça, V. P. Lima, M. F. F. Mediano, P. H. S. Figueiredo, M. O. C. Rocha, and H. S. Costa. 2022. The health-related quality of life in patients with post-COVID-19 after hospitalization: A systematic review. Revista de Sociedade Brasileira de Medicina Tropical 55:e0741. Fogh, K., T. G. Larsen, C. B. Hansen, R. B. Hasselbalch, A. R. R. Eriksen, H. Bundgaard, R. Frikke-Schmidt, L. M. Hilsted, L. Østergaard, I. S. Johansen, I. Hageman, P. Garred, and K. Iversen. 2022. Self-reported long COVID and its association with the presence of SARS-CoV-2 antibodies in a Ddanish cohort up to 12 months after infection. Microbiol- ogy Spectrum 10(6):e02537–e02522. Ford, N. D., D. Slaughter, D. Edwards, A. Dalton, C. Perrine, A. Vahratian, and S. Saydah. 2023. Long COVID and significant activity limitation among adults, by age—United States, June 1–13, 2022, to June 7–19, 2023. Morbidity and Mortality Weekly Report 72(32):866–870. Ford, N., A. Agedew, A. F. Dalton, J. Singleton, C. G. Perrine, and S. Saydah. 2024. Notes from the field: Long COVID prevalence among adults—United States, 2022. Morbidity and Mortality Weekly Report 73(6):135–136. Franco, J. V. A., L. I. Garegnani, G. V. Oltra, M. I. Metzendorf, L. F. Trivisonno, N. Sgarbossa, D. Ducks, K. Heldt, R. Mumm, B. Barnes, and C. Scheidt-Nave. 2022. Long-term health symptoms and sequelae following SARS-CoV-2 infection: An evidence map. International Journal of Environmental Research and Public Health 19(1):9915. Garmoe, W., K. Rao, B. Gorter, and R. Kantor. 2024. Neurocognitive impairment in post- COVID-19 condition in adults: Narrative review of the current literature. Archives of Clinical Neuropsychology acae017. Gavrilova, N., L. Soprun, M. Lukashenko, V. Ryabkova, T. V. Fedotkina, L. P. Churilov, and Y. Shoenfeld. 2022. New clinical phenotype of the post-COVID syndrome: Fibromyalgia and joint hypermobility condition. Pathophysiology 29(1):24-29. https://doi.org/10.3390/ pathophysiology29010003. Goldenberg, D. L. 2023. Applying lessons from rheumatology to better understand long CO- VID. Arthritis Care & Research (Hoboken) 76(1):49–56. Government of Canada. 2023. Second report: Spring 2023. https://health-infobase.canada.ca/ covid-19/post-covid-condition/spring-2023-report.html (accessed March 3, 2024). Hägglöf, E., M. Bell, E. Zettersten, L. Engerström, and E. Larsson. 2023. Long-term survival after intensive care for COVID-19: A nationwide cohort study of more than 8000 pa- tients. Annals of Intensive Care 13(1):76. Hampshire, A., A. Azor, C. Atchison, W. Trender, P. J. Hellyer, V. Giunchiglia, M. Husain, G. S. Cooke, E. Cooper, A. Lound, C. A. Donnelly, M. Chadeau-Hyam, H. Ward, and P. El- liott. 2024. Cognition and memory after COVID-19 in a large community sample. New England Journal of Medicine 390(9):806–818. Haslam, A., T. Olivier, and V. Prasad. 2023. The definition of long COVID used in interven- tional studies. European Journal of Clinical Investigations 53(8):e13989. Hayes, L. D., J. Ingram, and N. F. Sculthorpe. 2021. More than 100 persistent symptoms of SARS-CoV-2 (long COVID): A scoping review. Frontiers in Medicine (Lausanne) 8:750378. Hedberg, P., and P. Nauclér. 2024. Post-COVID-19 condition after SARS-CoV-2 infections dur- ing the Omicron surge vs the Delta, Alpha, and Wild type periods in Stockholm, Sweden. The Journal of Infectious Diseases 229(1):133–136. Heesakkers, H., J. G. van der Hoeven, S. Corsten, I. Janssen, E. Ewalds, K. S. Simons, B. Wester- hof, T. C. D. Rettig, C. Jacobs, S. van Santen, A. J. C. Slooter, M. C. E. van der Woude, M. van den Boogaard, and M. Zegers. 2022. Clinical outcomes among patients with 1-year survival following intensive care unit treatment for COVID-19. JAMA 327(6):559–565. PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 95 HHS (U.S. Department of Health and Human Services). 2020. Health equity in Healthy People 2030. https://health.gov/healthypeople/priority-areas/health-equity-healthy-people-2030 (accessed March 31, 2024). HHS. 2021. Guidance on “Long COVID” as a disability under the ADA section 504, and sec- tion 1557. https://www.hhs.gov/civil-rights/for-providers/civil-rights-covid19/guidance- long-covid-disability/index.html (accessed April 24, 2024). HHS/OASH (Department of Health and Human Services and Office of the Assistant Secretary for Health). 2022. HHS releases Long COVID report providing insights and opportuni- ties to support patient community. https://www.hhs.gov/about/news/2022/11/21/hhs- releases-long-covid-report-providing-insights-and-opportunities-support.html (accessed April 1, 2024). HHS. 2022. National Research Action Plan on Long COVID. https://www.covid.gov/sites/ default/files/documents/National-Research-Action-Plan-on-Long-COVID-08012022.pdf (accessed April 4, 2024). Hill, E. L., H. B. Mehta, S. Sharma, K. Mane, S. K. Singh, C. Xie, E. Cathey, J. Loomba, S. Russell, H. Spratt, P. E. DeWitt, N. Ammar, C. Madlock-Brown, D. Brown, J. A. McMurry, C. G. Chute, M. A. Haendel, R. Moffitt, E. R. Pfaff, and T. D. Bennett. 2023. Risk factors associated with post-acute sequelae of SARS-CoV-2: An N3C and NIH RECOVER study. BMC Public Health 23(1):2103. Huerne, K., K. B. Filion, R. Grad, P. Ernst, A. S. Gershon, and M. J. Eisenberg. 2023. Epi- demiological and clinical perspectives of long COVID syndrome. American Journal of Medicine Open 9:100033. IOM (Institute of Medicine). 2014. Chronic multisymptom illness in Gulf War veterans: Case definitions reexamined. Washington, DC: The National Academies Press. IOM. 2015. Beyond myalgic encephalomyelitis/chronic fatigue syndrome: Redefining an ill- ness. Washington, DC: The National Academies Press. Ireson, J., A. Taylor, E. Richardson, B. Greenfield, and G. Jones. 2022. Exploring invisibility and epistemic injustice in Long COVID—A citizen science qualitative analysis of patient stories from an online COVID community. Health Expectations 25(4):1753-1765. Jackson, J. C., J. Clune, H. Hoenig, M. Morey, V. Anderson, L. Denne, N. Williams, C. Seibert, R. Torres, D. R. Janz, B. Levine, B. Pun, K. Archer, E. Schiro, J. Jones, J. Zoz, J. Thompson, and E. W. Ely. 2010. The Returning to Everyday Tasks Utilizing Rehabilitation Networks (RETURN) trial: A pilot, feasibility trial including in-home cognitive rehabilitation of ICU survivors. American Journal of Respiratory and Critical Care Medicine 181:A5359. Jackson, J. C., E. W. Ely, M. C. Morey, V. M. Anderson, L. B. Denne, J. Clune, C. S. Seibert, K. R. Archer, R. Torres, D. Janz, E. Schiro, J. Jones, A. K. Shintani, B. Levine, B. T. Pun, J. L. Thompson, N. E. Brummel, and H. Hoenig. 2012. Cognitive and physical rehabilitation of intensive care unit survivors: Results of the RETURN randomized controlled pilot investigation. Critical Care Medicine 40(4):1088-1097. Jacobs, E. T., C. J. Catalfamo, P. M. Colombo, S. M. Khan, E. Austhof, F. Cordova-Marks, K. C. Ernst, L. V. Farland, and K. Pogreba-Brown. 2023. Pre-existing conditions associated with post-acute sequelae of COVID-19. Journal of Autoimmunity 135:102991. Jacobs, M. M., E. Evans, and C. Ellis. 2023. Racial, ethnic, and sex disparities in the incidence and cognitive symptomology of long COVID-19. Journal of the National Medical As- sociation 115(2):233–243. Jiang, L., X. Li, J. Nie, K. Tang, and Z. A. Bhutta. 2023. A systematic review of persistent clinical features after SARS-CoV-2 in the pediatric population. Pediatrics 152(2):e2022060351. Kedor, C., H. Freitag, L. Meyer-Arndt, K. Wittke, L. G. Hanitsch, T. Zoller, F. Steinbeis, M. Haffke, G. Rudolf, B. Heidecker, T. Bobbert, J. Spranger, H. D. Volk, C. Skurk, F. Koni- etschke, F. Paul, U. Behrends, J. Bellmann-Strobl, and C. Scheibenbogen. 2022. A prospec- tive observational study of post-COVID-19 chronic fatigue syndrome following the first pandemic wave in Germany and biomarkers associated with symptom severity. Nature Communications 13(1):5104. PREPUBLICATION COPY—Uncorrected Proofs 

96 A LONG COVID DEFINITION Khullar, D., Y. Zhang, C. Zang, Z. Xu, F. Wang, M. G. Weiner, T. W. Carton, R. L. Rothman, J. P. Block, and R. Kaushal. 2023. Racial/ethnic disparities in post-acute sequelae of SARS- CoV-2 infection in New York: An EHR-based cohort study from the recover program. Journal of General Internal Medicine 38(5):1127–1136. Kim, C., B. Chen, S. Mohandas, J. Rehman, Z. A. Sherif, K. Coombs, RECOVER Mechanis- tic Pathways Task Force, and RECOVER Initiative. 2023. The importance of patient- partnered research in addressing long COVID: Takeaways for biomedical research study design from the RECOVER Initiative’s Mechanistic Pathways taskforce. Elife 12:e86043. Kioi, Y., H. Yorifuji, Y. Higami, and Y. Katada. 2023. Serositis and lymphopenia are common features of systemic lupus erythematosus following SARS-CoV-2 infection: A case report and literature review. International Journal of Rheumatic Diseases 26(11):2267–2271. Klein, J., J. Wood, J. Jaycox, R. M. Dhodapkar, P. Lu, J. R. Gehlhausen, A. Tabachnikova, K. Greene, L. Tabacof, A. A. Malik, V. Silva Monteiro, J. Silva, K. Kamath, M. Zhang, A. Dhal, I. M. Ott, G. Valle, M. Pena-Hernandez, T. Mao, B. Bhattacharjee, T. Takahashi, C. Lucas, E. Song, D. McCarthy, E. Breyman, J. Tosto-Mancuso, Y. Dai, E. Perotti, K. Akdu- man, T. J. Tzeng, L. Xu, A. C. Geraghty, M. Monje, I. Yildirim, J. Shon, R. Medzhitov, D. Lutchmansingh, J. D. Possick, N. Kaminski, S. B. Omer, H. M. Krumholz, L. Guan, C. S. Dela Cruz, D. van Dijk, A. M. Ring, D. Putrino, and A. Iwasaki. 2023. Distinguishing fea- tures of long COVID identified through immune profiling. Nature 623(7985):139–148. Kleinman, A. 1988. The illness narratives: Suffering, healing and the human condition. New York: Basic Books. Komaroff, A. L., and W. I. Lipkin. 2023. ME/CFS and Long COVID share similar symptoms and biological abnormalities: road map to the literature. Frontiers in medicine, 10, 1187163. Korte, W., M. Buljan, M. Rösslein, P. Wick, V. Golubov, J. Jentsch, M. Reut, K. Peier, B. Nohynek, A. Fischer, R. Stolz, M. Cettuzzi, O. Nolte. 2021. SARS-CoV-2 IgG and IgA antibody response is gender dependent; and IgG antibodies rapidly decline early on. Journal of Infection 82(1):e11–e14. Kromydas, T., E. Demou, R. Edge, M. Gittins, S. V. Katikireddi, N. Pearce, M. van Tongeren, J. Wilkinson, and S. Rhodes. 2023. Occupational differences in the prevalence and severity of long-COVID: Analysis of the coronavirus (COVID-19) infection survey. Occupational and Environmental Medicine 80(10):545–552. Kucirka, L. M., S. A. Lauer, O. Laeyendecker, D. Boon, and J. Lessler. 2020. Variations in false- negative rate of reverse transcriptase polymerase chain reaction based SARS-CoV-2 tests by time since exposure. Annals of Internal Medicine 173(4):262–267. Lai, Y. J., S. H. Liu, S. Manachevakul, T. A. Lee, C. T. Kuo, and D. Bello. 2023. Biomarkers in long COVID-19: A systematic review. Frontiers in Medicine 10:1085988. Larsen, N. W., L. E. Stiles, R. Shaik, L. Schneider, S. Muppidi, C. T. Tsui, L. N. Geng, H. Bo- nilla, and M. G. Miglis. 2022. Characterization of autonomic symptom burden in Long COVID: A global survey of 2,314 adults. Frontiers in Neurology 13:1012668. Lim, E. J., and C. G. Son. 2020. Review of case definitions for myalgic encephalomyelitis/ chronic fatigue syndrome (ME/CFS). Journal of Translational Medicine 18(1):289. Lim, S. H., H. J. Ju, J. H. Han, J. H. Lee, W.-S. Lee, J. M. Bae, and S. Lee. 2023. Autoimmune and autoinflammatory connective tissue disorders Following COVID-19. JAMA Netw Open 6(10):e2336120. https://doi.org/10.1001/jamanetworkopen.2023.36120. Linh, T. T. D., D. K. N. Ho, N. N. Nguyen, C. J. Hu, C. H. Yang, and D. Wu. 2023. Global prevalence of post-COVID-19 sleep disturbances in adults at different follow-up time points: A systematic review and meta-analysis. Sleep Medicine Reviews 71:101833. Lippi, G., B. M. Henry, J. Favresse, and M. Plebani. 2023. Addressing standardized definitions of post-COVID and long-COVID. Clinical Chemistry and Laboratory Medicine (CCLM) 61(8):1361–1362. PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 97 Louie, P., and C. Wu. 2023. Race, socioeconomic status, and long COVID. Social Currents https://doi.org/10.1177/23294965231215081. Ma, Y., J. Deng, Q. Liu, M. Du, M. Liu, and J. Liu. 2023. Long-term consequences of asymp- tomatic SARS-CoV-2 infection: A systematic review and meta-analysis. International Journal of Environmental Research and Public Health 20(2):1613. Malkova, A., I. Kudryavtsev, A. Starshinova, D. Kudlay, Y. Zinchenko, A. Glushkova, P. Yablonskiy, and Y. Shoenfeld. 2021. Post COVID-19 syndrome in patients with asymp- tomatic/mild form. Pathogens 10(11):1408. Malone, L. A., A. Morrow, Y. Chen, D. Curtis, S. D. de Ferranti, M. Desai, T. K. Fleming, T. M. Giglia, T. A. Hall, E. Henning, S. Jadhav, A. M. Johnston, D. R. C. Kathirithamby, C. Kokorelis, C. Lachenauer, L. Li, H. C. Lin, T. Locke, C. MacArthur, M. Mann, S. A. McGrath-Morrow, R. Ng, L. Ohlms, S. Risen, S. C. Sadreameli, S. Sampsel, S. K. S. Tejtel, J. K. Silver, T. Simoneau, R. Srouji, S. Swami, S. Torbey, M. V. Gutierrez, C. N. Williams, L. A. Zimmerman, and L. E. Vaz. 2022. Multi-disciplinary collaborative consensus guid- ance statement on the assessment and treatment of postacute sequelae of SARS-CoV-2 infection (PASC) in children and adolescents. PM&R 14(10):1241–1269. Mancini, D. M., D. L. Brunjes, A. Lala, M. G. Trivieri, J. P. Contreras, and B. H. Natelson. 2021. Use of cardiopulmonary stress testing for patients with unexplained dyspnea post-coronavirus disease. JACC Heart Fail Dec;9(12):927-937. https://doi.org/10.1016/j. jchf.2021.10.002. Marshall, J. C., S. Murthy, J. Diaz, N. K. Adhikari, D. C. Angus, Y. M. Arabi, K. Baillie, M. Bauer, S. Berry, B. Blackwood, M. Bonten, F. Bozza, F. Brunkhorst, A. Cheng, M. Clarke, V. Q. Dat, M. de Jong, J. Denholm, L. Derde, J. Dunning, X. Feng, T. Fletcher, N. Foster, R. Fowler, N. Gobat, C. Gomersall, A. Gordon, T. Glueck, M. Harhay, C. Hodgson, P. Horby, Y. Kim, R. Kojan, B. Kumar, J. Laffey, D. Malvey, I. Martin-Loeches, C. McArthur, D. McAuley, S. McBride, S. McGuinness, L. Merson, S. Morpeth, D. Needham, M. Netea, M.-D. Oh, S. Phyu, S. Piva, R. Qiu, H. Salisu-Kabara, L. Shi, N. Shimizu, J. Sinclair, S. Tong, A. Turgeon, T. Uyeki, F. van de Veerdonk, S. Webb, P. Williamson, T. Wolf, and J. Zhang. 2020. A minimal common outcome measure set for COVID-19 clinical research. The Lancet Infectious Diseases 20(8):e192–e197. McGowan, K. 2023. Family caregiver of people with Long COVID bear an extra burden. National Public Radio, February 6. https://www.npr.org/sections/health-shots/2023/02/ 06/1153800857/family-caregivers-of-people-with-long-covid-bear-an-extra-burden (accessed April 19, 2024). Meenakshisundaram, C., A. Moustafa, M. Ranabothu, A. Maraey, and B. Grubb. 2024. Im- pact of COVID-19 infection on baseline autonomic symptoms in patients with preexist- ing postural tachycardia syndrome and orthostatic intolerance: A retrospective study. American Journal of the Medical Sciences 367(5):P323-P327. https://doi.org/10.1016/j. amjms.2023.12.011. Moberg, J., A. D. Oxman, S. Rosenbaum, H. J. Schünemann, G. Guyatt, S. Flottorp, C. Glen- ton, S. Lewin, A. Morelli, G. Rada, P. Alonso-Coello, J. Moberg, A. Oxman, P. A. Coello, H. Schünemann, G. Guyatt, S. Rosenbaum, A. Morelli, E. Akl, C. Glenton, M. Gulmezo- glu, S. Flottorp, S. Lewin, R. A. Mustafa, G. Rada, J. Singh, E. von Elm, J. Vogel, J. Watine, and P. Alonso-Coello for the GRADE Working Group. 2018. The GRADE evidence to decision (EtD) framework for health system and public health decisions. Health Research Policy and Systems 16(1):45. Mohammad, K. O., A. Lin, and J. B. C. Rodriguez. 2022. Cardiac manifestations of post-acute COVID-19 infection. Current Cardiology Reports 24(12):1775–1783. Mohandas, S., P. Jagannathan, T. J. Henrich, Z. A. Sherif, C. Bime, E. Quinlan, M. A. Portman, M. Gennaro, J. Rehman, and R. M. P. T. Force. 2023. Immune mechanisms underlying COVID-19 pathology and post-acute sequelae of SARS-CoV-2 infection (PASC). Elife 12:e86014. PREPUBLICATION COPY—Uncorrected Proofs 

98 A LONG COVID DEFINITION Monika, M., T. D. Andrew, P. H. Jacqueline, H. Chuan, H. Yuk-Lam, T. Vidisha, W. C. Alicia, P. Vidul Ayakulangara, W. Xuan, G. Z. Harrison, Y. Doris, S. Malarkodi Jebathilagam, M. Michele, V. Shyam, R. B.-J. Brendin, B. R. Rachel, M. Sumitra, J. G. Michael, P. L. Kath- erine, X. Zongqi, A. B. Gabriel, C. Tianxi, and C. Kelly. 2023. Characterization of long COVID definitions and clinical coding practices. medRxiv 2023.2010.2004.23296301. Morrow, A. K., L. A. Malone, C. Kokorelis, L. S. Petracek, E. F. Eastin, K. L. Lobner, L. Neuen- dorff, and P. C. Rowe. 2022. Long-term COVID 19 sequelae in adolescents: The overlap with orthostatic intolerance and ME/CFS. Current Pediatrics Reports 10(2):31–44. Munblit, D., M. E. O’Hara, A. Akrami, E. Perego, P. Olliaro, and D. M. Needham. 2022. Long COVID: Aiming for a consensus. Lancet Respiratory Medicine 10(7):632–634. Nanwani-Nanwani, K., L. López-Pérez, C. Giménez-Esparza, I. Ruiz-Barranco, E. Carrillo, M. S. Arellano, D. Díaz-Díaz, B. Hurtado, A. García-Muñoz, M. Relucio, M. Quintana- Díaz, M. R. Úrbez, A. Saravia, M. V. Bonan, F. García-Río, M. L. Testillano, J. Villar, A. García de Lorenzo, and J. M. Añón. 2022. Prevalence of post-intensive care syndrome in mechanically ventilated patients with COVID-19. Science Reports 12(1):7977. NASEM (National Academies of Sciences, Engineering, and Medicine). 2020. Framework for equitable allocation of COVID-19 vaccine. Washington, DC: The National Academies Press. NASEM. 2024. Long-term health effects stemming from COVID-19 and implications for social security administration. https://www.nationalacademies.org/our-work/long-term- health-effects-stemming-from-covid-19-and-implications-for-the-social-security-adminis- tration (accessed April 24, 2024). NIH (National Institutes of Health). 2024. What is Long COVID? Building our understanding about recovery. https://recovercovid.org/long-covid (accessed April 19, 2024). Neville, T. H., R. D. Hays, C. H. Tseng, C. A. Gonzalez, L. Chen, A. Hong, M. Yamamoto, L. Santoso, A. Kung, K. Schwab, S. Y. Chang, N. Qadir, T. Wang, and N. S. Wenger. 2022. Survival after severe COVID-19: Long-term outcomes of patients admitted to an intensive care unit. Journal of Intensive Care Medicine 37(8):1019–1028. NICE (Institute for Health and Care Excellence). 2022. COVID-19 rapid guideline: Managing the longterm effects of COVID-19. https://www.nice.org.uk/guidance/ng188/chapter/5- Management (accessed March 30, 2024). Nuzzo, D., S. Vasto, L. Scalisi, S. Cottone, G. Cambula, M. Rizzo, D. Giacomazza, and P. Picone. 2021. Post-acute COVID-19 neurological syndrome: A new medical challenge. Journal of Clinical Medicine 10(9). Olsen, J. 2021. Caregivers are missing for the Long COVID conversation. STAT, February 18. https://www.statnews.com/2021/02/18/caregivers-missing-long-covid-conversation/ (accessed April 19, 2024). O’Mahoney, L. L., A. Routen, C. Gillies, W. Ekezie, A. Welford, A. Zhang, U. Karamchandani, N. Simms-Williams, S. Cassambai, A. Ardavani, T. J. Wilkinson, G. Hawthorne, F. Cur- tis, A. P. Kingsnorth, A. Almaqhawi, T. Ward, D. Ayoubkhani, A. Banerjee, M. Calvert, R. Shafran, T. Stephenson, J. Sterne, H. Ward, R. A. Evans, F. Zaccardi, S. Wright, and K. Khunti. 2023. The prevalence and long-term health effects of Long COVID among hospitalised and non-hospitalised populations: A systematic review and meta-analysis. EClinicalMedicine 55:101762. Ong, I. Z., D. L. Kolson, and M. K. Schindler. 2023. Mechanisms, effects, and management of neurological complications of post-acute sequelae of COVID-19 (NC-PASC). Biomedi- cines 11(2):377. Orban, Z. S., L. Visvabharathy, G. S. Perez Giraldo, M. Jimenez, and I. J. Koralnik. 2023. SARS-CoV-2-specific immune responses in patients with postviral syndrome after sus- pected COVID-19. Neurology, Neuroimmunology, & Neuroinflammation 10(6):e200159. PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 99 Pagen, D. M. E., C. J. A. van Bilsen, S. Brinkhues, M. Van Herck, K. Konings, C. D. J. den Heijer, H. L. G. ter Waarbeek, M. A. Spruit, C. J. P. A. Hoebe, and N. H. T. M. Dukers- Muijrers. 2023. Prevalence of long-term symptoms varies when using different post- COVID-19 definitions in positively and negatively tested adults: The PRIME post-COVID study. Open Forum Infectious Diseases 10(10):ofad471. Pan, D., and M. Pareek. 2023. Toward a universal definition of post–COVID-19 condition— How do we proceed? JAMA Network Open 6(4):e235779. Park, J. W., X. Wang, and R.-H. Xu. 2022. Revealing the mystery of persistent smell loss in long COVID patients. International Journal of Biological Sciences 18(12):4795–4808. Pavli, A., M. Theodoridou, and H. C. Maltezou. 2021. Post-COVID syndrome: Incidence, clini- cal spectrum, and challenges for primary healthcare professionals. Archives of Medical Research. 52(6):575-581. Peluso, M. J., T. M. Deveau, Munter, S. E., D. Ryder, A. Buck, G. Beck-Engeser, F. Chan, S. Lu, S. A. Goldberg, R. Hoh, V. Tai, L. Torres, N. S. Iyer, M. Deswal, L. H. Ngo, M. Bu- itrago, A. Rodriguez, J. Y. Chen, B. C. Yee, A. Chenna, J. W. Winslow, C. J. Petropoulos, A. N. Deitchman, J. Hellmuth, M. A. Spinelli, M. S. Durstenfeld, P. Y. Hsue, J. D. Kelly, J. N. Martin, S. G. Deeks, P. W. Hunt, and T. J. Henrich. 2022. Impact of pre-existing chronic viral infection and reactivation on the development of Long COVID. medRxiv 06.21.22276660. Peluso, M. J., M. Abdel-Mohsen, T. J. Henrich, and N. R. Roan. 2024a. Systems analysis of innate and adaptive immunity in long COVID. Seminars in Immunology 72:101873. Peluso, M. J., Z. N. Swank, S. A. Goldberg, S. Lu, T. Dalhuisen, E. Borberg, Y. Senussi, M. A. Luna, C. Chang Song, A. Clark, A. Zamora, M. Lew, B. Viswanathan, B. Huang, K. Anglin, R. Hoh, P. Y. Hsue, M. S. Durstenfeld, M. A. Spinelli, D. V. Glidden, T. J. Henrich, J. D. Kelly, S. G. Deeks, D. R. Walt, and J. N. Martin. 2024b. Plasma-based antigen persistence in the post-acute phase of COVID-19. The Lancet Infectious Diseases S1473-3099(24)00211-1. Pfaff, E. R., C. Madlock-Brown, J. M. Baratta, A. Bhatia, H. Davis, A. Girvin, E. Hill, E. Kelly, K. Kostka, J. Loomba, J. A. McMurry, R. Wong, T. D. Bennett, R. Moffitt, C. G. Chute, M. Haendel, N. C. Consortium, and R. Consortium. 2023. Coding long COVID: Character- izing a new disease through an ICD-10 lens. BMC Medicine 21(1):58. Piquet, V., C. Luczak, F. Seiler, J. Monaury, A. Martini, A. B. Ward, J.-M. Gracies, and D. Mo- tavasseli, on behalf of the COVID Rehabilitation Study Group. 2021. Do patients with COVID-19 benefit from rehabilitation? Functional outcomes of the first 100 patients in a COVID-19 rehabilitation unit. Archives of Physical Medicine and Rehabilitation 102:1067–1074. Premraj, L., N. V. Kannapadi, J. Briggs, S. M. Seal, D. Battaglini, J. Fanning, J. Suen, C. Robba, J. Fraser, and S.-M. Cho. 2022. Mid and long-term neurological and neuropsychiatric manifestations of post-COVID-19 syndrome: A meta-analysis. Journal of the Neurologi- cal Sciences 434:120162. Pretorius, E., C. Venter, G. J. Laubscher, M. J. Kotze, S. O. Oladejo, L. R. Watson, K. Rajarat- nam, B. W. Watson, and D. B. Kell. 2022. Prevalence of symptoms, comorbidities, fibrin amyloid microclots and platelet pathology in individuals with long COVID/post-acute sequelae of COVID-19 (PASC). Cardiovascular Diabetology 21(1):148. Pollack, B., E. von Saltza, L. McCorkell, L. Santos, A. Hultman, A. K. Cohen, and L. Soares. 2023. Female reproductive health impacts of Long COVID and associated illnesses including ME/CFS, POTS, and connective tissue disorders: a literature review. Front Rehabil Sci Apr 28(4):1122673. https://doi.org/10.3389/fresc.2023.1122673. Proal, A. D., M. B. VanElzakker, S. Aleman, K. Bach, B. P. Boribong, M. Buggert, S. Cherry, D. S. Chertow, H. E. Davies, and C. L. Dupont. 2023. SARS-CoV-2 reservoir in post-acute sequelae of COVID-19 (PASC). Nature Immunology 24:1616–1627. PREPUBLICATION COPY—Uncorrected Proofs 

100 A LONG COVID DEFINITION Qeadan, F., B. Tingey, J. Egbert, M. G. Pezzolesi, M. R. Burge, K. A. Peterson, and T. Honda. 2022. The associations between COVID-19 diagnosis, type 1 diabetes, and the risk of diabetic ketoacidosis: A nationwide cohort from the U.S. using the Cerner Real-World Data. PLOS One 17(4):e0266809. Rando, H. M., T. D. Bennett, J. B. Byrd, C. Bramante, T. J. Callahan, C. G. Chute, H. E. Davis, R. Deer, J. Gagnier, F. M. Koraishy, F. Liu, J. A. McMurry, R. A. Moffitt, E. R. Pfaff, J. T. Reese, R. Relevo, P. N. Robinson, J. H. Saltz, A. Solomonides, A. Sule, U. Topaloglu, and M. A. Haendel. 2021. Challenges in defining long COVID: Striking differences across literature, electronic health records, and patient-reported information. medRxiv Mar 26:2021.03.20.21253896. Rao, S., R. S. Gross, S. Mohandas, C. R. Stein, A. Case, B. Dreyer, N. M. Pajor, H. T. Bun- nell, D. Warburton, E. Berg, J. B. Overdevest, M. Gorelik, J. Milner, S. Saxena, R. Jhaveri, J. C. Wood, K. E. Rhee, R. Letts, C. Maughan, N. Guthe, L. Castro-Baucom, and M. S. Stockwell. 2024. Postacute sequelae of SARS-CoV-2 in children. Pediatrics 153(3):e2023062570. Rapaport, L. 2024. Long COVID has caused thousands of U.S. deaths: New CDC data. Med- scape, January 3. https://www.medscape.com/viewarticle/long-covid-has-caused-thou- sands-us-deaths-new-cdc-data-2024a100006l?form=fpf (accessed February 29, 2024). Raveendran, A. V. 2021. Long COVID-19: Challenges in the diagnosis and proposed diagnostic criteria. Diabetes & Metabolic Syndrome 15(1):145–146. Reis Carneiro, D., I. Rocha, M. Habek, R. Helbok, J. Sellner, W. Struhal, G. Wenning, and A. Fanciulli. 2023. Clinical presentation and management strategies of cardiovascular au- tonomic dysfunction following a COVID-19 infection—A a systematic review. European Journal of Neurology 30(5):1528–1539. Roedl, K., D. Jarczak, O. Boenisch, G. de Heer, C. Burdelski, D. Frings, B. Sensen, A. Nierhaus, S. Kluge, and D. Wichmann. 2022. Chronic critical illness in patients with COVID-19: Characteristics and outcome of prolonged intensive care therapy. Journal of Clinical Medicine 11(4):1049. Savin, E., Rosenn, G., Tsur, A.M., Hen, O., Ehrenberg, S., Gendelman, O., Buskila, D., Halp- ert, G., Amital, D., Amital, H. 2023. The possible onset of fibromyalgia following acute COVID-19 infection. PLoS One 18(2):e0281593. https://doi.org/10.1371/journal. pone.0281593. Saydah, S. H., J. T. Brooks, and B. R. Jackson. 2022. Surveillance for post-COVID conditions is necessary: Addressing the challenges with multiple approaches. Journal of General Internal Medicine 37(7):1786–1788. Schultheiss, C., E. Willscher, L. Paschold, C. Gottschick, B. Klee, L. Bosurgi, J. Dutzmann, D. Sedding, T. Frese, M. Girndt, J. I. Holl, M. Gekle, R. Mikolajczyk, and M. Binder, 2023. Liquid biomarkers of macrophage dysregulation and circulating spike protein illustrate the biological heterogeneity in patients with post-acute sequelae of COVID-19. J Med Virol 95(1):e28364. Seeley, M. C., C. Gallagher, E. Ong, A. Langdon, J. Chieng, D. Bailey, A. Page, H. S. Lim, and D. H. Lau. 2023. High incidence of autonomic dysfunction and postural orthostatic tachycardia syndrome in patients with Long COVID: Implications for management and health care planning. The American Journal of Medicine S0002-9343(23)00402-3. Sevin, C. M., and E. W. Ely. 2022. Outcomes among patients with 1-year survival after inten- sive care unit treatment for COVID-19. JAMA 327(21):2149–2150. SeyedAlinaghi, S., A. Bagheri, A. Razi, P. Mojdeganlou, H. Mojdeganlou, A. M. Afsahi, A. Afzalian, P. Paranjkhoo, R. Shahidi, P. Mirzapour, Z. Pashaei, M. A. Habibi, P. Shahbazi, S. Nooralioghli Parikhani, N. S. Farizani Gohari, Y. Popoola, E. Mehraeen, and D. Hack- ett. 2023. Late complications of COVID-19: An umbrella review on current systematic reviews. Archives of Academic Emergency Medicine 11(1):e28. PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 101 Shah, B., M. N. Ahmad, M. Khalid, A. Minhas, R. Ali, Z. Sarfraz, and A. Sarfraz. 2023. Long COVID and wavering incidence of pulmonary embolism: A systematic review. Journal of Community Hospital Internal Medicine Perspectives 13(5):23–31. Sherif, Z. A., C. R. Gomez, T. J. Connors, T. J. Henrich, W. B. Reeves, and RECOVER Mecha- nistic Pathway Task Force. 2023. Pathogenic mechanisms of post-acute sequelae of SARS- CoV-2 infection (PASC). Elife 12:e86002. Silva, J., T. Takahashi, J. Wood, P Lu, S. Tabachnikova, J. Gehlhausen, K. Greene, B. Bhat- tacharjee, V. Silva Monteiro, C. Lucas, R. Dhodapkar, L. Tabacof, M. Pena-Hernandez, K. Kamath, T. Mao, D. Mccarthy, R. Medzhitov, D. van Dijk, H. Krumholz, L. Guan, D. Putrino, and A. Iwasaki. 2024. Sex differences in symptomatology and immune profiles of long COVID. medRxiv. https://doi.org/10.1101/2024.02.29.24303568. SSA (Social Security Administration). 2023. Long COVID: A guide for health professionals on providing medical evidence for social security disability claims. https://www.ssa.gov/ disability/professionals/documents/EN-64-128.pdf. (accessed May 3, 2024). Soriano, J. B., S. Murthy, J. C. Marshall, P. Relan, J. V. Diaz, and WHO Clinical Case Working Group on Post-COVID-19 Condition. 2022. A clinical case definition of post-COVID-19 condition by a Delphi consensus. Lancet Infectious Diseases 22(4):e102–e107. Stephenson, T., B. Allin, M. D. Nugawela, N. Rojas, E. Dalrymple, S. Pinto Pereira, M. Soni, M. Knight, E. Y. Cheung, I. Heyman, and R. Shafran. 2022. Long COVID (post-COVID-19 condition) in children: A modified Delphi process. Archives of Disease in Childhood 107(7):674–680. Strang, S., M. Osmanovic, C. Hallberg, and P. Strang. 2018. Family caregivers’ heavy and overloaded burden in advanced chronic obstructive pulmonary disease. Journal of Pal- liative Medicine 21(12):1768–1772. Subramanian, A., K. Nirantharakumar, S. Hughes, P. Myles, T. Williams, K. M. Gokhale, T. Taverner, J. S. Chandan, K. Brown, N. Simms-Williams, A. D. Shah, M. Singh, F. Kidy, K. Okoth, R. Hotham, N. Bashir, N. Cockburn, S. I. Lee, G. M. Turner, G. V. Gkoutos, O. L. Aiyegbusi, C. McMullan, A. K. Denniston, E. Sapey, J. M. Lord, D. C. Wraith, E. Leggett, C. Iles, T. Marshall, M. J. Price, S. Marwaha, E. H. Davies, L. J. Jackson, K. L. Matthews, J. Camaradou, M. Calvert, and S. Haroon. 2022. Symptoms and risk factors for long COVID in non-hospitalized adults. Nature Medicine 28(8):1706–1714. Sylvester, S. V., R. Rusu, B. Chan, M. Bellows, C. O’Keefe, and S. Nicholson. 2022. Sex dif- ferences in sequelae from COVID-19 infection and in long COVID syndrome: A review. Current Medical Research and Opinion 38(8):1391–1399. Szukiewicz, D., P. Wojdasiewicz, M. Watroba, and G. Szewczyk. 2022. Mast cell activation syndrome in COVID-19 and female reproductive function: Theoretical background vs. accumulating clinical evidence. Journal of Immunological Research 9534163. Thaweethai, T., S. E. Jolley, E. W. Karlson, E. B. Levitan, B. Levy, G. A. McComsey, L. Mc- Corkell, G. N. Nadkarni, S. Parthasarathy, U. Singh, T. A. Walker, C. A. Selvaggi, D. J. Shinnick, C. C. M. Schulte, R. Atchley-Challenner, G. A. Alba, R. Alicic, N. Altman, K. Anglin, U. Argueta, H. Ashktorab, G. Baslet, I. V. Bassett, L. Bateman, B. Bedi, S. Bhat- tacharyya, M. A. Bind, A. L. Blomkalns, H. Bonilla, P. A. Bush, M. Castro, J. Chan, A. W. Charney, P. Chen, L. B. Chibnik, H. Y. Chu, R. G. Clifton, M. M. Costantine, S. K. Cribbs, S. I. Davila Nieves, S. G. Deeks, A. Duven, I. F. Emery, N. Erdmann, K. M. Erlandson, K. C. Ernst, R. Farah-Abraham, C. E. Farner, E. M. Feuerriegel, J. Fleurimont, V. Fonseca, N. Franko, V. Gainer, J. C. Gander, E. M. Gardner, L. N. Geng, K. S. Gibson, M. Go, J. D. Goldman, H. Grebe, F. L. Greenway, M. Habli, J. Hafner, J. E. Han, K. A. Hanson, J. Heath, C. Hernandez, R. Hess, S. L. Hodder, M. K. Hoffman, S. E. Hoover, B. Huang, B. L. Hughes, P. Jagannathan, J. John, M. R. Jordan, S. D. Katz, E. S. Kaufman, J. D. Kelly, S. W. Kelly, M. M. Kemp, J. P. Kirwan, J. D. Klein, K. S. Knox, J. A. Krishnan, A. Kumar, A. O. Laiyemo, A. A. Lambert, M. Lanca, J. K. Lee-Iannotti, B. P. Logarbo, M. T. Longo, PREPUBLICATION COPY—Uncorrected Proofs 

102 A LONG COVID DEFINITION C. A. Luciano, K. Lutrick, J. H. Maley, J. G. Marathe, V. Marconi, G. D. Marshall, C. F. Martin, Y. Matusov, A. Mehari, H. Mendez-Figueroa, R. Mermelstein, T. D. Metz, R. Morse, J. Mosier, C. Mouchati, J. Mullington, S. N. Murphy, R. B. Neuman, J. Z. Nikol- ich, I. Ofotokun, E. Ojemakinde, A. Palatnik, K. Palomares, T. Parimon, S. Parry, J. E. Patterson, T. F. Patterson, R. E. Patzer, M. J. Peluso, P. Pemu, C. M. Pettker, B. A. Plunkett, K. Pogreba-Brown, A. Poppas, J. G. Quigley, U. Reddy, R. Reece, H. Reeder, W. B. Reeves, E. M. Reiman, F. Rischard, J. Rosand, D. J. Rouse, A. Ruff, G. Saade, G. J. Sandoval, S. M. Schlater, F. Shepherd, Z. A. Sherif, H. Simhan, N. G. Singer, D. W. Skupski, A. Sowles, J. A. Sparks, F. I. Sukhera, B. S. Taylor, L. Teunis, R. J. Thomas, J. M. Thorp, P. Thuluvath, A. Ticotsky, A. T. Tita, K. R. Tuttle, A. E. Urdaneta, D. Valdivieso, T. M. VanWagoner, A. Vasey, M. Verduzco-Gutierrez, Z. S. Wallace, H. D. Ward, D. E. Warren, S. J. Weiner, S. Welch, S. W. Whiteheart, Z. Wiley, J. P. Wisnivesky, L. M. Yee, S. Zisis, L. I. Horwitz, A. S. Foulkes, and R. Consortium. 2023. Development of a definition of postacute sequelae of SARS-CoV-2 infection. JAMA 329(22):1934–1946. Tran, V. T., R. Porcher, I. Pane, and P. Ravaud. 2022. Course of post COVID-19 disease symptoms over time in the ComPaRe long COVID prospective e-cohort. Nature Com- munications 13(1):1812. Tsampasian, V., H. Elghazaly, R. Chattopadhyay, M. Debski, T. K. P. Naing, P. Garg, A. Clark, E. Ntatsaki, and V. S. Vassiliou. 2023. Risk factors associated with post-COVID-19 condi- tion: Asystematic review and meta-analysis. JAMA Internal Medicine 183(6):ioi230017. Turner, S., M. A. Khan, D. Putrino, A. Woodcock, D. B. Kell, and E. Pretorius. 2023. Long COVID: Pathophysiological factors and abnormalities of coagulation. Trends in Endo- crinology & Metabolism 34(6):321–344. Tyson, A. and G. Pasquini. 2024. How Americans View the Coronavirus, COVID-19 Vaccines Amid Declining Levels of Concern. https://www.pewresearch.org/science/2024/03/07/ how-americans-view-the-coronavirus-covid-19-vaccines-amid-declining-levels-of- concern/ (accessed March 30, 2024). Vashisht, R., A. Patel, and B. O. Crews. 2021. Age- and sex-associated variations in the sensi- tivity of serological tests among individuals infected with SARS-CoV-2. JAMA Network Open 4(2):e210337. Vivaldi, G., P. E. Pfeffer, M. Talaei, T. J. Basera, S. O. Shaheen, and A. R. Martineau. 2023. Long-term symptom profiles after COVID-19 vs other acute respiratory infections: An analysis of data from the COVIDENCE UK study. EClinicalMedicine Oct 6:65:102251. Vu, Q. M., A. L. Fitzpatrick, J. R. Cope, J. Bertolli, N. Sotoodehnia, T. E. West, and E. R. Unger. 2024. Estimates of incidence and predictors of fatiguing illness after SARS-CoV-2 infec- tion. Emerging Infectious Diseases 30(3):539-547. Walker, A. J., B. MacKenna, P. Inglesby, L. Tomlinson, C. T. Rentsch, H. J. Curtis, C. E. Morton, J. Morley, A. Mehrkar, S. Bacon, G. Hickman, C. Bates, R. Croker, D. Evans, T. Ward, J. Cockburn, S. Davy, K. Bhaskaran, A. Schultze, E. J. Williamson, W. J. Hulme, H. I. Mc- Donald, R. Mathur, R. M. Eggo, K. Wing, A. Y. S. Wong, H. Forbes, J. Tazare, J. Parry, F. Hester, S. Harper, S. O’Hanlon, A. Eavis, R. Jarvis, D. Avramov, P. Griffiths, A. Fowles, N. Parkes. I. J. Douglas, S. J. W. Evans, L. Smeeth, and B. Goldacre. 2021. Clinical coding of Long COVID in English primary care: A federated analysis of 58 million patient records in situ using OpenSAFELY. British Journal of General Practice 71(712):e806-e814. Walker, S., H. Goodfellow, P. Pookarnjanamorakot, E. Murray, J. Bindman, A. Blandford, K. Bradbury, B. Cooper, F. L. Hamilton, J. R. Hurst, H. Hylton, S. Linke, P. Pfeffer, W. Ricketts, C. Robson, F. A. Stevenson, D. Sunkersing, J. Wang, G. Gomes, W. Henley, and the Living With COVID Recovery Collaboration. 2023. Impact of fatigue as the primary determinant of functional limitations among patients with post-COVID-19 syndrome: A cross-sectional observational study. British Medical Journal Open 13(6):e069217. PREPUBLICATION COPY—Uncorrected Proofs 

DEFINING LONG COVID 103 Wander, P. L., A. Baraff, A. Fox, K. Cho, M. Maripuri, J. P. Honerlaw, Y. Ho, A. T. Dey, A. M. O’Hare, A. S. B. Bohnert, E. J. Boyko, M. L. Maciejewski, E. Viglianti, T. J. Iwashyna, D. M. Hynes, T. F. Osborne, and G. N. Ioannou. 2023. Rates of ICD-10 Code U09.9 docu- mentation and clinical characteristics of VA patients with post–COVID-19 Condition. Journal of American Medical Association 6(12):e2346783. Weidman, K., E. LaFond, K. L. Hoffman, P. Goyal, C. N. Parkhurst, H. Derry-Vick, E. Schenck, and L. Lief. 2022. Post-intensive care unit syndrome in a cohort of COVID-19 survivors in New York City. Annals of the American Thoracic Society 19(7):1158–1168. WHO (World Health Organization). 2015. World Health Organization best practices for the naming of new human infectious diseases. https://www.who.int/publications/i/item/ WHO-HSE-FOS-15.1 WHO. 2023. A clinical case definition for post-COVID-19 condition in children and adoles- cents by expert consensus. WHO/2019-nCoV/Post_COVID-19_condition/CA/Clinical_ case_definition/2023.1 Wingrove, J., J. Makaronidis, F. Prados, B. Kanber, M. C. Yiannakas, C. Magee, G. Castellazzi, L. Grandjean, X. Golay, C. Tur, O. Ciccarelli, E. D’Angelo, C. A. M. Gandini Wheeler- Kingshott, and R. L. Batterham. 2023. Aberrant olfactory network functional connectiv- ity in people with olfactory dysfunction following COVID-19 infection: An exploratory, observational study. EClinicalMedicine 58:101883. Woodrow, M., C. Carey, N. Ziauddeen, R. Thomas, A. Akrami, V. Lutje, D. C. Greenwood, and N. A. Alwan. 2023. Systematic review of the prevalence of Long COVID. Open Forum Infectious Diseases 10(7):ofad233. Wulf Hanson, S., C. Abbafati, J. G. Aerts, Z. Al-Aly, C. Ashbaugh, T. Ballouz, O. Blyuss, P. Bobkova, G. Bonsel, S. Borzakova, D. Buonsenso, D. Butnaru, A. Carter, H. Chu, C. De Rose, M. M. Diab, E. Ekbom, M. El Tantawi, V. Fomin, R. Frithiof, A. Gamirova, P. V. Glybochko, J. A. Haagsma, S. Haghjooy Javanmard, E. B. Hamilton, G. Harris, M. H. Heijenbrok-Kal, R. Helbok, M. E. Hellemons, D. Hillus, S. M. Huijts, M. Hultström, W. Jassat, F. Kurth, I. M. Larsson, M. Lipcsey, C. Liu, C. D. Loflin, A. Malinovschi, W. Mao, L. Mazankova, D. McCulloch, D. Menges, N. Mohammadifard, D. Munblit, N. A. Nekliudov, O. Ogbuoji, I. M. Osmanov, J. L. Peñalvo, M. S. Petersen, M. A. Puhan, M. Rahman, V. Rass, N. Reinig, G. M. Ribbers, A. Ricchiuto, S. Rubertsson, E. Samitova, N. Sarrafzadegan, A. Shikhaleva, K. E. Simpson, D. Sinatti, J. B. Soriano, E. Spiridonova, F. Steinbeis, A. A. Svistunov, P. Valentini, B. J. van de Water, R. van den Berg-Emons, E. Wallin, M. Witzenrath, Y. Wu, H. Xu, T. Zoller, C. Adolph, J. Albright, J. O. Amlag, A. Y. Aravkin, B. L. Bang-Jensen, C. Bisignano, R. Castellano, E. Castro, S. Chakrabarti, J. K. Collins, X. Dai, F. Daoud, C. Dapper, A. Deen, B. B. Duncan, M. Erickson, S. B. Ewald, A. J. Ferrari, A. D. Flaxman, N. Fullman, A. Gamkrelidze, J. R. Giles, G. Guo, S. I. Hay, J. He, M. Helak, E. N. Hulland, M. Kereselidze, K. J. Krohn, A. Lazzar-Atwood, A. Lindstrom, R. Lozano, D. C. Malta, J. Månsson, A. M. Mantilla Herrera, A. H. Mokdad, L. Monasta, S. Nomura, M. Pasovic, D. M. Pigott, R. C. Reiner, Jr., G. Reinke, A. L. P. Ribeiro, D. F. Santomauro, A. Sholokhov, E. E. Spurlock, R. Walcott, A. Walker, C. S. Wiysonge, P. Zheng, J. P. Bettger, C. J. L. Murray, and T. Vos. 2022. Estimated global proportions of individuals with persistent fatigue, cognitive, and respiratory symptom clusters following symptomatic COVID-19 in 2020 and 2021. JAMA 328(1):1604–1615. Xie, Y., B. Bowe, and Z. Al-Aly. 2021. Burdens of post-acute sequelae of COVID-19 by severity of acute infection, demographics and health status. Nature Communications 12(1):6571. Xie, Y., E. Xu, B. Bowe, and Z. Al-Aly. 2022. Long-term cardiovascular outcomes of CO- VID-19. Nature Medicine 28(3):583–590. Xu, E., Y. Xie, and Z. Al-Aly. 2022. Long-term neurologic outcomes of COVID-19. Nature Medicine 28(11):2406–2415. PREPUBLICATION COPY—Uncorrected Proofs 

104 A LONG COVID DEFINITION Xu, E., Y. Xie, and Z. Al-Aly. 2023. Risks and burdens of incident dyslipidaemia in long COVID: a cohort study. Lancet Diabetes Endocrinol Feb;11(2):120-128. https://doi. org/10.1016/S2213-8587(22)00355-2. Yin, K., M. J. Peluso, X. Luo, R. Thomas, M. G. Shin, J. Neidleman, A. Andrew, K. Young, T. Ma, R. Hoh, K. Anglin, B. Huang, U. Argueta, M. Lopez, D. Valdivieso, K. Asare, T. M. Deveau, S. E. Munter, R. Ibrahim, L. Standker, S. Lu, S. A. Goldberg, S. A. Lee, K. L. Lynch, J. D. Kelly, J. N. Martin, J. Munch, S. G. Deeks, T. J. Henrich, and N. R. Roan. 2023. Long COVID Manifests with T cell dysregulation, inflammation, and an uncoordinated adaptive immune response to SARS-CoV-2. bioRxiv https://doi. org/10.1101/2023.02.09.527892. Zhang, Y., H. Hu, V. Fokaidis, C. Lewis V, J. Xu, C. Zang, Z. Xu, F. Wang, M. Koropsak, J. Bian, J. Hall, R. L. Rothman, E. A. Shenkman, W. Q. Wei, M. G. Weiner, T. W. Carton, and R. Kaushal. 2023. Identifying environmental risk factors for post-acute sequelae of SARS-CoV-2 infection: An EHR-based cohort study from the RECOVER Program. Environmental Advances 11:100352. Zhao, S., S. Toniolo, A. Hampshire, and M. Husain. 2023. Effects of COVID-19 on Cognition and Brain Health. Trends in Cognitive Science 27(11):1053-1067. Zheng, Y.-B., N. Zeng, K. Yuan, S.-S. Tian, Y.-B. Yang, N. Gao, X. Chen, A.-Y. Zhang, A. L. Kondratiuk, P.-P. Shi, F. Zhang, J. Sun, J.-L. Yue, X. Lin, L. Shi, A. Lalvani, J. Shi, Y.-P. Bao, and L. Lu. 2023. Prevalence and risk factor for long COVID in children and ado- lescents: A meta-analysis and systematic review. Journal of Infection and Public Health 16(5):660–672. Zuin, M., G. Rigatelli, V. Battisti, G. Costola, L. Roncon, and C. Bilato. 2022. Increased risk of acute myocardial infarction after COVID-19 recovery: A systematic review and meta- analysis. International Journal of Cardiology 372:138–143. PREPUBLICATION COPY—Uncorrected Proofs