Evaluation of Post-Acute Sequelae of SARS-CoV-2 (PASC) Index and a Recent Long COVID Criteria in Korean Long COVID-19 Cohort

Abstract

Background

To compare the post-acute sequelae of severe acute respiratory syndrome-coronavirus-2 (SARS-CoV-2) infection (PASC) index and the National Academies of Sciences, Engineering, and Medicine (NASEM) criteria in identifying long coronavirus disease (COVID) among adults with confirmed coronavirus disease 2019.

Methods

A prospective cohort study was conducted from November 2022 to February 2025 at a single tertiary care hospital in Seoul, Korea. Adults aged 18 years or older with confirmed SARS-CoV-2 infection were enrolled, yielding a total of 183 participants. Follow-up assessments took place at 1-, 3-, 6-, and 12-months post-infection. The primary outcome was the prevalence of long COVID at 12 months, measured using the PASC index (≥ 12 points across specified symptoms) and the NASEM criteria (≥ 1 symptom persisting for at least 3 months).

Results

Of 183 participants, 26.2% (48/183) met the PASC index, whereas 47.5% (87/183) fulfilled the NASEM criteria. Of the 48 patients who met the PASC index, 44 (91.7%) also met the NASEM criteria.

Conclusion

The NASEM criteria classified nearly half of participants with long COVID and covered those with PASC index, while the PASC index identified about one quarter. Although the NASEM criteria capture a broader range of persistent symptoms, the PASC index may offer a more stringent threshold, potentially informing targeted research and clinical decision-making.

Graphical Abstract

INTRODUCTION

According to a report by the WHO, over 700 million people have been infected with the virus since the coronavirus disease 2019 (COVID-19) pandemic.1 Despite the global decline in COVID-19 infections, the disease continues to exert long-term effects on healthcare systems.2,3,4 One such long-term effect is “long COVID,” also referred to as post-acute sequelae of severe acute respiratory syndrome-coronavirus-2 (SARS-CoV-2) infection (PASC) or post-COVID conditions.5 This condition affects more than 400 million individuals worldwide and places a substantial burden on both healthcare resources and society.6,7 Nonetheless, its clinical manifestations remain highly heterogeneous and multifaceted, leaving the underlying pathophysiological mechanisms, largely unexplained.8,9,10

The complexity of long COVID is evident in the use of multiple terms and definitions.11 Without a gold-standard definition, establishing consistent diagnostic and classification criteria becomes challenging, which in some cases has led to dismissing symptoms.12,13 Such skepticism and delayed recognition of the condition have inevitably impeded research progress.14 Compounding this issue, many studies investigating long COVID adopt their own definitions, complicating efforts to develop unified diagnostic criteria or identify reliable biomarkers and therapeutic agents.15,16,17 In response, the National Institutes of Health’s Researching COVID to Enhance Recovery (RECOVER) initiative introduced the PASC index, and the National Academies of Sciences, Engineering, and Medicine (NASEM) also published a novel criteria of long COVID.18,19 The PASC index was the first scoring system developed to objectively assess long COVID, but it does not account for symptom duration and may be ambiguous for patients whose scores fall just below the diagnostic threshold. The NASEM criteria diagnose long COVID based on the persistence of symptoms for at least three months without requiring diagnostic confirmation of COVID-19 infection and has been adopted by the U.S. Centers for Disease Control and Prevention for public health policy purposes.20 Although this definition is broad and inclusive, this broad nature has been reported to limit clinical acceptance among clinicians.21 Both approaches have their advantages and limitations, comparative clinical data remain limited.

Therefore, the present study aimed to classify long COVID cases using both the PASC index and the NASEM criteria, and to assess the clinical profiles associated with each method.

METHODS

Study design and population

This prospective cohort study was conducted at a single tertiary care hospital in Seoul, Korea, from November 2022 to February 2025. Adults (≥ 18 years of age) with COVID-19 confirmed via rapid antigen testing or polymerase chain reaction were eligible for enrollment. Participants were recruited through in-hospital and community advertisements, as well as from outpatient clinics and inpatient wards, following informed consent.

Data collection and follow-up

The study protocol included follow-up visits at 1, 3, 6, and 12 months post-acute COVID-19 infection. At each visit, participants underwent clinical evaluations and completed surveys assessing long COVID symptoms. While clinical assessments were conducted in-person, surveys were administered online. At the baseline visit, participants provided demographic information through a questionnaire, including age, sex, vaccination history, number of COVID-19 episodes, and body mass index. Additional information on comorbidities, such as smoking status, chronic lung diseases (asthma and chronic obstructive pulmonary disease), diabetes mellitus, chronic liver disease, and other medical history were obtained through a review of electronic medical records.

In May 2023, the RECOVER Initiative introduced criteria for the PASC index. Consequently, from July 2023 onwards, an additional questionnaire for PASC index calculation was incorporated into the study protocol. To compensate for the lack of PASC index data for participants enrolled before July 2023, the study also recruited individuals who were in the post-acute phase of COVID-19 (Fig. 1).

Fig. 1. Study flow diagram of long COVID assessment in a prospective cohort.

NASEM = National Academies of Sciences, Engineering, and Medicine, PASC = post-acute sequelae of SARS-CoV-2 infection, COVID = coronavirus disease.

Outcomes and definitions

The primary outcomes of this study were the prevalence of long COVID at 12 months post-infection, assessed using two distinct definitions: the PASC index and the NASEM criteria (Supplementary Tables 1 and 2). The PASC index, developed as part of the RECOVER Initiative, classifies long COVID based on a symptom score. Participants with a score of 12 or higher across 12 specific symptoms one month after infection were considered to have long COVID according to this definition. The NASEM criteria, in contrast, define long COVID as the presence of at least one common symptom or diagnosable condition persisting intermittently or continuously for three months or longer. For this study, we focused solely on symptoms, excluding diagnosable conditions to maintain consistency with the symptom-based PASC index. To establish symptom persistence, we applied the NASEM long COVID classification when identical symptoms were reported at a minimum of two time points among the one-, three-, six-, and twelve-month assessments. Cases meeting the criteria only at the one- and twelve-month points were excluded to ensure a more robust definition of persistence (Fig. 1).

Statistical analysis

For continuous variables, data were summarized as the median and interquartile range, whereas categorical variables were described using counts and percentages. To compare groups, independent t-tests, Mann-Whitney U tests, or χ² tests were applied, as appropriate. To investigate risk factors for long COVID as defined by each criterion, we conducted a comprehensive statistical analysis. Variables demonstrating statistical significance in univariate analyses, along with those previously identified as risk factors in the literature, were incorporated into multivariate logistic regression models. The final model was refined using a backward stepwise selection process. We assessed the robustness of the final model using the Hosmer-Lemeshow goodness-of-fit test. Potential multicollinearity among predictor variables was examined using the variance inflation factor. Statistical significance was defined as a two-tailed P value below 0.05. As a sensitivity analysis, we evaluated a complete-case cohort of participants who completed all four follow-up visits at 1, 3, 6, and 12 months (n = 39). In this subgroup, we described the proportions of long COVID according to the NASEM criteria and PASC index definitions and compared these findings with those of the primary analysis to assess robustness. Analyses were conducted using RStudio (version 4.4.1).

Ethics statement

This study was approved by the Institutional Review Board (IRB) of Asan Medical Center (IRB No. 2022–1477). Written informed consent was obtained from all participants.

RESULTS

Clinical characteristics of study participants

A total of 183 participants were enrolled in this cohort study. As summarized in Table 1, the median age of the cohort was 48.0 years, and 38.3% of the participants were male. At initial presentation, 38.5% required hospitalization. 80.8% of the participants had completed full vaccination.

Table 1. Baseline and clinical characteristics of participants in the study.

Characteristics		Involved participants (N = 183)
Age, yr, median (IQR)		48.0 (34.5–62.0)
Male gender		70 (38.3)
Body mass index, kg/m2, median (IQR)		22.0 (20.1–24.8)
No. of COVID-19 episodesa
	1	144 (79.6)
	2	35 (19.3)
	3	2 (1.1)
Vaccination status for COVID-19b,c
	Fully vaccinated	13 (7.1)
	Partially vaccinated	22 (12.1)
	Unvaccinated	147 (80.8)
Smokingb
	Non-smoker	138 (75.8)
	Smoker	44 (24.2)
Charlson comorbidity index, median (IQR)b		1.0 (0.0–4.0)
Underlying diseaseb		89 (48.0)
	COPD	6 (3.3)
	Asthma	4 (2.2)
	Diabetes mellitus	37 (20.3)
	CKD	27 (14.8)
	Chronic liver diseases	5 (2.7)
	Solid tumor	24 (13.2)
	Hematologic malignancy	9 (4.9)
	Hematopoietic stem cell transplant	4 (2.2)
	Solid organ transplant	29 (15.9)
	Mental health problems	6 (3.3)
Immunosuppressantb		29 (15.9)
Hospitalization for COVID-19b		70 (38.5)
Hospital treatmentb
	Oxygen therapy	41 (22.5)
	Steroid	44 (24.2)
	Remdesivir	71 (39.0)
	Tocilizumab	10 (5.5)
	Baricitinib	9 (4.9)

Data are presented as number (%) of patients, unless otherwise indicated.

IQR = Interquartile range, COVID-19 = coronavirus disease 2019, COPD = chronic obstructive pulmonary disease, CKD = chronic kidney disease.

^aData were missing for 2 patients.

^bData were missing for 1 patient.

^cDefinition of vaccination: unvaccinated (0 doses), partially vaccinated (1–2 doses), and fully vaccinated (≥ 3 doses).

During the follow-up period, response rates were recorded at time points as follows: at 1 month, 63 of 183 participants (34.4%); at 3 months, 90 of 183 (49.2%); at 6 months, 106 of 183 (57.9%); and at 12 months, 154 of 183 (84.2%). When classified according to the PASC index and the NASEM criteria, 48 participants (26.2%) met the PASC index, while 87 (47.5%) fulfilled the NASEM criteria. Among these, 44 participants satisfied both criteria (Supplementary Fig. 1). Among the four individuals who were diagnosed with long COVID based only on the PASC index, three were lost to follow-up and one died.

Factors associated with the PASC index and NASEM criteria

A comparison of baseline characteristics according to the PASC index and NASEM criteria is summarized in Supplementary Tables 3 and 4. The results of the univariate analysis and the multivariate findings are summarized in Tables 2 and 3. In the multivariable analysis, when long COVID was defined according to the PASC index, the only significant risk factor for long COVID was age (adjusted odds ratio [aOR], 1.08; 95% confidence interval [CI], 1.05–1.12; P < 0.001).

Table 2. Univariate and multivariate analysis of long COVID based on the PASC index.

Variables		Univariate analysis			Multivariate analysis
		Odds ratio	95% CI	P value	Adjusted odds ratio	95% CI	P value
Age		1.09	1.06–1.13	< 0.001***	1.08	1.05–1.12	< 0.001***
Sex, male		3.75	1.89–7.62	< 0.001***
Vaccination status
	Unvaccinated	1.00	-	-	1.00	-	-
	Partially vaccinated	0.25	0.04–1.27	0.103	0.16	0.02–1.12	0.074
	Fully vaccinated or boosted	0.59	0.19–2.05	0.377	0.58	0.13–2.62	0.469
Smoking		2.55	1.23–5.29	0.011*
Charlson comorbidity index		1.54	1.33–1.82	< 0.001***
Body mass index		0.99	0.91–1.01	0.615
Asthma		9.00	1.09–185.0	0.059
COPD		15.7	2.45–306.0	0.013*	6.14	0.80–128.0	0.123
Diabetes mellitus		3.97	1.84–8.64	< 0.001***
CKD		3.25	1.39–7.62	0.006
Solid organ transplant		2.8	1.21–6.39	0.014*	2.38	0.90–6.40	0.081
Immunosuppressant		2.34	1.00–5.35	0.049*
Hospitalization for COVID-19		4.67	2.33–9.69	< 0.001***
Oxygen use		2.57	1.22–5.4	0.012*
Steroid use		2.93	1.41–6.08	0.004**
Remdesivir use		4.5	2.25–9.33	< 0.001***

COVID-19 = coronavirus disease 2019, PASC = post-acute sequelae of SARS-CoV-2 infection, CI = confidence interval, COPD = chronic obstructive pulmonary disease, CKD = chronic kidney disease.

^*P < 0.05, ^**P < 0.01, ^***P < 0.001.

Table 3. Univariate and multivariate analysis of long COVID based on the NASEM criteria.

Variables		Univariate analysis			Multivariate analysis
		Odds ratio	95% CI	P value	Adjusted odds ratio	95% CI	P value
Age		1.03	1.01–1.05	0.004**	1.04	1.02–1.07	< 0.001***
Sex, male		1.04	0.57–1.90	0.898	0.52	0.23–1.13	0.107
No. of COVID-19 episodes		2.20	1.11–4.58	0.028*	3.19	1.54–7.00	0.003**
Vaccination status
	Unvaccinated	1.00	-	-
	Partially vaccinated	1.17	0.29–4.73	0.826
	Fully vaccinated or boosted	1.03	0.33–3.34	0.959
Charlson comorbidity index		1.25	1.11–1.43	< 0.001***
Body mass index		0.99	0.93–1.01	0.475
Diabetes mellitus		2.72	1.28–6.03	0.01*
CKD		3.87	1.61–10.4	0.004**	4.14	1.62–11.7	0.004**
Solid organ transplant		2.94	1.29–7.19	0.013*
Immunosuppressant		2.94	1.29–7.19	0.013*
Hospitalization for COVID-19		2.85	1.54–5.57	< 0.001***
Steroid use		2.43	1.21–4.98	0.013*
Remdesivir use		2.70	1.47–5.05	0.002**

COVID = coronavirus disease, NASEM = National Academies of Sciences, Engineering, and Medicine, CI = confidence interval, COVID-19 = coronavirus disease 2019, CKD = chronic kidney disease.

^*P < 0.05, ^**P < 0.01, ^***P < 0.001.

Using the NASEM criteria to define long COVID, significant risk factors for long COVID were identified as age (aOR, 1.04; 95% CI, 1.02–1.07; P < 0.001), the number of COVID-19 episodes (aOR, 3.19; 95% CI, 1.55–7.00; P = 0.003), and chronic kidney disease (CKD; aOR, 4.14; 95% CI, 1.62–11.7; P = 0.004).

Long COVID-defined symptom frequency over time

A summary of symptom frequency at each time point, based on two definitions of long COVID, was presented (Fig. 2). Participants with the PASC index (≥ 12) most commonly reported brain fog, fatigue, and thirst, with frequency ranging from 16% to 30%. A similar pattern was observed in the non-PASC group. However, alterations in smell or taste and post-exertional malaise were observed in fewer than 2% of non-PASC participants. In the NASEM-defined group, brain fog, fatigue, and headache were the most frequent symptoms, with frequency ranging from 17% to 48% across the assessed time points.

Fig. 2. Temporal symptom distributions by PASC index and NASEM criteria. Each cell in the heatmap represents the frequency (%) at the corresponding time point. The shading in each column reflects the frequency, ranging from 0% up to 50%.

PASC = post-acute sequelae of SARS-CoV-2 infection, GI = gastrointestinal, NASEM = National Academies of Sciences, Engineering, and Medicine.

Sensitivity analysis

Among the 39 complete responders, 31 (79.5%) met the NASEM criteria and 14 (35.9%) met the PASC index. All participants classified as PASC-positive were also NASEM-positive, resulting in 17 participants (43.6%) classified exclusively by the NASEM criteria and none classified exclusively by the PASC index.

DISCUSSION

In this prospective study, we enrolled patients and monitored their symptoms to compare two recently proposed definitions of long COVID. The findings demonstrated that nearly half of the participants were classified under the NASEM criteria, while around one quarter fell into the PASC index. Most individuals who met the PASC index also satisfied the NASEM criteria. Thus, although the NASEM criteria reflects a duration of at least three months, it remains broader and more inclusive in scope. Furthermore, Bello-Chavolla et al.22 previously reported that the PASC index ≥ 12 represents a severe phenotype of long COVID. These findings suggest that patients presenting with multiple symptoms meeting the PASC index threshold after COVID-19 infection are likely to experience persistent and clinically significant manifestations. Therefore, a high PASC index score may serve as an early indicator of severe long COVID, warranting close clinical monitoring, particularly in the absence of long-term follow-up.

Previous studies have reported that the prevalence of long COVID among individuals diagnosed with COVID-19 varies considerably, ranging from approximately 7% to 50%.23,24,25,26,27 Such discrepancies are thought to stem from differing definitions of long COVID used across various investigations.14 In the RECOVER cohort study, Thaweethai et al.19 described that 23% of COVID infected individuals met the PASC index, aligning with the present study.

In the present study, participants with a PASC index of 12 or higher exhibited a persistent prevalence of symptoms such as brain fog and fatigue over time (Fig. 2). Notably, these symptoms remained significantly higher in subjects not classified by the PASC index, which is the opposite of the NASEM classification. Given that these symptoms are a key feature of long COVID, this suggests that some patients below the PASC index threshold persist with symptoms that would be classified as long COVID and would be diagnosed with long COVID if the NASEM criteria were applied.24,28,29,30 When the PASC index falls below a certain threshold but symptoms are present (PASC index 1-11), it does not adequately capture these persistent symptoms. This suggests that the NASEM criteria, which includes such cases, may provide a more inclusive and clinical framework for diagnosing long COVID. However, given that the PASC index ≥ 12 is predominantly incorporated into the NASEM criteria and the PASC index is a metric that limits the probability of exceeding the threshold to approximately 5% in uninfected individuals, it possesses the capacity to mitigate false positives in individuals who align with the NASEM definition.19 Therefore, by providing an objective and granular set of criteria, the PASC index is expected to greatly facilitate future research on long COVID.

According to the multivariable regression analyses of long COVID risk factors, older age was identified as an independent risk factor in both models, and the number of COVID-19 episodes emerged as an independent risk factor based on the NASEM criteria of long COVID. Previous studies have consistently shown that older age and repeated COVID-19 infection increase the risk of long COVID, whereas the role of CKD remains controversial.23,31,32,33 In the present cohort, 19 patients with CKD (66.7%) had undergone solid organ transplantation and were receiving immunosuppressive therapy (data not shown). Because previous research has suggested that immunosuppressed status is a risk factor for long COVID, we presume that this underlying immunosuppression may have contributed to CKD being identified as an independent risk factor in our study.33 Moreover, repeated COVID-19 infection was identified as a significant factor only according to the NASEM criteria of long COVID. This finding suggests that repeated infection may contribute to the persistence of long COVID, underscoring the importance of strategies aimed at preventing reinfection to reduce the burden of long COVID.

Our study has several limitations. First, although this study employed a prospective design, the sample size was relatively small and was not a multicenter study, and the response rate at the 1-month follow-up was especially low. Such limited data may restrict the validity of the findings. However, to address potential bias arising from incomplete follow-up, we conducted a sensitivity analysis restricted to participants who completed all four follow-up visits. The distribution of long COVID definitions in this complete-case cohort closely mirrored that observed in the full cohort, supporting the robustness of our findings despite the limited response rate at early follow-up. Second, because the study was performed in a single tertiary care center primarily treating patients with severe acute illness, the proportion of hospitalized cases was relatively high, potentially limiting the generalizability of our findings. Third, the study results were limited to assessments at 1, 3, 6, and 12 months, and the NASEM criteria of “symptom persistence for three months or more” was applied solely by checking whether the same symptom was present at two distinct time points. Potential fluctuations in symptoms between these intervals may not have been fully captured. Therefore, there is a risk that symptom severity and duration may have been either under- or overestimated. Fourth, for comparison with the PASC index, we excluded all diagnosable conditions based on the NASEM criteria and did not include individuals who were never diagnosed with COVID-19. To address these limitations, further large-scale prospective studies following both infected and uninfected populations are needed. Lastly, the PASC index was updated in 2024.34 Although the report indicated that the original index and its refined version are largely consistent, there may still be subtle differences that represent a potential limitation of this study.

In conclusion, this study compared two updated definitions of long COVID within a prospective cohort. According to the PASC index, 26% of participants were identified as having long COVID, whereas the NASEM criteria identified 48%. Notably, the NASEM criteria largely encompassed those classified by the PASC index. While the broader scope of the NASEM criteria may capture more individuals with long COVID, the PASC index offers a more specific metric that can provide researchers with clearer insights.

SUPPLEMENTARY MATERIALS

Supplementary Table 1

The survey questions and a scoring system consisting of the post-acute sequelae of SARS-CoV-2 infection (PASC) index

Supplementary Table 2

The survey questions and common symptoms consisting of the NASEM criteria of long COVID

Supplementary Table 3

Baseline and clinical characteristics of participants according to NASEM criteria

Supplementary Table 4

Baseline and clinical characteristics of participants according to PASC index

Supplementary Fig. 1

Venn diagram illustrating overlap between the PASC index and NASEM criteria.