Identifying commonalities and differences between EHR representations of PASC and ME/CFS in the RECOVER EHR cohort

Abstract

Background: Shared symptoms and biological abnormalities between post-acute sequelae of SARS-CoV-2 infection (PASC) and myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) could suggest common pathophysiological bases and would support coordinated treatment efforts. Empirical studies comparing these syndromes are needed to better understand their commonalities and differences.

Methods: We analyzed electronic health record data from 6.5 million adult patients from the National COVID Cohort Collaborative. PASC and ME/CFS diagnostic groups were defined based on recorded diagnoses, and other recorded conditions within the two groups were used to train separate machine learning-driven computable phenotypes (CPs). The most predictive conditions for each CP were examined and compared, and the overlap of patients labeled by each CP was examined. Condition records from the diagnostic groups were also used to statistically derive condition clusters. Rates of subphenotypes based on these clusters were compared between PASC and ME/CFS groups.

Results: Approximately half of patients labeled by one CP are also labeled by the other. Dyspnea, fatigue, and cognitive impairment are the most-predictive conditions shared by both CPs, whereas other most-predictive conditions are specific to one CP. Recorded conditions separate into cardiopulmonary, neurological, and comorbidity clusters, with the cardiopulmonary cluster showing partial specificity for the PASC groups.

Conclusions: Data-driven approaches indicate substantial overlap in the condition records associated with PASC and ME/CFS diagnoses. Nevertheless, cardiopulmonary conditions are somewhat more commonly associated with PASC diagnosis, whereas other conditions, such as pain and sleep disturbances, are more associated with ME/CFS diagnosis. These findings suggest that symptom management approaches to these illnesses could overlap.

Fig. 1. Longitudinal trends in index dates of diagnostic groups.

a Counts of index dates for diagnostic groups over time, binned by quarter. In this panel, “PASC_ME/CFS” (red dashed and dotted line) indicates patients who met criteria for the PASC and ME/CFS diagnostic groups, while “PASC” (blue solid line) and “ME/CFS” (green dashed line) indicate patients who met criteria only for that given group. b Cumulative counts of index dates for diagnostic groups over time, binned by quarter. In this panel, lines indicate all patients who met criteria for that diagnostic group, i.e., patients in both groups are included in both lines.

Fig. 2. Overlap of patient groups.

a Patient overlap in diagnostic groups, with index dates restricted to October 1, 2022 onward. b Patient overlap in CP groups, with index dates restricted to July 1, 2020 onward. Blue shading = PASC only; red shading = overlap / in both groups; green shading = ME/CFS only.

Fig. 3. Comparison of most-predictive conditions for the PASC and ME/CFS CPs.

An importance score threshold of 0.05 was used to define most-predictive conditions for each CP. Red bubbles indicate conditions that meet this criterion for both CPs, blue bubbles for PASC only, and green bubbles for ME/CFS only. Note different ranges on axes. Bubble sizes are proportional to the percentage of patients in the cluster definition group with a record of the given condition. Condition names are abbreviated. Abbreviations follow, but full names can also be found in Supplementary Table 4. Attention = Finding related to attentiveness; C. = chronic; Depressive = Depressive disorder; GAD = Generalized anxiety disorder; Hypothyr. = Hypothyroidism; Immune = Disorder of immune function; OSAS = Obstructive sleep apnea syndrome; Smell/taste = Sensory disorder of smell and/or taste; syn. = syndrome; Vit B. = Vitamin B deficiency; Vit. D = Vitamin D deficiency.

Fig. 4. Condition clusters.

Clusters indicate patterns of condition record co-occurrence in a combined PASC and ME/CFS group. Clusters are denoted by box color with blue boxes corresponding to the neurological cluster, yellow boxes to the cardiopulmonary cluster, and brown boxes to the comorbidity cluster. Box sizes correspond to the weight of each condition within its cluster. Condition names are abbreviated. Abbreviations follow, but full names can also be found in Supplementary Data 2. Ab. ECG = Electrocardiogram abnormal; Ab. lung imaging = Abnormal findings on diagnostic imaging of lung; Anxiety = Anxiety disorder; Asthma = Uncomplicated asthma; Atherosclerosis = Atherosclerosis of coronary artery without angina pectoris; Attention = Finding related to attentiveness; Blood chem. = Blood chemistry outside reference range; C. = chronic; COPD = Chronic obstructive lung disease; Depression episode = Major depression, single episode; Depressive = Depressive disorder; Dizziness = Dizziness and giddiness; GAD = Generalized anxiety disorder; GERD w/o esophagitis = Gastroesophageal reflux disease without esophagitis; Hyperlipidemia = Mixed hyperlipidemia; Hypertension = Essential hypertension; Hypothyr. = Hypothyroidism; OSAS = Obstructive sleep apnea syndrome; Postop. state = Postoperative state; Type 2 DM = Type 2 diabetes mellitus; Type 2 DM w/o complication = Type 2 diabetes mellitus without complication; Vit. D = Vitamin D deficiency.

Fig. 5. Condition cluster and subphenotype rates across patient groups.

a Proportion of patients in each diagnostic group meeting criteria for each condition cluster. b Proportion of patients in each diagnostic group meeting criteria for each subphenotype. c Proportion of patients in each CP group meeting criteria for each condition cluster. d Proportion of patients in each CP group meeting criteria for each subphenotype. “PASC_ME/CFS” indicates patients in the PASC and ME/CFS groups, while “PASC” and “ME/CFS” indicate patients only in that group. Subphenotype categories are applied exclusively, e.g., a subphenotype of “1” indicates patients who meet criteria for cluster 1 only. In contrast, the cluster heatmaps indicate all patients who meet criteria for each cluster, e.g., cluster 1 includes patients from the “1,” “1 & 2,” “1 & 3,” and “All” subphenotypes. Darker shading corresponds to higher proportions. Cluster 1 = neurological cluster; Cluster 2 = comorbidity cluster; Cluster 3 = cardiopulmonary cluster.