Post-COVID-19 Condition and Health Status

Abstract

Background: Observational studies of the long-term effects of COVID-19 infection generally focus on individual symptoms rather than health status. Objective: Longitudinal assessment of general health status following COVID-19 infection. Design: Observational study, with data collected from two telephone surveys at 32 ± 10 and 89 ± 25 days after discharge from the hospital or emergency department (ED) for a COVID-19 infection. Medicaid or no insurance was our marker of low socioeconomic status (SES). Acute disease severity was determined by summing 10 severity markers (yes-no) from the health encounter. Baseline comorbidity was a modified Charlson Index. Participants: 40 patients. Mean age was 54 ± 15 years, 50% were female, and 40% had low socioeconomic status. Main Measures: (1) the 20-item Medical Outcomes Study Short-Form General Health Survey (SF-20); (2) Dyspnea (modified Medical Research Council); (3) Psychological symptoms (Patient Health Questionnaire for Anxiety and Depression); (4) Cognitive function (Cognitive Change Questionnaire); (5) Fatigue (Short Fatigue Questionnaire); (6) A 10-item review of systems (ROS) questionnaire. Key Results: Percentages with abnormal symptoms at the first and second surveys were (respectively): Dyspnea (40, 33), Fatigue (53, 50), Anxiety (33, 18), Depression (20, 10), PHQ-4 Composite (25, 13), and Cognitive (18, 10). Mean scores on the SF-20 subscales, Physical Functioning, Role Functioning, Social Functioning, Health Perception, Mental Health, and Pain were numerically lower than means from a published study of elderly outpatients. With the exception of Pain, all SF-20 subscale scores improved significantly by the second survey. In multivariable analyses, dyspnea was predictive of impairment in all SF-20 subscales at the second survey. Conclusions: COVID-19 infection causes persistent abnormality across multiple patient-reported outcome areas, including health status. The persistence of impairment in each health status component is influenced by baseline dyspnea.