Predictors of Long-COVID-19 and its Impact on Quality of Life: Longitudinal Analysis at 3, 6 and 9 Months After Discharge from a Portuguese Centre

Abstract

Introduction: Long-COVID-19 impacts health-related quality of life (HR-QoL) but data is scarce. The aim of this study was to describe and prospectively assess the prevalence and risk factors for long-COVID-19 after hospital discharge, and to evaluate its impact on patient HR-QoL. Methods: Single-centre longitudinal study including all COVID-19 patients discharged between December 2020 and February 2021. Patients were contacted remotely at three, six and nine months. Data were collected as follows: 1) Long-COVID-19 symptoms were self-reported; 2) HR-QoL were assessed using the 3-level EuroQoL-5D (EQ-5D-3L) questionnaire. Pregnant women, demented, bedridden, and non-Portuguese-speaking patients were excluded. Results: The three-, six- and nine-month assessments were completed by 152, 117 and 110 patients (median age: 61 years; male sex: 56.6%). Long- COVID-19 (= 1 symptom) was reported by 66.5%, 62.4% and 53.6% of patients and HR-QoL assessment showed impairment of at least some domain in 65.8%, 69.2% and 55.4% of patients at three, six and nine months, respectively. Fatigue was the most common long-COVID-19 symptom. Anxiety/ depression domain was the most frequently affected in all three time-points, peaking at six months (39%), followed by pain/discomfort and mobility domains. Long-COVID-19 was associated with the impairment of all EQ-5D-3L domains except for self-care domain at each time-point. Neither intensive care unit admission nor disease severity were associated with long-COVID-19 nor with impairment of any EQ-5D-3L domain. After adjusting for sex, age, frailty status, and comorbid conditions, long-COVID-19 remained significantly associated with HR-QoL impairment at three (OR 4.27, 95% CI 1.92 - 9.52, p < 0.001), six (OR 3.46, 95% CI 1.40 - 8.57, p = 0.007) and nine months (OR 4.13, 95% CI 1.62 - 10.55, p = 0.003) after hospital discharge. In a longitudinal analysis, patients reporting long-COVID-19 at three months had an EQ-5D-3L index value decreased by 0.14 per visit (p < 0.001) compared to those without long-COVID-19 and both groups had a non-significant change in mean EQ-5D-3L index over the nine-month period (time-point assessment, Z = 0.91, p = 0.364). Conclusion: Clinical sequelae associated with long-COVID-19 can persist for at least nine months after hospital discharge in most patients and can impair long-term HR-QoL in more than half of patients regardless of disease severity, and clinicodemographic characteristics.