Dear Editor,
As the pandemic advanced, interest grew regarding the long‐term course of Coronavirus Disease 2019 (COVID‐19) [1, 2]. Previous work described many symptoms that some people experience for months after the acute infection—varying from 13% in population‐based studies to 72% in samples of hospitalised patients [3, 4]. It led the World Health Organization (WHO) to define the post‐COVID condition—also known as ‘long COVID’—characterised by persisting symptoms for at least 2 months, usually occurring within 3 months of the SARS‐CoV‐2 infection [5].
While the burden of long COVID is alarming worldwide [6], we have scarce data from populations in lower‐and‐middle‐income countries (LMICs) [7, 8]. Therefore, we estimated the prevalence of COVID‐19 symptoms upon admission and 1 year after discharge using data from regular interviews with patients who survived hospitalisation for COVID‐19 in Brazil. We also examined the prevalence and predictors of long COVID.
This cohort study is part of the CO‐FRAIL (COVID‐19 and Frailty) Study conducted at Hospital das Clinicas, the largest academic medical centre in Latin America [9]. This hospital served as a major medical centre for COVID‐19 in Brazil, receiving patients from a diverse socioeconomic area in Sao Paulo, encompassing 85 cities (Fig. S1). We included consecutive patients aged ≥50 years who survived hospitalisation for confirmed COVID‐19 between March and December 2020, as this older population is at higher risk for the severe consequences of the disease [9]. We excluded admissions lasting <24 h and patients with baseline missing data or lost to follow‐up (Fig. S2).
Trained physicians collected baseline data (sociodemographic factors, COVID‐19 symptoms, clinical measures, laboratory findings and in‐hospital treatments and adverse events) in parallel with the admissions using detailed electronic hospital record reviews complemented with telephone interviews with close family members whenever needed. Of note, standardised medical forms specially designed for the pandemic contained information on 14 COVID‐19 symptoms reported upon admission. Additionally, we conducted follow‐up interviews 1, 3, 6, 9 and 12 months after discharge using structured telephone assessments with patients or close family members (when patients were unable to respond) to achieve information on 15 COVID‐19 symptoms (Table S1 details the list of COVID‐19 symptoms) [10]. We defined long COVID as present when participants reported at least one COVID‐related symptom within 3 months of hospital discharge and lasting at least 2 months [5].
Our sample comprised 1042 patients with a median age of 63 years and 45% females (Table 1). Figure S3 shows the prevalence of each COVID‐19 symptom on admission and during follow‐up interviews. The most prevalent symptoms after discharge were fatigue, myalgia, dyspnea, coughing and paresthesia. We observed a decline in the median number of COVID‐19 symptoms over the follow‐up, particularly after 9 months (Fig. S4). Nonetheless, 354 (34%) patients had at least one remaining COVID‐19 symptom at the end of 1 year.
Table 1.
Characteristics of patients who survived hospitalisation for COVID‐19 according to long COVID
| Long COVID | |||||
|---|---|---|---|---|---|
| Total | No | Yes | |||
| Characteristics, median (IQR) or N (%) | (N = 1042) | (N = 454) | (N = 588) | p‐Value a | OR (95% CI) b |
| Sociodemographic measures | |||||
| Age (years) | 63 (57, 72) | 64 (57, 73) | 63 (57, 70) | 0.06 | 0.99 (0.98–1.01) |
| Female sex | 464 (45) | 187 (41) | 277 (47) | 0.06 | 1.49 (1.10–2.04) |
| Race/ethnicity | 0.77 | ||||
| White | 653 (63) | 280 (62) | 373 (63) | (Reference) | |
| Black | 369 (35) | 166 (37) | 203 (35) | 0.86 (0.64–1.15) | |
| Other | 20 (2) | 8 (2) | 12 (2) | 0.97 (0.71–1.79) | |
| Married | 637 (61) | 295 (65) | 342 (58) | 0.03 | 0.66 (0.49–0.88) |
| Literacy (years) | 5 (4, 10) | 5 (4, 8) | 5 (4, 10) | 0.16 | 1.02 (0.99–1.06) |
| Pre‐COVID‐19 characteristics | |||||
| Diabetes mellitus | 457 (44) | 186 (41) | 271 (46) | 0.10 | 1.04 (0.78–1.40) |
| Hypertension | 763 (73) | 317 (70) | 446 (76) | 0.03 | 1.35 (0.96–1.90) |
| Cerebrovascular disease | 94 (9) | 50 (11) | 44 (7) | 0.05 | 0.67 (0.41–1.09) |
| Cardiovascular disease | 284 (27) | 121 (27) | 163 (28) | 0.70 | 1.09 (0.77–1.53) |
| Pulmonary disease | 123 (12) | 45 (10) | 78 (13) | 0.10 | 1.11 (0.72–1.73) |
| Chronic kidney disease | 171 (16) | 81 (18) | 90 (15) | 0.27 | 0.89 (0.60–1.34) |
| Liver disease | 35 (3) | 20 (4) | 15 (3) | 0.10 | 0.63 (0.29–1.35) |
| Cancer | 141 (14) | 80 (18) | 61 (10) | <0.001 | 0.73 (0.48–1.12) |
| Smoker | 0.42 | ||||
| Never | 633 (61) | 286 (63) | 347 (59) | (Reference) | |
| Current | 57 (5) | 24 (5) | 33 (6) | 1.42 (0.75–2.67) | |
| Previous | 352 (34) | 144 (32) | 208 (35) | 1.32 (0.97–1.79) | |
| Dementia or CIND | 133 (13) | 64 (14) | 69 (12) | 0.26 | 1.03 (0.66–1.60) |
| Frailty (CFS ≥5) | 191 (18) | 76 (17) | 115 (20) | 0.24 | 0.99 (0.61–1.60) |
| Body mass index (kg/m2) | 27.7 (24.2, 32) | 26.4 (23.4, 30.5) | 28.3 (25.1, 33.2) | <0.001 | 1.03 (1.01–1.05) |
| Admission characteristics | |||||
| COVID‐19 symptoms (days) | 8 (5, 11) | 7 (5, 10) | 8 (5, 11) | 0.04 | 1.01 (0.98–1.05) |
| SOFA score | 5 (3, 7) | 5 (2, 6) | 6 (4, 8) | <0.001 | – |
| In‐hospital characteristics | |||||
| Lowest haemoglobin (mg/dl) | 10.6 (8.8, 12.2) | 10.8 (9.2, 12.3) | 10.5 (8.5, 12) | 0.02 | 1.03 (0.95–1.12) |
| Highest neutrophil/lymphocyte ratio | 4.7 (3.1, 6.7) | 4.6 (3.0, 6.6) | 4.8 (3.2, 6.7) | 0.29 | 0.98 (0.93–1.02) |
| Highest C‐reactive protein (mg/dl) | 160 (81, 258) | 137 (74, 215) | 187 (88, 285) | <0.001 | 1.00 (0.99–1.01) |
| Intensive care admission | 540 (52) | 180 (40) | 360 (61) | <0.001 | 1.68 (1.16–2.44) |
| Invasive ventilation | 318 (31) | 97 (21) | 221 (38) | <0.001 | 0.97 (0.57–1.65) |
| Dialysis | 95 (9) | 32 (7) | 63 (11) | 0.04 | 1.04 (0.58–1.85) |
| Corticosteroids | 597 (57) | 223 (49) | 374 (64) | <0.001 | 1.25 (0.93–1.68) |
| Immunobiologicals | 8 (1) | 3 (1) | 5 (1) | 0.73 | 0.96 (0.20–4.58) |
| Anticoagulation | 221 (21) | 83 (18) | 138 (23) | 0.04 | 0.83 (0.54–1.28) |
| Septic shock | 163 (16) | 49 (11) | 114 (19) | <0.001 | 0.82 (0.49–1.38) |
| Venous/pulmonary thromboembolism | 116 (11) | 33 (7) | 83 (14) | <0.001 | 1.72 (0.94–3.13) |
| Delirium | 353 (34) | 141 (31) | 212 (36) | 0.09 | 1.02 (0.71–1.46) |
| Length of stay (days) | 13 (8, 23) | 11 (7, 18) | 15 (9, 27) | <0.001 | 1.02 (1.01–1.03) |
Abbreviations: CFS, Clinical Frailty Scale; CI, confidence interval; CIND, cognitive impairment not dementia; IQR, interquartile range; OR, odds ratio; SOFA, Sequential Organ Failure Assessment.
Comparisons between patients with versus without long COVID using the Mann–Whitney test for interval variables and the chi‐squared test for categorical variables. We ran complete case analyses after excluding 47 potential cases with baseline missing data and 14 lost to follow‐up, as described in Fig. S2 (flowchart of the study sample).
Estimates were computed using a multivariable logistic model containing all the characteristics as predictors and long COVID as the outcome; we excluded the SOFA score from the model to avoid overfitting.
Overall, 588 (56%) patients presented long COVID. This condition was associated with hypertension, higher body mass index, greater disease severity, lower haemoglobin and in‐hospital complications (Table 1). In multivariable analyses, female sex, higher body mass index, admission to intensive care unit and longer length of stay were independent predictors of long COVID (Table 1).
We demonstrated that long‐lasting complaints of COVID‐19 were common, with one in three patients experiencing at least one symptom at the end of a 1‐year follow‐up. Fatigue followed by myalgia, dyspnea, paresthesia and cough were the most common symptoms after discharge. We also found that over half of patients evolved with long COVID, particularly those with complicated hospitalisations. Our study corroborates recent research findings and advances knowledge on the long course of COVID by reporting a detailed symptoms trajectory among patients living in an LMIC [1, 2, 3].
Our study had limitations. First, the findings were built on a university hospital, impacting their generalisability, particularly for community settings. Second, our data from the first waves could not investigate the effects of newer SARS‐CoV‐2 variants and evidence‐based therapies on long COVID. Third, although WHO declares long COVID usually occurs within 3 months of disease onset, symptoms may occur even after our definition of within 3 months after discharge [5]. As we did not have a control group, our ability to distinguish the specific long‐term symptoms of COVID‐19 from those that might result from hospitalisation or other causes was limited, particularly for symptoms that appeared long after discharge. Our study also has remarkable strengths. We conducted one of the largest cohorts of COVID‐19 survivors living in LMICs. Trained physicians obtained the main data prospectively using structured assessments. We assessed data from early in the hospitalisations to a long time after discharge, characterising better the long course of COVID‐19 compared to previous work [7, 8].
In sum, we demonstrated that recovery from COVID‐19 can extend far beyond the resolution of the acute infection, demanding comprehensive post‐acute care, especially rehabilitation.
Funding
This work was supported by Hospital das Clinicas HCFMUSP, Faculdade de Medicina, Universidade de Sao Paulo, Brazil, from donations to the #HCComvida campaign. Donations were directed to the institution's initiatives related to fighting COVID‐19, including research projects. Márlon J. R. Aliberti is supported by a post‐doc scholarship from HCFMUSP with funds donated by Nubank under the #HCCOMVIDA scheme. The funders had no role in the study's conceptualisation, methodology, data curation, analysis, writing or supervision.
Conflict of interest
All the authors declare no conflict of interest.
Author contributions
Murilo B. Dias: Conceptualisation; Data curation; Methodology; Writing – original draft; Writing – review and editing. Ana Paula V. Medeiros: Conceptualisation; Data curation; Methodology; Writing – original draft; Writing – review and editing. Sarah S. de Melo: Conceptualisation; Writing – review and editing. Cecilia S. Fonseca: Conceptualisation; Writing – review and editing. Wilson Jacob‐Filho: Supervision; Conceptualisation; Writing – review and editing. Thiago J. Avelino‐Silva: Conceptualisation; Methodology; Formal Analysis; Data curation; Writing – original draft; Writing – review and editing. Márlon J. R. Aliberti: Supervision; Conceptualisation; Methodology; Data curation; Writing – original draft; Writing – review and editing.
Ethics statement
The Committee for Ethics in Research of the University of Sao Paulo Medical School approved the study (CAAE: 32037120.6.0000.006) and authorised investigators to secure verbal consent in the study's follow‐up interviews. The CO‐FRAIL Study is registered in the Brazilian Clinical Trials Registry, accessible at http://www.ensaiosclinicos.gov.br/rg/RBR‐7w5zhr/.
Supporting information
Supplemental Figure 1: Distribution of the study population in a diverse socioeconomic area in Sao Paulo, Brazil, according to postal code and the Human Development Index districts.
Supplemental Figure 2: Flowchart of the study sample.
Supplemental Figure 3: Bubble chart indicating the prevalence of COVID‐19 symptoms in survivors of hospitalization during one year after discharge.
Supplemental Figure 4: Number of COVID‐19 symptoms over a 1‐year follow‐up among survivors of hospitalization.
Supplemental Table 1: Assessment form of COVID‐19 symptoms.
Dias MB, Medeiros APV, de Melo SS, Fonseca CS, Jacob‐Filho W, Avelino‐Silva TJ, et al. The long and winding road of COVID‐19 in survivors of hospitalisation: Symptoms trajectory and predictors of long COVID. J Intern Med. 2023;293:264–268.
Murilo B. Dias and Ana Paula V. Medeiros contributed equally to this work.
Contributor Information
Murilo B. Dias, Email: murilodias96@gmail.com.
Márlon J. R. Aliberti, Email: maliberti@usp.br.
Data availability statement
The datasets used and analysed during the current study are available from the corresponding author upon reasonable request.
References
- 1. Nehme M, Braillard O, Chappuis F, Courvoisier DS, Kaiser L, Soccal PM, et al. One‐year persistent symptoms and functional impairment in SARS‐CoV‐2 positive and negative individuals. J Intern Med. 2022;292(1):103–15. 10.1111/joim.13482 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Huang L, Li X, Gu X, Zhang H, Ren L, Guo L, et al. Health outcomes in people 2 years after surviving hospitalisation with COVID‐19: a longitudinal cohort study. Lancet Respir Med. 2022. 10.1016/S2213-2600(22)00126-6 [Epub ahead of print]. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Ballering AV, van Zon SKR, Olde Hartman TC, Rosmalen JGM; Lifelines Corona Research Initiative . Persistence of somatic symptoms after COVID‐19 in the Netherlands: an observational cohort study. Lancet 2022;400(10350):452–61. 10.1016/S0140-6736(22)01214-4 [Epub ahead of print]. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Chen C, Haupert SR, Zimmermann L, Shi X, Fritsche LG, Mukherjee B. Global prevalence of post COVID‐19 condition or long COVID: a meta‐analysis and systematic review. J Infect Dis. 2022:jiac136. 10.1093/infdis/jiac136 [Epub ahead of print]. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Soriano JB, Murthy S, Marshall JC, Relan P, Diaz JV; WHO Clinical Case Definition Working Group on Post‐COVID‐19 Condition . A clinical case definition of post‐COVID‐19 condition by a Delphi consensus. Lancet Infect Dis. 2022;22(4):e102–7. 10.1016/S1473-3099(21)00703-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Mannucci PM, Nobili A, Tettamanti M, D'Avanzo B, Galbussera AA, Remuzzi G, et al. Impact of the post‐COVID‐19 condition on health care after the first disease wave in Lombardy. J Intern Med. 2022;292(3):450–62. 10.1111/joim.13493 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Sudre CH, Murray B, Varsavsky T, Graham MS, Penfold RS, Bowyer RC, et al. Attributes and predictors of long COVID. Nat Med. 2021;27(4):626–31. 10.1038/s41591-021-01292-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Carfì A, Bernabei R, Landi F; Gemelli Against COVID‐19 Post‐Acute Care Study Group . Persistent symptoms in patients after acute COVID‐19. JAMA. 2020;324(6):603–5. 10.1001/jama.2020.12603 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Aliberti MJR, Szlejf C, Avelino‐Silva VI, Suemoto CK, Apolinario D, Dias MB, et al. COVID‐19 is not over and age is not enough: using frailty for prognostication in hospitalised patients. J Am Geriatr Soc. 2021;69(5):1116–27. 10.1111/jgs.17146 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Poco PCE, Aliberti MJR, Dias MB, de Fatima Takahashi S, Leonel FC, Altona M, et al. Divergent: age, frailty, and atypical presentations of COVID‐19 in hospitalized patients. J Gerontol A Biol Sci Med Sci. 2021;76(3):e46–51. 10.1093/gerona/glaa280 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental Figure 1: Distribution of the study population in a diverse socioeconomic area in Sao Paulo, Brazil, according to postal code and the Human Development Index districts.
Supplemental Figure 2: Flowchart of the study sample.
Supplemental Figure 3: Bubble chart indicating the prevalence of COVID‐19 symptoms in survivors of hospitalization during one year after discharge.
Supplemental Figure 4: Number of COVID‐19 symptoms over a 1‐year follow‐up among survivors of hospitalization.
Supplemental Table 1: Assessment form of COVID‐19 symptoms.
Data Availability Statement
The datasets used and analysed during the current study are available from the corresponding author upon reasonable request.