Meta-Analysis Identifies 5 Risk Factors for Post-COVID-19 Condition

Patients hospitalized for COVID-19 who were female, experienced severe acute illness requiring advanced respiratory support, were admitted to the intensive care unit (ICU), had incomplete or absent vaccination, or had hypertension faced significantly higher odds of developing post-COVID-19 conditions, according to a systematic review and meta-analysis published in Frontiers in Public Health.¹

The review pooled data from 11 studies conducted in the United Kingdom, Russia, Brazil, Italy, Thailand, France, South Africa, and China, following patients hospitalized with confirmed or probable COVID-19 for periods ranging from approximately 54 days to 2 years after acute infection or discharge. Investigators identified these 5 factors as the most consistent predictors of post-COVID-19 outcomes in a population they describe as clinically important, yet understudied relative to broader, largely nonhospitalized long COVID cohorts.¹

How the Review Was Conducted

The review followed Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 principles, searching PubMed/MEDLINE, Scopus, Web of Science Core Collection, and the World Health Organization (WHO) COVID-19 Global Literature Database for studies published between January 1, 2020, and February 24, 2026. The search identified 573 records from databases and 18 more through reference lists and citation tracking. After 183 duplicates were removed, 408 records were screened, and 51 full-text reports were assessed for eligibility.¹

Eleven studies met criteria for the qualitative synthesis, and 7 of those contributed adjusted odds ratios to the quantitative meta-analysis, yielding 9 pooled study-factor comparisons across 5 risk-factor groups. Eligible studies enrolled hospitalized patients and assessed post-acute outcomes at least 4 weeks after acute infection, hospitalization, recovery, or discharge.¹

Risk of bias, assessed using a Newcastle-Ottawa Scale-based approach, was rated low-moderate for 3 studies, moderate for 5, and moderate-high for 3, with loss to follow-up, self-reported outcome assessment, and broad outcome definitions cited as the main concerns. Pooled estimates were calculated using a random-effects model, and leave-one-out sensitivity analyses showed the direction of each association held after excluding any single study.¹

Five Factors With Consistent Associations

Acute disease severity or respiratory support, including invasive mechanical ventilation or high-intensity respiratory support, showed the strongest pooled association with post-COVID-19 conditions (OR, 2.55; 95% CI, 1.77-3.68; based on 3 studies). Female sex was the most frequently pooled predictor, drawing on 4 studies, with a pooled OR of 1.94 (95% CI, 1.60-2.35).¹

ICU admission carried a pooled OR of 1.88 (95% CI, 1.24-2.83; 2 studies), incomplete or absent vaccination a pooled OR of 1.90 (95% CI, 1.33-2.72; 2 studies), and hypertension a pooled OR of 1.60 (95% CI, 1.16-2.20; 2 studies).¹

Several additional candidate factors, including obesity, age, smoking, dysgeusia, initial symptom load, and neurologic or psychiatric comorbidity, were reported in individual cohorts but could not be pooled because the available effect estimates were incomplete, differently defined, or insufficiently comparable. These were summarized narratively rather than statistically.¹

Defining a Moving Target

Outcome terminology varied considerably across the included studies, a heterogeneity the review's authors called a central methodological challenge. Some studies applied the WHO definition of post-COVID-19 condition, under which the condition is usually assessed 3 months from COVID-19 onset and requires symptoms lasting at least 2 months without an alternative explanation, and others defined the outcome as at least 1 persistent symptom at a fixed follow-up point, long COVID by structured interview, or patient-perceived incomplete recovery.¹

The CDC similarly defines long COVID as a chronic condition present for at least 3 months after SARS-CoV-2 infection, noting it occurs more often after severe illness but can affect anyone who contracts COVID-19, including children.²

The Mayo Clinic cites researcher estimates that 10% to 35% of people who have had COVID-19 went on to develop long COVID, with more than 200 symptoms linked to the condition across body systems. Because no approved laboratory test can confirm a long COVID diagnosis, both organizations note that clinicians must rely on patient history and symptom pattern rather than a single diagnostic marker.^2,3

Implications for Post-Discharge Pharmacy Care

The review's authors say the findings support a risk-based approach to post-discharge follow-up. Patients with severe acute disease, ICU admission, or advanced respiratory support may warrant more structured post-discharge assessment, and vaccination status can help contextualize risk alongside age, comorbidity, and variant period. Female patients and those with hypertension or cardiometabolic comorbidity may also need closer monitoring, particularly if they report persistent fatigue, dyspnea, cognitive symptoms, psychological symptoms, or reduced functional capacity.¹

That risk stratification dovetails with a role pharmacists have already begun to fill. A commentary in the Journal of Pharmacy Practice describes pharmacists as well-positioned to screen discharged patients for emerging long COVID symptoms, optimize medication regimens, and coordinate transitions of care between discharging and receiving pharmacists to help reduce readmissions.⁴

Through a collaborative practice agreement and where state law allows, community and ambulatory care pharmacists can order laboratory tests and initiate treatment for identified symptoms, such as albuterol for persistent dyspnea, and can counsel patients on vaccination, addressing a factor the new meta-analysis ties directly to post-COVID-19 risk.^1,4

The review's authors caution that the pooled estimates rest on a small number of studies per risk factor, meaning the results should be read as risk signals rather than definitive prediction parameters. Outcome-definition heterogeneity, residual confounding by age and comorbidity burden, and a restriction to English- and Russian-language studies further limit how broadly the findings can be applied.¹

The authors call for future studies to use standardized outcome definitions and report adjusted estimates for a common set of predictors, including age, comorbidities, vaccination status, and length of hospital stay, to support development of validated risk-prediction tools for post-discharge care.¹

REFERENCES

1. Kassymbek S, Abduldayeva A, Safonov N, Doszhanova G. Risk factors for post-COVID-19 condition among previously hospitalized COVID-19 patients: a systematic review and meta-analysis. Front Public Health. 2026;14:1911482. Published 2026 Aug 14. doi:10.3389/fpubh.2026.1911482

2. Centers for Disease Control and Prevention. Long COVID basics. CDC. Updated May 6, 2026. Accessed September 1, 2026. https://www.cdc.gov/long-covid/about/index.html

3. Mayo Clinic Staff. Long COVID: lasting effects of COVID-19. Mayo Clinic. Updated August 23, 2024. Accessed September 1, 2026. https://www.mayoclinic.org/diseases-conditions/coronavirus/in-depth/coronavirus-long-term-effects/art-20490351

4. Gosser R, Anderson S, Blaszczyk A, Jen C, Bhatt S. In it for the long haul: post-acute sequelae of severe acute respiratory syndrome coronavirus 2. J Pharm Pract. 2022;0(0):1-4. doi:10.1177/08971900221088799