COVID-19 vaccination, clinical presentation and 14-day outcomes in Cumilla, Bangladesh: a cross-sectional study

Background Community evidence from Bangladesh on how vaccination, clinical presentation and short-term outcomes intersected during the early COVID-19 vaccine rollout is limited. We assessed test positivity, clinical characteristics, vaccination status and 14-day outcomes among people tested for suspected COVID-19 in six upazilas of Cumilla district. Methods Cross-sectional study of 320 adults tested between 1 April and 31 May 2021, with a retrospective day-by-day 14-day symptom history and outcome assessment for the 86 test-positive cases. Test results (reverse transcription polymerase chain reaction [RT-PCR], n = 286; rapid antigen, n = 34) came from laboratory records; other data were collected by structured telephone interview. We estimated odds ratios, fitted a multivariable logistic regression for positivity, and described symptom trajectories and outcomes. Results Eighty-six participants tested positive (26.9%; 95% CI 22.3–32.0); cases had a mean age of 42.5 years, 54.7% were men and 90.7% were symptomatic. In the adjusted model (n = 297), symptomatic presentation (adjusted odds ratio [aOR] 8.18; 95% CI 3.64–18.41) and recent contact with a confirmed case (aOR 2.73; 1.50–4.97) were associated with positivity; age, sex, comorbidity and smoking were not. At least one vaccine dose was reported by 38.4% of cases and 51.7% of test-negative participants (aOR 0.55; 0.31–0.97), but this estimate was imprecise, attenuated in sensitivity analyses (RT-PCR-only aOR 0.75; 0.40–1.38) and does not represent vaccine effectiveness. At 14 days, 56 cases (65.1%) remained in home isolation, 15 (17.4%) were hospitalized without oxygen, 8 (9.3%) needed oxygen, 5 (5.8%) intensive care and 2 (2.3%; 95% CI 0.6–8.1) died. Conclusions Symptoms and recent exposure were the strongest correlates of a positive test, but symptoms alone had poor specificity, so testing remained essential. About one in three cases was hospitalized or died within 14 days. The observational design, mixed testing methods and recalled symptom histories preclude causal inference about vaccination.