Is rotavirus aetiology in young children with acute diarrhoea associated with sociodemographic and clinical factors, including rotavirus vaccination status? A secondary cross-sectional analysis of the ABCD trial

Abstract

Introduction: One of the leading causes of global child mortality continues to be diarrhoea where rotavirus contributed to about 24% of deaths among all diarrhoeal deaths, mostly in low-income and middle-income countries. Rotavirus vaccination programmes have contributed to the reduction of mortality from 24% to 19% in rotavirus infections among hospitalised children, but the burden of rotaviral diarrhoea remains high, especially in settings with undernutrition. We aimed to determine the association of rotaviral diarrhoea aetiology with prior vaccination, socioeconomic status and clinical factors in children to see their utility in clinical settings.

Methods: We analysed secondary data from a multicentre clinical trial on antibiotic impact in children with diarrhoea and increased risk of mortality. We used stored stool samples of 6697 children aged 2-23 months old, presenting to a health facility with diarrhoea and increased risk of mortality. We determined rotavirus aetiology prevalence using quantitative PCR (qPCR) and looked at its association with the patient's rotaviral vaccination status, clinical symptoms and sociodemographic characteristics. Prevalence ratios (PR) were calculated with log-binomial regression models; if they did not converge, log-Poisson models were used.

Results: Rotavirus prevalence of 21.1% was observed. There was a weak and statistically non-significant inverse association between rotavirus vaccination and rotaviral diarrhoea aetiology (adjusted PR: 0.71, 95% CI 0.49 to 1.03). Of the five tested clinical symptoms, shorter diarrhoea duration was associated with rotaviral aetiology (PR: 2.65; 95% CI: 1.29 to 5.45). Of the seven tested socioeconomic characteristics, only maternal and paternal secondary education compared with no formal education were associated with rotaviral aetiology (PR: 0.86; 95% CI: 0.74 to 1.00, PR: 0.87, 95% CI: 0.75 to 1.00 respectively).

Conclusion: Rotaviral diarrhoea aetiology cannot accurately be determined with prior receipt of rotavirus vaccination among children presenting to facilities with diarrhoea and increased risk of mortality. Short diarrhoea duration and parental secondary education were associated with increased prevalence of rotaviral aetiology; however, their utility in clinical care remains unclear.

Keywords: Child health; Cross-sectional survey; Global Health; Vaccines.

Figure 1. Flow chart of 2–23-month-old children presenting with moderate to severe diarrhoea. ‘Likely’ diarrhoea-associated aetiology was determined by Ct value cut-offs for specific enteropathogens. Ct values greater than these cut-offs but <35 were considered ‘possible’ aetiologies, while Ct value >35 was considered ‘Unlikely’ aetiology. ‘Likely’ rotaviral diarrhoea aetiology was determined by Ct value cut-off of 32.0. Ct, cycle threshold.