A comparison of diarrheal severity scores in the MAL-ED multisite community-based cohort study

Gwenyth O. Lee, Stephanie A. Richard, Gagandeep Kang, Eric R. Houpt, Jessica C. Seidman, Laura L. Pendergast, Zulfiqar A. Bhutta, Tahmeed Ahmed, Estomih R. Mduma, Aldo A. Lima, Pascal Bessong, Mats Steffi Jennifer, Md Iqbal Hossain, Ram Krishna Chandyo, Emanuel Nyathi, Ila F. Lima, John Pascal, Sajid Soofi, Bodhidatta Ladaporn, Richard L. GuerrantLaura E. Caulfield, Robert E. Black, Margaret N. Kosek*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

27 Scopus citations


Objectives: There is a lack of consensus on how to measure diarrheal severity. Within the context of a multisite, prospective cohort study, we evaluated the performance of a modified Vesikari score (MAL-ED), 2 previously published scores (Clark and CODA [a diarrheal severity score (Community DiarrheA) published by Lee et al]), and a modified definition of moderate-to-severe diarrhea (MSD) based on dysentery and health care worker diagnosed dehydration. Methods: Scores were built using maternally reported symptoms or fieldworker-reported clinical signs obtained during the first 7 days of a diarrheal episode. The association between these and the risk of hospitalization were tested using receiver operating characteristic analysis. Severity scores were also related to illness etiology, and the likelihood of the episode subsequently becoming prolonged or persistent. Results: Of 10,159 episodes from 1681 children, 143 (4.0%) resulted in hospitalization. The area under the curve of each score as a predictor of hospitalization was 0.84 (95% confidence interval: 0.81, 0.87) (Clark), 0.85 (0.82, 0.88) (MAL-ED), and 0.87 (0.84, 0.89) (CODA). Severity was also associated with etiology and episode duration. Although families were more likely to seek care for severe diarrhea, approximately half of severe cases never reached the health system. Conclusions: Community-based diarrheal severity scores are predictive of relevant child health outcomes. Because they require no assumptions about health care access or utilization, they are useful in refining estimates of the burden of diarrheal disease, in estimating the effect of disease control interventions, and in triaging children for referral in low- and middle-income countries in which the rates of morbidity and mortality after diarrhea remain high.

Original languageEnglish
Pages (from-to)466-473
Number of pages8
JournalJournal of Pediatric Gastroenterology and Nutrition
Issue number5
StatePublished - 2016
Externally publishedYes


  • Diarrhea
  • Epidemiology
  • Pediatric


Dive into the research topics of 'A comparison of diarrheal severity scores in the MAL-ED multisite community-based cohort study'. Together they form a unique fingerprint.

Cite this