Impact of Rotavirus Vaccine Introduction and Postintroduction Etiology of Diarrhea Requiring Hospital Admission in Haydom, Tanzania, a Rural African Setting

Abstract

Background: No data are available on the etiology of diarrhea requiring hospitalization after rotavirus vaccine introduction in Africa. The monovalent rotavirus vaccine was introduced in Tanzania on 1 January 2013. We performed a vaccine impact and effectiveness study as well as a quantitative polymerase chain reaction (qPCR)-based etiology study at a rural Tanzanian hospital.

Methods: We obtained data on admissions among children <5 years to Haydom Lutheran Hospital between 1 January 2010 and 31 December 2015 and estimated the impact of vaccine introduction on all-cause diarrhea admissions. We then performed a vaccine effectiveness study using the test-negative design. Finally, we tested diarrheal specimens during 2015 by qPCR for a broad range of enteropathogens and calculated pathogen-specific attributable fractions (AFs).

Results: Vaccine introduction was associated with a 44.9% (95% confidence interval [CI], 17.6%-97.4%) reduction in diarrhea admissions in 2015, as well as delay of the rotavirus season. The effectiveness of 2 doses of vaccine was 74.8% (95% CI, -8.2% to 94.1%) using an enzyme immunoassay-based case definition and 85.1% (95% CI, 26.5%-97.0%) using a qPCR-based case definition. Among 146 children enrolled in 2015, rotavirus remained the leading etiology of diarrhea requiring hospitalization (AF, 25.8% [95% CI, 24.4%-26.7%]), followed by heat-stable enterotoxin-producing Escherichia coli (AF, 18.4% [95% CI, 12.9%-21.9%]), Shigella/enteroinvasive E. coli (AF, 14.5% [95% CI, 10.2%-22.8%]), and Cryptosporidium (AF, 7.9% [95% CI, 6.2%-9.3%]).

Conclusions: Despite the clear impact of vaccine introduction in this setting, rotavirus remained the leading etiology of diarrhea requiring hospitalization. Further efforts to maximize vaccine coverage and improve vaccine performance in these settings are warranted.

Figure 1.

Monthly diarrhea admissions to Haydom Lutheran Hospital, January 2010–December 2015. Rotavirus vaccine was added to the national immunization program on 1 January 2013. The vaccine effectiveness study was performed from August 2014 to December 2015. Diarrhea etiology was assessed by quantitative polymerase chain reaction testing of all isolates from 2015. Inset (top right) shows the proportion of enrolled subjects who were positive for rotavirus by enzyme immunoassay (EIA) for each month.

Figure 2.

Monthly incidence of hospital admissions for diarrhea during 2010–2012, prior to vaccine introduction, and 2013–2015, after vaccine introduction. Incidence (solid lines) and 95% confidence intervals (dotted lines) were derived from Poisson regression models.

Figure 3.

Flowchart for vaccine effectiveness and diarrhea etiology study. Rotavirus diarrhea was defined by for the primary analysis as detection of rotavirus by enzyme immunoassay and confirmed by quantitative polymerase chain reaction (qPCR).

Figure 4.

Etiology of diarrhea requiring hospitalization in 2015 by quantitative polymerase chain reaction (qPCR) (n = 146). Attributable fraction estimates were derived from qPCR detection of each pathogen combined with odds ratios derived from the Global Enteric Multicenter Study (GEMS). Abbreviations: C. coli, Campylobacter coli; C. jejuni, Campylobacter jejuni; E. histolytica, Entamoeba histolytica; EIEC, enteroinvasive Escherichia coli; H. pylori, Helicobacter pylori; ST-ETEC, heat-stable enterotoxin-producing Escherichia coli; tEPEC, typical enteropathogenic Escherichia coli; V. cholerae, Vibrio cholerae.

Figure 5.

Heat-stable enterotoxin-producing Escherichia coli colonization factor profile for all highly-diarrhea associated detections in 2015 by quantitative polymerase chain reaction. Abbreviations: CFA, colonization factor antigen; CS, coli surface antigen.