Praziquantel and Upper Gastrointestinal Bleeding in Hepatic Schistosomiasis: A Quasi-Experimental Study.

BACKGROUND

There is a general consensus that widespread use of praziquantel in populations where schistosomiasis is endemic prevents development of hepatic schistosomiasis and its complications. However, a few studies have reported discordant findings linking praziquantel to the occurrence of upper gastrointestinal bleeding (UGIB) in some patients with hepatic schistosomiasis and varices.

OBJECTIVE

We explored if there was any causal association between recent praziquantel use (rPZQ) and upper gastrointestinal bleeding in hepatic schistosomiasis in rural Africa.

PATIENTS AND METHODS

A quasi-experimental, retrospective case-controlled study was performed. It involved adult patients with past or acute UGIB, varices, periportal fibrosis, and/or cirrhosis. Cases had acute variceal bleeding while controls did not. The outcome was the frequency of lifetime episodes of UGIB and exposure was rPZQ (received praziquantel in the last 11 months from the date of enrollment). The data analysis included 2 × 2 tables, logistic regression, and propensity-score matching. Odds ratios (ORs), average treatment effects (ATEs), and their 95% confidence intervals (CIs) were used for inference.

RESULTS

Over 6 weeks, we enrolled 19 cases with 92 lifetime episodes of UGIB, and 66 controls with 192 lifetime episodes of UGIB. Cases were more likely to experience UGIB than controls following rPZQ (92% vs. 62%; OR 7.6; 95% CI 3.4-17). Factors predictive of more lifetime episodes of UGIB at multivariable analysis included rPZQ (adjusted OR 13; 95% CI 2.9-53), relative leukocytosis (adjusted OR 26; 95% CI 7.6-89), large varices (adjusted OR 5.0; 95% CI 1.7-15), a family member with hepatosplenic schistosomiasis (adjusted OR 19; 95% CI 7.4-51), advanced periportal fibrosis (adjusted OR 8.0; 95% CI 2.6-22), ascites (adjusted OR 14; 95% CI 4.3-47), and jaundice (adjusted OR 32; 95% CI 7.8-128). While the ATE following rPZQ among the treated was 0.40 (95% CI 0.33-0.48).

CONCLUSIONS

Our findings suggest the presence of a plausible causal association between recent praziquantel use and increased frequency of UGIB in our study population.