Algorithm for diagnosis of early Schistosoma haematobium using prodromal signs and symptoms in pre-school age children in an endemic district in Zimbabwe.

Abstract

IntroductionPrompt diagnosis of acute schistosomiasis benefits the individual and provides opportunities for early public health intervention. In endemic areas schistosomiasis is usually contracted during the first 5 years of life, thus it is critical to look at how the infection manifests in this age group. The aim of this study was to describe the prodromal signs and symptoms of early schistosomiasis infection, correlate these with early disease progression and risk score to develop an easy to use clinical algorithm to identify early Schistosoma haematobium infection cases in resource limited settings.MethodologyTwo hundred and four, preschool age children who were lifelong residence of a schistosomiasis endemic district and at high risk of acquiring schistosomiasis were followed up from July 2019 to December 2019, during high transmission season. The children received interval and standard full clinical evaluations and laboratory investigations for schistosomiasis by clinicians blinded from their schistosomiasis infection status. Diagnosis of S. haematobium was by urine filtration collected over three consecutive days. Signs and symptoms of schistosomiasis at first examination visit were compared to follow-up visits. Signs and symptoms common on the last schistosomiasis negative visit (before a subsequent positive) were assigned as early schistosomiasis infection (ESI), after possible alternative causes were ruled out. Logistic regression identified clinical predictors. A model based score was assigned to each predictor to create a risk for every child. An algorithm was created based on the predictor risk scores and validated on a separate cohort of 537 preschool age children.ResultsTwenty-one percent (42) of the participants were negative for S. haematobium infection at baseline but turned positive at follow-up. The ESI participants at the preceding S. haematobium negative visit had the following prodromal signs and symptoms in comparison to non-ESI participants; pruritic rash adjusted odds ratio (AOR) = 21.52 (95% CI 6.38-72.66), fever AOR = 82 (95% CI 10.98-612), abdominal pain AOR = 2.6 (95% CI 1.25-5.43), pallor AOR = 4 (95% CI 1.44-11.12) and a history of facial/body swelling within the previous month AOR = 7.31 (95% CI 3.49-15.33). Furthermore 16% of the ESI group had mild normocytic anaemia, whilst 2% had moderate normocytic anaemia. A risk score model was created using a rounded integer from the relative risks ratios. The diagnostic algorithm created had a sensitivity of 81% and a specificity of 96.9%, Positive predictive value = 87.2% and NPV was 95.2%. The area under the curve for the algorithm was 0.93 (0.90-0.97) in comparison with the urine dipstick AUC = 0.58 (0.48-0.69). There was a similar appearance in the validation cohort as in the derivative cohort.ConclusionThis study demonstrates for the first time prodromal signs and symptoms associated with early S. haematobium infection in pre-school age children. These prodromal signs and symptoms pave way for early intervention and management, thus decreasing the harm of late diagnosis. Our algorithm has the potential to assist in risk-stratifying pre-school age children for early S. haematobium infection. Independent validation of the algorithm on another cohort is needed to assess the utility further.