Urogenital schistosomiasis elimination in Zanzibar: accuracy of urine filtration and haematuria reagent strips for diagnosing light

23 October 2018
Stefanie Knopp, Shaali M Ame, Jan Hattendorf, Said M Ali, Iddi S Khamis, Faki Bakar, Mwanaidi A Khamis, Bobbie Person, Fatma Kabole and David Rollinson

BACKGROUND:

Urine filtration and microhaematuria reagent strips are basic standard diagnostic methods to detect urogenital schistosomiasis. We assessed their accuracy for the diagnosis of light intensity infections with Schistosoma haematobium as they occur in individuals living in Zanzibar, an area targeted for interruption of transmission.

METHODS:

Urine samples were collected from children and adults in surveys conducted annually in Zanzibar from 2013 through 2016 and examined with the urine filtration method to count S. haematobium eggs and with the reagent strip test (Hemastix) to detect microhaematuria as a proxy for infection. Ten percent of the urine filtration slides were read twice. Sensitivity was calculated for reagent strips, stratified by egg counts reflecting light intensity sub-groups, and kappa statistics for the agreement of urine filtration readings.

RESULTS:

Among the 39,207 and 18,155 urine samples examined from children and adults, respectively, 5.4% and 2.7% were S. haematobium egg-positive. A third (34.7%) and almost half (46.7%) of the egg-positive samples from children and adults, respectively, had ultra-low counts defined as 1-5 eggs per 10 ml urine. Sensitivity of the reagent strips increased significantly for each unit log10 egg count per 10 ml urine in children (odds ratio, OR: 4.7; 95% confidence interval, CI: 4.0-5.7; P < 0.0001) and adults (OR: 2.6; 95% CI: 1.9-3.7, P < 0.0001). Sensitivity for diagnosing ultra-light intensity infections was very low in children (50.1%; 95% CI: 46.5-53.8%) and adults (58.7%; 95% CI: 51.9-65.2%). Among the 4477 and 1566 urine filtration slides read twice from children and adults, most were correctly identified as negative or positive (kappa = 0.84 for children and kappa = 0.81 for adults). However, 294 and 75 slides had discrepant results and were positive in only one of the two readings. The majority of these discrepant slides (76.9% of children and 84.0% of adults) had counts of 1-5 eggs per 10 ml urine.

CONCLUSIONS:

We found that many individuals infected with S. haematobium in Zanzibar excrete > 5 eggs per 10 ml urine. These ultra-light infections impose a major challenge for accurate diagnosis. Next-generation diagnostic tools to be used in settings where interruption of transmission is the goal should reliably detect infections with ≤ 5 eggs per 10 ml urine.