Banca de DEFESA: LANA CRISTINA EVANGELISTA FERREIRA SA

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : LANA CRISTINA EVANGELISTA FERREIRA SA
DATE: 16/02/2023
TIME: 14:00
LOCAL: NÚCLEO DE MEDICINA TROPICAL
TITLE:

Disentangling the effects of irregular parasite shedding and imperfect test sensitivity on the microscopy-based detection of Giardia duodenalis in stool samples

KEY WORDS:

Intestinal protozooses, stool parasitological examination, diagnosis, irregular parasite shedding, hierarchical modeling.

PAGES: 110
BIG AREA: Ciências da Saúde
AREA: Medicina
SUBÁREA: Clínica Médica
SPECIALTY: Doenças Infecciosas e Parasitárias
SUMMARY:

The parasitic protozoan Giardia duodenalis (‘Giardia’ hereafter) is a leading global cause of diarrhea. As with many other intestinal pathogens, Giardia diagnosis typically involves detecting parasite presence in stool samples. However, not all stool samples drawn from infected individuals contain parasites (due to irregular parasite shedding), and no test can detect the target parasites in 100% of the samples that indeed contain those parasites (due to imperfect test sensitivity). Disentangling the effects of irregular shedding and imperfect sensitivity on parasite detection would help us better understand both transmission dynamics (which heavily depend on parasite-shedding probabilities) and diagnostic-test performance (of which sensitivity is a critical component). Here, we illustrate an approach to separately estimating, under the assumption of no false-positives, the probabilities of Giardia shedding in the stool of infected hosts (denoted θ) and of Giardia detection in Giardia-positive stool samples (true test sensitivity, denoted p). With this information, we then derive corrected estimates of host-infection frequency (Ψ). We collected 1–3 stool samples, in consecutive weeks, from 276 children (4–73 months) attending 8 kindergartens in 4 urban sectors of the Federal District, Brazil. Samples were processed by the spontaneous-sedimentation method and examined via optical microscopy by two independent observers – an expert parasitologist and a graduate student. Using these replicate test results and multi-level hierarchical models, we estimated the probability of Giardia shedding: θ=0.44. We found no evidence that shedding changed with child gender or age, or between stool sample-collection days; that test sensitivity depended on child age; or that infection frequency changed with child age or varied among urban sectors. The average probability that an expert observer, using a single test, detects Giardia in a stool sample drawn from an infected child was estimated at  0.28. Therefore, reaching upper limit of the probability requires increasing θ, which in turn requires drawing, then pooling, replicate stool samples. Our analyses suggest, in sum, that ~56% of stool samples drawn from infected children did not contain Giardia. As no specific test can (or should) detect parasites in parasite-free samples, θ =0.44 measures the upper limit of the probability of detecting Giardia in a single stool sample drawn from an infected child. The only strategy capable of taking this probability beyond this limit is to draw-and-pool replicate samples; with 4 pooled samples. By allowing estimation (and modeling) of pathogen-shedding probabilities (θ), the approach we illustrate paves the way to studying pathogen transmission cycles and performance of microscopy-based diagnostics used in routine practice in unprecedented detail. 

BANKING MEMBERS:
Presidente - 3343228 - RODRIGO GURGEL GONCALVES
Interna - 1335413 - FABIANA BRANDAO ALVES SILVA
Externa ao Programa - 2088007 - CARLA NUNES DE ARAUJO
Externa à Instituição - ROSÂNGELA MARIA RODRIGUES - UFG