Commercial enzyme-linked immunosorbent assay versus polymerase chain reaction for the diagnosis of chronic Chagas disease: a systematic review and meta-analysis

Chronic Chagas disease diagnosis relies on laboratory tests due to its clinical characteristics. The aim of this research was to review commercial enzyme-linked immunosorbent assay (ELISA) and polymerase chain reaction (PCR) diagnostic test performance. Performance of commercial ELISA or PCR for the diagnosis of chronic Chagas disease were systematically searched in PubMed, Scopus, Embase, ISI Web, and LILACS through the bibliography from 1980-2014 and by contact with the manufacturers. The risk of bias was assessed with QUADAS-2. Heterogeneity was estimated with the I2 statistic. Accuracies provided by the manufacturers usually overestimate the accuracy provided by academia. The risk of bias is high in most tests and in most QUADAS dimensions. Heterogeneity is high in either sensitivity, specificity, or both. The evidence regarding commercial ELISA and ELISA-rec sensitivity and specificity indicates that there is overestimation. The current recommendation to use two simultaneous serological tests can be supported by the risk of bias analysis and the amount of heterogeneity but not by the observed accuracies. The usefulness of PCR tests are debatable and health care providers should not order them on a routine basis. PCR may be used in selected cases due to its potential to detect seronegative subjects.

PREVALENCE AND PREDICTORS OF VITAMIN A DEFICIENCY AMONG INFANTS IN WESTERN KENYA USING A CROSS-SECTIONAL ANALYSIS

Vitamin A (VA) deficiency (VAD) is a major nutritional public health problem among children under-5-years-old in the developing world including Kenya. A community-based cross-sectional survey among 1,630 children (aged 6-23 mos) was undertaken in Western Kenya. A questionnaire was administered to collect demographic, socio-economic and dietary intake information. Prevalence of low retinol-binding protein (RBP) concentrations was assessed using Dried Blood Spot (DBS) methodology. Analysis of RBP was carried out using rapid enzyme immunoassay (EIA) and C-reactive protein (CRP) was carried out using enzyme linked immunosorbent assay (ELISA) to estimate VA and sub-clinical inflammation statuses, respectively. Values were adjusted for influence of inflammation using CRP (CRP >5 mg/L) and population prevalence of VAD (RBP <0.825 μmol/L, biologically equivalent to 0.70 μmol/L retinol) estimated. Anthropometric data gave three indices: stunting, wasting and underweight—all of which took age and sex into consideration. Mean (geometric± SD) concentration of RBP was adequate (1.56±0.79μmol/L) but the inflammation-adjusted mean (±SE) prevalence of VAD was high (20.1±1.1%) in this population. The level of CRP was within normal range (1.06±4.95 mg/L) whilst 18.4±0.9% of the children had subclinical inflammation (CRP>5 mg/L). Intake of VA capsule (VAC) by a child was a predictor of VAD with children who have not taken VA during the past 1 year prior to the survey having a 30% increased risk of VAD (OR (CI): 1.3 (1.1-1.7); p=0.025. Additionally, age of the child was a predictor with older children (18-23 mos) having a 30 % increased risk of VAD (OR (CI): 1.3 (1.1-1.9); p=0.035); the caretaker’s knowledge on VA and nutrition was also a predictor of VAD with children whose caretaker’s had poor knowledge having a 40 % increased risk of VAD (OR (CI): 1.4 (1.0-1.9); p=0.027. A child’s district of residence was also a significant predictor of VAD. Prevalence of VAD in this sample of infants was high. Predictors of VAD included child intake of VAC in the last 1 year before the survey, older children, children whose caretakers had poor VA and nutritional knowledge and a child’s district of residence. There is a need to improve knowledge on nutrition and VA of caretakers; undertake a targeted VAC distribution, particularly in children older than 1 year and above and use a sustainable food-based intervention in the areas with severe VAD.