P0248 - Bedside Expedited Testing for Anemia in Tanzania: the BETA Study

Background: Anemia causes significant morbidity and mortality in low-resource settings and is a common problem in sickle cell disease (SCD). Timely, accurate diagnosis of anemia enables appropriate triage and treatment. In the Bedside Expedited Testing for Anemia and Sickle Cell Disease (BETA) study, we aimed to evaluate the accuracy of two simple point-of-care diagnostic devices in the hands of the intended end-users: one for anemia and one for SCD. Here we present the results of the Anemocheck-LRS assay, a simple, inexpensive, color-based test that quickly quantifies hemoglobin concentration to diagnose anemia.

Methods: BETA was performed at Bugando Medical Centre in northwest Tanzania. Patients under 21 years old with a concern for very severe anemia or SCD were recruited from the hospital’s wards and clinics. After a finger-prick, 10 uL of blood were collected in a capillary tube and added to a vial of premixed reagents. Within 2 minutes, the solution was compared to a color scale representing hemoglobin concentrations from 2-8 g/dL and interpreted independently by both an expert reader familiar with the assay (a member of the study team), and a novice reader (any adult at the bedside). A HemoCue 201+ was used to measure the gold-standard hemoglobin concentration. Interobserver agreement between the expert and novice readers, correlation between the AnemoCheck-LRS and the HemoCue 201+, and the sensitivity and specificity of the AnemoCheck-LRS to determine the presence or absence of very severe anemia (Hb≤5 g/dL and Hb≤6 g/dL) were calculated.

Results: BETA recruited 233 participants (mean age 5.8±4.3 years, range 0-20 years). The majority were recruited from the inpatient ward (87%) or the sickle cell clinic (9%). Three expert readers each interpreted a mean of 77.7 tests. Novice readers interpreted a mean of 1.4 tests each and included 164 intern doctors (31%), nurses (26%), family members (16%), medical students (12%), medical officers (7%), and others (7%). Interobserver agreement was excellent (r = 0.91, p < 0.001). The absolute difference (mean ± SD) between the AnemoCheck-LRS and the HemoCue 201+ was 0.7±0.6 g/dL for experts and 0.8±0.6 g/dL for novices. There was a strong correlation between AnemoCheck-LRS and the HemoCue 201+ hemoglobin concentration (r=0.78 and 0.72 respectively). AnemoCheck-LRS had 96% sensitivity and 51% specificity for detection of Hb ≤6 g/dL, and 70% sensitivity and 91% specificity for detection of Hb ≤ 5 g/dL.

Conclusion: BETA provides essential data regarding the use of the AnemoCheck-LRS to diagnose anemia in a real-world setting within sub-Saharan Africa. The optimized, second-generation AnemoCheck-LRS demonstrated good correlation with the gold standard results and excellent interobserver agreement. The AnemoCheck-LRS could be used as a screening tool to quickly identify children who may benefit from blood transfusion or may have undiagnosed SCD.