Abstract
With chronic disease prevalence on the rise globally, surveillance and monitoring are critical to improving health outcomes. Point-of-care (POC) testing can facilitate epidemiological research and enhance surveillance systems in limited resource settings, but previous research has identified bias between POC devices and laboratory testing. We compared the performance of two POC blood analyzers, the iSTAT handheld (Abbott, Princeton, NJ, USA) and the StatSensor Creatinine (Nova Biomedical, Waltham, MA, USA) to concurrent blood samples analyzed at a local laboratory that were collected from 89 agricultural workers in Guatemala. We measured creatinine and other measures of underlying health status with the POC and the lab blood samples. Pearson correlation coefficients, Bland-Altman plots, no intercept linear regression models and two-sample t-tests were used to evaluate the agreement between the POC and lab values collected across three study days and to assess differences by study day in a field setting. On average there was no observed difference between the iSTAT and lab creatinine measurements (p = 0.91), regardless of study day. Using lab creatinine as the gold standard, iSTAT creatinine results were more accurate compared to the Statsensor, which showed some bias, especially at higher values. The iSTAT had good agreement with the lab for sodium and blood urea nitrogen (BUN), but showed differences for potassium, anion gap, bicarbonate (TCO2), glucose, and hematocrit. In this tropical field setting, the research team devised a protocol to prevent the devices from overheating. In limited resource settings, POC devices carry advantages compared to traditional lab analyses, providing timely results to patients, researchers, and healthcare systems to better evaluate chronic health conditions. Technical challenges due to use of POC devices in high heat and humidity environments can be addressed using a standard protocol for transporting and operating the devices.