A strategic, system-based knowledge management approach to dealing with high error rates in the deployment of point-of-care devices
Thesis (MBA)--Stellenbosch University, 2014.
There is a growing trend towards the use of point of care testing in resource poor settings, in particular in the diagnosis and treatment of infectious diseases such as Human Immunodeficiency Virus (HIV), Tuberculosis (TB) and Malaria. The Alere PIMA CD4 counter is widely used as a point of care device in the staging and management of HIV. While the instrument has been extensively validated and shown to be comparable to central laboratory testing, little is known about the error rates of these devices, as well as the factors that contribute to error rates. This research was a retrospective analysis of error rates from 61 PIMA point of care devices deployed in nine African countries belonging to Medisciens Sans Frontiers. The data was collected between January 2011 and June 2013. The objectives of the study were to determine the overall error rate and, where possible, determine the root cause. Thereafter the study aimed to determine the variables that contribute to the root causes and make recommendations to reduce the error rate. The overall error was determined to be 13.2 percent. The errors were further divided into four root causes and error rates assigned to each root cause based on the error codes generated by the instrument. These error rates were found to be operator error (48.4%), instrument error (2.0%), reagent/cartridge error (1%) and sample error (4.3%). It was found that a high percentage of the errors were ambiguous (44.3%), meaning that they had more than one possible root cause. A systems-based knowledge management approach was used to create a qualitative politicised influence diagram, which described the variables that affect each of the root causes. The influence diagram was subjected to loop analysis where individual loops were described in terms of the knowledge type (tacit or explicit), the knowing type (know-how, know-who, know-what and know-why), and the actors involved with each variable. Where possible, the variable was described as contributing to pre-analytical, analytical or post-analytical error. Recommendations to reduce the error rates for each of the variables were then made based on the findings.