E-ISSN 2231-3206 | ISSN 2320-4672

2015, Vol:4,Issue:12

Research Articles
  • Indi J Medic Science and P Health.2015; Volume:4(12):1640-1654 doi : 10.5455/ijmsph.2015.15032015336
  • Measurement System Analysis (MSA) of Empirically-derived Composite Measure of Preventive Care Counseling Practices of HIV Medical Care Providers
  • Osaro Mgbere,, Mamta Singh, Raouf Arafat

Abstract

Background: Although preventive care services and risk reduction counseling are offered to HIV patients during their routine clinic visits, assessment of clinicians practice performance has been difficult because of the use of many separate indicators.
Objective: To assess the measurement system of empirically-derived composite measure of preventive care practices of HIV medical care providers (overall preventive care counseling index, OPCi) in managed care clinical settings using the following statistical properties: precision (gauge reproducibility and repeatability, gauge R&R), accuracy (bias and linearity), stability, and process capability.
Materials and Methods: Data used were obtained from the cross-sectional survey of HIV medical care providers in 13 HIV care facilities in Houston/Harris County, Texas, USA, that participated in the Centers for Disease Control and Pre- vention (CDC) Medical Monitoring Project in 2009. Result: Domain-specific preventive care counseling indices developed produced significant ( p ≤ 0.05) Cronbach’s alpha coefficients that ranged from 0.64 to 0.91. The variance components for OPCi attributed to the provider and patient statuses were 29.7% and 31.4%, respectively, while the gauge R&R was 68.6%. A high percent gauge R&R precision to tolerance ratio (99.13%) and precision to total variation ratio (82.85%) were obtained. With measurement system intraclass correlations without bias factors (r ) of 0.45 and with bias factors (r ) of 0.31, the system was classified as pe b “third class monitor,” indicating the need for process improvement. However, the system expanded uncertainty was within set range at 99% confidence level making it statistically “stable” and “capable” at a process capability ratio (C ) of 0.999 p (95% CI: 0.793–1.205) and long-term sigma yield of 2.292. Conclusion: Reduction in measurement variations would enhance providers’ performance output and improve both the quality of preventive care counseling received and the overall health outcome of HIV patients.