The influence of productivity goals, job stress, and perceived workload on healthcare providers' laboratory test order error rates in a military medical treatment facility

A one month cross-sectional non-experimental study was conducted to investigate what effect a healthcare provider's expected productivity, measured in relative value units (RVUs), self-reported job stress, and perception of workload had on their laboratory test ordering error rates (LTOERs). Seventy-eight credentialed healthcare providers (physicians, residents, general medical officers, nurse practitioners, and one physician's assistant) participated in the study which took place in a military teaching hospital. Participant consent, demographic data and laboratory test ordering information was collected at recruitment. Computer reports were used to obtain study participants' RVU data and total number of laboratory test orders; and laboratory error logs were used to collect subjects' test order error data for the one month study period. After the data collection period, an 18 question, likert-scaled survey was distributed to obtain participants' level of job stress (10 questions, Cronbach's alpha = .93) and their perceived amount of workload (8 questions, Cronbach's alpha = .81). A Pearson product-moment correlation revealed a statistically significant relationship between healthcare provider's RVUs and their Laboratory Test Ordering Error Rates (r = .29, n = 77, p < .05), and an independent samples t-test identified a statistically significant difference [ t(77 = -2.06, p < .05] of medium effect (eta squared = .05) in mean LTOERs between males (M=.5, SD=1.1) and females (M=1.0, SD=1.4) showing that females made twice as many ordering errors than males. Analysis of Variance (ANOVA) identified a statistically significant difference between provider types [F(2,74) = 5.39, p < .05] with General Medical Officers (non-specialty physicians including Residents and Interns) having the lowest mean LTOER (M = .07), Nurse Practitioners with a mean LTOER of .952 and Medical Doctors (Physicians with post-residency training) having the highest mean LTOER of 1.06. Civilian healthcare providers' ordering error rates were higher (M=.965, SD=1.06) than providers in uniform (M=.635, SD=1.30). Standard regression analysis revealed that the Male category was the only variable to have significant ability to predict Laboratory Test Ordering Errors in the model used for this study [F (9, 65) = 2.39; p < .05], and the Beta Coefficient achieved from the analysis indicated male LTOERs were negatively related to the outcome. A second independent variable, Total Relative Value Units (RVUs), approached significance with a p value of .07. Previous studies have not been found that reported computerized laboratory test order error rates for healthcare providers or what influence demographic characteristics, productivity goals, job stress or perceived workload may have on the HCPs' ability to accurately enter laboratory test orders into an established computer system; therefore, further research is needed to fully develop a better understanding of these factors.