387 Comparison of Physician Versus Nurse Assessment of Clinical Opiate Withdrawal Scale Evaluation Using Video Simulation

Abstract

The Clinical Opiate Withdrawal Scale (COWS) was first described by Wesson et al in 1999 for use in initiating treatment with the partial opioid agonist, buprenorphine. The components of COWS were previously validated in other scoring systems, but the COWS tool was never fully evaluated prospectively. Despite this, it is currently widely utilized to determine degree of withdrawal symptoms and eligibility for treatment with buprenorphine. The original description of COWS states that it is “clinician performed”; however, in practice the score is frequently calculated by physicians, mid-level providers as well as nurses. In response to the opioid epidemic, many emergency departments are now initiating buprenorphine in selected patients if they demonstrate significant signs and symptoms of opioid withdrawal. It is important to determine if there is consistency in the scores obtained by nursing as compared to physicians. In order to evaluate the reliability and accuracy of the participant’s COWS score, 4 simulated, standardized patient encounters were pre-recorded by the study team. Two nurses and 2 physicians not participating in the study validated the videos for overall clarity and to ensure that each video had all components required to calculate the COWS. The information given in the videos corresponded to a COWS score predetermined by the study team. The videos were designed to assess the lower end of the COWS as this affects clinical decision to initiate treatment with buprenorphine. Individuals that are participating in the study watched and evaluated all 4 pre-recorded standardized patients and completed a COWS scoring sheet. The study subjects returned the scoring sheets and the data were de-identified. The primary outcome of interest was comparing differences in total COWS score as calculated by physicians and nurses. Secondary outcomes of interest included assessing individual components of the COWS for variation. Comparative statistics were used to evaluate differences in the outcome of interest. 23 physicians and 26 nurses completed the study procedures. The median scores are presented in the table. The median scores for each patient encounter were similar as calculated by nurses versus physicians. There was a statistical difference found only for patient 3 although the clinical significance is likely negligible (Median score by nurses = 4 versus median score by physicians = 5; P=0.03). The ranges of scores showed a lack of precision of this scoring instrument. For example, the range of scores obtained by nurses for Pt. 1 was 9-25 and by physicians was 8-22. No individual component of the COWS was consistently noted to have significantly different scores as determined by nurses versus physicians. Comparison of total COWS scores obtained by nurses versus physicians demonstrates similar overall accuracy. This data suggests that it is reasonable to have nurses perform the COWS assessment. A concerning finding is the lack of overall precision of the instrument. The wide ranges of scores obtained implies that there could be clinical implications such as whether or not a patient is eligible to receive buprenorphine.