Session number: 1075
Abstract title: Perceived Impact of CPRS on Quality and Safety of Patient Care
Author(s):
PA Glassman - VA HSR & D Center of Excellence for the Study of Healthcare Provider Behavior
B Simon - VA HSR & D Center of Excellence for the Study of Healthcare Provider Behavior
P Belperio - VA HSR & D Center of Excellence for the Study of Healthcare Provider Behavior
A Lanto - VA HSR & D Center of Excellence for the Study of Healthcare Provider Behavior
Objectives: How clinicians perceive computerized patient care systems may impact on whether they are effectively used. Our objective was to explore clinicians’ attitudes and experiences with the computerized patient record system and, in particular, with physician order entry and automated drug alerts.
Methods: Cross-sectional survey. Eligible providers included 263 clinicians practicing within a Southern California Veterans Affairs Healthcare System. We developed a 67-item survey (19 questions) to elicit information in four domains: (1) Computer use for patient-related activities; (2) Recognition of common drug-drug and drug-disease interactions; (3) Recognition of adverse drug events; and (4) Attitudes about pharmacist tasks and drug alerts.
Results: Response rate was 64%. Participating clinicians had positive perceptions of the Computerized Patient Record System (CPRS) in regards to improving medical record retrieval (85%), prescribing more safely (55%) and requesting laboratory tests with less error (52%). However, they felt it also increased the time needed to complete a patient encounter (67%) and to write prescriptions (50%) compared to written methods. Clinicians recognized an average of 45% of 10 possible drug-drug interactions (range 11%-64%) and 56% of 11 possible drug-condition (disease) interactions (range 24-87%). About 90% of clinicians felt that drug alerts would help them to identify interactions but the biggest reported barrier to using drug alerts was the number of non-relevant alerts (55% of providers). Further statistical analyses found no association between physician demographics and attitudes but did find that general internists did better than non-medical specialists in recognizing drug interaction pairs, that younger physicians were more likely to recognize drug condition pairs, and that clinicians who spent more time in clinic recognized more drug-condition interactions.
Conclusions: CPRS is perceived to improve quality and safety but may decrease clinician efficiency. Clinicians have some difficulty recognizing important drug pairs though they are willing to accept and act on drug interaction alerts. The biggest obstacle to using these, however, was perceived to be too many irrelevant alerts.
Impact statement: VA centers should be sensitive to increased workload and the effect on appointment duration when implementing, or expanding functions in, CPRS. Further refinement of drug alerts is necessary to increase specificity and usability.