Magnus D et al. 2002 "GPs' views on computerized drug interaction alerts: questionnaire survey."
Reference
Magnus D, Rodgers S, Avery AJ. GPs' views on computerized drug interaction alerts: questionnaire survey. J Clin Pharm Ther 2002;27(5):377-382.
Abstract
"Background: There is evidence that patients are being prescribed potentially hazardous drug-drug combinations in general practice despite the use of computerized drug interaction alert systems. One reason for this may be that general practitioners (GPs) are overriding these alerts without properly checking them.
Aims: (i) To assess GPs views on the relevance of information provided by alert systems; (ii) To determine the proportion of GPs that admit to frequently overriding interaction alerts without properly checking them; and (iii) To explore factors that might be associated with a tendency to override alerts.
Methods: Questionnaire survey of GPs in four primary care trusts in the Nottingham area of the UK.
Results: The response rate was 70% (236 of 336) and detailed analysis was conducted on the 220 respondents who had drug interaction alert systems on their practice computers. Of these, 22% (49 of 220) admitted to frequently or very frequently overriding drug interaction alerts without properly checking them. Potential reasons for overriding alerts included the perception that the alerts were frequently irrelevant. Nevertheless 90% (198 of 219) agreed that it should be more difficult to override alerts for potentially lethal drug combinations. Logistic regression analysis showed that users of the [the most common EHR] computer system were much less likely to admit to overriding alerts without properly checking them than users of other computer systems. Conclusion: A minority of GPs admit to frequently overriding their drug interaction alert systems without properly checking them. The type of computer system used by GPs may make it more or less likely that they override alerts."
Aims: (i) To assess GPs views on the relevance of information provided by alert systems; (ii) To determine the proportion of GPs that admit to frequently overriding interaction alerts without properly checking them; and (iii) To explore factors that might be associated with a tendency to override alerts.
Methods: Questionnaire survey of GPs in four primary care trusts in the Nottingham area of the UK.
Results: The response rate was 70% (236 of 336) and detailed analysis was conducted on the 220 respondents who had drug interaction alert systems on their practice computers. Of these, 22% (49 of 220) admitted to frequently or very frequently overriding drug interaction alerts without properly checking them. Potential reasons for overriding alerts included the perception that the alerts were frequently irrelevant. Nevertheless 90% (198 of 219) agreed that it should be more difficult to override alerts for potentially lethal drug combinations. Logistic regression analysis showed that users of the [the most common EHR] computer system were much less likely to admit to overriding alerts without properly checking them than users of other computer systems. Conclusion: A minority of GPs admit to frequently overriding their drug interaction alert systems without properly checking them. The type of computer system used by GPs may make it more or less likely that they override alerts."
Objective
There were three aims: "(i) To assess GPs views on the relevance of information provided by [drug] alert systems; (ii) To determine the proportion of GPs that admit to frequently overriding [drug] interaction alerts without properly checking them; and (iii) To explore factors that might be associated with a tendency to override alerts."
Tools Used
Type Clinic
Primary care
Size
Small and/or medium
Geography
Urban and suburban
Other Information
Four clinics in the Nottingham, United Kingdom area were involved in the study.
Type of Health IT
Decision support system
Type of Health IT Functions
The alerts differ by electronic health record system. The majority of physicians surveyed use a system in which "alerts are based on drug interactions highlighted in the British National Formulary (rather than another data source) and the system does not grade the alerts in terms of potential severity."
Context or other IT in place
Electronic health records (EHR)
Workflow-Related Findings
A "minority of GPs [22.3%] admit to frequently or very frequently overriding their drug interaction alert systems without properly checking them."
Question: "How often do drug interaction alerts change your initial prescribing decisions?" Never or Rarely: 30.1%, Sometimes: 56.2%, Frequently or Very Frequently: 13.7%.
Question: "How often do drug interaction alerts provide you with information that you already know?" Never or Rarely: 1.4%, Sometimes: 23.2%, Frequently or Very Frequently: 75.5%.
Question: "How often do drug interaction alerts provide information that is irrelevant to the patient?" Never or Rarely: 24.2%, Sometimes: 40.2%, Frequently or Very Frequently: 35.6%.
Question: "How often do you feel annoyed by the drug interaction alert messages?" Never or Rarely: 49.6%, Sometimes: 28.2%, Frequently or Very Frequently: 22.3%.
Most respondents agree that "[t]he drug interaction alert system is good at alerting me to significant interactions" (88.1%). Most also agree that "[t]here should be a greater distinction between important and less important interactions" (85.9%) and that "[d]rug interaction alerts would be more useful if they were clearly graded according to severity" (85.0%). Some agree that the alerts frequently provide information the physician was not aware of (40.0%), yet most agree that drug interaction alerts are sometimes irrelevant (73.5).
The reasons provided for overriding the alerts include that the alert is irrelevant (87.1%), having more faith in their own knowledge (41.4%), having more faith in other information (51.6%) and being too busy to read the alert (19.4%).
Users of the most common system were less likely to override alerts "without properly checking them ... The most likely [reason] comes from the fact that alerts on [this] system are not graded in terms of severity. This may encourage GPs to try to check all of them properly to ensure that they are not missing anything important."
Study Design
Only postintervention (no control group)
Study Participants
Two hundred thirty-six general practitioners responded to the survey. Two hundred twenty of these respondents had computerized drug interaction alert systems.