Physicians' decisions to override computerized drug alerts in primary care

From Clinfowiki
Jump to: navigation, search

Although there is sufficient evidence that computerized physician order entry (CPOE) reduces medication errors in the inpatient setting, little is known about whether such systems are effective in an ambulatory care setting.

Introduction

The researchers studied 3,481 alerts created by the CPOE system at 5 adult primary care practices affiliated with Beth Israel Deaconess Medical Center in Boston. In total, 94.2% (3,280) were overridden by the physician. There was a slight variation between the rate of overrides of drug allergy alerts (91.2%) and drug interaction alerts (94.6%). Even the level 1 drug interaction alerts (high severity, strong evidence) were overridden 89.4% of the time.

Method

A random sample of 189 alerts was selected for detailed review to identify the factors associated with physicians’ decisions to override an alert and to determine if the override resulted in an adverse drug event (ADE). The researchers found 3 ADEs among the 122 patients with alert overrides, a statistically insignificant finding, and additionally judged only one of the events as potentially preventable.

Some reasons cited for overriding of alerts included: out-of-date information, interactions that were not clinically significant, failure to note patient tolerance of medication combinations and the inability to weight risk versus benefit.

Comment

It has been proven that alerts can provide significant benefit in preventing medical errors. However, they also have the potential to become inefficient time-wasters and annoyances in a system that is not properly programmed. Alerts must be informed by the most up to date research and, ideally, a CPOE system should have to ability to allow alerts to be fine-tuned at the facility, physician, medication and patient levels.

Although there is sufficient evidence that computerized physician order entry (CPOE) reduces medication errors in the inpatient setting, little is known about whether such systems are effective in an ambulatory care setting. The researchers studied 3,481 alerts created by the CPOE system at 5 adult primary care practices affiliated with Beth Israel Deaconess Medical Center in Boston. In total, 94.2% (3,280) were overridden by the physician. There was a slight variation between the rate of overrides of drug allergy alerts (91.2%) and drug interaction alerts (94.6%). Even the level 1 drug interaction alerts (high severity, strong evidence) were overridden 89.4% of the time.

A random sample of 189 alerts was selected for detailed review to identify the factors associated with physicians’ decisions to override an alert and to determine if the override resulted in an adverse drug event (ADE). The researchers found 3 ADEs among the 122 patients with alert overrides, a statistically insignificant finding, and additionally judged only one of the events as potentially preventable.

Some reasons cited for overriding of alerts included: out-of-date information, interactions that were not clinically significant, failure to note patient tolerance of medication combinations and the inability to weight risk versus benefit.

Comment

It has been proven that alerts can provide significant benefit in preventing medical errors. However, they also have the potential to become inefficient time-wasters and annoyances in a system that is not properly programmed. Alerts must be informed by the most up to date research and, ideally, a CPOE system should have to ability to allow alerts to be fine-tuned at the facility, physician, medication and patient levels.