May 8, 2006
Using Safety Alerts in Electronic Medical Records Reduces Rates
of Potentially Dangerous Medication Interactions
Using Safety Alerts in Electronic Medical Records Reduces Rates of
Potentially Dangerous Medication Interactions
(PORTLAND, Ore.) - Creating safety alerts that appear in electronic medical records (EMRs) when
clinicians co-prescribe medications with potentially dangerous interactions
reduces the number of co-prescriptions by nearly 15 percent. The results of the
Safety in Prescribing study, conducted by Kaiser Permanente's Center for Health
Research, appear in the May 8th issue of the Archives of Internal Medicine.
"This study shows that safety alerts in electronic medical records are a powerful tool to help
clinicians make decisions about what drugs they should prescribe for their
patients," says Adrianne Feldstein, MD, MS, lead author of the study and an
investigator at Kaiser Permanente's Center for Health Research. "We compared
the effects of two interventions in outpatient clinical settings - safety
alerts in EMRs versus safety alerts plus clinician education courses. We found
that adding clinician education to EMR safety alerts did not reduce the rates
of co-prescribing medications that interact any more than EMR safety alerts
alone. The reason is unclear, but it may be that the "just in time" information
provided by safety alerts is more useful to clinicians than education courses."
The study focused on patients who were prescribed warfarin, which is an anticoagulant that
prevents blood clots from forming or growing larger. Warfarin is prescribed to
patients with blood clots in the lower extremities; patients with atrial
fibrillation, artificial heart valves, and coronary artery stents; and patients
who have knee or hip replacements. It was chosen for the study because it is
widely used - over one million prescriptions annually - and because so many
drugs - prescription and non-prescription - interact with warfarin to increase
its blood-thinning effects. Results of a report from a large pharmacy benefits
manager database show that more than two-thirds of patients receiving warfarin
during a one-year period received at least one interacting medication that
increased bleeding risk.
To conduct the study, the researchers tracked the co-prescribing rates of clinicians at 15 primary
care clinics before the alerts from January 2000 through November 2002. They
analyzed the electronic medical records of 9910 patients taking warfarin to
find out the prescription rates of warfarin and five interacting medications -
acetaminophen, non-steroidal anti-inflammatory medications, fluconazole,
metronidazole, and sulfamethoxasole. During a four-month period from December
2002 to March 2003, safety alerts were randomly implemented in eight clinics,
and safety alerts and education courses were implemented in the other seven
clinics. The safety alerts remained active in all 15 clinics for 18 months from
April 2003 through August 2004.
In the 12 months prior to the alerts, warfarin patients received 3,288 interacting prescriptions
for every 10,000 warfarin patients. Interacting prescriptions for warfarin
patients decreased immediately after the alerts were implemented, with an
estimated drop in the first month of 329.7 interacting prescriptions for every
10,000 warfarin patients. Twelve months after the alerts were implemented,
warfarin patients received 2804.2 interacting prescriptions for every 10,000
warfarin users. Compared to an co-prescription without safety alerts estimated
at 3294.0, the alerts led to an estimated reduction of 489.8 interacting
prescriptions for every 10,000 warfarin patients, a 14.9% relative reduction.
The effect was predominantly due to reductions in the co-prescribing of
warfarin and acetaminophen.
"Clinicians face difficult choices when ordering prescription drugs for warfarin patients," Dr.
Feldstein says. "Because so many drugs interact with warfarin and have the
potential for harm, ideal choices are seldom available. For example, if a
warfarin patient is having a lot of pain, should the clinician prescribe
acetaminophen - which can cause excessive bleeding - or should the clinician
prescribe codeine - which can cause drowsiness and could lead to an older
patient tripping or falling and breaking a hip? There is not always an easy
answer. What safety alerts do is help clinicians make these difficult decisions
by alerting them about the known interactions among the medications they've
entered into the electronic medical record and by indicating alternative
medications they might prescribe."
This study was funded through cooperative agreement U18 HS11843 for the Agency for healthcare
Research and Quality (AHRQ). This study is one of a series of studies being
conducted by researchers at ten HMOs throughout the United States as part of
AHRQ's Centers for Education and Research on Therapeutics and Patient Safety
research programs.
Kaiser Permanente's Center for Health Research, founded in 1964, is a non-profit research
institution whose mission is advancing knowledge to improve health.
Kaiser Permanente is the nation's leading integrated health care
organization. Founded in 1945, it serves more than 8.4 million people
nationwide, including more than 485,000 in Oregon and Southwest
Washington.
For more infomation contact:
Terry Fitzpatrick - 503-335-6602 or
Brad Brokaw 503-813-4820