Even in primary-care physician offices where health information technology has seldom been used, introducing personal digital assistants as prescribing tools can help lower the potential for medication errors, according to a study by Kimberly Galt, associate dean for research and professor of pharmacy practice, and her colleagues at Creighton University in Omaha.
Using PDAs led to significant reductions in errors of legibility and omissions, and less use of abbreviations and symbols, the report concluded. The report was included in a four-volume set of 140 articles published by the U.S. Agency for Healthcare Research and Quality released May 4. The findings also were presented at the National Patient Safety Foundation Congress held May 4-6 in Orlando, Fla.
In a three-year study funded by a $900,000 AHRQ grant, Galt's team first analyzed the hand-written orders generated by 78 prescribers at 31 physician offices in eastern Nebraska and western Iowa between June and December 2002. Half of the physicians were randomly selected to be trained in using PDAs for writing prescriptions. Data were collected on the 14,378 prescriptions they wrote, both by hand and with the PDA, between March and August of 2003 and compared with the 19,372 prescriptions previously written.
Omission of patient age or birthdate on prescriptions fell from 95.5% to 59.2%, and significant declines were also seen in omitting refill status and indication for use, according to the report. The use of symbols also declined from 76.7% to 47.4%, while illegible prescriptions decreased from 9.1% to 2.7%.
"Each of these independently are an individual potential source of a mistake," Galt said, adding that knowing a patient's age improves correct dosing and patient identity verification.
The physician offices chosen for the study all used computers for billing, but none had computers in patient-care rooms. The physicians randomly chosen for the PDA intervention were given training on using the devices to write prescriptions and to use drug-information software for decisionmaking support.
The PDAs were not integrated into an electronic medical records system. Even though almost all of the offices were in urban areas, high-speed Internet access was not readily available, Galt said. "The technological readiness was not there for the offices to receive a more sophisticated system."
Galt said that although a freestanding PDA could be seen as a "suboptimal model," there were still meaningful reductions in error rates. Use of the PDAs, however, varied widely.
Nine of the 39 physicians were considered "adopters" and used the PDAs to write between 88% and 100% of prescriptions. Thirteen were identified as "potential adopters" who used the PDA for between 16% and 61% of prescriptions.
The remaining 17 physicians were classified as nonadopters and included the "techno terrified" group that didn't generate any prescriptions via the PDA.
Use of the drug information application also varied widely, with 21% using it daily, 18% weekly, 24% monthly, and 37% using it quarterly or less frequently.
Galt said the thinking behind the study was to find simple things that make a substantial difference, and that demonstrating how physicians can reduce errors by using a printer and a freestanding PDA with a downloaded drug-information application met this standard.
"It has value," she said. "It's realistic and practical for a physician office that doesn't have access to anything else."