Pharmacy, Lab Data Link Reveals Errors
Researchers found that 1.5 percent of prescriptions for oral potassium supplements dispensed by one Midwestern hospitals outpatient pharmacy went to patients with serum potassium levels of 5.3 mEq/L or more. According to recently published clinical practice guidelines (PDF) for potassium replacement therapy, normal serum potassium levels range from about 3.6 to 5.0 mEq/L.
The study, reported in the Oct. 15 American Journal of Medicine, examined computerized laboratory data and records of outpatient prescriptions filled at Chicagos Cook County Hospital in 1995. Although the laboratory and outpatient pharmacy maintained electronic records, the two sets of data did not link into a single system. The researchers merged the data to reveal problems with potassium prescribing.
The researchers also reviewed the manual laboratory records for 1995 and October through December1994. Records for 1994 were used to find serum potassium test results for patients who received oral supplements in early 1995.
According to the computerized records, the pharmacy in 1995 dispensed 32,563 prescriptions for oral potassium supplements to 12,825 different patients. Laboratory data showed 9,790 incidents in which serum potassium levels were at least 5.3 mEq/L; these incidents involved 4,188 patients. Nearly 500 of the potassium prescriptions belonged to patients with a high serum potassium level.
Eighty-five percent of the seemingly inappropriate potassium prescriptions were written less than three months after the laboratory reported a high serum level for the patients. More than half of the inappropriate prescriptions were written on the same day that the patients serum potassium test was run.
The study also examined how prescribers, who might not have known a patients potassium level when writing the prescription, reacted after the laboratory test revealed a high value. Review of the medical records for 25 percent of the patients with high potassium levels revealed no evidence that prescribers asked these patients to decrease or stop taking their supplement.
The study did not examine whether patients who inappropriately received a potassium supplement actually had an adverse event.
According to the researchers, all prescriptions at the hospital were manually written by hospital physicians and then entered into the pharmacy computer system. Laboratory test results were printed out and placed in each patients medical record. These results were also accessible through computer terminals in the hospital.
The researchers suggested that inappropriate prescribing of oral potassium supplements to outpatients could be avoided if physicians used computerized order-entry systems that directly link laboratory test results with pharmacy data. To buttress their view, the researchers cited three reports on how that type of computer system reduces the occurrence of similar medication errors in the inpatient setting.