Computerized decisionmaking systems improve physician prescribing for long-term-care residents
Research Activities, January 2010, No. 353
Patients living in long-term-care facilities often have compromised kidney function. Since their kidneys cannot process medications properly, these individuals are at increased risk for adverse drug events. Using computer systems to calculate appropriate dosing of medications can improve physician drug prescribing for these patients, concludes a new study.
The researchers randomly assigned 22 long-stay units so that prescribing physicians would receive alerts (intervention units) or not receive alerts (control units) provided by a computerized clinical decision support system with prescriber order entry. Researchers could track alerts in the control units, but the alerts were hidden from the physicians. Alerts related to recommended drug doses and frequencies, as well as when to avoid using a particular drug for residents with renal insufficiency. Physicians were also prompted to order serum creatinine tests (that determine kidney function) when this information was not readily available. The researchers compared the proportion of appropriate final drug orders in both the intervention and control units.
The rates of alerts were nearly equal at 2.5 per 1,000 resident days in the intervention units and 2.4 in the control units. The proportion of dose alerts for which the final drug orders included an appropriate dose were similar between the intervention and control units. However, significantly higher proportions of final drug orders were appropriate in the intervention units for the remaining alert categories. For example, intervention units were 2.4 times more likely to have final drug orders calling for medication administration at an appropriate frequency, 2.6 times more likely to have final orders that omitted drugs that should be avoided, and 1.8 times more likely to order creatinine tests when this information was missing. Overall, final drug orders were appropriate 20 percent more often in the intervention units. Drugs that triggered alerts most often were levofloxacin, nitrofurantoin, cephalexin, metformin, gabapentin, and glyburide. The study was supported in part by the Agency for Healthcare Research and Quality (HS10481 and HS15430).
See "Computerized clinical decision support during medication ordering for long-term care residents with renal insufficiency," by Terry S. Field, D.Sc., Paula Rochon, M.D., M.P.H., Monica Lee, R.Ph., and others, in the July/August 2009 Journal of the American Medical Informatics Association 16(4), pp. 480-485.