Published in the November 2005 issue of Today’s Hospitalist
Studies look at fallout of discrepancies in medication histories
Two recent studies shed new light on the frequency and impact of discrepancies in the medication reconciliation process in hospitalized patients.
One of the studies, which was published in the Sept. 12 Archives of Internal Medicine, focused on patients over 65 admitted to the hospital with one of nine conditions. After discharge, a geriatric nurse practitioner visited patients at home to check what medications patients were taking. They then compared the results to the patients’ pre- and post-hospital drug regimens.
Researchers found one or more discrepancies in more than 14 percent of patients. Of those patients, 62 percent experienced a single discrepancy; 25 percent experienced two discrepancies; 8 percent experienced three discrepancies; and 5 percent experienced four or more discrepancies.
Upon further examination, researchers found that about half of the discrepancies were due to patients, while half were due to system problems. When discrepancies were due to patient factors, researchers found that patients often were not using the medication, but that their noncompliance was unintentional. System problems were typically due to incomplete, inaccurate or illegible discharge instructions, either as a result of handwriting or abbreviations.
Five classes of medications accounted for half of all medications, a fact that researchers said was probably due to the prevalence of the drugs on the list.
Medication discrepancies were also more likely when congestive heart failure was present. Researchers hypothesize that’s because medications are adjusted so frequently in these patients, or because these patients are typically treated with multiple medications and experience problems when their medications are adjusted in new health care settings.
Discrepancies appeared to have an impact on patient care. Just under 15 percent of patients in whom there was a medication discrepancy were rehospitalized at 30 days. By comparison, only 6 percent of patients in whom there was no discrepancy were rehospitalized.
In a second study in the Aug. 30 issue of the Journal of the Canadian Medical Association, researchers conducted a meta-analysis on the incidence, source and type of medication history discrepancies.
Researchers found that up to 67 percent of patients experienced at least one type of medication discrepancy.
Up to 61 percent of patients experienced at least one error of omission, defined as the deletion of a drug that was used before admission. Up to 22 percent experienced an error of commission, defined as the addition of a drug not used before admission.
On a positive note, the meta-analysis noted that one study found that a comprehensive medication history yielded twice as many prescription medications as physician- acquired medication histories.
Pay-for-performance outcomes, potential problems
A study examining a pay-for-performance plan found only marginal improvement on three measures, but potential problems in the way that incentives were awarded to the participating physician groups.
The study, which appeared in the Oct. 12 Journal of the American Medical Association, compared the performance of 300 large physician groups on three measures: cervical cancer screening, mammography and A1c.
The physicians being paid for improvements in performance showed greater improvement on measures for cervical cancer screening (5.3 percent improvement vs. 1.7 percent), for an overall improvement of 3.6 percent.
There was considerably less improvement, however, on the other two measures. Mammography rates climbed 1.9 percent for the participating groups, compared to 0.2 percent for the control groups. A1c rates improved 2.1 percent in both groups.
Among the groups participating in the pay-for-performance plan, researchers found some potential problems with the way quality bonuses were distributed.
Groups whose baseline performance was at or above the targets when the project began improved the least, in part because they were already doing well on the indicators. While they showed the least improvement, those groups received the largest share of bonus payments.
Groups that had low baseline levels of performance at the outset and showed dramatic improvements in performance “but still lagged behind on overall performance scores “received smaller shares of the bonuses. That finding led the study’s authors to conclude that pay-for-performance systems need to reward not only performance, but improvement.
FDA issues warning about confusion over soundalike drugs
The FDA has released a warning about two soundalike drugs that are creating confusion among physicians.
In a Sept. 30 advisory, the FDA said that several prescribing errors had involved the following two drugs:
“¢ The extended-release tablet form of Toprol-XL, or metoprolol succinate, a drug from AstraZeneca used to treat hypertension, has been confused with Topamax, or topiramate, from Ortho-McNeil. The latter drug is used to treat epilepsy and migraines.
“¢ Toprol-XL has also been confused with Tegretol and Tegretol-XR, a drug from Novartis Pharmaceuticals. The latter drug is used for complex partial seizures, generalized tonic-clonic seizures and trigeminal neuralgia.
The FDA report is online.
Sleep deprivation in housestaff similar to having a few drinks
If you wonder how too much work and too little sleep affect the performance of housestaff, new data say it is comparable to having a few drinks.
A study in the Sept. 6 Journal of the American Medical Association found that housestaff who work 80 to 90 hours a week display signs of impairment similar to residents who have worked a shorter week and had a few drinks.
Researchers had pediatric residents undergo tests to measure any impairment after working light call (44 hours a week) and heavy call (80 to 90 hours a week). They also ranked performance by giving residents who had worked light call three to four drinks.
On tests designed to look for impairment, residents who had worked heavy call performed at about the same level as residents who had worked light call and had a blood alcohol concentration of 0.04 percent. (Most states define intoxication as a level of 0.08 percent.)
Housestaff who had worked heavy call had reaction times that were 7 percent slower than residents who had been on light call, and they made 40 percent more mistakes. During a simulated driving test, the heavy call residents drove faster and had difficulty maintaining a constant speed.
Many hospitalists seek career guidance; all seek job satisfaction. Watch for the forthcoming feature: 2008 Hospitalist Salary Survey appearing in October 2008.