SAN FRANCISCO – Clinically important medication errors occurred in 51% of patients within 3 months of discharge from hospitalization for acute coronary syndrome and/or decompensated heart failure, a study of 851 patients showed.
In addition, the randomized controlled trial found no benefit from a predischarge pharmacist intervention and counseling incorporating tools designed for patients with low health literacy levels, compared with usual care in which the treating health care provider reconciled medications with the help of support software before patient discharge.
Although the trial focused on comparing the pharmacist intervention with usual care, "the biggest take-home message is that medication errors are very common" in these patients after hospitalization, Dr. Cecelia N. Theobald said at the annual meeting of the American College of Physicians.
Patients came from Vanderbilt University Medical Center in Nashville, Tenn., and from Brigham and Women’s Hospital in Boston. The study randomized 423 patients to the intervention and 428 to usual care.
In the intervention group, pharmacists reconciled medications and provided in-depth patient counseling before discharge, including a review of potential drug side effects. The pharmacists had specialized aids to help low-literacy patients adhere to their medication regimens, and they provided tailored follow-up to patients via postdischarge phone calls. In the usual-care group, the treating provider reconciled medications and pharmacist consultation was available on request. The usual-care group did not use the low-literacy aids or phone follow-up.
To assess outcomes, two clinicians reviewed patient records and patients were interviewed by phone 30 days after discharge.
Thirty percent of patients had one or more adverse drug events that were considered to be preventable or ameliorable. Another 30% of patients had at least one potential adverse drug event, she said. Adverse drug events occurred in 47% of patients on cardiovascular agents other than diuretics, 21% of patients on diuretics, and 5% of patients on opioids. Potential adverse drug events were seen in 43% of patients on cardiovascular agents other than diuretics, and in 12% of patients on diuretics (Ann. Intern. Med. 2012;157:1-10).
On a per-patient basis, 0.87 clinically important medication errors occurred in the intervention group, compared with 0.95 events per patient in the usual-care group, a difference that was not statistically significant, reported Dr. Theobald of Vanderbilt University and her associates.
These events included adverse drug events (0.43 per patient in the intervention group and 0.40 per patient in the control group) and potential adverse drug events (0.44 per patient in the intervention group and 0.55 in the control group), rates of which did not differ significantly between groups.
"If we can’t figure out a way to talk to our patients immediately after discharge, these problems will continue," one physician in the audience said during the discussion session after the presentation.
Dr. Theobald noted that the Vanderbilt University system still has pharmacists available to counsel patients before discharge, but only at the request of clinicians, not routinely.
At the start of the study, 41% in the intervention group and 42% in the control group were female, and 61% in both groups had acute coronary syndrome. Congestive heart failure was diagnosed in 32% of the intervention group and 31% of the control group, and both diagnoses were present in 7% of the intervention group and 8% of the control group. Before admission, patients in the intervention group were on a median of eight medications, and those in the control group were taking a median of seven medications.
Health literacy levels were considered marginal in 9% of each group, and inadequate in 12% of the intervention group and 9% of the control group. Twelve percent of patients in the intervention group and 11% of patients in the control group had cognitive impairment.
There were suggestions of benefit from the intervention, compared with usual care, in three prespecified subgroups, but these did not reach statistical significance. With the intervention, clinically important medication errors were 32% less likely in patients with inadequate health literacy, 38% less likely in cognitively impaired patients, and 17% less likely in patients treated at Vanderbilt.
Further work would be needed to determine if high-risk subgroups should be targeted for this kind of intervention, Dr. Theobald said.
In general, 13%-17% of patients develop clinically important medication errors after hospitalization, according to previous studies. An estimated 50%-75% of those errors are preventable or ameliorable, other data suggest. Some studies report that postdischarge medication errors may be more common in patients who are older, are on complex regimens of multiple medications, are cognitively impaired, or have low health literacy. Previous trials of interventions to reduce posthospitalization medication errors have produced mixed results.