Rating recommendations for consumers about patient safety: sense, common sense, or nonsense?

BACKGROUND: Although many organizations offer advice about the consumer's role in improving patient safety, little is known about these recommendations. METHODS: The Internet and medical literature were searched to identify patient safety recommendations for consumers. Recommendations were classified by type and tabulated by frequency. Nine investigators rated each recommendation for the quality of supporting empirical evidence, magnitude of benefit, and likely patient adherence. For a consumer perspective, 22 relatives of the investigators who were also mothers rated each recommendation. RESULTS: Twenty-six organizations identified 160 distinct recommendations; 115 (72%) addressed medication safety, 37 (23%) advised patients about preparation for hospitalization or surgery, and 18 (11%) offered general advice. Organizations most frequently advised patients to make a list of their medications (92% of organizations), to ask questions about their health and treatment (81%), to enlist an advocate (77%), and to learn about possible medication side effects (77%). Investigators assigned high scores to 11 of the 25 most frequently cited recommendations and to 4 of the 25 least common recommendations. There was little association between the frequency with which recommendations were promulgated and investigators' ratings (r = 0.27, p < .001). Investigators' scores correlated with those of the mothers (r = 0.71, p < .001). DISCUSSION: Contrary to expectation, there was little overlap among the 160 recommendations offered by the 26 organizations. Health care organizations offer many patient safety recommendations of limited value. These organizations should offer a concise and coherent set of recommendations on the basis of evidence, magnitude of benefit, and likely adherence.

Medication reconciliation in ambulatory oncology.

BACKGROUND: Few models for medication reconciliation in ambulatory primary or specialty care have been described, perhaps because of the special challenges posed by this environment. METHODS: Dana-Farber Cancer Institute (Boston) created a reconciliation program that was designed as a patient-clinician partnership intervention. Policies that require clinicians to review and update medication lists at regular appointments were augmented. Clinic assistants printed patients' medication lists from the electronic medical record and distributed lists to established patients for review. Patients provided updated lists to their oncology clinicians. Clinicians then entered the information or indicated changes to be entered by a pharmacist. RESULTS: At baseline, 81% of patients' medication lists included at least one error or omission. With medication reconciliation, 90% of incorrect medication lists were updated. In contrast, only 2% of medication lists were corrected among patients who received "usual" care (p < .001). DISCUSSION: From the program's inception in November 2005 through August 2007, patients and staff reconciled 24,148 medication lists by making 53,040 changes to 168,475 listed drugs, a rate of 31 changes per 100 medications. Implementation required broad staff engagement and ongoing attention to operational issues.