BACKGROUND/RATIONALE:
Patient transfer between sites of care is associated with prescribing lapses, including medication discrepancies and adverse drug events (ADEs). To improve medication safety for patients transferred between healthcare sites, the Joint Commission made medication reconciliation a 2006 hospital patient safety goal.

OBJECTIVE(S):
1) To conduct a controlled trial to evaluate the effectiveness of a CPRS medication reconciliation tool compared with usual care for veterans transferred between sites of care; 2) To identify system and provider factors that impede and facilitate use of the tool; 3) To characterize providers' prescribing processes at the time of patient transfer and to identify providers' residual unmet decision-support needs.

METHODS:
This was a mixed quantitative and qualitative evaluation study. In the quantitative study we compared use of the medication reconciliation tool compared with usual care among veterans admitted to the 2 general medical units at the J.J. Peters VA between October 1, 2005 and February 28, 2006. Medication reconciliation was implemented in a unit-by-unit staggered fashion in anticipation of hospital-wide implementation by January 1, 2006. The study design permitted detection of changes in usual care over time, and concurrent and pre-/post- comparisons for intervention effectiveness. We examined differences between intervention and control in medication discrepancy risk as primary outcome, and occurrence of ADE, provider drug prescribing, and healthcare use as secondary outcomes. All measurements were performed retrospectively by chart review. In the qualitative evaluation, we recorded providers' actions and verbalized thoughts as they applied the tool to test cases, and performed focus group interviews with representative tool users to identify factors that facilitated or hindered use.

FINDINGS/RESULTS:
For the quantitative study (n=795 hospital admissions) medication reconciliation was significantly associated with fewer ADEs caused by medication changes that were prescribing errors (odds ratio (OR) 0.57; 95% CI 0.34-0.98; p=0.041), but not overall ADEs caused by medication changes (OR 1.04; 95% CI 0.68-1.61; p=0.855). Medication reconciliation was marginally associated with lower medication discrepancy risk (coefficient -0.31; 95% CI -0.64-0.01; p=0.057). For the qualitative study, internal medicine house staff physicians (n=23) and inpatient staff pharmacists (n=12) were interviewed. Participants agreed that a central goal of medication reconciliation is to prevent prescribing errors, but disagreed about whether it achieves this goal. Computerization facilitated the task, but computers and patients can be unreliable sources of information. When time was limited, physicians considered other responsibilities higher priority. Both physicians and pharmacists expressed low self-efficacy; i.e., low perceived capability to achieve the process' objectives.

IMPACT:
The study has filled important gaps in knowledge. First, medication reconciliation upon hospital admission is associated with a reduction in ADEs caused by medication changes that are prescribing errors but not overall ADEs caused by medication changes. Second, key barriers to medication reconciliation are unreliable sources of medication information and tasks that compete for providers' time and attention. Study findings provide a clearer picture of the size and type of benefit that can be expected from medication reconciliation, while showing that addressing barriers might improve the process and its outcomes.

---

External Links for this Project

---

PUBLICATIONS:

Journal Articles

1. Kushniruk AW, Santos SL, Pourakis G, Nebeker JR, Boockvar KS. Cognitive analysis of a medication reconciliation tool: applying laboratory and naturalistic approaches to system evaluation. Studies in health technology and informatics. 2011 Jun 22; 164:203-7. [view]
2. Boockvar KS, Blum S, Kugler A, Livote E, Mergenhagen KA, Nebeker JR, Signor D, Sung S, Yeh J. Effect of admission medication reconciliation on adverse drug events from admission medication changes. Archives of internal medicine. 2011 May 9; 171(9):860-1. [view]
3. Boockvar KS, Santos SL, Kushniruk A, Johnson C, Nebeker JR. Medication reconciliation: barriers and facilitators from the perspectives of resident physicians and pharmacists. Journal of hospital medicine. 2011 Jul 1; 6(6):329-37. [view]

Conference Presentations

1. Boockvar K. Effect of admission medication reconciliation on adverse drug events in hospitalized Veterans. Poster session presented at: VA HSR&D National Meeting; 2011 Feb 16; Washington, DC. [view]
2. Boockvar K. Medication management during transitions. Paper presented at: Nurse Practitioner Association of New York State Annual Conference; 2011 Oct 29; Saratoga Springs, NY. [view]
3. Boockvar KS. Medication reconciliation for reducing adverse drug events. Paper presented at: VA Bronx VAMC Annual Research Week; 2007 May 14; Bronx, NY. [view]
4. Boockvar KS. Medication reconciliation for reducing drug discrepancy adverse events. Paper presented at: VA Career Development Award Program Annual Conference; 2007 Feb 21; Arlington, VA. [view]
5. Boockvar KS. Medication reconciliation in a computerized system: study of provider and computer interface. Presented at: American Geriatrics Society Annual Meeting; 2008 May 4; Washington, DC. [view]
6. Boockvar K. Optimizing medication reconciliation during care transitions. Paper presented at: American Medical Directors Association Annual Symposium; 2012 Mar 8; San Antonio, TX. [view]
7. Nebeker JR, Boockvar KS, Kushniruk A, Santos S, Signor D. Providers approach to medication reconciliation in a computerized system. Poster session presented at: VA HSR&D National Meeting; 2009 Feb 12; Washington, DC. [view]
8. Boockvar K. Providers' approach to medication reconciliation in a computerized system. Poster session presented at: VA Career Development Award Program Annual Conference; 2009 Feb 12; Baltimore, MD. [view]

DRA: Health Systems Science
DRE: Prevention
Keywords: Adverse events, Informatics, Safety
MeSH Terms: none