Table 1 Differences between MARQUIS1 and MARQUIS2

Site Selection

How and when sites were recruited

Informal process, sites identified prior to submission of grant application

Widespread search, formal application process, most sites identified at beginning of study period

Toolkit

Best possible medication history (BPMH) training

Didactic materials only, including slide presentations and videos

Didactic materials plus simulation materials with standardized cases and role-playing to enhance learning and verify competency

Role of staff taking BPMH

Agnostic to type of personnel

Increased emphasis on the value of pharmacy technicians as “medication reconciliation assistants” trained to take accurate medication histories

Return on investment

Rudimentary calculations

More precise calculations based on MARQUIS1 data

Patient counseling tools

Didactic materials, including slide presentations and videos

Enhanced didactic materials plus scripts and worksheets developed with Patient and Family Advisory Council (PFAC) input

Implementation Approach

Site team training

Webinars

Webinars + 4 regional workshops

Site visits, number and timing

2 site visits: first visit in months 5–10, second in months 16–19

1 site visit within first 6 months

Patient-family engagement

No formalized program

Established and engaged PFAC in monthly discussions

Inter-site sharing

No formalized sharing

3 peer-to-peer webinars featuring sites’ stories of successes and challenges

Health information technology (HIT)

Discovered significant challenges exist with the design, implementation, and use of HIT during medication reconciliation processes that, together with health systems issues, impacted medication safety

Provided guidance on how best to work with existing HIT—e.g. allowing pharmacists to make changes to medication lists, documenting the quality of the medication history taken and its sources, and customizing discharge instructions to make medication changes clear to patients and clinicians

Analyses

Intervention assessment

Scoring system of interventions; categorization of site-level intervention components based on meeting minutes analyzed retrospectively; no data on receipt of interventions at the patient level

Prospective collection of site-level interventions based on monthly site surveys; prospective collection of patient-level interventions as part of data collection on discrepancies

Outcome assessment

Total medication discrepancies with potential for harm, involving adjudication; total medication discrepancies

Total medication discrepancies per medication per patient, as adopted by the Leapfrog Group [22]

Program evaluation

Surveys, direct observation, interviews, focus groups of contextual factors, intervention fidelity

RE-AIM^a framework