Our evidence review on the use of health information technology (health IT) in enabling medication management derives from a summary of 428 studies assessing the use of integrated health IT that assisted with at least one phase of the medication management process, and associated aspects of education and reconciliation. We define the medication management process as having five phases; prescribing, order communication, dispensing, administering, and monitoring. For this report we also included medication reconciliation, education, and adherence.

Key Question (KQ) 1: Effectiveness assessed the effectiveness of health IT on changes in process and intermediate, cost, and clinical outcomes. We limited our studies to those that used qualitative methods or included comparison groups, hypothesis testing, and appropriate statistical analysis. We did this in an effort to limit inclusion to studies that used research methods and had data that we could use to draw conclusions. The majority (378 of 428 studies) of the evaluated studies were included in KQ1: Effectiveness. Even so, our findings indicate that RCTs and other methodologies with controlled populations are lacking in adequate details and robust methods, which result in an only incremental addition to the evidence base for the use of MMIT. Most of the studies in this evidence report are quantitative observational assessments, often using historical controls.

The evidence from these studies indicate positive effects on improving process, often measured as improvements in medication orders during the prescribing and monitoring phases. The bulk of this evidence of improvement is shown in studies set in hospitals. We also found improvements associated with MMIT systems related to use, usability, knowledge, skills, and attitudes. These cumulated changes can, but may not always, lead to efficiency and cost gains. On the other hand, little work is being done on the other phases of medication management with integrated health IT. Some IT applications used in dispensing and administering are stand-alone technologies and, by definition, not included in this report.

We found little evidence of significant effects on clinical outcomes; possibly because of the small number of events, the outcomes being far removed from the application of the technology; or that they were often not the main endpoints of the studies included in the review. We do not know if MMIT applications are clinical- and cost effective because of a lack of sound economic data.

The qualitative literature highlighted positive and negative perceptions and satisfaction with the integrated health IT applications, supporting much of the literature on the importance of the effects of the technologies on workflow and the working relationships of the users.

We have identified a number of gaps in the evidence of the effects of MMIT applications: most notably the order communication, dispensing, and administering, as well as reconciliation and education. Inpatient care is well-studied, followed by ambulatory care. A low number of studies assessed long-term care and effects on pharmacies, especially those outside the hospital setting. Gaps in research also exist for studies that evaluate MMIT that are not computerized decision support systems (CDSSs) or computerized provider order entry (CPOE) systems. The domain of patient and informal caregiver access to MMIT applications, especially those applications that are integrated with such existing clinical applications as electronic medical record systems (EMRs), will be an exciting and promising new domain of study. One major gap is the assessment of MMIT tools that are used by nonphysicians.

The value of the MMIT systems needs to be summed up across financial, clinical, and organizational components. The values proposition for each stakeholder will be different based on their own value set, and what is important to each has not been well-studied. Though some evidence suggests positive financial and organizational gains, these gains are not universal and will depend on the technology, the setting, and the impact on the stakeholders using them. Clinical benefit is proving difficult to assess. Rigorous studies are needed to truly assess economic and other values.

This broad scoped review sought to include a large number of health ITs across an array of settings and users. As such, the literature was heterogeneous and difficult to synthesize. Based on our findings, we feel that it is important to note that the burden of proof of the value of MMIT is somewhat limited, but promising. We feel also that decisionmakers must be aware of the potential for negative impacts of the technologies and carefully consider these possibilities during implementation and provide for continued monitoring across all stakeholder groups. . The evidence base of MMIT applications is strong and varied, and it can be further strengthened by using multicentered studies, building an integrated body of evidence which will demonstrate the transferability and applicability of the systems, and multidisciplinary teams of researchers, or at least consultation input from clinicians; methodologists, including biostatisticians; informaticians; particular health IT users and representatives of various stakeholders. We can also learn much from those who work with complex interventions, pragmatic trials, research syntheses, knowledge translation or translation research challenges, and health technology assessment studies to enhance the construction, conduct, and communication of health IT implementation and use research findings.