This study assesses whether an electronic medication management support system improves quality, safety, and cost-effectiveness of the drug therapy in adult patients with polypharmacy compared to usual care during the observation period of 15 months from baseline per practice (from 2017 4th quarter to 2020 3rd quarter).

Title:

Application of an electronic medication management support system.

Objectives:

The aim of the study is to assess the effectiveness of an electronic medication management support system on hospitalization and death, its cost-effectiveness and its effects on the quality and safety of prescribing in patients with polypharmacy.

Methods:

A parallel, cluster-randomised controlled trial will be conducted in about 1,080 general practices located in the Westphalia-Region (Germany), which care for adult patients covered by BARMER health insurance. Per group, 17,200 clustered-randomized patients (about 32 patients per practice) will be included, if they take ≥ 5 long-term prescriptions (at least 2 consecutive quarters).

In the intervention group, family physicians will use an user-initiated clinical decision support system (CDSS) which provides drug-therapy relevant information (e.g. on diagnoses and treatments) and alerts in case of drug-drug, drug-disease and drug-age interactions. Based on that information, the family physician systematically assesses the appropriateness of patient's medication and optimizes it. Patients receive an updated medication plan in their mother's language, if needed.

In the control group, patients continue to receive usual care where there is no access to the CDSS.

The primary outcome is the combined endpoint of all-cause mortality and all-cause hospital admissions after the observation period of 12 quarters or 15 months per practice. The outcome measures will be based on secondary data (i.e. claims data). Primary and secondary outcomes will be measured at patient level. The primary analysis will be performed adhering to the intention-to-treat (ITT) principle.

Amendment (07/04/2020):

The parallel group comparison will be conducted, as planned. However, the design of the primary analysis will be changed into a cluster-randomised stepped wedge design with an open cohort, as the recruitment to the target of N=1,080 practices was not reached. The changes of the design will allow to safe the power of the trial of 80% to detect differences of the combined primary outcome between both groups. The primary outcome as well as the ITT-analysis were not changed. The changes of the design implied the following subsequent adaptations of the protocol:

1. Arms and assigned interventions:

   The GP practices of the intervention group will continue to provide the experimental intervention as described in the experimental arm for the entire time in the trial.

   The GP practices of the control arm (usual care, no intervention) will switch to the experimental arm after five quarters control time and provide the experimental intervention for the remaining time in the trial.

2. Case finding:

   The BARMER creates lists of potentially eligible patients ("Potenzialpatienten") in GP practices participating in the AdAM trial based on the fulfillment of inclusion/exclusion criteria in claims data at each quarter between the first quarter 2018 and the third quarter of 2020.

   Intervention practices receive the lists quarterly from the quarter after randomization and control practices quarterly starting on the date of switch.

3. Informed consent:

The GPs in practices of the experimental arm (either they were randomized and assigned as intervention practices at the beginning of the trial or they switch after the end of five quarters control time) invite potentially eligible patients from the lists and ask for a written informed consent.

Inclusion Criteria:

• Patients ≥ 18 years of age with polypharmacy (5 or more drugs taken for at least two quarters).
• Informed consent.

Exclusion Criteria:

• Incapacitated patients without legal representative

Due to the character of the intervention, blinding is not possible.