Validation of a Patient-Level Medication Regimen Complexity Index as a Possible Tool to Identify Patients for Medication Therapy Management Intervention

Authors

  • Jan D. Hirsch,

    Corresponding author
    1. Skaggs School of Pharmacy and Pharmaceutical Sciences, University of California, San Diego, La Jolla, California
    2. Veterans Affairs of San Diego Healthcare System, San Diego, California
    • Address for correspondence: Jan D. Hirsch, 9500 Gilman Drive MC 0714, La Jolla, CA 92093; e-mail: janhirsch@ucsd.edu.

    Search for more papers by this author
  • Kelli R. Metz,

    1. Department of Clinical Pharmacy, Skaggs School of Pharmacy and Pharmaceutical Sciences, University of Colorado Anschutz Medical Campus, Aurora, Colorado
    Search for more papers by this author
  • Patrick W Hosokawa,

    1. Colorado Health Outcomes Program, School of Medicine, University of Colorado Anschutz Medical Campus, Aurora, Colorado
    Search for more papers by this author
  • Anne M. Libby

    1. Department of Clinical Pharmacy, Skaggs School of Pharmacy and Pharmaceutical Sciences, University of Colorado Anschutz Medical Campus, Aurora, Colorado
    2. Colorado Health Outcomes Program, School of Medicine, University of Colorado Anschutz Medical Campus, Aurora, Colorado
    Search for more papers by this author

Abstract

Background

The Medication Regimen Complexity Index (MRCI) is a 65-item instrument that can be used to quantify medication regimen complexity at the patient level, capturing all prescribed and over-the-counter medications. Although the MRCI has been used in several studies, the narrow scope of the initial validation limits application at a population or clinical practice level.

Purpose

To conduct a MRCI validation pertinent to the desired clinical use to identify patients for medication therapy management interventions.

Methods

An expert panel of clinical pharmacists ranked medication regimen complexity for two samples of cases: a single-disease cohort (diabetes mellitus) and a multiple-disease cohort (diabetes mellitus, hypertension, human immunodeficiency virus infection, geriatric depression). Cases for expert panel review were selected from 400 ambulatory clinic patients, and each case description included data that were available via claims or electronic medical records (EMRs). Construct validity was assessed using patient-level MRCI scores, medication count, and additional patient data. Concordance was evaluated using weighted κ agreement statistic, and correlations were determined using Spearman rank-order correlation coefficient (ρ) or Kendall τ.

Results

Moderate to good concordance between patient-level MRCI scores and expert medication regimen complexity ranking was observed (claims data, consensus ranking: single-disease cohort 0.55, multiple disease cohort 0.63). In contrast, only fair to moderate concordance was observed for medication count (single-disease cohort 0.33, multiple-disease cohort 0.48). Adding more-detailed administration directions from EMR data did not improve concordance. MRCI convergent validity was supported by strong correlations with medication count (all cohorts 0.90) and moderate correlations with morbidity measures (e.g., all cohorts; number of comorbidities 0.46, Chronic Disease Score 0.46). Nonsignificant correlation of MRCI scores with age and gender (all cohorts 0.08 and 0.06, respectively) supported MRCI divergent validity.

Limitations

This study used cross-sectional, retrospective patient data for a small number of patients and clinical pharmacists from only two universities; therefore, results may have limited generalizability.

Conclusions

The patient-level MRCI is a valid tool for assessing medication regimen complexity that can be applied by using data commonly found in claims and EMR databases and could be useful to identify patients who may benefit from medication therapy management.

Ancillary