Multilevel interventions: measurement and measures

J Natl Cancer Inst Monogr. 2012 May;2012(44):67-77. doi: 10.1093/jncimonographs/lgs011.

Abstract

Background: Multilevel intervention research holds the promise of more accurately representing real-life situations and, thus, with proper research design and measurement approaches, facilitating effective and efficient resolution of health-care system challenges. However, taking a multilevel approach to cancer care interventions creates both measurement challenges and opportunities.

Methods: One-thousand seventy two cancer care articles from 2005 to 2010 were reviewed to examine the state of measurement in the multilevel intervention cancer care literature. Ultimately, 234 multilevel articles, 40 involving cancer care interventions, were identified. Additionally, literature from health services, social psychology, and organizational behavior was reviewed to identify measures that might be useful in multilevel intervention research.

Results: The vast majority of measures used in multilevel cancer intervention studies were individual level measures. Group-, organization-, and community-level measures were rarely used. Discussion of the independence, validity, and reliability of measures was scant.

Discussion: Measurement issues may be especially complex when conducting multilevel intervention research. Measurement considerations that are associated with multilevel intervention research include those related to independence, reliability, validity, sample size, and power. Furthermore, multilevel intervention research requires identification of key constructs and measures by level and consideration of interactions within and across levels. Thus, multilevel intervention research benefits from thoughtful theory-driven planning and design, an interdisciplinary approach, and mixed methods measurement and analysis.

Publication types

  • Research Support, U.S. Gov't, Non-P.H.S.
  • Review

MeSH terms

  • Cancer Care Facilities / standards
  • Cancer Care Facilities / statistics & numerical data
  • Confounding Factors, Epidemiologic
  • Continuity of Patient Care* / standards
  • Continuity of Patient Care* / trends
  • Delivery of Health Care, Integrated / standards
  • Delivery of Health Care, Integrated / trends
  • Health Services Research* / methods
  • Health Services Research* / trends
  • Humans
  • Interdisciplinary Communication*
  • Neoplasms / diagnosis*
  • Neoplasms / prevention & control
  • Neoplasms / therapy*
  • Outcome Assessment, Health Care
  • Outcome and Process Assessment, Health Care*
  • Patient Care Team / standards
  • Patient Care Team / trends
  • Quality of Health Care / standards
  • Quality of Health Care / trends
  • Research Design*
  • United States