Review Article
Recommended methods for determining responsiveness and minimally important differences for patient-reported outcomes

https://doi.org/10.1016/j.jclinepi.2007.03.012Get rights and content

Abstract

Objective

The objective of this review is to summarize recommendations on methods for evaluating responsiveness and minimal important difference (MID) for patient-reported outcome (PRO) measures.

Study Design and Setting

We review, summarize, and integrate information on issues and methods for evaluating responsiveness and determining MID estimates for PRO measures. Recommendations are made on best-practice methods for evaluating responsiveness and MID.

Results

The MID for a PRO instrument is not an immutable characteristic, but may vary by population and context, and no one MID may be valid for all study applications. MID estimates should be based on multiple approaches and triangulation of methods. Anchor-based methods applying various relevant patient-rated, clinician-rated, and disease-specific variables provide primary and meaningful estimates of an instrument's MID. Results for the PRO measures from clinical trials can also provide insight into observed effects based on treatment comparisons and should be used to help determine MID. Distribution-based methods can support estimates from anchor-based approaches and can be used in situations where anchor-based estimates are unavailable.

Conclusion

We recommend that the MID is based primarily on relevant patient-based and clinical anchors, with clinical trial experience used to further inform understanding of MID.

Introduction

Patient-reported outcomes (PROs) are frequently incorporated in clinical trials comparing health interventions for chronic diseases. These PROs include measures of health-related quality of life (HRQL), symptoms, and treatment satisfaction. PROs provide the patient's perspective and help us understand the effects of disease and treatment on symptoms, functioning, and other outcomes [1], [2], [3]. For many chronic diseases, PROs represent one of the most important health outcomes for evaluating the effectiveness of treatments and changes in disease trajectory. As far back as Hippocrates, listening to the patient has been considered an integral part of medical science [4]. Therefore, the patient's perspective of her health is integral to understanding health outcomes. The application of relevant and psychometrically sound PROs in clinical trials assists patients, their family members, and clinicians in understanding the comprehensive impact of treatment on patient symptoms, functioning, treatment preferences, and general well being.

To be useful in clinical trials evaluating new health interventions, PROs, similar to other health outcomes, must have acceptable reliability and validity [1], [2], [5], [6]. Responsiveness is an aspect of construct validity [7] and is determined by evaluating the relationship between changes in clinical and patient-based endpoints and changes in the PRO scores over time, or based on the application of a treatment of known and demonstrated efficacy [2], [5], [8]. Responsiveness can be evaluated based on observational studies or in clinical trials. Evidence supporting responsiveness and for interpreting PRO results is critical for clinical trial settings. Information on the interpretation of changes or differences in PRO scores is based on the minimal important difference (MID). Demonstrating a MID is also important evidence for achieving successful PRO claims through regulatory agencies [9], [10]. Nonetheless, virtually all instruments found to differentiate among clinically distinct groups are also found to be responsive to change.

Although responsiveness and interpretation of PRO measures have been discussed for the past 15 years or more [2], [5], [7], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], recommendations about the best approach for evaluating responsiveness and determining MIDs for PRO instruments are still needed. For example, the FDA requested further information and guidance on methods for determining responsiveness and MID [9]. Although there is an evolving consensus as to the best approach to evaluating responsiveness and MID [25], [26], [29], there is no clear statement about the recommended methods and about important issues underlying responsiveness and MID.

This report focuses on issues and recommendations for evaluating responsiveness and MID for PRO measures in chronic disease. These issues are especially germane given that for most chronic diseases cure is not feasible, and that the main objective of treatment is to maintain or improve patient functioning and well being. The remainder of this report will cover (1) conceptual issues and definitions; (2) methods for evaluating responsiveness and MID; (3) recommended decision criteria for determining MID; and (4) summary and conclusions. We will illustrate methods and concepts using published health outcomes literature.

Section snippets

Interpretation of PROs: conceptual issues and definitions

PROs require the patient to assign a response to questions (or statements) about their perceptions or activities, such as symptoms, capabilities, or performance of roles or responsibilities. These responses are typically combined in some way to create summary scores that can be used to measure concepts such as physical, psychological, or social functioning and well being, or symptom burden or severity. Symptoms can be rated based on frequency, severity, duration, degree of bother, or impact on

Methods of evaluating responsiveness and clinical significance

Longitudinal studies are needed to determine whether a PRO instrument is responsive to changes or differences. These studies may be randomized clinical trials comparing treatments of known efficacy or observational studies where patients are treated with usual medical care and followed over relevant periods of time. For clinical trial designs, there needs to be some evidence that the treatment is effective and that the expected changes in clinical status are linked to expected changes in the

Recommended decision criteria for determining MID

The application of multiple methods to determine the MID for a PRO instrument in a specific patient population will almost always result in a range of values for the MID. This is the essence of triangulation, that is, examining multiple values from different approaches and hopefully converging on a small range of values (or one single value). It is recommended that the different MID estimates be graphed to visually depict the range of estimates. Figure 1 provides a summary of MID estimates from

Summary and conclusions

For PRO endpoint data to be accepted as evidence of treatment efficacy there must be evidence documenting the instrument's conceptual framework, content validity, and psychometric qualities. For responsiveness, it is necessary to demonstrate that the PRO scores are sensitive to actual changes in health status. Although demonstrating responsiveness is a key component to establishing an instrument's construct validity, it is also important to determine the MID to assist in interpreting

Acknowledgments

This paper was supported in part by Genentech, South San Francisco, California, the UCLA/DREW Project EXPORT, National Institutes of Health, National Center on Minority Health & Health Disparities, (P20-MD00148-01), the UCLA Center for Health Improvement in Minority Elders/Resource, Centers for Minority Aging Research, National Institutes of Health, National Institute of Aging, (AG-02-004), and the National Institute of Aging (AG20679-01).

References (52)

  • E.F. Juniper et al.

    Determining a minimal important chance in the disease-specific quality of life questionnaire

    J Clin Epidemiol

    (1994)
  • K. Niebauer et al.

    Impact of omalizumab on quality-of-life outcomes in patients with moderate-to-severe allergic asthma

    Ann Allergy Asthma Immunol

    (2006)
  • K.W. Wyrwich et al.

    Further evidence supporting an SEM-based criteria for identifying meaningful intra-individual changes in health-related quality of life

    J Clin Epidemiol

    (1999)
  • D.A. Revicki et al.

    Recommendations on health-related quality of life research to support labeling and promotional claims in the United States

    Qual Life Res

    (2000)
  • R.J. Wilke et al.

    Measuring treatment impact: a review of patient-reported outcomes and other efficacy endpoints in approved product labels

    Contr Clin Trials

    (2004)
  • Hippocrates. On decorum. Hippocrates, with an English translation. In: Jones WH, translator. Cambridge, MA: Harvard...
  • R.D. Hays et al.

    Reliability and validity (including responsiveness)

  • P.M. Fayers et al.

    Quality of life: Assessment, analysis and interpretation

    (2000)
  • R.D. Hays et al.

    Responsiveness to change: an aspect of validity, not a separate dimension

    Qual Life Res

    (1992)
  • FDA

    Guidance for industry—patient-reported outcome measures: Use in medical product development to support labeling claims

    (2006)
  • Committee for Medicinal Products for Human use

    Reflection paper on the regulatory guidance for the use of health-related quality of life (HRQL) measures in the evaluation of medicinal products

    (2005)
  • N.S. Jacobson et al.

    Clinical significance: a statistical approach to defining meaningful change in psychotherapy research

    J Clin Consult Psychol

    (1991)
  • M.J. Liang

    Evaluating measurement responsiveness

    J Rheumatol

    (1995)
  • E. Lydick et al.

    Interpretation of quality of life changes

    Qual Life Res

    (1993)
  • M. Testa et al.

    Interpreting pharmacoeconomic and quality-of-life clinical trial data for use in therapeutics

    Pharmacoeconomics

    (1992)
  • D. Osoba et al.

    Interpreting the significance of changes in health-related quality of life scores

    J Clin Oncol

    (1998)
  • Cited by (1596)

    View all citing articles on Scopus
    View full text