Original Article
The PROMIS Physical Function item bank was calibrated to a standardized metric and shown to improve measurement efficiency

https://doi.org/10.1016/j.jclinepi.2013.10.024Get rights and content

Abstract

Objective

To document the development and psychometric evaluation of the Patient-Reported Outcomes Measurement Information System (PROMIS) Physical Function (PF) item bank and static instruments.

Study Design and Setting

The items were evaluated using qualitative and quantitative methods. A total of 16,065 adults answered item subsets (n > 2,200/item) on the Internet, with oversampling of the chronically ill. Classical test and item response theory methods were used to evaluate 149 PROMIS PF items plus 10 Short Form-36 and 20 Health Assessment Questionnaire-Disability Index items. A graded response model was used to estimate item parameters, which were normed to a mean of 50 (standard deviation [SD] = 10) in a US general population sample.

Results

The final bank consists of 124 PROMIS items covering upper, central, and lower extremity functions and instrumental activities of daily living. In simulations, a 10-item computerized adaptive test (CAT) eliminated floor and decreased ceiling effects, achieving higher measurement precision than any comparable length static tool across four SDs of the measurement range. Improved psychometric properties were transferred to the CAT's superior ability to identify differences between age and disease groups.

Conclusion

The item bank provides a common metric and can improve the measurement of PF by facilitating the standardization of patient-reported outcome measures and implementation of CATs for more efficient PF assessments over a larger range.

Introduction

What is new?

  • This article describes the development of a comprehensive physical function (PF) item bank. This item bank can improve the measurement of PF by standardizing the metric and enabling short and precise computerized adaptive test assessment through readily available software, thus facilitating the use of patient self-assessment in clinical practice and research.

Measurement of patient-reported outcomes (PROs) in clinical studies has steadily increased in frequency, as has its importance in evaluating therapies and developing treatment plans. The plethora of outcomes tools available today allows for increasing the specificity of measurement over a wide range of domains. However, most widely used PRO tools have well-described shortcomings, including high respondent burden and lack of measurement precision. Moreover, results from different instruments can be hard to compare, which limits the interpretability of PRO data.

To address these shortcomings, the National Institutes of Health funded an initiative to build a comprehensive Patient-Reported Outcomes Measurement Information System (PROMIS) [1], [2]. The PROMIS uses item response theory (IRT) and computerized adaptive tests (CATs), which are believed to be promising solutions to the most important limitations of current measurement tools. An IRT item bank consists of a set of items measuring the same construct and parameters that describe the items' measurement properties [3]. Item banks provide the foundation for CATs, which make it possible to administer the most informative items to an individual respondent [4]. Thus, higher precision can be achieved, while respondent burden can be reduced [5], [6], [7].

One aim of the PROMIS initiative was to build an improved item bank for the Physical Function (PF) construct, which has been evaluated using IRT methods for more than a decade [8], [9], [10], [11], [12]. Items covering a wide range of functioning, from self-care to strenuous activities, have been calibrated using IRT models, and some of the first CATs were developed for PF construct [13], [14]. We presented results from a rigorous evaluation of IRT methods in preparation for development of the PROMIS PF item bank earlier in this journal [6]. The current article reports on the development and initial psychometric evaluation of the PROMIS PF item bank.

Section snippets

Methods

Item bank development and evaluation followed the general PROMIS approach, described in detail elsewhere [2], [6], [15], [16]. Issues specific to the PROMIS PF item bank are discussed in the following sections.

Sample

A total of 52% were female. Age ranged from 18 to 95 years with a mean age of 54 years. A total of 82% were white, 9% black, and 8% multiracial; and 9% Hispanic or Latino. Education ranged from less than high school (3%) to an advanced degree (19%), with 24% having a college degree, 38% some college experience, and 16% a high school diploma. Most reported at least one chronic condition, but most also reported no limitations in carrying out daily activities (see Table A-1 at www.jclinepi.com).

Skewness

Discussion

The PROMIS aims to make a major contribution to improved measurement of PROs. The development of the PF item bank is one part of the project.

An important contribution of PROMIS is its extensive qualitative work. The literature search used to build the PF item bank [17] defines this construct based on the body of instruments that have emerged over the last three decades. The resulting item bank contributes to the long-term goal of substituting an instrument-defined measurement system with a

Acknowledgments

The Patient-Reported Outcomes Measurement Information System (PROMIS; www.nihpromis.org) is a National Institutes of Health (NIH) Roadmap initiative to develop a computerized system measuring patient-reported outcomes in respondents with a wide range of chronic diseases and demographic characteristics. The opening PROMIS initiative was funded by cooperative agreements to a Statistical Coordinating Center (Northwestern University, Principal Investigator [PI]: David Cella, PhD, U01AR52177) and

References (40)

  • J.E. Ware et al.

    Practical implications of item response theory and computerized adaptive testing: a brief summary of ongoing studies of widely used headache impact scales

    Med Care

    (2000)
  • J.B. Bjorner et al.

    Calibration of an item pool for assessing the burden of headaches: an application of item response theory to the headache impact test (HITTM)

    Qual Life Res

    (2003)
  • C.V. Granger et al.

    Performance profiles of the functional independence measure

    Am J Phys Med Rehabil

    (1993)
  • W.P. Fisher et al.

    Equating the MOS SF36 and the LSU HSI physical functioning scales

    J Outcome Meas

    (1997)
  • F. Wolfe et al.

    Development and validation of the health assessment questionnaire II: a revised version of the health assessment questionnaire

    Arthritis Rheum

    (2004)
  • S.M. Haley et al.

    Activity outcome measurement for postacute care

    Med Care

    (2004)
  • J.E. Ware et al.

    Item response theory and computerized adaptive testing: implications for outcomes measurement in rehabilitation

    Rehabil Psychol

    (2005)
  • B.B. Reeve et al.

    Psychometric evaluation and calibration of health-related quality of life item banks: plans for the Patient-Reported Outcomes Measurement Information System (PROMIS)

    Med Care

    (2007)
  • D.A. Dewalt et al.

    Evaluation of item candidates: the PROMIS qualitative item review

    Med Care

    (2007)
  • B. Bruce et al.

    Better assessment of physical function: item improvement is neglected but essential

    Arthritis Res Ther

    (2009)
  • Cited by (281)

    View all citing articles on Scopus
    View full text