Exploring the use of patient-reported outcomes in studies of patients with rare diseases: a systematic survey of the literature

Article type
Authors
Mulla S1, Miller P1, Agarwal A1, Khokhar R1, Mohiuddin M2, Sadeghirad B3, Couban R1, Adams-Webber T4, Crawford M4, Guyatt G1, Johnston B5
1McMaster University, Canada
2American Academy of Orthopaedic Surgeons, United States
3Kerman University of Medical Sciences, Iran
4The Hospital for Sick Children, Canada
5SickKids Research Institute, Canada
Abstract
Background:
Patient-reported outcomes (PROs) are reports of a patient’s health condition that come directly from the patient without interpretation by a clinician or other proxy. Systematic reviews and meta-analyses of clinical trials of treatments often consider information from PROs. Confidence in PRO results requires the use of instruments with established psychometric properties, i.e. validity, reliability, and responsiveness (sensitivity to change).

Objectives:
To explore the use of PROs in peer-reviewed literature involving patients with six rare diseases: Fabry Disease, Gaucher Disease, Mucopolysaccharidoses Types I and II, Niemann-Pick Type B, and Pompe Disease. To investigate the extent to which the PROs used within studies are psychometrically sound.

Methods:
We systematically searched six databases (CENTRAL, MEDLINE, EMBASE, CINAHL, PsychINFO, and Web of Science) for experimental and non-experimental studies, and studies that reported on the development of PROs for the rare diseases of interest. Pairs of trained reviewers independently screened potentially eligible articles and subsequently extracted data from the eligible studies.

Results:
Our search yielded 2028 studies; 996 underwent full text screening. Of the 141 studies that met our eligibility criteria and underwent data extraction, the majority (n = 74) involved patients with Fabry Disease. The SF-36 was the most frequently utilized PRO (n = 48). Pain, the most common symptom examined, was measured using the Brief Pain Inventory in 29 studies. Only four studies addressed the development and testing of condition-specific PROs. There were 37 studies in which investigators used a PRO without established psychometric properties.

Conclusions:
The majority of researchers have relied on generic instruments to evaluate quality of life and symptoms in patients with rare diseases. Investigators are wise to consider the use of condition-specific instruments in addition, but the use of an untested instrument is problematic. Additional research to develop and investigate the psychometric properties of PROs used in studies with patients with rare diseases is warranted.