Collection of Multiple Patient-Reported Outcome Measures (CRAM-PROMs) in orthopaedic trauma: a randomized trial to assess the impact of quantity on quality - 29/11/24

Doi : 10.1016/j.otsr.2024.104076

Francesc A. Marcano-Fernández ^a,⁎ , Carlos Prada ^b, Sheila Sprague ^b,^c, Sofia Bzovsky ^b, Jodi Gallant ^b, Gina Del Fabbro ^b, Herman Johal ^b
^a Department of Orthopaedic Surgery, Parc Taulí Hospital Universitari. Institut d’Investigació i Innovació Parc Taulí (I3PT-CERCA), Universitat Autònoma de Barcelona, Sabadell, Spain
^b Division of Orthopaedic Surgery, Department of Surgery, McMaster University, Hamilton, Ontario, Canada
^c Department of Health Research Methods, Evidence, and Impact, McMaster University, Hamilton, Ontario, Canada

^⁎Corresponding author.

En prensa. Pruebas corregidas por el autor. Disponible en línea desde el Friday 29 November 2024

Abstract

Background

We rely on sound clinical research to ensure the safety and effectiveness of medical interventions. We are nowadays experiencing an increased utilization of Patient Reported Outcome Measures (PROMs) in research, for which there is good evidence on validity and reliability. However, these measures are often validated in isolation and not in a real-life setting where they are used in conjunction with other questionnaires and measures. Our study objective was to determine if the number of PROMs questionnaires completed by fracture patients affected their reliability; more specifically, the internal consistency of the EuroQuol Five Dimension-Five Level (EQ-5D-5L) questionnaire.

Hypothesis

Our hypothesis was that there is a progressive decline in reliability as we increase the number of questionnaires given to participants.

Patients and method

This is a randomized study carried out in a single Level I academic trauma center. Patients presenting to the fracture clinic for follow-up were screened. Participants were randomly assigned to 4 groups: group 1 only answered the EuroQol Five Dimension–Five Level (EQ-5D-5L) questionnaire, group 2 answered one additional questionnaire before the EQ-5D-5L, group 3 answered two and group 4 answered three additional questionnaires before the EQ-5D-5L. The primary outcome measured was the internal consistency of the EQ-5D-5L as a measure of reliability.

Results

Enrolment was 115 participants. Twenty-eight participants were randomized to Group 1, 29 to Group 2, 29 to Group 3, and 29 to Group 4. There was a progressive decline of reliability as number of questionnaires increased except for a sudden rise in group 4. (Group 1: 0.83, 95% Confidence Interval (CI) 0.70 to 0.91; Group 2: 0.74, 95% CI 0.56 to 0.87; Group 3: 0.68, 95% CI 0.44 to 0.83; Group 4: 0.81, 95% CI 0.68 to 0.90). Completeness was 100% for Group 1, 98.5% for Group 2, 100% for Group 3, and 92% for Group 4.

Discussion

These results cannot strongly support our initial hypothesis. Although there is an initial decrease in Cochran’s alpha for Groups 2 and 3 consecutively, the sudden rise in group 4 limits the validity our results. Notwithstanding, researchers should consider these findings when designing their research trials to avoid potential misleading results. Hence, the number of given questionnaires in research should be limited in order for these to maintain their reliability.