Combining Scores from Different Patient Reported Outcome Measures in Meta-Analyses
Author Information
Author(s): Puhan Milo A, Soesilo Irene, Guyatt Gordon H, Schünemann Holger J
Primary Institution: Horten Centre, University of Zurich, Switzerland
Hypothesis
Is it justified to combine scores from different patient-reported outcome measures in meta-analyses?
Conclusion
Pooling results from different instruments in meta-analysis should be approached with caution due to differences in responsiveness.
Supporting Evidence
- The correlation between CRQ and SGRQ change scores was found to be 0.88.
- Standardized response means of the CRQ were significantly higher than for the SGRQ.
- High correlation does not imply similar responsiveness between the two instruments.
- Pooling results from different instruments can lead to biased meta-analyses.
Takeaway
This study looked at two questionnaires used for patients with lung disease and found that even if they seem to measure the same thing, they can show different results.
Methodology
The study systematically reviewed longitudinal studies using both the Chronic Respiratory Questionnaire (CRQ) and the St. George's Respiratory Questionnaire (SGRQ) to assess their correlation and responsiveness.
Potential Biases
There is a risk of underestimating treatment effects if less responsive instruments are used in meta-analyses.
Limitations
The study lacked individual patient data to explore the association between CRQ and SGRQ change scores in greater detail.
Participant Demographics
Patients with chronic obstructive pulmonary disease (COPD).
Statistical Information
P-Value
< 0.001
Confidence Interval
95% CI 0.62–1.00
Statistical Significance
p < 0.001
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website