Psychometric Properties of the Patient Reported Outcomes Measurement Information System (PROMIS) Scales in Acute Intermittent Porphyria Patients

Background: Acute Intermittent Porphyria (AIP) is a rare inborn error of heme biosynthesis characterized clinically by life-threatening acute neurovisceral attacks. Few studies have assessed quality of life (QoL) tools in the AIP population, and none are validated for use in AIP. In this study, the Patient Reported Outcomes Measurement Information System 57 (PROMIS-57) scales were assessed in AIP patients to determine the factor structure of PROMIS items and to compare this to previously proposed factor structures in the general population.Methods: Baseline data from the Porphyrias Consortium’s Longitudinal Study of the Porphyrias was obtained for 259 AIP patients. This included detailed disease and medical history information as well as PROMIS-57 data. Exploratory and confirmatory factor analyses (EFA and CFA, respectively) were conducted on baseline PROMIS-57 data. CFA was used to test the hypothesized three-factor structure from the PROMIS literature. Results: Internal consistency was high for all the PROMIS scales assessed. The EFA revealed a five-factor model, each consisting of a separate domain: pain interference, anxiety, depression, sleep disturbance, and fatigue. Model fit was good overall. A CFA of the hypothesized three-factor model did not converge suggesting an inappropriate structural model. Conclusion: The EFA showed a five-factor model that fit the data well, however a CFA of the three-factor model observed in the general population did not fit the data well. These findings, together with previous studies assessing correlations with clinical features, indicate that the PROMIS scales may be valid and reliable measures for AIP patients. However, further analyses are needed to confirm this. Further studies should assess correlations with other tools, whether the PROMIS scales are responsive to treatment, if they capture QoL longitudinally, and a CFA in a second sample.