Background/Purpose: Patient-reported outcomes are routinely used in rheumatology research and clinical care. Yet, often outcomes cannot be compared across studies or diseases because a variety of measures are used in these assessments, and many important health domains are not assessed because of lack of measures or concerns about questionnaire burden. Further, many “traditional” patient-reported outcomes measures are available only in English, which is an increasingly limiting factor. The NIH Patient-Reported Outcomes Measurement Information System (PROMIS) consists of a group of patient-reported outcome measures that span a wide array of physical, social, and emotional health outcomes; are applicable across health conditions; and are available free of charge and in multiple languages. A test of the PROMIS measures has not been undertaken in patients with RA, OA, or fibromyalgia. This analysis presents an initial psychometric evaluation of PROMIS measures in a large cohort of these patients.
Methods: Data were from a subset of respondents to a single administration of a questionnaire that included the PROMIS 29-item profile. The sample included 528 individuals (RA: 323; OA: 109; fibromyalgia: 96). Using short questionnaires, the PROMIS-29 assesses 7 of the PROMIS health domains: Physical Function, Pain Interference, Fatigue, Depression, Anxiety, Sleep Disturbance, and Ability to Participate in Social Roles. Each section was scored and converted to t-scores, with mean = 50 and SD = 10. Analyses examined correlations of PROMIS measures with scales measuring related constructs (SF-36 subscales, Health Assessment Questionnaire [HAQ], and numeric rating scales for pain, fatigue, and sleep problems), for the total sample and within disease groups. Analyses also examined ability of PROMIS measures to discriminate among levels of satisfaction with health.
Results: PROMIS scales exhibited moderate (r<0.7) correlations with most of the comparison measures, with some correlations slightly higher, indicating that similar constructs were being measured (Table 1). Results were similar for each disease group. All PROMIS scales discriminated among levels of satisfaction with health, yielding significant overall ANOVA results and significant non-parametric tests of trend (Table 2).
Conclusion: These PROMIS short forms exhibited strong psychometric properties. Use of PROMIS offers important expansions of current measures so that aspects of health, social functioning, and quality of life that are important to patients can be included without increasing questionnaire burden.
Table 1. Correlation of PROMIS Scales with Related Measures |
|||||||
|
PROMIS 29-Item Profile scales* |
||||||
|
PF |
PI |
FAT |
DEP |
ANX |
SLP |
SocR |
SF-36 Physical Function |
0.83 |
|
|
|
|
|
|
SF-36 Role Physical |
|
|
|
|
|
|
0.62 |
SF-36 Role Emotional |
|
|
|
|
|
|
0.43 |
SF-36 Vitality |
|
|
-0.77 |
|
|
|
|
SF-36 Bodily Pain |
|
-0.73 |
|
|
|
|
|
SF-36 Mental Health |
|
|
|
-0.69 |
-0.67 |
|
|
HAQ |
-0.74 |
|
|
|
|
|
|
NRS-pain |
|
0.62 |
|
|
|
|
|
NRS-fatigue |
|
|
0.71 |
|
|
|
|
NRS-sleep |
|
|
|
|
|
0.71 |
|
* For all PROMIS scales, higher scores reflect “more” of the construct being measured. E.g., Higher Physical Function scores reflect better functioning; higher Pain Interference scores reflect greater pain interference PF = Physical Functioning; PI = Pain Interference; FAT = Fatigue; DEP = Depression; ANX = Anxiety; SLP = Sleep disturbance; SocR = Ability to Participation in Social Roles. HAQ = Health Assessment Questionnaire |
Table 2. PROMIS 29-Profile Scores* by Levels of Satisfaction with Health |
|||||||
|
PROMIS 29-Item Profile scales* |
||||||
Satisfaction with health: |
PF |
PI |
FAT |
DEP |
ANX |
SLP |
SocRole |
Very satisfied (n=61) |
46.9 |
52.7 |
48.0 |
45.0 |
44.6 |
47.5 |
52.5 |
Somewhat satisfied (n=173) |
41.1 |
58.1 |
53.8 |
47.9 |
47.9 |
51.4 |
45.8 |
Neither satisfied nor dissatisfied (n=81) |
37.9 |
61.1 |
57.9 |
51.3 |
51.3 |
55.0 |
40.9 |
Somewhat dissatisfied (n=139) |
35.8 |
63.6 |
60.9 |
52.7 |
52.8 |
56.0 |
38.3 |
Very dissatisfied (n=89) |
33.0 |
66.7 |
67.2 |
60.3 |
58.8 |
61.7 |
32.2 |
* Table presents mean scores for individuals in each satisfaction with health rating group. Differences were tested with ANOVA and non-parametric test for trend. All were significant p<.001. |
Disclosure:
P. P. Katz,
None;
S. Pedro,
None;
K. Michaud,
None.
« Back to 2014 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/test-of-the-nih-patient-reported-outcomes-measurement-information-system-promis-29-item-profile-in-a-large-cohort-of-rheumatic-disease-patients/