Patient reported outcomes (PROs) are important treatment endpoints in inflammatory bowel diseases (IBD). The gastrointestinal (GI) PRO Measurement Information System (PROMIS) has been validated in the general population. We aimed to assess these measures in an IBD cohort.Methods:
Crohn's and Colitis Foundation of America's Partners is an Internet-based cohort of IBD subjects. We used univariate and bivariate analyses to assess GI-PROMIS domains relative to psychosocial PROMIS domains, validated disease activity indices (short Crohn's disease activity index [sCDAI] and simple clinical colitis activity index [SCCAI]) and quality of life (QoL) indices (short IBD questionnaire [SIBDQ]) in a cross-sectional sample. GI-PROMIS domains were reported as percentiles of severity among symptomatic individuals.Results:
Two thousand three hundred seventy-eight Crohn's Disease (CD) and 1455 ulcerative colitis (UC) respondents had a median age of 41 years. Median disease duration was 11 years for CD subjects and 8 years for UC subjects. Fifty-seven percent of CD subjects and 42% of UC subjects were in remission. All comparisons of symptoms are only among those with the relevant GI symptoms. CD subjects reported diarrhea at the 54th percentile and belly pain at the 55th percentile; UC patients reported diarrhea at the 50th percentile and belly pain at the 49th percentile. Compared to CD subjects who did not have surgery, those who had surgery reported significantly more diarrhea (57th percentile versus 50th percentile, P <0.01), belly pain (58th percentile versus 53rd percentile, P = 0.02), gas (54th percentile versus 51st percentile, P = 0.04) and bowel incontinence (60th percentile versus 50th percentile, P < 0.01). Stratified by surgical status, UC subjects had no significant differences. Compared to CD subjects in remission, those with active disease reported significantly worse symptoms on all 8 GI-PROMIS domains; the same was observed for UC subjects with the exception of disrupted swallowing. CD subjects taking narcotics reported significantly worse symptoms than those not taking narcotics for all 8 GI-PROMIS domains. The same was observed for UC subjects with the exception of disrupted swallowing, bowel incontinence and constipation. Compared to CD subjects not taking steroids, those taking steroids reported more diarrhea (62nd percentile versus 53rd percentile, P < 0.01), nausea and vomiting (65th percentile versus 55th percentile, P = 0.01), belly pain (64th percentile versus 54th percentile, P < 0.01) and gas (62nd percentile versus 52nd percentile, P < 0.01). Compared to UC subjects not taking steroids, those taking steroids reported more diarrhea (65th percentile versus 48th percentile, P < 0.01), belly pain (67th percentile versus 46th percentile, P < 0.01) and gas (60th percentile versus 47th percentile, P < 0.01). Subjects with worse QoL reported significantly worse symptoms on all 8 GI-PROMIS domains compared to those with the best QoL. Subjects with more severe diarrhea, belly pain and gas had significantly poorer psychosocial PROs.Conclusions:
To date, this is the largest description of general GI-PROMIS domains in IBD subjects. IBD subjects did not report worse GI symptoms compared to the symptomatic population. There were strong associations between worse GI-PROMIS scales and narcotic use, steroid use, worse QoL, disease activity and psychosocial symptoms. GI-PROMIS holds potential as important PRO measures and correlate with other validated indices in IBD.