Accuracy of Physician's Assessment of IBS Disease Severity

Abstract

Listen

Translate article

Introduction:Irritable bowel syndrome (IBS) is a chronic, functional gastrointestinal disorder comprising about 25% of all referrals to gastroenterologists. While physician-determined IBS severity guides clinical decision-making, it is unclear what determines the gastroenterologist's judgment of severity and the concordance between MD and patient ratings. Previously, our group (Lackner, Ma et al., 2009) identified factors predictive of patient rating of IBS severity. In this study, we aimed to determine the degree of agreement between physician and patient assessment of disease severity and to identify clinical factors that influenced MD ratings. Subjects:82 Rome II-diagnosed IBS adults (mean age 46.7 yrs, 86.6% F) completed baseline assessments for an NIH-funded RCT. Methods:Patient questionnaires included:the 2 bloating/ distension and 7 defecation of the GSRS-IBS items and UCLA Pain Scale to assess GI symptoms (abdominal discomfort,defecation, pain); Anxiety Sensitivity Inventory, Brief Symptom Inventory and SF-36 to evaluate emotions (visceral anxiety, depression, fatigue, somatization); IBS Locus of Control (LOC), Coping Strategies Questionnaire, State Trait Anxiety Inventory and Penn State Worry Questionnaire to assess cognitions (LOC, catastrophizing, anxiety, worry). The dependent variables were IBS severity assessed with 2 patient reported measures (UCLA IBS Symptom Severity [UCLA SS], IBS Symptom Severity Scale [IBS-SSS]) and MD rating (CGI Severity) with responses from 'not ill' to 'most extremely ill' on a 7-point Likert scale. Results: Bivariate analysis revealed only a modest correlation between MD rating and patient-reported UCLA SS and IBS-SSS (r=0.49 and 0.45 respectively, p<0.01). In 53.7% of cases, MD rating corresponded with patients' rating while the physician underestimated severity in 31.7% and overestimated in 14.6% of subjects. The disagreement with patients was most prevalent when MD rated IBS as mild. In stepwise linear regression, the blocks of demographic variables, GI symptoms, emotions and cognitions accounted for only 35% of the variance in MD rating in contrast to 57% and 69% on patient scores (UCLA SS and IBS-SSS, respectively). GI symptoms alone explained 16% of variance in MD assessment, while accounting for 49% and 64% of variance in the 2 patient scales (Table 1). Conclusion:Our main finding is a lack of correspondence between MD and patient ratings of severity. This discordance is particularly apparent with IBS patients whom MD rates as being 'mildly ill'. Although GI symptoms, emotions, cognitions and demographic variables, are reasonably good predictors of patient ratings, these only accounted for a third of the variance in physician's rating of IBS severity. Further studies to identify additional factors that influence physician rating of IBS severity are needed. Supported by NIH/NIDDK grants 67878 and 77738 Table 1: Contribution of various factors in physician and patient rating of IBS disease severity