P178 Variation in assessment of outcomes for inflammatory bowel disease in routine clinical practice: a mixed-methods study
V. Razanskaite, P. Williamson, B. Young, K. Bodger
University of Liverpool, Biostatistics, Liverpool, UK
Background
A number of global initiatives have sought to standardise the assessment of outcomes for inflammatory bowel disease (IBD), both for clinical trials (Core Outcome Sets1) and routine practice (ICHOM2). Our aim was to explore variability in the assessment and recording of clinical IBD outcomes during routine practice in England, evaluating the range of individual outcomes elicited and the extent of objective quantification of symptoms.
Methods
We performed ethnographic observations of 102 IBD clinic consultations conducted by 24 IBD clinicians (10 consultants, 10 IBD nurses and 4 trainees) in six acute hospitals in the North West region of England. We analysed 909 retrospective clinic records from outpatient IBD clinic visits recorded for observed patients in routine electronic patient records (EPRs). Audio-recordings of observed consultations and clinic records were analysed for pre-defined IBD outcomes, including those items required for Crohn’s Disease Activity Index (CDAI), Harvey–Bradshaw Index (HBI), Mayo Clinic Score (MCS) and Simple Clinical Colitis Activity Index (SCCAI). In addition, we performed 24 semi-structured interviews with IBD clinicians to explore the barriers and facilitators to capturing structured IBD outcomes in EPRs.
Results
The most commonly elicited and recorded outcomes were general well-being (95% consultations [C] and 61% clinic records [R]), stool frequency [SF] (87% C, 68% R), abdominal pain [AP] (79% C, 46% R), blood in stool [BS] (71% C, 57% R) and loose stool (64% C, 51% R). Clinical disease activity indices (HBI and SCCAI) were collected in only 15 (14.7%) observed consultations and recorded in 51 (5.6%) clinic letters. There was marked variation in the quantification of individual symptom items of MCS, SCCAI, CDAI and HBI in analysed consultations and records. Both PRO2 Crohn’s disease outcomes [AP and SF] were collected in 29/50 (58%) observed consultations and recorded in 169/484 (35%) records. Both PRO2 ulcerative colitis outcomes [SF and BS] were collected in 42/52 (81%) consultations and 253/425 (60%) records. Selected outcomes were recorded significantly more frequently by IBD nurses compared with doctors. Clinicians reported a range of IT barriers and issues with psychometric properties of currently available clinical disease activity indices in practice.
Conclusion
There is significant variability in the breadth, depth and quantification of IBD clinical outcomes during routine clinical assessments. Although most domains of clinical disease activity indices were elicited in consultations, formal scoring and assessment over fixed time periods was rare. There is an urgent need for practical alternatives to clinician reported disease activity indices in routine care.