Patient-reported outcomes of colorectal cancer surgery: A systematic review of outcome measures


Year:

Session type:

R. Whistance1, R. Forsythe2, J. Jones3, G. Smith3, A. Pullyblank4, A. McNair1, S. Brookes1, K. Avery1, J. Blazeby1
1Academic Unit of Surgical Research, School of Social and Community Medicine, University of Bristol, Bristol, UK, 2Division of Surgery Head and Neck, University Hospitals Bristol NHS Foundation Trust, Bristol, UK, 3Colorectal Cancer Patient Representative, North Bristol NHS Trust, Bristol, UK, 4Department of General Surgery, North Bristol NHS Trust, Bristol, UK

Background

Surgical innovation in colorectal cancer (CRC) requires evaluation with outcomes from the patients' perspective. Inconsistent use of patient-reported outcome (PRO) questionnaires prevents cross-trial comparisons and hinders data synthesis in meta-analyses. This study aimed to identify PRO questionnaires used in CRC surgical studies, extract individual questions (items) and group items pertaining to similar consequences of CRC surgery into domains. This study will inform the development of a core outcome set for CRC surgery (an agreed set of endpoints reported, as a minimum, in all trials).

Method

A systematic review identified original research articles (from 2009 to 2010) and systematic reviews (from 2001 to 2010) reporting PROs for CRC patients treated with curative intent. Excluded were articles of non-biomedical interventions; screening; treatment of CRC metastases; and non-curative, non-surgical therapies. Copies of identified PRO questionnaires were obtained and items extracted. Two researchers categorized the items into domains. The domains were checked by two CRC patients and a colorectal surgeon.

Results

Of 5644 abstracts screened, 31 original research articles and 6 systematic reviews were included. 58 different PRO questionnaires were reported on 184 occasions in 103 publications. The most frequently reported questionnaires were the EORTC QLQ-C30 (n=50) and the EORTC QLQ-CR38 (n=33). 40 questionnaires (69.0%) were reported once. Full versions of 50 questionnaires (86.2%) were available in English language, pertaining to 920 individual items. These items were grouped into 51 domains. The domains comprising most items were ‘Anxiety' (n=85), ‘Fatigue' (n=67) and ‘Physical Function' (n=63).

Conclusion

There is evidence for inconsistency in the choice of PRO questionnaires for CRC surgical studies. This may hinder comparisons between trials and data synthesis in meta-analyses. Work is ongoing to develop a core outcome set for CRC surgery and consideration to whether this will require separate short and longer term outcomes is being given.