Objectives: To produce a valid, reliable instrument to gauge the extent to which GPs document relevant signs, symptoms and risk factors in referral letters to colorectal surgeons.
Design: GPs and colorectal surgeons were invited to participate in a two-part questionnaire survey about the ideal contents of a referral letter. In the second round participants were asked to reconsider the questionnaire in the light of the group's collective replies in the first round. The instrument was tested for predictive validity and inter-rater reliability.
Setting: GPs in North Nottinghamshire Health Authority and colorectal surgeons in North Trent.
Participants: 125 GPs registered in two districts with North Nottinghamshire Health Authority and nine colorectal surgeons in North Trent were invited to participate.
Main outcome measures: Mean scores in the second round of the questionnaire were used to produce an instrument in which marks could be ascribed to each item mentioned on a GP referral letter.
Results: There was a 68.6% response rate to the questionnaire survey. The instrument had substantial inter-rater reliability (r= 0.77). Higher scores predicted cases that would be offered urgent appointments by the specialist (OR = 1.06, 95% CI = 1.01 to 1.10). Cases with pathology were not referred with more thorough documentation of pre-referral assessment (score 33 vs. 31, mean difference 2.3, p = 0.06 (t-test), 95% Cl = -0.07 to 4.02).
Conclusions: In some cases, patients with pathology are entering secondary care with communications from GPs in which the relevant history and examination are not fully documented. Explicit documentation of GP assessment prior to referral may have a significant impact on how cases might be managed in secondary care.