UCSF

Rationale, aims and objectives

Specialty care referrals have doubled in the last decade. Optimization of the pre-referral workup by a primary care doctor can lead to a more efficient first specialty visit with the patient. Guidance regarding pre-referral laboratory testing is a first step towards improving the specialty referral process. Our aim was to establish consensus regarding appropriate pre-referral workup for common gastrointestinal and liver conditions.

Methods

The Delphi method was used to establish local consensus for recommending certain laboratory tests prior to specialty referral for 13 clinical conditions. Seven conditions from The University of Michigan outpatient referral guidelines were used as a baseline. An expert panel of three PCPs and nine gastroenterologists from three academic hospitals participated in three iterative rounds of electronic surveys. Each panellist ranked each test using a 5-point Likert scale (strongly disagree to strongly agree). Local panellists could recommend additional tests for the initial diagnoses, and also recommended additional diagnoses needing guidelines: iron deficiency anaemia, abdominal pain, irritable bowel syndrome, fatty liver disease, liver mass and cirrhosis. Consensus was defined as ≥70% of experts scoring ≥4 (agree or strongly agree).

Results

Applying Delphi methodology to extrapolate externally developed referral guidelines for local implementation resulted in considerable modifications. For some conditions, many tests from the external group were eliminated by the local group (abdominal bloating; iron deficiency anaemia; irritable bowel syndrome). In contrast, for chronic diarrhoea, abnormal liver enzymes and viral hepatitis, all/most original tests were retained with additional tests added. For liver mass, fatty liver disease and cirrhosis, there was high concordance among the panel with few tests added or eliminated.

Conclusions

Consideration of externally developed referral guidelines using a consensus-building process leads to significant local tailoring and adaption. Our next steps include implementation and dissemination of these guidelines and evaluating their impact on care efficiency in clinical practice.