A quantitative and qualitative evaluation of the effect of computerised decision support for the initial assessment (triple as

Objectives: (i) To compare the quality of clinicians’ decisions made with and without computerised triple assessment decision support (TADS) during assessment of women referred to specialist breast cancer units, (ii) To investigate the opinions of clinicians about the decision support software used. Design: Balanced-block crossover experiment with simulated cases; questionnaire study and semi-structured interviews. Setting: Secondary care in 15 specialist breast clinics in the south east of England. Participants: 36 clinicians were opportunistically sampled (24 agreed to participate). Intervention: A web-based computerised decision support system (DSS): TADS. Main outcome measures: The number of deviations from current best practice guidelines; opinions on the applicability of decision support to triple assessment. Results: Clinicians made significantly more deviations from guidelines (60/120 without DSS, 16/120 with DSS, P<0.001), significantly more deviations that could not be explained by local procedural variations (41 of 120 without DSS, 10 of 120 with DSS, P<0.001), significantly more potentially critical deviations (16/120 without DSS, 1/120 with DSS, P=0.001), and significantly more potentially critical deviations that were irretrievable (10/120 without DSS, 1/120 with DSS, P=0.02) without decision support. Clinicians were significantly more likely to agree with the statement: “patient care in triple assessment would benefit from computerised decision support” after using the system (P<0.025). Conclusions: To the best of our knowledge, this is the first evaluation of a DSS that supports multiple decisions in a care pathway. These results suggest that TADS could improve patient care and that it would be acceptable to many breast cancer clinicians. Care pathways constitute a core element of the NHS Care Records Service (CRS) of the National Programme for IT (NpfIT) and the NHS is currently looking into decision support technologies capable of delivering evidence-based care wherever and whenever it is needed. Within this context, although significant investment in IT infrastructure would be required in a number of hospitals, we recommend that a clinical trial of TADS be conducted in real clinical settings.