Information sharing between intensive care and primary care after an episode of critical illness; A mixed methods analysis\

Introduction Poor quality communication between hospital doctors and GPs at the time of hospital discharge is associated with adverse patient outcomes. This may be more marked after an episode of critical illness, the complications of which can persist long after hospital discharge. Aims 1. to evaluate information sharing between ICU staff and GPs after a critical illness 2. to identify factors influencing the flow and utilisation of this information. Methods Parallel mixed methods observational study in an Irish setting, with equal emphasis on quantitative and qualitative data. Descriptive analysis was performed on quantitative data derived from GP and ICU consultant questionnaires. Qualitative data came from semi-structured interviews with GPs and consultants, and were analysed using directed content analysis. Mixing of data occurred at the stage of interpretation. Results GPs rarely received information about an episode of critical illness directly from ICU staff, with most coming from patients and relatives. Information received from hospital sources was frequently brief and incomplete. Common communication barriers reported by consultants were insufficient time, low perceived importance and difficulty establishing GP contact. When provided information, GPs seldom actioned specific interventions, citing insufficient guidance in hospital correspondence and poor knowledge about critical illness complications and their management. A majority of all respondents thought that improved information sharing would benefit patients. Cultural influences on practice were identified in qualitative data. A priori qualitative themes were: (1) perceived benefits of information sharing, (2) factors influencing current practice and (3) strategies for optimal information sharing. Emergent themes were: (4) the central role of the GP in patient care, (5) the concept of the "whole patient journey" and (6) a culture of expectation around a GP's knowledge of hospital care. Conclusions Practical and cultural factors contribute to suboptimal information sharing between ICU and primary care doctors around an episode of critical illness in ICU. We propose a three-milestone strategy to improve the flow and utilisation of information when patients are admitted, discharged or die within the ICU.