Improving communication of patient issues on transfer out of intensive care

The written medical handover document is frequently poor in quality and highly variable which raises concerns about patient safety. Intensive care unit (ICU) patients have complex medical and social issues which increases the risk of errors during ongoing hospital treatment. Our project team of four doctors and two nurses aimed to improve the documentation of patient problems as they leave the ICU. A literature review and process mapping of both medical and nursing transfer documentation helped in understanding the current process. Current problems (CP) were defined as any patient issues which require ongoing thought, management or follow-up. Our progress was tracked using a measure of the number of CPs listed in the free-text field titled ‘Current Problems’ in 50 medical transfer documents. This was graphed on a control chart showing a process in statistical control. Means and control limits were recalculated whenever a process shift occurred. There was no relationship between the number of CPs listed and length of ICU stay, age of patient, or severity of illness on presentation (Acute Physiologic Assessment and Chronic Health Evaluation II score). An inter-relationship graph identified the key drivers which were amenable to change: (1) the doctors completing the clinical summary at the time of discharge did not have all the information readily available to them and (2) the doctors were uncertain of the types of problem which should be communicated. Improvements were designed and trialled using Plan-Do-Study-Act cycles to address these two key drivers. At baseline, the average number of CPs per patient was 1.8. After implementation of a paper problem list at the patient bedside, with supporting education, the average increased to 2.7. This was further improved by the addition of a checklist of common patient problems. This increased the average to 3.85. These improvements were permanently implemented and ongoing audits have shown sustained improvement using statistical process control methods.