PP25 Using a linked health dataset to identify what happens to patients following prehospital care

Background

UK prehospital services receive poor information about what happens to patients beyond the prehospital phase of care; this limits the opportunity to assess patient outcomes following non-conveyance. We aimed to identify what happens to patients after prehospital discharge, as part of two studies 1) to develop better ways of measuring prehospital quality and performance (Prehospital Outcomes for Evidence Based Evaluation, (PhOEBE)) and 2) to assess Variation in Ambulance Non-conveyance decisions (VAN).

Method

We developed a patient level dataset, linking ambulance call and clinical data to subsequent ED, hospital and mortality data. Data for 6 months of 2013 was provided by one ambulance service and linked to national datasets using NHS Digital's data-linking service. We identified a sub-group of non-conveyed patients and identified recontacts (EMS, ED, hospital admission or death) within 3 days of the call. We traced 24% (2514/10634) of hear and treat calls and 84% (42,796/50894) of see and treat calls.

Results

Within our sample, 20.6% of see and treat calls had a recontact within 3 days, 6.3% (2694/42.796) were admitted to hospital and 0.3% (129/42796) died. Poor hear and treat linkage rates limited data reliability, therefore we conducted a sensitivity analysis using traced calls only and traced plus non-traced calls as the denominator. We found that hear and treat hospital admissions within 3 days range from 2.5%–10.5% and deaths within 3 days range from 0.06%–0.24%

We developed a new indicator to monitor the quality and safety of non-transport decisions "Proportion of calls left at home which result in hospital admission or death within 3 days".

Conclusions

We linked multiple data sources to identify what happens to patients following prehospital care and identified low rates of hospital admission and death following non-conveyance. We developed a new non-conveyance indicator to assess safety and monitor non-conveyance rates over time.

http://ift.tt/2wYLVr8