BACKGROUND: People experiencing homelessness are known to have complex health needs and to be high users of hospital accident and emergency (A&E) departments. It is unclear whether access to a day-time specialist homeless medical practice, as opposed to routine general practice, influences A&E attendance rates.

AIM: This study investigated whether registration with a specialist homeless service would alter A&E attendance rates in a single geographical region in Scotland.

DESIGN & SETTING: A health board area with a specialist service for people experiencing homelessness was selected. Data were obtained from the hospital records of 4408 A&E attendances by people experiencing homelessness at NHS Lothian (based on a broad definition of homelessness and including those in temporary accommodation) between January 2015 and July 2017.

METHOD: The attendances were compared between people registered with a specialist service and those registered with a mainstream GP.

RESULTS: The reasons for attendance and urgency of attendance were broadly similar between the two groups. Repeat attendance was similarly high in both groups. Almost 70% in both groups attended with problems deemed urgent, very urgent, or requiring immediate resuscitation. The patients registered with the specialist homeless service were more likely to be older and male; however, this did not affect the frequency of attendance.

CONCLUSION: People experiencing homelessness attending A&E mainly do so for urgent or very urgent problems. This was not related to the type of day-time primary care service they had access to. Strategies to reduce attendances, such as out-of-hours mobile medical units, should be explored.

How this fits in? People experiencing homelessness are known to have high attendance to A&E departments. It has been suggested poor access to primary care services may be a cause for this increased attendance, but it was found that access to primary care services did not affect attendance rate in this group. As most attendances were urgent and frequently related to drugs and alcohol, it may be that harm reduction strategies are needed to reduce the frequency of emergency attendance and resulting morbidity.