Community wards: An evaluation of the virtual ward services
Christine McGregor | Scottish Government
As pressure on hospital demand increases from an ageing population with multiple co-morbities, long term conditions and frailty a number of approaches are been developed to examine if Intermediate Care models can be both clinically and cost effective, as well as popular with both patients and acceptable to other key stakeholders within health and social care.
A retrospective matched control evaluation using patient level activity and costs was carried out for three cohorts across three community ward programmes before and after intervention. The same analysis was also carried out for three matched control cohorts. Matching was carried out on the basis of age, sex, locality and SPARRA score (Scottish Patients At Risk of Readmission and Admission). All three CW cohorts achieved greater reductions in Accident and Emergency attendances and emergency admissions compared to the control groups. This is illustrated as follows:
• For Accident and Emergency attendances (also incorporating costs of ambulance journeys): the Community Ward cohort achieved reductions of between 45%-55%; the control group 24%-37%.
• For emergency admissions: the CW cohort achieved reductions in number of episodes 45% - 50%; the control group between 30% - 39%. For length of stay: the CW cohort reduced between 26%-38% ; the control group 18%-53%.
• Overall, the resource use reduction across activity between the two time periods of six months: was £401,363 for the CW cohort; and £227,256 for the control.
• This gives a resource use saving per capita: of £2,826 for the CW cohort; and £1,600 for the control.
The impact was greatest in one CW where there was full GP engagement and strict adherence to those who would benefit most from CW admission (adults with multiple long term conditions). Resource use (costs) were lower across the CW cohorts compared to the control. The service was also deemed cost effective (reduction in resource use was more than the cost of running the service).
The sample size across cohorts were relatively small at 142. In addition, while everything was done to try and limit factors that would impact on results in the control group, it is not possible to say with complete certainty all factors were adequately controlled for. In terms of external validity, care would need to be taken if the model was replicated elsewhere to account for differences such as underlying service structure and population. The study only examined Accident and Emergency attendances and emergency (unscheduled) admissions. Ideally, if regional data could have been obtained at patient level this would have been extended to GP activity, community health activity and also social care activity.
Overall, the study is a valuable contribution to the literature on Intermediate Care models. It demonstrates that when managed effectively a Community Ward can not only be patient centred but also cost effective.