Cost analysis of a hospital-at-home initiative using discrete event simulation

OBJECTIVES: To examine whether, from a National Health Service (NHS) and local authority social services' viewpoint, a hospital-at-home service was cost saving compared with conventional inpatient care. METHODS: The subjects of this part-retrospective and part-prospective cost analysis were 51 elderly medical and orthopaedic surgical patients assessed at Hillingdon Hospital, West London, as being suitable for hospital-at-home care. Thirty patients received hospital-at-home care, provided for up to 14 days, while 21 patients remained in hospital and received standard inpatient care. All direct costs to the NHS hospital, community health services' provider and social services' department during the initial episode of care and the three months after discharge were collected for each group of patients. Costs and clinical event data were entered in a discrete event simulation model which generated baseline results. Uncertainty surrounding the model's parameters was explored using sensitivity analysis. RESULTS: The baseline simulation performed with 1000 patients in each group showed the mean cost per patient for hospital-at-home care and three-month follow-up to be around three-fifths the mean cost per patient of inpatient care and follow-up. Most of the excess cost in the inpatient group was attributable to the initial period of hospitalisation. Under all assumptions used in the sensitivity analysis, the hospital-at-home service was less costly. CONCLUSIONS: For elderly patients assessed as needing no more than 14 days of hospital care, hospital-at-home care is cost saving to health and social care agencies when compared with conventional inpatient care.