Steps towards cost-benefit analysis of regional neurosurgical care.

Abstract

OBJECTIVE: To determine the cost of averting death or severe disability by neurosurgical intervention. DESIGN: Retrospective analysis of one year's admissions for neurosurgery; comparison of outcome with expected outcome in the absence of neurosurgical intervention and with the cost of neurosurgery. SETTING: Wessex Neurological Centre. PATIENTS: 1026 Patients were admitted to the neurosurgical service in 1984. Of 1185 admissions, 978 case records were available and outcome was known in 919. MAIN OUTCOME MEASURES: Outcome was assessed with the Glasgow outcome scale, modified as necessary, from the case notes, or by letter follow up to the general practitioner. Expected outcomes for each of the 54 diagnoses were derived from both published reports where available and an expert panel of 18 consultant neurosurgeons. The cost of the neurosurgical service for 1983-4 was known from a separate study and the cost per patient was calculated using the length of stay. RESULTS: The cost of neurosurgery in 1983-4 was 1.8 million pounds. In all, 243 deaths or severe disabilities were estimated to have been averted at an average cost of 7325 pounds (range 5000 pounds to 70,000 pounds). The overall cost per quality adjusted life year (QALY) was 350 pounds (range 34 pounds to greater than 400,000 pounds). The cost of long term care for severely disabled survivors is at least 18-fold greater than the cost of neurosurgical intervention to avert such disability. CONCLUSIONS: In Britain neurosurgery is not expensive in comparison with the costs and benefits of other areas of medicine, and the cost per QALY is unexpectedly low except for severe diffuse head injury, malignant brain tumors, and cerebral metastases. The neurosurgical budget should be assessed in the context of managing a patient in hospital and subsequently in the community.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 94 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.