OBJECTIVES: To measure, in a service setting, the effect of magnetic resonance imaging on diagnosis, diagnostic certainty, and patient management in the neurosciences; to measure the cost per patient scanned; to estimate the marginal cost of imaging and compare this with its diagnostic impact; to measure changes in patients' quality of life; and to record the diagnostic pathway leading to magnetic resonance imaging. DESIGN: Controlled observational study using questionnaires on diagnosis and patient management before and after imaging. Detailed costing study. Quality of life questionnaires at the time of imaging and six months later. Diagnostic pathways extracted from medical records for a representative sample. SETTING: Regional superconducting 1.5 T magnetic resonance service. SUBJECTS: 782 consecutive neuroscience patients referred by consultants for magnetic resonance imaging during June 1988-9; diagnostic pathways recorded for 158 cases. MAIN OUTCOME MEASURES: Costs of magnetic resonance imaging and preliminary investigations; changes in planned management and resulting savings; changes in principal diagnosis and diagnostic certainty; changes in patients' quality of life. RESULTS: Average cost of magnetic resonance imaging was estimated at 206.20/patient pounds (throughput 2250 patients/year, 1989-90 prices including contrast and upgrading). Before magnetic resonance imaging diagnostic procedures cost 164.40/patient pounds (including inpatient stays). Management changed after imaging in 208 (27%) cases; saving an estimated 80.90/patient pounds. Confidence in planned management increased in a further 226 (29%) referrals. Consultants' principal diagnosis changed in 159 of 782 (20%) referrals; marginal cost per diagnostic change was 626 pounds. Confidence in diagnosis increased in 236 (30%) referrals. No improvement in patients' quality of life at six month assessment. CONCLUSIONS: Any improvement in diagnosis with magnetic resonance imaging is achieved at a higher cost. Techniques for monitoring the cost effectiveness of this technology need to be developed.
OBJECTIVES: To measure, in a service setting, the effect of magnetic resonance imaging on diagnosis, diagnostic certainty, and patient management in the neurosciences; to measure the cost per patient scanned; to estimate the marginal cost of imaging and compare this with its diagnostic impact; to measure changes in patients' quality of life; and to record the diagnostic pathway leading to magnetic resonance imaging. DESIGN: Controlled observational study using questionnaires on diagnosis and patient management before and after imaging. Detailed costing study. Quality of life questionnaires at the time of imaging and six months later. Diagnostic pathways extracted from medical records for a representative sample. SETTING: Regional superconducting 1.5 T magnetic resonance service. SUBJECTS: 782 consecutive neuroscience patients referred by consultants for magnetic resonance imaging during June 1988-9; diagnostic pathways recorded for 158 cases. MAIN OUTCOME MEASURES: Costs of magnetic resonance imaging and preliminary investigations; changes in planned management and resulting savings; changes in principal diagnosis and diagnostic certainty; changes in patients' quality of life. RESULTS: Average cost of magnetic resonance imaging was estimated at 206.20/patient pounds (throughput 2250 patients/year, 1989-90 prices including contrast and upgrading). Before magnetic resonance imaging diagnostic procedures cost 164.40/patient pounds (including inpatient stays). Management changed after imaging in 208 (27%) cases; saving an estimated 80.90/patient pounds. Confidence in planned management increased in a further 226 (29%) referrals. Consultants' principal diagnosis changed in 159 of 782 (20%) referrals; marginal cost per diagnostic change was 626 pounds. Confidence in diagnosis increased in 236 (30%) referrals. No improvement in patients' quality of life at six month assessment. CONCLUSIONS: Any improvement in diagnosis with magnetic resonance imaging is achieved at a higher cost. Techniques for monitoring the cost effectiveness of this technology need to be developed.