Abstract
Although cost-effectiveness is becoming the foremost evaluative criterion within health service management of spine surgery, scientific knowledge about cost-patterns and cost-effectiveness is limited. The aims of this study were (1) to establish an activity-based method for costing at the patient-level, (2) to investigate the correlation between costs and effects, (3) to investigate the influence of selected patient characteristics on cost-effectiveness and, (4) to investigate the incremental cost-effectiveness ratio of (a) posterior instrumentation and (b) intervertebral anterior support in lumbar spinal fusion. We hypothesized a positive correlation between costs and effects, that determinants of effects would also determine cost-effectiveness, and that posterolateral instrumentation and anterior intervertebral support are cost-effective adjuncts in posterolateral lumbar fusion. A cohort of 136 consecutive patients with chronic low back pain, who were surgically treated from January 2001 through January 2003, was followed until 2 years postoperatively. Operations took place at University Hospital of Aarhus and all patients had either (1) non-instrumented posterolateral lumbar spinal fusion, (2) instrumented posterolateral lumbar spinal fusion, or (3) instrumented posterolateral lumbar spinal fusion + anterior intervertebral support. Analysis of costs was performed at the patient-level, from an administrator's perspective, by means of Activity-Based-Costing. Clinical effects were measured by means of the Dallas Pain Questionnaire and the Low Back Pain Rating Scale at baseline and 2 years postoperatively. Regression models were used to reveal determinants for costs and effects. Costs and effects were analyzed as a net-benefit measure to reveal determinants for cost-effectiveness, and finally, adjusted analysis (for non-random allocation of patients) was performed in order to reveal the incremental cost-effectiveness ratios of (a) posterior instrumentation and (b) anterior support. The costs of non-instrumented posterolateral spinal fusion were estimated at DKK 88,285(95% CI 81,369;95,546), instrumented posterolateral spinal fusion at DKK 94,396(95% CI 89,865;99,574) and instrumented posterolateral lumbar spinal fusion + anterior intervertebral support at DKK 120,759(95% CI 111,981;133,738). The net-benefit of the regimens was significantly affected by smoking and functional disability in psychosocial life areas. Multi-level fusion and surgical technique significantly affected the net-benefit as well. Surprisingly, no correlation was found between treatment costs and treatment effects. Incremental analysis suggested that the probability of posterior instrumentation being cost-effective was limited, whereas the probability of anterior intervertebral support being cost-effective escalates as willingness-to-pay per effect unit increases. This study reveals useful and hitherto unknown information both about cost-patterns at the patient-level and determinants of cost-effectiveness. The overall conclusion of the present investigation is a recommendation to focus further on determinants of cost-effectiveness. For example, patient characteristics that are modifiable at a relatively low expense may have greater influence on cost-effectiveness than the surgical technique itself--at least from an administrator's perspective.