Patterns-of-care and health economic analysis of robot-assisted radical prostatectomy in the Australian public health system

Objectives To compare patterns of care and peri-operative outcomes of robot-assisted radical prostatectomy (RARP) with other surgical approaches, and to create an economic model to assess the viability of RARP in the public case-mix funding system. Patients and Methods We retrospectively reviewed all radical prostatectomies (RPs) performed for localized prostate cancer in Victoria, Australia, from the Victorian Admitted Episode Dataset, a large administrative database that records all hospital inpatient episodes in Victoria. The first database, covering the period from July 2010 to April 2013 (n = 5 130), was used to compare length of hospital stay (LOS) and blood transfusion rates between surgical approaches. This was subsequently integrated into an economic model. A second database (n = 5 581) was extracted to cover the period between July 2010 and June 2013, three full financial years, to depict patterns of care and make future predictions for the 2014-2015 financial year, and to perform a hospital volume analysis. We then created an economic model to evaluate the incremental cost of RARP vs open RP (ORP) and laparoscopic RP (LRP), incorporating the cost-offset from differences in LOS and blood transfusion rate. The economic model constructs estimates of the diagnosis-related group (DRG) costs of ORP and LRP, adds the gross cost of the surgical robot (capital, consumables, maintenance and repairs), and manipulates these DRG costs to obtain a DRG cost per day, which can be used to estimate the cost-offset associated with RARP in comparison with ORP and LRP. Economic modelling was performed around a base-case scenario, assuming a 7-year robot lifespan and 124 RARPs performed per financial year. One- and two-way sensitivity analyses were performed for the four-arm da Vinci SHD, Si and Si dual surgical systems (Intuitive Surgical Ltd, Sunnyvale, CA, USA). Results We identified 5 581 patients who underwent RP in 20 hospitals in Victoria with an open, laparoscopic or robot-assisted surgical approach in the public and private sector. The majority of RPs (4 233, 75.8%), in Victoria were performed in the private sector, with an overall 11.5% decrease in the total number of RPs performed over the 3-year study period. In the most recent financial year, 820 (47%), 765 (44%) and 173 patients (10%) underwent RARP, ORP and LRP, respectively. In the same timeframe, RARP accounted for 26 and 53% of all RPs in the public and private sector, respectively. Public hospitals in Victoria perform a median number of 14 RPs per year and 40% of hospitals perform <10 RPs per year. In the public system, RARP was associated with a mean (±sd) LOS of 1.4 (±1.3) days compared with 3.6 (±2.7) days for LRP and 4.8 (±3.5) days for ORP (P < 0.001). The mean blood transfusion rates were 0, 6 and 15% for RARP, LRP and ORP, respectively (P < 0.001). The incremental cost per RARP case compared with ORP and LRP was A$442 and A$2 092, respectively, for the da Vinci S model, A$1 933 and A$3 583, respectively, for the da Vinci Si model and A$3 548 and A$5 198, respectively for the da Vinci Si dual. RARP can become cost-equivalent with ORP where ~140 cases per year are performed in the base-case scenario. Conclusions Over the period studied, RARP has become the dominant approach to RP, with significantly shorter LOS and lower blood transfusion rate. This translates to a significant cost-offset, which is further enhanced by increasing the case volume, extending the lifespan of the robot and reductions in the cost of consumables and capital.