The effect of the approach to radical prostatectomy on the profitability of hospitals and surgeons

Authors


Yair Lotan, Department of Urology, The University of Texas Southwestern Medical Center, 5323 Harry Hines Boulevard, Dallas, TX 75390-9110, USA.
e-mail: yair.lotan@utsouthwestern.edu

Abstract

Study Type – Therapy (economics analysis)
Level of Evidence 2b

OBJECTIVE

To evaluate the profit margins for radical retropubic prostatectomy (RRP), laparoscopic radical prostatectomy (LRP) and robot-assisted laparoscopic prostatectomy (RALP), and the effect on the reimbursement to the urologist, as there has been a dramatic increase in use of RALP, with the cost of the robot borne by hospitals.

METHODS

Data on costs and payments to hospital and surgeon from 2003 to 2008 for RRP, LRP and RALP were obtained from the hospital and urology department. We determined the profit based on the difference between payments received and total cost.

RESULTS

Between 2000 and 2008, 1279 RPs were performed at our private hospital. The introduction of RALP increased total number of RPs and replaced most RRPs. RRP represents the only procedure where payments exceed total costs. For RRP there was a significantly higher profit for patients with comorbidities. The type of payer had a large effect on profit. Medicare provides a small profit for RRP but a significant loss of >US$4000 for RALP. While all insurance companies resulted in losses for LRP and RALP, there was variability of almost $600/case for LRP and >$1400/case for RALP. RALP provided the highest reimbursement for the surgeon due to additional reimbursement for the S2900 code (use of robot).

CONCLUSIONS

The introduction of RALP has increased the case volume at our hospital and improved profits for the surgeon. The hospital loses money on each LRP and RALP case compared with RRP, which provides a small profit.

Abbreviations
(R)(L)RP

(retropubic) (laparoscopic) radical prostatectomy

RALP

robot-assisted LRP

UTSW

University of Texas Southwestern Medical Center

DRG

Diagnosis-related Group.

INTRODUCTION

National healthcare expenditures are increasing rapidly, leading to greater scrutiny of medical decision-making based on economic concerns [1]. The introduction of new technologies requires a comparative analysis to determine if the purported benefits justify new costs. In the medical field a prospective evaluation of cost-effectiveness is rare, as new medications and products are introduced primarily based on their ability to advance patient care. While this model might allow the most expeditious introduction of new products, it does not necessarily lead to the most cost-effective outcome. In fact, some new technologies can offer added expense but no additional benefit, as in the case of newer generation lithotripters for managing stone disease [2]. One major financial consideration is that most new technologies need to be purchased by hospitals, and thus the main financial burden is placed on hospital systems and not third-party payers such as Medicare or insurance companies. Indeed, there is rarely added reimbursement for the acquisition of capital equipment, such as facility fees with shock-wave lithotripsy representing one exception. In addition, to remain competitive, hospitals often feel pressured to acquire costly technology above and beyond what the population size or market would require, just to continue to attract both admitting physicians and potential patients.

The introduction of robotics and the rapid adoption of this technology for radical prostatectomy (RP) highlights some of the concerns related to the costs associated with technological advances. Prostate cancer is the most common cancer in men, with an estimated 186 320 new cases in the USA in 2008 [3]. Most cases of newly diagnosed prostate cancer are clinically localized and RP is one of the main treatments offered to patients. Furthermore, robotic-assisted laparoscopic RP (RALP) has been rapidly adopted by urologists, resulting in it becoming the main procedure currently used to treat prostate cancer surgically in the USA [4,5].

Surprisingly, there is a paucity of research into the financial effects of capital equipment acquisition on hospital systems. Nor are there evaluations of the effect of adopting new techniques on the income of the surgeon. Due to the limitations in resources in the healthcare system, it is important to understand how adopting expensive technologies can affect the profitability of procedures for a hospital. As such, we evaluated the cost and reimbursement for RP at our hospital between 2003 and 2008. Surgeons at our institution started performing laparoscopic RP (LRP) in September 2003 and RALP in January of 2006. We also evaluated the effect of adopting new technologies on the reimbursement to the urologist.

METHODS

This study was approved by the institutional review board at University of Texas Southwestern Medical Center (UTSW). We acquired data on the number of procedures performed of each technique for RP, including retropubic RP (RRP), LRP and RALP, from 2000 to 2008 at our private hospital from our billing department. We also obtained reimbursement to the surgeon for procedures performed at our private hospital from September 2005 to September 2008.

Data were acquired from the hospital billing department for total cost, direct and indirect costs (all in US$), and payments for RRP, LRP and RALP from 2003 to 2008. The date of surgery, surgeon, payer and Diagnosis-related Group (DRG) were also included. DRG 334 represents patients with comorbidities or complications, while DRG 335 represents patients without comorbidities or complications. While there were changes in the DRG system in 2008, this was not included as most patients were treated before 2008. Cost data were available for 645 procedures between 2003 and 2008; this included patients who had RRP (161), LRP (220) and RALP (264). Patients with erroneous hospital accounts were excluded from further analysis, to avoid biasing one or the other technology, resulting in 617 evaluable patients including RRP (157), LRP (214) and RALP (246). Cost data rather than charge data were used because it is a better reflection of resource use. The cost of the robot and maintenance were not included in the analysis, as it would become the single most important determinant of profitability. Reasons for this decision and the effects of such cost are enumerated in the discussion section. Based on these data, we determined the profit based on the difference between payments received and total cost. We stratified data based on procedure type, DRG and payer.

RESULTS

Between 2000 and 2008, 1279 RPs were performed at our private hospital; Fig. 1 shows the distribution and number of cases. While the number of LRPs remained stable, the introduction of RALP both increased the total number of RPs and replaced most RRPs at our hospital. The total cost, payments and profits are shown in Table 1. Direct costs were highest for RALP, at $6988, $5670 and $4021 for RALP, LRP and RRP, respectively. The indirect cost represented 32%, 34% and 38% of the total cost of RALP, LRP and RRP. RRP was the only procedure where payments exceed total costs. Table 1 also shows the differences in profit for each procedure based on DRG code and type of payer. Except for RRP, the costs of LRP and RALP exceeded payments, resulting in negative profits. For RRP there was a significantly higher profit for DRG 334. LRP was not profitable but DRG 334 resulted in a better reimbursement. There was a negligible difference in profits based on DRG for RALP. DRG 334 represented 39% of RRP, 21% of LRP and 10% of RALP. The type of payer had a large effect on the profit. Comparing insurance companies 1 and 2 for RRP, there is a difference of >$5000. Medicare provides a small profit for RRP but a significant loss of >$4000 for each RALP. While all insurance companies resulted in losses for LRP and RALP, there was variability of almost $600/case for LRP and >$1400/case for RALP. There was no consistent pattern of profit based on year of procedure (Table 1). Medicare reimbursement for DRG 334 was over $1500–2000 more per case than DRG 335 (Table 2). There was a slight but inconsistent increase in reimbursement over time.

Figure 1.

Distribution and number of RRP, LRP and RALP between 2000 and 2008.

Table 1.  Total cost, payments and profits for RPs, and profit based on DRG and type of payer
Cost, US$RRPLRPRALP
  • *

    Only the three insurance companies with most patients were included.

  • †In 2008, the hospital increased the cost of the operating room for robotic cases to account for the specialized equipment and added cost of using the robot. na, not available.

N  157  214   246
Mean total (range) 6473 (3677–16490) 8557 (6074–13 239)10 269 (5 494–40 401)
Direct 4021 5670 6 988
Indirect 2452 2887 3 281
Mean total payment (range) 6893 (2000–17820) 6805 (1103–20 431) 7 616 (1 457–27 210)
Mean profit (range)  419 (−10404, 11663)−1752 (−9433, 11 994)−2 653 (−30 398, 17 900)
Profit*
DRG 334, n   60   44    25
 Profit  1103−1237−2 587
DRG 335, n   95  170   221
 Profit  −15−1885−2 660
Type of payer
Medicare, n   36   34    60
 Profit  209−1301−4 355
Insurance 1, n   19   30    29
 Profit−2870−1688−3 614
Insurance 2, n   23   40    43
 Profit 2339−1099−2 175
Insurance 3, n   36   30    42
 Profit  945−1679−2 481
By year
2003−1231  
2004   171−1163 
2005  503−2299 
2006 −454−1908−2 968
2007  273 −1710−2 323
2008nana−3 343
Table 2.  Hospital reimbursement from Medicare by year
Costs, US$: YearLRP/RALPRRP
334335334335
  1. na, not available.

20037950586479505816
20048466594179956086
2005na614176596179
200685946534nana
20078510668976486798
2008na6626nana

The payment to the surgeon was based on the payment from insurance (private or Medicare) and any co-payment by the patient. With the exception of insurance 2 for RRP, all insurance companies reimbursed the surgeon better than Medicare (Table 3). UTSW has a joint approach for negotiating with insurance companies and health maintenance organizations. A group of administrators and physician leaders meet with the insurance companies and negotiate simultaneously the hospital reimbursement structure and physician payments, which makes the observed disparity particularly interesting. Insurance 1, while causing a significant loss of money to the hospital for RRP, provided the surgeon with the best reimbursement for the same procedure. RALP provided the highest reimbursement and while the LRP paid better than open RRP, the main difference was the additional reimbursement for the S2900 code (use of robot) which contributed to the overall payment (Table 4). As can be seen in Table 4, Medicare does not pay for the S2900 code (use of a robot during surgery) and the payment for the code is variable between insurance companies.

Table 3.  Payment to the surgeon
PaymentRRPLRPRALP
N  61 189 291
Mean total (range)2250 (1298–5384)2662 (1080–8480)3007 (1422–10560)
Mean insurance (range)1992 (745–3350)2173 (641–5400)2154 (671–5026)
Payer
Medicare162818041711
Insurance 1252619522375
Insurance 2160321402976
Insurance 3213819132207
Table 4.  Payment for S2900 code
PayerMean reimbursementPaid, % (n/N)
Medicare  0 0
Insurance 129224 (10/41)
Insurance 263582 (41/50)
Insurance 3 8736 (26/72)

DISCUSSION

Prostate cancer is the most common cancer in men and RP is one of the primary treatments for this disease. Medical expenditures for prostate cancer treatment in the USA totalled $1.3 billion in 2000, based on data from the Urologic Disease of America Project [6]. RP rates were 108/100 000 in 2000, with rates decreasing over time in patients aged >65 years, but increasing in patients aged 40–54 years [6]. The use of RALP has increased, and with it the economic burden of adopting the technology [4,5]. The da Vinci robot (Intuitive Surgical Inc., Sunnyvale, CA, USA) costs $1.5–1.75 million, with a maintenance fee of $112 000–150 000/year, depending on model and year of purchase [7]. Furthermore, there is a cost of surgical instruments that have a limited usage and can add >$1000/case [8]. In essence, a RALP requires the use of at least two needle holders, one pair of scissors and one or more grasping devices. Each of these devices will last for 10 uses and cost $2200, making the cost per instrument per case of $220. The added cost of the robot is not reimbursed by Medicare or insurance companies, and as such is left to hospital systems. The goal of our study was to evaluate the effects of adopting RALP on the profitability of RP at our hospital.

We found that adopting RALP increased the overall number of RPs at our hospital, with a significant reduction in RRPs. The hospital generated a profit for RRP primarily due to patients with comorbidities or complications (DRG 334). The hospital lost money on LRP and RALP, but more for RALP. There was an effect of the payer on the hospital system, and there was no significant increase in reimbursement over time. From a surgeon’s standpoint, RALP offered advantages both from increased volume and reimbursement based in part on the S2900 code.

The profit or loss of a procedure is based on the total cost and payments. The cost of RRP was less than LRP and RALP, resulting in a higher profit margin. The higher cost of RALP was primarily due to operative instruments and their effect on direct costs. The proportion of indirect costs was similar for all procedures. In fact, the analysis did not consider the added cost of acquiring and maintaining the robot, as there is no facility fee for the robot and this would add >$1000/case even amortizing the robot over 7 years and performing 150 cases per year. One limitation of the analysis is that costs of the procedures can vary between institutions, but our costs were similar to those reported by other publications focusing on the cost of RP [9–11]. The other consideration is that payment for RP has not increased significantly over time, especially based on Medicare reimbursement. Another limitation of the analysis is that it is based on the payer mix of our hospital and the contracts that our hospital has negotiated. Medicare was the payer for <25% of patients. Although there is regional variation of Medicare payments, this is unlikely to have a dramatic effect on the profit margin for LRP and RALP. However, if a hospital has very good contracts with insurance companies or many self-pay patients, then it is possible that each procedure would be profitable. It is also clear that the contracts that hospitals make with insurance companies can have a significant effect. As profit margins are so small, a difference of $500 per case can significantly change the balance of a procedure. Similarly a reduction in cost of equipment or other aspects of hospitalization would improve profit margins.

In many hospitals the robot itself is a gift or donation, often attached to a centre of some kind, thus eliminating the acquisition cost. However, even when applying the $150 000 annual maintenance cost only, it would add $1000 per case if 150 cases were performed, and still $500 for 300 cases, which probably is the logistical case number limit of one robot. The additional $500 would further reduce any potential profit margin. Lastly, even if a robot is donated, it would be prudent for a hospital or hospital system to plan for replacement and thus, over perhaps 10 years, enough money would need to be saved to purchase a new robot after such period. This would add an additional $175 000 annually and increase the cost per case by $1000 or more, assuming 150 cases were being done.

There are several important issues raised by this analysis. New technology should be adopted to improve outcomes. Unfortunately, there have been no randomized trials comparing outcomes for different surgical approaches to RP, and several reviews have shown no significant advantages in cancer control, continence or potency for the RALP approach [5,12,13], The main advantage currently reported consistently for the RALP and ‘pure’ LRP is a decrease in blood loss, but transfusion rates for open RRP tend to be low [5,13]. The decrease in pain and length of stay has been questioned by several studies which used similar pathways to discharge for RALP and RRP [14,15]. Finally, the hope for improved cancer control due to better visualization has been dampened by a recent report of increased use of salvage radiotherapy or hormone therapy in patients who had a minimally invasive RP, but this might not apply to high-volume surgeons [16,17]. As such, the clinical benefits of RALP have yet to be confirmed, but there is a recognized yet poorly defined cost to using a robot for RP. Furthermore, each case is associated with added equipment costs [8]. The few studies comparing RALP and RRP found that the robotic approach is more expensive [8, 18]. These studies did not specifically evaluate the effect of acquiring a robot on hospital profits. Finally, it is ironic that one benefit of LRP and RALP is to reduce intraoperative blood loss, as anaemia represents one complication that allows a hospital to use a DRG with greater reimbursement. This is not an insignificant difference, representing >$1500 per case.

While the profit for a procedure can be examined, and the absolute gain or loss to a hospital determined, it is an important point that hospitals do not exist in a vacuum. In the real world there are numerous factors that affect a hospital’s decision to adopt new technologies. Hospitals exist in a competitive marketplace, competing for not only patients but providers. They need to provide comprehensive services and if there is a demand by patients or physicians for a technology, this drives acquisition even if the product is not cost-effective. It is difficult to make objective financial considerations to the marketing advantages of offering a new technology such as a robot, or the damage from not having this technology. However, as more hospitals acquire a robot, the uniqueness of having a robot decreases, as does the marginal benefit from an advertising standpoint. There are other considerations for an institution, as there can be a downstream benefit to having a robot. Patients who want a RALP might end up choosing radiotherapy and this could increase other financial benefits that are not quantified by this type of analysis. Furthermore, there are other issues that can affect a hospital. A hospital has fixed costs, including administrative, billing, security, nursing, electricity, etc. These costs exist whether the hospital beds are occupied or not. As the reimbursement for all procedures is higher than the direct costs, a hospital that is not fully occupied might perform better fiscally with a procedure that has a small loss, as opposed to having an empty bed with no income. This might not be the case for a hospital that is fully occupied, assuming that it can generate a profit with other diagnoses.

The introduction of RALP has increased the current costs of RP and affected the potential profit accrued by hospitals. Prospective studies should be conducted to determine if RALP offers a significant advantage over other approaches to RP. If this is the case, then one could consider a facility fee or other reimbursement for hospitals to offset the cost of acquiring and maintaining a robot. A decrease in cost of the robot and instruments could also reduce losses from RALP. As there are other technologies under development and in stages of adoption, it would be ideal to evaluate the financial effects of these technologies on the cost of care of prostate cancer before their widespread use. One such example is the use of proton-beam therapy, which costs >$100 million to build and has not yet shown a significant advantage over other radiation treatments [19,20].

In conclusion, the introduction of RALP has increased the case volume at our hospital and improved profits for the surgeon. However, the hospital loses money on each LRP and RALP case compared with RRP, which provides a small profit. There is a significant effect of the payer on the payment for procedures.

CONFLICT OF INTEREST

None declared.

Ancillary