Page 136 - An Evidence Review of Active Surveillance in Men With Localized Prostate Cancer

236
$56,725 for EBRT and $32,795 and $35,037 for RP. Notably, each treatment had a distinct
cost pattern when total costs were separated into those related to medications, office visits and
hospitalizations (please see Appendix Table 4.5 for details).
The fourth study was an ancillary investigation from the SPCG-4 trial, a randomized study of
RP versus WW for men with localized prostate cancer. 237 The study reported cost estimates
(based on Swedish prices and converted to euros, €) for a subset of patients (n=212; 105
managed with WW and 107 with RP) participating in the trial. After a median followup of 11.8
years for the WW group and 12.2 years for the RP group, the total mean cost of WW was
€18,124 for WW compared with €24,147 for RP. After adjustment for age, Gleason score and
PSA, the difference between treatments remained and was statistically significant (P = 0.003).
The applicability of Swedish cost estimates to the United States setting is likely to be limited.

Findings From Model-Based Economic Evaluations
In addition to the primary cost studies discussed above, we also considered model-based cost
estimates, including the cost-minimization or cost-consequence analyses and cost components of
cost-effectiveness or cost-utility analyses of AS (but not WW). We did not review the cost-
effectiveness or cost-utility analyses reported in these documents and focused only on cost
information.
Two economic evaluations conducted by the Institute for Clinical and Economic Reviews
(ICER) 238,239 reported cost-utility analyses comparing AS with active treatment options (IMRT,
e
BT, and open retropubic nerve-sparing RP). An associated publication based on these reports
was also reviewed. 240 Cost estimates were obtained from multiple sources including outpatient
costs from the 2008 Red Book, Hospital Outpatient Prospective Payment System, Physician Fee
Schedule, Centers for Medicare & Medicaid Services Lab Fees and Durable Medical Equipment
Schedules; and inpatient payments from the Hospital Inpatient Prospective Payment System, the
2008 Anesthesia Conversion Factor and American Society of Anesthesiologists payment
information. The information obtained was incorporated into a disease history Markov model to
determine the cost (and effectiveness) of each treatment strategy. The AS protocol included
regular physical examinations, PSA testing, and rebiopsies (one year following diagnosis and
every 3 years thereafter). In the model, 61.1 percent of all patients originally assigned to AS
received treatment, using a lifetime horizon; 28.3 percent of all patients received treatment
within the first 5 years. In the base case analysis (cohort of 65 year old men diagnosed with low-
risk clinically localized prostate cancer, lifetime model horizon, 3 percent annual discount rate)
comparing AS (followed by IMRT with short term ADT in cases of progression, or IMRT alone
in cases of patient preference for active treatment) and RP, the total cost of an open RP
management strategy was estimated at $28,348 and the total cost of an AS strategy was
estimated at $30,422. In a similar model (cohort of 65 year old men diagnosed with low-risk
clinically localized prostate cancer, additional life expectancy of 16 years, lifetime model
horizon, 3 percent annual discount rate) the total costs were $30,422 for AS, $23,348 for RP,
$25,484 for BT, $37,861 for intensity-modulated RT and $53,828 for proton beam RT.
Sensitivity analyses produces largely consistent results.
241
One study reported a cost comparison of AS and RP. Costs were obtained from 2008
Medicare reimbursement rates. The costs for RP were based on Medicare reimbursements in
2008 for patients treated by a single, high-volume surgeon (volume was not reported). Several
AS protocols were modeled: they differed by biopsy frequency (within a year of diagnosis and

e ICER reports are available online at http://www.icer-review.org; last accessed September 30, 2011.

82