Clinical Inputs
We searched the PubMed MEDLINE database for BCS outcomes related to average (invasive and DCIS) re-excision rates in the United States, published since 2014.5-8,13,14,23-34 We also searched reduction in re-excision as a result of utilizing FCS13,14,34,35 and reduction in re-excision as a result of utilizing MarginProbe16-22,36,37during the same timeframe. The ratio of reoperations as re-excision BCS vs conversion to mastectomy was researched for both single and multiple reoperations after BCS.28,38-46 Further, we reviewed the literature for rate of mastectomy that is bilateral, as well as the average rate of implant reconstruction after mastectomy.47-49 The American Society of Plastic Surgeons 2020 Plastic Surgery Statistics Report provided details on timing of reconstruction (immediate or delayed).50 All mastectomy data was limited to the United States to avoid bias as a result of different decision guidelines in other healthcare systems. Our own institutional data was analyzed to supplement these references.
Using these clinical inputs, a decision tree was created of the three possible surgical care pathways: BCS alone, BCS+FCS, or BCS+MarginProbe. (Figure 1)
Financial Inputs
Fairhealth.org was used to gather commercial cost data. Medicare data was collected from CMS.gov. To complement this comparison, we gathered data on the number of people who have healthcare insurance from the most recent U.S. Census Bureau report on healthcare coverage in the United States51 which included a breakdown of types of insurance. The CPT codes used for the purposes of this study are found in Table 1.
Table 1. CPT Codes used for cost-effectiveness comparison of FCS vs MarginProbe
CPT Code
|
Procedure
|
19301
|
Mastectomy, partial (e.g., lumpectomy, tylectomy, quadrantectomy, segmentectomy)
|
19303
|
Mastectomy, simple, complete
|
88307
|
Surgical pathology, gross and microscopic examination
|
19340
|
Insertion of breast implant on same day of mastectomy (i.e., immediate)
|
19342
|
Insertion or replacement of breast implant on separate day from mastectomy
|
Fairhealth.org does not provide national average costs, therefore 11 ZIP codes were chosen from 11 different states geographically covering all regions of the United States. The average cost between these states provided a surrogate for a national average cost. Cost variation by state was also calculated. The states and corresponding ZIP codes used are depicted in Table 2.
Table 2. States and ZIP codes comprising national average cost calculations
ZIP Code
|
State
|
ZIP Code
|
State
|
ZIP Code
|
State
|
ZIP Code
|
State
|
99164
|
Washington
|
80523
|
Colorado
|
55959
|
Minnesota
|
10001
|
New York
|
93505
|
California
|
66101
|
Kansas
|
49254
|
Michigan
|
21231
|
Maryland
|
59634
|
Montana
|
77590
|
Texas
|
32313
|
Florida
|
|
|
MarginProbe costs were provided directly by the device manufacturer. A one-time cost for the MarginProbe is priced at $50,000. The list price for the disposable probe is $995 per procedure. Actual pricing for console and disposable probes may be lower based on IDN contract pricing and negotiated procedure volume discounts. All cost-effectiveness analyses use device list price unless stated otherwise.
Price comparison formula
Using the clinical and financial inputs in combination with the Decision Tree, a cost formula was created for each surgical pathway (Figure 2) for input into a Pro-Forma Model.
N = Number of Patients
|
BCS = Breast Conserving Surgery
|
Recon = Reconstruction
|
R = Rate
|
FCS = Full Cavity Shave
|
IR = Immediate Reconstruction
|
C = Cost
|
MP = MarginProbe
|
DR = Delayed Reconstruction
|
Re = Reoperation
|
TM = Total Mastectomy
|
BI = Bilateral Mastectomy
|
Cost-Effectiveness Analysis
We referenced the Cost-Effectiveness Economic Evaluation framework provided by the Office of the Associate Director for Policy and Strategy on the cdc.gov website52 to perform three analyses for output on the Pro-Forma: 1) The cost-effectiveness of adding FCS to BCS, 2) the cost-effectiveness of adding MarginProbe to BCS, and 3) the cost-effectiveness of MarginProbe and FCS compared to each other.
Re-excision rate reduction as a result of implementing either MarginProbe or FCS to BCS alone were set by default to 50%. The rate of reduction can be adjusted in the Pro-Forma. The MarginProbe console and disposable probe costs are set at list price by default but can be adjusted in the Pro-Forma.
Net costs/net savings were calculated by subtracting the formula-calculated pathway cost from the pathway cost of its comparator. i.e., the net cost of adding FCS to BCS is the pathway cost of FCS minus the pathway cost of BCS alone. Net costs/net savings are presented on both a per patient basis and total annual basis based on BCS volume inputs in the Pro-Forma. When the intervention improves re-excision rates but is more costly, the cost-effectiveness is also presented as a ratio, reported as “cost per re-excision prevented.”