Impact of High-Deductible Health Plans on Breast Reconstruction: Considerations for Financial Toxicity

Abstract

Background

High-deductible health plans (HDHP) are used within the US to curb unnecessary healthcare spending; however, the resulting increased out-of-pocket (OOP) costs may be associated with financial toxicity. The aim was to assess the impact of HDHPs on use and seasonality of mastectomy and breast reconstruction procedure. The hypothesis is that the high OOP costs of HDHPs will lead to decreased overall service use and greater fourth quarter use after the deductible has been met.

Methods

MarketScan was queried from 2014–2017 for episodes of mastectomy, BR (immediate and delayed), breast revision and reduction. Only patients continuously enrolled for the full calendar year after the index operation were included. HDHP and low deductible health plans (LDHP) were compared based on OOP cost-sharing. Outcomes included surgery utilization rates, seasonality of operations, and median/mean OOP costs.

Results

Annual mastectomy and BR utilization rates varied little between LDHP and HDHP. Mastectomies, delayed BR, and elective breast procedures (p<0.001) all showed significant increase in 4^th quarter utilization whereas immediate breast reconstruction did not. Regardless of timing and reconstruction method, HDHP had significantly greater median OOP compared to LDHP (all p< 0.001).

Conclusion

Mastectomy and breast reconstruction rates did not differ between LDHPs and HDHPs, but seasonality for all breast procedures was measured with the exception of immediate breast reconstruction, suggesting that women are rational economic actors. Regardless of service timing and reconstruction modality, HDHP patients had greater OOP costs compared to LDHP, and serves as a good starting point for provider engagement in financial toxicity.

Introduction

Financial toxicity describes the adverse effects of financial burden and stress put on patients that result from healthcare treatments and expenses [1, 2]. This phenomenon is particularly relevant in oncology due to high treatment and diagnostic costs, prevalence of multimodal therapies (e.g. chemotherapy, radiation, surgery), unemployment during prolonged treatment regimens, and difficulties with return to the workforce.

In an effort to restrict unsustainable healthcare expenditures, a number of changes have been made to US health policy and insurance design [3]. The term moral hazard describes the notion that insurance coverage may be associated with greater health care use because patients do not bear full financial responsibility for treatments received [4]. As such high-deductible health plans (HDHP) have been introduced, these insurance instruments place greater financial responsibility or “first dollar risk” on patients via a higher deductible, working under the premise that unnecessary health care utilization will be minimized by incentivizing greater discretion among enrollees [5]. For the year 2020, the Internal Revenue Service defined HDHP as any plan with an annual deductible limit of at least $1,400 for an individual and $2,800 for a household[6]. While evidence suggests HDHPs may curb some unnecessary healthcare consumption [7, 8], higher out-of-pocket costs can also lead to avoidance of preventative service and potentially worsen health [9]. Moreover, greater healthcare expenditures can be burdensome to patients leading to the unintended consequence of financial toxicity.

Breast cancer is a costly illness often requiring multimodal treatments and reconstructive services. Breast reconstruction (BR) has demonstrated significant benefits in health-related quality of life [10]; however, it is preference sensitive and as a result some patients may elect to forego it for a variety of reasons including advanced disease stage, need for adjuvant radiation, extended recovery, anticipated costs[11, 12], and time off from work. When performed in a delayed fashion, breast reconstruction may lengthen the episode of cancer care, accruing additional costs that may be passed onto the patient.

The impact of HDHPs on utilization patterns of mastectomy and BR procedures has not been evaluated. The aim of the current study is to measure associations between mastectomy and breast reconstruction service utilization with the type of health plan deductible. The hypothesis is that the high out-of-pocket costs of HDHPs will lead to decreased overall service utilization and greater fourth quarter use after the deductible has been met.

Methods

Data Source

MarketScan is a large, national multi-payer database that captures the health claims data on employer-insured patients. It is well represented in the health services research literature and dataset features have been described elsewhere[13]. In brief, the MarketScan databases include data on healthcare utilization across a variety of settings including physician office visits and hospital stays. Out-of-pocket (OOP) healthcare expenditures such as insurance copays, deductibles, and coinsurance are included. Diagnoses are coded on 99% of all claims while procedures are coded on 85% of physician claims making it a robust dataset with little missing data [14]. MarketScan years used in the current study were 2014–2017.

Plan Types

MarketScan has eight plan types: exclusive provider organizations (EPO), health maintenance organizations (HMO), point of service (POS) with Capitation, POS, Comprehensive, preferred provider organizations (PPO), consumer-directed health plans (CDHP), and high deductible health plans (HDHP). These plans were categorized as low, intermediate, and high OOP cost share groups where OOP expenditure was defined as payment towards a patient’s deductible, coinsurance, and copayment. The median value and approximated 95% confidence intervals (95% CI) for OOP were obtained and used to define OOP cost-sharing groups [15]. For this study, EPO, HMO and POS with capitation were considered “low cost-sharing” or “low deductible health plans” (LDHP), Comprehensive, PPO and POS were “intermediate”, and HDHP and CDHP were determined to be “high cost-sharing” or “high deductible health plans” (HDHP). Comparisons in the current study are made between the LDHP and HDHP groups.

Study Population

MarketScan was queried for mastectomy codes (Procedure codes: ‘19303’,’19304’,’19305’,’19306’,’19307’,’8534’,’8536’,’8541’,’8542’,’8543’,’8544’,’8545’,’8546’,’8547’,’8548’,’0HTT0ZZ’,’0HTU0ZZ’,’0HTV0ZZ’,’19301’,’19302’,’19303’,’19304’,’19305’,’19306’,’19307’,’8533’,’8534’,’8535’,’8536’,’8541’,’8542’,’8543’,’8544’,’8545’,’8546’,’8547’,’8548’,’0H0T0JZ’,’0H0T3JZ’,’0H0U0JZ’,’0H0U3JZ’,’0H0V0JZ’,’0H0V3JZ’,’0HRT0JZ’,’0HRT3JZ’,’0HRU0JZ’,’0HRU3JZ’,’0HRV0JZ’,’0HRV3JZ’,’0HTT0ZZ’,’0HTU0ZZ’,’0HTV0ZZ’) and breast reconstruction codes (CPT: ‘S2068’,’S2066’,’S2067’,’19367’,’19364’,’19361’,’19340’,’19357’, ‘19318’, ‘19380’) from January 2014 to December 2017. Only female patients who were 18 years or older were included. Patients were excluded if they were not continuously enrolled for the full calendar year in which the procedure occurred. To understand the impact of OOP cost-sharing on mastectomy and BR rates, mastectomy alone patients were defined as patients with only a mastectomy and no BR. Immediate BR patients were defined as patients with a mastectomy and BR on the same day. Delayed reconstruction patients were defined as patients with a mastectomy and a reconstructive procedure that occurred in the next calendar year or beyond (e.g., mastectomy in 2014 and reconstruction in 2015) in order to isolate the cost of reconstruction in the absence of mastectomy and other cancer therapies.

Outcomes of Interest

Primary outcomes of interest were OOP cost per procedure which was defined as the total expenditure (copay, coinsurance, and deductible) incurred by a patient on the date of service if the procedure happened in an outpatient visit, or the full stay for inpatient visits. Service utilization rates were defined as the rate of procedure among all insured persons per plan group and reported as the rate per 100,000. Rates were compared between LDHP versus HDHP for each procedure. Seasonality was determined by comparing utilization rates during the last quarter compared to the first nine months of the year between LDHP and HDHP groups[13]. For these analyses, procedures considered somewhat more discretionary or elective such as breast reduction and breast revision were also evaluated (Procedure codes ‘19318’, ‘19380’)

Secondary outcomes of interest were comparing OOP for microvascular free flap reconstruction (Procedure codes: ‘S2068’,’S2066’,’S2067’,’19367’,’19364’,’19361’) versus tissue expander (TE)/implant reconstruction (Procedure codes: ‘19340’,’19357’) by timing of reconstruction.

Statistical Analysis

OOP estimates were adjusted to inflation, age, and geographical area while service utilization rates were adjusted by age and geographical area. Comparisons of service utilization rates by plan group were analyzed by a two-proportion p-test where the null hypothesis was that there is no difference between the proportions. Regarding seasonality, all four years were aggregated and weighted. The average utilization rate for the fourth quarter was compared to the weighted, average rate in the remaining nine months with a two-proportion p test. Median OOP was compared using a Kruskal Wallis Rank Sum test. All tests were two sided and p < 0.01 were considered significant. All analyses were conducted using SQL and R Statistical Software (packages: stats, tidyverse).

Results:

A total of 32,832,272 covered lives were included in this study where 3,358,167 were included in the LDHP group, 21,683,679 were included in the intermediate OOP cost share group, and 7,790,426 were included in the HDHP group. Both the average age and median income increased when moving from the high (33.13 years; $44,587.87) to low deductible (34.67 years; $51,074.62) groups.

Figure 1 displays density plots for each plan group for all continuously enrolled female patients in MarketScan who were continuously enrolled in one health insurance plan for the entire calendar year. Mean (SD) and median OOP expenditures increased when moving from low (mean: $416.88 [$715.35]; median: $146.40) to intermediate (mean: $912.95 [$1,274.89]; median: $395.30) to high (mean: $1,178.46 [$1,591.05]; median: $483.97) OOP cost-sharing groups.

Figure 1: Density plots of Out-of-pocket Expenditure for Low, Intermediate, and High Cost-Sharing Groups with Mean and Median Values.

Low (in red), intermediate (in green), and high (in blue) represent cost-sharing or plan group. Median values are represented by the dashed lines and mean values are represented by the solid lines.

Service Utilization Rates and OOP Cost: Mastectomy and Reconstruction Procedures

Table 1 describes service utilization rates and OOP cost (mean and median) by plan group. Overall utilization rates were greatest for mastectomy followed by immediate and thereafter by delayed reconstruction. Mastectomy rates were greater in the LDHP cohort at 45.8 cases per 100,000 persons, compared to 43.9 cases per 100,00 persons for the HDHP cohort. However, this was noted to be non-significant (p = 0.09). Immediate reconstruction service utilization was significantly greater for patients in the HDHP group at 17.4 cases per 100,000 compared with 15.6 cases per 100,000 persons in the LDHP group (p = 0.007). Rates of delayed reconstruction were lowest at 9.4 cases per 100,000 persons in the LDHP versus 8.9 cases per 100,000 persons in the HDHP group.

Table 1:

Count, Mean Out-of-pocket Expenditure (in $), Median Out-of-pocket Expenditure, and Service Utilization Rate per 100,000 by Cost-Sharing Group for Mastectomy, Immediate Reconstruction, and Delayed Reconstruction Procedures

Plan Group	Mastectomy				Immediate Reconstruction				Delayed Reconstruction
	N	Mean	Median	Rate	N	Mean	Median	Rate	N	Mean	Median	Rate
Low	8,723	$435.00	$102.40	45.79	2,961	$695.07	$128.00	15.57	1,773	$580.23	$87.00	9.38
High	12,102	$1,080.94	$697.95	43.90	4,815	$1,628.73	$741.71	17.44	2,463	$1,792.27	$1,132.25	8.89
P value	−			0.093	−			0.007	−			0.354

Abbreviations: N Count

P value calculated by 2-proportion test comparing rates of procedure by cost-sharing group

Seasonality: Mastectomy, Reconstruction, and Elective Breast Procedures

Figures 2–5 display seasonality of service utilization for mastectomy, breast reconstruction, and elective breast procedures by quarter for HDHP versus LDHP groups. For the LDHP group, the average mastectomy rate did not differ significantly throughout the year; however, the HDHP group had a significantly greater average fourth quarter rate of mastectomy use compared to the rest of the year (p < 0.001). For immediate reconstruction there was no seasonal differences in the average service utilization for either the LDHP or HDHP groups. For delayed reconstructions, the HDHP group also had significantly higher fourth quarter utilization rates compared to the rest of the year (p < 0.001), whereas the LDHP cohort did not differ (p = 0.230). Seasonality could also be seen for more elective breast procedures (Figure 5). For both the LDHP and HDHP groups, rates of breast reduction and revision were significantly greater in the fourth quarter compared to the rest of the year (all p < 0.001).

Figures 2: Seasonality of Service Utilization Rates per Quarter from 2014–2017 by Low versus High Cost-Sharing Groups for Mastectomy Procedures.

The top panel demonstrates seasonality of service utilization rates for mastectomy procedures in the low cost-sharing insurance group while the bottom shows these rates for the high cost-sharing insurance group

Figures 5: Seasonality of Service Utilization Rates per Quarter from 2014–2017 by Low versus High Cost-Sharing Groups for Elective Breast Procedures.

The top panel demonstrates seasonality of service utilization rates for breast reduction and breast revision procedures in the low cost-sharing insurance group while the bottom shows these rates for the high cost-sharing insurance group

OOP Cost: Flap versus TE/Implant

For immediate flap reconstructions (Table 2), the median OOP costs were greater in the HDHP (Flap: $477.57) compared to the LDHP group (Flap: $148.56; p < 0.001). Immediate tissue expander/implant procedures also had a significantly greater median OOP cost in the HDHP (tissue expander/implant, $281.32) versus the LDHP (tissue expander/implant, $60.78; P < 0.001) group.

Table 2:

Mean and Median Out-of-pocket Expenditure (in $) with 95% Confidence Intervals by Cost-Sharing Group for Immediate Flap and Implant Procedures

Plan Group	Flap				Implant
	Mean	Mean 95% CI (Lower, Upper)	Median	Median 95% CI (Lower, Upper)	Mean	Mean 95% CI (Lower, Upper)	Median	Median 95% CI (Lower, Upper)
Low	$687.83	(592.67, 782.98)	$148.56	(121.55, 268.02)	$563.54	(521.08, 606)	$60.78	(40.2, 92.84)
High	$1,475.82	(1,347.91, 1,603.73)	$477.57	(228.93, 656.06)	$1,270.41	(1,220.76, 1,320.07)	$281.32	(187.73, 360)
P value *	−		<0.001		−		<0.001

Abbreviations: 95% CI 95% Confidence Interval

^*p value calculated using Kruskal-Wallis Rank Sum Test of Median Values

Among delayed flap and TE/implant reconstructions (Table 3), median OOP costs were significantly greater in the HDHP versus LDHP groups (all p < 0.001). Median OOP costs for flaps were $916.32 versus $225.16 in the HDHP and LDHP groups respectively. Similarly, for tissue expander/implants, median costs were $1120.94 and $67.00 for the HDHP and LDHP groups, respectively.

Table 3:

Mean and Median Out-of-pocket Expenditure (in $) with 95% Confidence Intervals by Cost-Sharing Group for Delayed Flap and Implant Procedures

Plan Group	Flap				Implant
	Mean	Mean 95% CI (Lower, Upper)	Median	Median 95% CI (Lower, Upper)	Mean	Mean 95% CI (Lower, Upper)	Median	Median 95% CI (Lower, Upper)
Low	$730.08	(604.59, 855.58)	$225.16	(121.55, 303.88)	$498.70	(442.27, 555.12)	$67.00	(57.88, 100.51)
High	$1,691.22	(1,534.12, 1,848.33)	$916.32	(551.68, 1,281.99)	$1,680.18	(1,597.69, 1,762.67)	$1,120.94	(968.54, 1,265.74)
P value *	−		< 0.001		−		<0.001

Abbreviations: 95% CI 95% Confidence Interval

^*p value calculated using Kruskal-Wallis Rank Sum Test of Median Values

Discussion

While a recent increase in total out-of-pocket spending for BR has been described [16], to the authors’ knowledge the current study is the first investigation that specifically addresses the interaction between insurance plan cost-sharing and utilization of breast oncology and reconstructive surgical services. Specifically, the absolute differences in the incidence of mastectomies and delayed breast reconstruction were small and did not differ significantly between the high and low deductible plans. For immediate reconstruction, the rates of services were greater for HDHP compared to LDHP group, but the absolute differences were small, 17.4 versus 15.6 reconstructions per 100,000 insured lives. Importantly the direction of this finding does not suggest an economic disincentive. Taken together, the findings suggest that women do not forgo mastectomy or breast reconstruction even if greater financial obligations are present in the form of HDHPs. Recent survey evidence showed that 28% of women with breast cancer considered cost when making decisions regarding surgical treatment of their disease [12]; however, the current data does not translate into fewer overall procedures.

In contrast, the timing or periodicity of breast oncology and reconstructive services was impacted by insurance plan deductible. For mastectomy and delayed reconstruction, the incidence of surgery demonstrated seasonality in HDHP groups, but not LDHP groups, with a significant rise in these operations during the fourth quarter of the calendar year when the annual deductible was likely to have been met (Figs. 2 and 4). For more elective breast procedures, such as breast revision and reductions (figure 5), seasonality was demonstrated in both the low and high deductible groups. Overall, the findings suggest that women are rational economic actors, whereby they wait for their deductible obligation to be met prior to proceeding with surgical services. For more discretionary procedures, any deductible seems to contribute to periodicity whereas only high deductibles disincentivize the more essential procedures of mastectomy or delayed reconstruction.

Figures 4: Seasonality of Service Utilization Rates per Quarter from 2014–2017 by Low versus High Cost-Sharing Groups for Delayed Reconstruction Procedures.

The top panel demonstrates seasonality of service utilization rates for delayed reconstruction procedures in the low cost-sharing insurance group while the bottom shows these rates for the high cost-sharing insurance group

Initially, it would seem curious that mastectomy would demonstrate seasonality especially in absence of any evidence to suggest a temporal distribution for breast cancer. However, existing evidence indirectly supports this finding with delayed breast cancer diagnostics for women in HDHPs [9, 17]. Specifically, HDHP members experienced delays in receipt of imaging, biopsy, and early-stage breast cancer diagnosis compared with LDHP enrollees. Moreover, low-income women in HDHPs waited 8.7 months longer to receive chemotherapy than high income earners suggesting that low-income women are a vulnerable population in the setting of HDHPs. The net effect is that cost procrastination is magnified along the entire care continuum in HDHPs, culminating in increased fourth quarter mastectomy surgery rates.

The only procedure in the current study in which seasonality was not observed is immediate breast reconstruction. The reasons for this are unclear but may stem from deductible fulfillment as part of the oncologic care episode, especially for women who receive neoadjuvant chemotherapy. Alternatively, women may strongly desire immediate reconstruction after committing to a mastectomy, regardless of the associated deductible costs. These hypothesis generating concepts require further exploration.

Research examining correlations between HDHPs and surgical services is lacking relative the large number of studies which show decrease of preventative care, office visits, emergency department visits, and prescription drug use [8]. However, a recent analysis by Chhabra et al. [13] evaluated the utilization of bariatric surgery in HDHPs. In contrast to the current study, patients with higher cost-sharing were found to be significantly less likely to undergo surgery (e.g., 5 fewer operations for a $1000 increase in cost sharing). Similar to the current study though, there was significant seasonality of surgery, such that the incidence of bariatric surgery was greater in the 4th quarter compared to the 1st quarter of the year for all plan types; these variations were most pronounced in high cost-sharing plans. The concordant findings with the current study support the impact of insurance design on economic behavior whereby patients desire a procedure, yet postpone until the year’s end to minimize the financial impact.

Recent studies on financial toxicity focus on a lack of provider engagement in the financial aspects of treatment[11]. For example, 73% of breast cancer patients surveyed [18], who expressed financial concern about treatment, felt that their providers did not address cost issues related to their cancer care. For women and surgeons concerned about financial toxicity related to breast reconstruction, the present data provide real-world numbers of the US commercially insured population for comparison and planning. Providers can inform patients in HDHPs that they will be responsible for a significantly greater fee than if they had a LDHP, but the difference in cost between any immediate method of reconstruction is small (Table 2). Moreover, patients can be informed that delayed, compared to immediate, reconstruction is likely to have a significant financial impact, especially in a HDHP (median cost ~$1100; Table 1), since deductible fulfillment may be taking place in calendar year separate from the time of the initial diagnosis. Healthcare financing conversations may be challenging for some providers but should be included as part of the shared decision-making for breast cancer surgery. Patients should also be guided towards financial counseling resources for more granular information.

Beyond the popularization of HDHPs as a means to stem rising US healthcare costs, the Affordable Care Act (ACA) introduced strategies to reduce patient healthcare expenses, including requiring all private health plans to provide preventive services (e.g. cancer screening) with no copays or deductibles [19]. Yet, the largest systematic review to date [8] demonstrated that healthcare consumers tend to limit spending globally which includes skipping high-value and continuing to use low-value services [20]. Moreover, some of the recommendations on preventative services were written as evidence-based clinical practices and not insurance coverage rules, leaving the details open to interpretation and patients subject to unintended fees. Another ACA provision intended to minimize patient financial toxicity was annual OOP maximums. The out-of-pocket limit for a 2021 Marketplace plan cannot be more than $8,550 for an individual and $17,100 for a family; however, according to the US Federal Reserve [21], 37% of Americans do not have enough savings to cover an urgent $400 expense. In summary, despite some of the well-recognized successes of the ACA, such as limits on preexisting conditions, US spending as a percentage of GDP reached an all-time high of 18% in 2020.

The current study has some limitations which are worth mentioning. The current dataset does not provide individual enrollee information on the deductible amount, therefore, correlates between behavior and deductible amounts cannot be directly measured. Second, data on insurance premiums paid through employers by means of paychecks- were not available. This expenditure may influence some behaviors measured in the current study as well as contribute to financial toxicity. The dataset does not include information on out-of-pocket maximums for any given patient, which may also impact healthcare decision making. For example, even after a deductible is met, an insurance plan may include additional copayments or other forms of cost-sharing that continue until the OOP maximum has been reached. OOP maximums are proprietary to individual health plans and not accessible through commercially available databases. October is breast cancer awareness month so may be an unmeasured confounder influencing the rate at which women get screened, have mastectomies, and ultimately have reconstruction in the fourth quarter. Lastly, the data used in this study may not fully reflect changes that have occurred in healthcare utilization as a result of the ACA which was ratified in 2010. A follow-up study with newer data would be useful to see if trends found in this study persist.

Conclusion

Compared to LDHP, HDHPs were not associated with differences in rates of breast oncology and reconstructive surgical services. However, there was a notable correlation in timing, such that women were much more likely to undergo surgery, both potentially life-saving and elective, in the 4th quarter of the fiscal year especially in HDHPs. Providers should engage patients specifically about deductible type and timing of reconstruction, in order to provide enhanced information on financial implications of breast cancer treatment.

Figures 3: Seasonality of Service Utilization Rates per Quarter from 2014–2017 by Low versus High Cost-Sharing Groups for Immediate Reconstruction Procedures.

The top panel demonstrates seasonality of service utilization rates for immediate reconstruction procedures in the low cost-sharing insurance group while the bottom shows these rates for the high cost-sharing insurance group