The COMPARE Study: Facility Costs Associated With Hysterectomy, Myomectomy, and the Sonata Procedure for Treatment of Uterine Fibroids


Elizabeth Brooks, PhD
TTi Health Research & Economics, Westminster, Md.
Linda Mihalov, MD
Benaroya Research Institute at Virginia Mason Medical Center, Seattle, Wash.
Dipak Delvadia, DO
Drexel University College of Medicine, Philadelphia,Pa.
Saifuddin Mama, MD
Cooper University Hospital, Camden, N.J.
Robert Bauserman, PhD
TTi Health Research & Economics, Westminster, Md.
Gretchen E. Makai, MD
Christiana Care Health Services Inc.
Matt Yuen, PhD
TTi Health Research & Economics, Westminster, Md.
Carter Little
TTi Health Research & Economics, Westminster, Md.
April Zambelli-Weiner, PhD
TTi Health Research & Economics, Westminster, Md.
David J. Levine, MD
Mercy Clinical Minimally Invasive Gynecology, St. Louis, Mo.

Abstract

Purpose: The COMPARE study examined facility costs of sonography-guided transcervical fibroid ablation (TFA) using the Sonata System, hysterectomy, and myomectomy for the treatment of symptomatic uterine fibroids.

Design and methodology: A facility-perspective cost analysis of hospital charge data in the perioperative and the 30-day postoperative period was conducted. Length of stay (LOS) and charge data for the TFA arm were collected in the SONATA trial. Hospital-specific cost-to-charge ratios (CCRs) were used to convert charges to facility costs. Charge data and LOS for comparator arms were collected retrospectively from the IBM MarketScan Commercial Database and converted to costs using Medicare national average CCRs.

Results: TFA patients had a statistically significant shorter mean LOS (5.1 hours) compared with hysterectomy (73 hours) or myomectomy (80 hours; all P<.001) patients. The mean facility cost of TFA was $7,701, lower than the $10,353 for hysterectomy and the $12,003 for myomectomy (all P<.001). TFA facility costs were also statistically significantly lower across all stratifications (P<.01) compared with hysterectomy and myomectomy subgroups for site of service (inpatient or outpatient), and/or use of robotic assistance. The lower facility costs were preserved in follow-up analyses that included route subtypes (e.g., laparoscopic, open, vaginal) and removal of high-cost outliers from comparator arms.

Conclusion: TFA using the Sonata System had a shorter LOS than either comparator arms regardless of the procedure route. Sonata was also associated with considerably lower facility costs compared with either hysterectomy or myomectomy regardless of procedure route, site of service, or use of robotic assistance.

Keywords: Uterine fibroids; Sonata; health care resource utilization; facility perspective; database analysis; transcervical fibroid ablation; TFA, uterine preserving; cost analysis; radiofrequency ablation.

Introduction

Uterine fibroids are benign smooth muscle tumors that are seen in more than 70% of women of childbearing age. While many women with uterine fibroids are asymptomatic, between 20% and 50% of women with uterine fibroids experience symptoms such as abnormal uterine bleeding, pelvic pressure, and subfertility (Wallach 1981, Evans 2007, Pritts 2001). Uterine fibroids place a substantial burden on the U.S. health care system and society — an estimated $4.1 billion to $9.4 billion annually in direct medical costs (2010 dollars) and further losses in work productivity (Cardozo 2012, Côté 2002).

The most common treatments for symptomatic uterine fibroids involve major surgical approaches. Hysterectomy is the most common. It is curative but involves the removal of the uterus and therefore childbearing potential. Myomectomy, the surgical removal of individual uterine fibroids, preserves the uterus (Lethaby 2001).

The adoption of new treatments no longer depends just on safety and effectiveness data. The net impact on health care resource utilization is becoming a large factor and is being scrutinized by payers and providers.

A new treatment alternative for symptomatic uterine fibroids is the Sonata procedure (Sonata System, Gynesonics Inc., Redwood City, Calif.), a minimally invasive, incisionless, transcervical fibroid ablation (TFA) treatment that preserves uterus. The heat-induced coagulative necrosis that results from the targeted ablation causes fibroids to shrink over time and is associated with symptom relief (Garza-Leal 2011, Bongers 2015). The pivotal SONATA investigational device exemption (IDE) trial demonstrated statistically significant symptom relief, improved quality of health outcomes, zero device-related adverse events, and high patient satisfaction with a low rate of surgical reintervention (Chudnoff 2019).

The FDA granted the Sonata System 510(k) clearance in August 2018. Because cost considerations are increasingly important factors in a new treatment’s adoption, a cost analysis (the COMPARE study) was conducted to compare facility costs of TFA with those of hysterectomy and myomectomy, using a combined prospective and retrospective cohort study design.

Methods

Data sources and study population

COMPARE is a mixed-mode (study data were obtained from clinical trial and Truven claims data bases), short-term (30-day), facility-perspective cost analysis of uterine fibroid treatment procedure costs. Length of stay (LOS) and charge data for the TFA arm were collected during the SONATA pivotal IDE clinical trial, a prospective, longitudinal, multi-center, interventional trial that en-rolled patients between April 2015 and October 2016.

For the hysterectomy and myomectomy arms, LOS and charge data were collected retrospectively for a commercially insured population from a nationally representative claims data base (Truven Health MarketScan Commercial Database) for the period from July 1, 2012, to June 30, 2014.

Criteria for participation in the SONATA clinical trial (the TFA arm of COMPARE) have been previously detailed (Chudnoff 2019). For the hysterectomy and myomectomy arms, the claims database analysis consisted of women, ages 25 to 50, who had one of the following: a claim with a uterine fibroid diagnosis on the same day as the index procedure; an inpatient claim with a uterine fibroid diagnosis at any time in the 12-month pre-index period; or at least two out-patient claims, occurring 30 days apart or more, with a uterine fibroid diagnosis at any time in the 12-month pre-index period. Hysterectomy and myomectomy procedures of interest were identified and categorized into procedure route subtypes using appropriate ICD-9-CM and CPT codes (Table 1). Patients who were included must have had no uterine fibroid treatment procedure in the 42-month pre-index period, no diagnosis associated with post-menopause in either the 42-month pre-index period or the 24-month post-index period, no diagnosis for an extrauterine pelvic mass or abdominal/pelvic malignancy in the 24-month post-index period, and 42 months of pre-index and 24 months of post-index continuous medical and pharmacy enrollment.

The hysterectomy arm included data from patients who underwent abdominal and laparoscopic procedures, both supracervical or total hysterectomies; or vaginal hysterectomies.

The myomectomy arm included data from patients who underwent laparoscopic or open myomectomies.

TABLE 1
Codes used for procedure identification
Procedure ICD-9 CPT
Myomectomy codes
Myomectomy, unspecified 68.29 58140, 58145, 58146
Laparoscopic   58545, 58546
Hysterectomy codes
Laparoscopic-assisted vaginal 68.51 58550, 58552, 58553, 58554
Abdominal supracervical 68.39 58180
Vaginal 68.59 58260, 58262, 58263, 58267, 58270, 58275, 58280, 58290, 58291, 58292, 58293, 58294
Laparoscopic supracervical 68.31 58541, 58542, 58543, 58544
Total abdominal 68.49 58150, 58152
Total laparoscopic 68.41 58570, 58571, 58572, 58573
Use of robotic assistance
Any 17.41, 17.42, 17.49 S2900
Outcomes

The primary outcomes of interest were LOS and the facility costs associated with the index procedure (i.e., TFA, hysterectomy, or myomectomy) and the index procedure-related hospitalization. As part of the SONATA trial protocol, participating sites submit-ted institution billing forms (UB-04) and/or provider billing forms (CMS 1500) for all charges accrued during the enrolled patients’ index procedure and hospitalization. The charges were converted into facility costs using the Medicare cost-to-charge ratio (CCR) for the facility where each patient’s procedure was performed. The CCRs were obtained from the American Hospital Directory (American Hospital Directory).

The CCRs of participating sites ranged from 0.07 to 1.05. For the hysterectomy and myomectomy arms, LOS and charges associated with the index procedure and the index procedure hospitalization were obtained from Truven Health Analytics as patient-level charge data. Charges were then converted into costs using the national average Medicare CCR of 0.30 (National Nurses United). All costs are presented in 2016 U.S. dollars.

Analyses

All analyses were conducted in Stata (version 13; StataCorp LLC, College Station, Texas). Patient characteristics were reported. A chi-square analysis was used to determine the difference between groups. Means for total payments, payments for the index procedure and LOS (defined as time of admission to time of discharge or time eligible to be discharged), and payments for post-operative health care resource utilization were reported for all study arms. The Welch t-test for unequal variances was used to compare mean facility costs between TFA and the comparator arms. For comparisons, a P value of <.05 was considered statistically significant. Additionally, the Bonferroni correction was made to account for multiple comparisons. For statistical comparisons where the Bonferroni correction was applied, the new significance level is indicated in the table footnote. For sensitivity analysis, high-cost outliers were removed from the hysterectomy and myomectomy arms. High-cost outliers were defined as procedures with costs more than 1.25 times the interquartile range (IQR).

Patient characteristics

Of the original 22 sites in the pivotal SONATA trial, five that had treated 45 patients in the IDE were able to participate in COMPARE. For the comparator arms, 35,463 patients who underwent hysterectomy and 5,228 patients who underwent myomectomy were included. Table 2 describes the characteristics of study patients. Women in the TFA arm were older (mean age, 45 ± 3.8 years) than those undergoing other procedures (P<.001). There were also statistically significant differences in the regions between TFA and the comparator arms.

TABLE 2
Characteristics of study participants
Characteristic TFA
(n=45)
Hysterectomy
(n=35,463)
Myomectomy
(n=5,228)
P valuea
Age (years)b 45 ± 3.8 42 ± ٥.٠ 37 ± 5.5 <.001
Region, n (%)
Northeast 11 (24.4%) 4,165 (11.7%) 1,063 (20.3%) <.001
North Central 22 (48.9%) 6,992 (19.7%) 691 (13.2%)  
South 3 (6.7%) 18,925 (53.4%) 2,724 (52.1%)  
West 9 (20.0%) 5,134 (14.5%) 721 (13.8%)  
Unknown 247 (0.7%) 29 (0.6%)  
aP value for age and LOS=t-test, assuming unequal variances; for region=chi-square test
bValues presented as mean ± standard deviation
TFA=transcervical fibroid ablation.

RESULTS

As summarized in Table 3 (page 43), patients treated with TFA had statistically significantly shorter mean LOS (5 ± 1.5 hours), compared with patients treated with hysterectomy (73 ± 39.1 hours) or myomectomy (80 ± 34.3 hours) (both P<.001). Furthermore, the mean LOS of TFA with Sonata was statistically significantly less than mean LOS for hysterectomy and myomectomy route subtypes with the shortest LOS. The hysterectomy route subtype with the shortest LOS was laparoscopic supracervical hysterectomy (57 ± 26.6 hours; P<.001). The myomectomy route subtype with the shortest LOS was laparoscopic myomectomy (67 ± 39.5 hours; P<.001).

TABLE 3
Length of hospital stay
Index procedure Mean LOS (hours)a P valueb
TFA (n=45) 5 ± 1.5 (Ref)
Hysterectomy (n=35,463) 73 ± 39.1 <.001
Abdominal supracervical hysterectomy (n=190) 80 ± 34.2 <.001
Laparoscopic supracervical hysterectomy (n=4,991) 57 ± 26.6 <.001
Total abdominal hysterectomy (n=5,975) 81 ± 42.1 <.001
Total laparoscopic hysterectomy (n=12,143) 58 ± 31.3 <.001
Vaginal hysterectomy (n=10,719) 57 ± 27.1 <.001
Hysterectomy, unspecified (n=1,445) 82 ± 38.3 <.001
Myomectomy (laparoscopic and unspecified) (n=5,228) 80 ± 34.3 <.001
Laparoscopic myomectomy (n=2,186) 67 ± 39.5 <.001
Myomectomy, unspecified (n=3,042) 80 ± 33.9 <.001
aValues presented as mean ± standard deviation
bt-test, assuming unequal variances
LOS=length of stay, TFA=transcervical fibroid ablation.
Facility costs

Results from the primary analysis are presented in Figure 1 (page 43). Mean costs for the facility to treat uterine fibroids were statistically significantly lower for TFA ($7,701 ± 2,558) than either hysterectomy ($10,353 ± 7,155; P<.001) or myomectomy ($12,003 ± 9,344; P<.001). Table 4 (page 43) provides an additional breakdown of procedure costs by site of service or use of robotic assistance. Regardless of stratification, the hysterectomy and myomectomy arms were associated with statistically significantly higher mean facility costs (P<.001) as compared with TFA.

FIGURE 1
Mean facility cost per procedure

Figure 1
Error bars=standard error
TFA=transcervical fibroid ablation.
TABLE 4
Breakdown of mean total costs by site of service or use of robotic assistancea
  TFA ($) Hysterectomy ($) Myomectomy ($)
Site of service Outpatient 7,701 ± 2,558
(n=45)
9,556 ± 6,393
(n=24,297)
11,103 ± 9,284
(n=3,007)
Inpatient 12,086 ± 8,324
(n=11,166)
13,221 ± 9,289
(n=2,221)
Robotic assistance None 7,701 ± 2,558
(n=45)
10,120 ± 6,996
(n=32,959)
11,683 ± 9,016
(n=4,936)
Assisted 13,412 ± 8,419
(n=2,504)
17,413 ± 12,597
(n=292)
aValues presented as mean ± standard deviation
Note: All P values for comparing hysterectomy and myomectomy subgroups to TFA <.001, using t-test, assuming unequal variances.
TFA=transcervical fibroid ablation.

When the combination of site of service and use of robotic assistance was used to stratify and compare, the costs of TFA remained statistically lower than either hysterectomy or myomectomy arms (P<.05) (Table 5, page 44). As expected, the least costly subgroups in hysterectomy and myomectomy arms were the outpatient non–robotic-assisted procedures (mean of $9,144 and $10,437, respectively), which were statistically significantly higher than the mean cost of the TFA treatment ($7,701, all P<.001).

Table 5
Breakdown of mean total costs by site of service and use of robotic assistance
Index procedure Site of service Robotic assistance Mean cost ($)a P valueb
TFA Outpatient None
(n=45)
7,701 ± 2,558 (Ref)
Hysterectomy Inpatient Assisted (n=142) 13,949 ± 8,214 <.001
None (n=11,024) 12,062 ± 8,323 <.001
Outpatient Assisted (n=2,362) 13,380 ± 8,432 <.001
None (n=21,935) 9,144 ± 5,989 <.001
Myomectomy Inpatient Assisted (n=11) 13,594 ± 5,113 .003
None (n=2,210) 13,219 ± 9,305 <.001
Outpatient Assisted (n=281) 17,563 ± 12,782 <.001
None (n=2,726) 10,437 ± 8,576 <.001
avalues presented as mean ± standard deviation
bt-test, assuming unequal variances; statistical significance based on Bonferroni correction of .005
TFA=Transcervical fibroid ablation

Removing high-cost outliers from hysterectomy and myomectomy arms verified the robustness of these findings (Figure 2). High-cost outliers were identified as hysterectomy and myomectomy procedures with greater than $20,803 or $25,475, respectively (i.e., costs >1.25 x IQR). When high-cost outliers were omitted, the mean facility cost for the index procedure decreased to $9,115 for hysterectomy and $10,475 for myomectomy. Even with high-cost outliers removed, TFA continued to have statistically significantly lower facility costs than hysterectomy and myomectomy procedures (P<.001).

FIGURE 2
Mean facility cost per procedure with highcost outliers excluded in hysterectomy and myomectomy groups

Figure 2
*Index costs >$20,803 omitted
**Index costs >$25,475 omitted
Error bars=standard error
TFA=transcervical fibroid ablation.

Table 6 summarizes facility costs stratified by procedure route. TFA was associated with statistically significantly lower mean facility costs ($7,701) across all stratifications of hysterectomy and myomectomy procedures (all P<.001).

Table 6
Comparison of TFA, hysterectomy, and myomectomy costs
Procedure Mean cost ($)a P valueb
TFA (n=45) 7,701 ± 2,558 (Ref)
Hysterectomy (n=35,463) 10,353 ± 7,155 <.001
Abdominal supracervical hysterectomy (n=190) 12,479 ± 6,655 <.001
Laparoscopic supracervical hysterectomy (n=4,991) 10,228 ± 6,564 <.001
Total abdominal hysterectomy (n=5,975) 11,429 ± 8,360 <.001
Total laparoscopic hysterectomy (n=12,143) 10,821 ± 7,031 <.001
Vaginal hysterectomy (n=10,719) 9,139 ± 6,527 <.001
Hysterectomy, unspecified (n=1,445) 11,116 ± 7,920 <.001
Myomectomy (n=5,228) 12,003 ± 9,344 <.001
Laparoscopic myomectomy (n=2,186) 13,245 ± 9,727 <.001
Myomectomy, unspecified (n=3,042) 11,110 ± 8,955 <.001
*Values presented as mean ± standard deviation
**t-test, assuming unequal variances
TFA=transcervical fibroid ablation.

DISCUSSION

This is the first study to compare the facility costs of the Sonata procedure with those of hysterectomy and myomectomy. The Sonata System has European Union and FDA clearance, and the multisite, longitudinal, prospective SONATA IDE pivotal trial provides the first cost data source for this treatment (Garza-Leal 2011, Bongers 2015, Chudnoff 2019). The use of the Truven Health MarketScan Commercial Database gave this study a higher statistical power with larger sample sizes. Using claims databases for outcomes comparisons has some drawbacks (Koch 2012). For example, with the comparator arms, certain demo graphic characteristics that may affect outcomes, such as race, socioeconomic status, education level, and health behaviors, (e.g., tobacco use), could not be collected. Additionally, Heisler (2009) showed that administrative claims data can be inconsistent, an issue that could have an effect on the results of this study. However, despite the potential shortfalls of administrative claims data, the results found in this analysis were similar to those reported in the peer-reviewed literature (Becker 2007). The hysterectomy and myomectomy results obtained in this analysis are consistent with studies that reported hysterectomy and myomectomy patient LOS ranging from two to three days (Becker 2007, Siskin 2006, Song 2017). The costs reported in these studies were also similar for total abdominal hysterectomy ($10,236), supracervical hysterectomies ($11,009), and myomectomies ($10,844) when converted from 2003 dollars to the 2016 dollar standard utilized for this analysis (Becker 2007).

CONCLUSION

TFA using the Sonata System had a statistically significant shorter LOS than either comparator arm, irrespective of procedure route. Sonata was also associated with statistically significantly lower facility costs compared with either hysterectomy or surgical myomectomy regardless of procedure route, site of service, or use of robotic assistance. Sonata treatment provided a 26% to 42% reduction in facility costs when compared with hysterectomy and myomectomy, respectively. The primary source of this comparative savings was the reduced usage of hospital resources.

Correspondence:
Elizabeth Brooks
1231 Tech Court Suite 201 Westminster, MD 21157
Tel: (800) 580-2990
ebrooks@tti-research.com

Disclosures:
Gynesonics, the manufacturer of the Sonata Sonography-Guided Transcervical Fibroid Ablation System (the Sonata System, previously known as the VizAblate System), which is discussed in this article, provided fund-ing (supporting Brooks, Yuen, Little, and Zambelli-Weiner) for the research presented in this manuscript. Mihalov reports receiving an honorarium or other financial benefit from Gynesonics, receiving research support from Gynesonics, and serving as a paid consultant or advisory board member for Gynesonics. Mama reports participating in the SONATA clinical trial. Delvadia, Makai, and Levine report no disclosures.

References

  • American Hospital Directory. https://www.ahd.com [registration required]. Accessed July 28, 2019.
  • Becker ER. National trends and determinants of hospitalization costs and lengths-of-stay for uterine fibroids procedures. J Health Care Finance. 2007;33(3):1–16.
  • Bongers M, Brölmann H, Gupta J, et al. Transcervical, intrauterine ultrasoundguided radiofrequency ablation of uterine fibroids with the VizAblate System: three- and six-month endpoint results from the FAST-EU study. Gynecol Surg. 2015;12(1):61–70.
  • Cardozo ER, Clark AD, Banks NK, et al. The estimated annual cost of uterine leiomyomata in the United States. Am J Obstet Gynecol. 2012;206(3):211.e1–9.
  • Chudnoff S, Guido R, Roy K, et al. Ultrasound-guided transcervical ablation of uterine leiomyomas. Obstet Gynecol. 2019;133(1):13–22.
  • Côté I, Jacobs P, Cumming D. Work loss associated with increased menstrual loss in the United States. Obstet Gynecol. 2002;100(4):683–687.
  • Evans P, Brunsell S. Uterine fibroid tumors: diagnosis and treatment. Am Fam Physician. 2007;75(10):1503–1508.
  • Garza-Leal JG, Toub D, León IH, et al. Transcervical, intrauterine ultrasoundguided radiofrequency ablation of uterine fibroids with the VizAblate System: safety, tolerability, and ablation results in a closed abdomen setting. Gynecol Surg. 2011;8(3):327–334.
  • Heisler CA, Melton LJ, Weaver AL, Gebhart JB. Determining perioperative complications associated with vaginal hysterectomy: code classification versus chart review. J Am Coll Surg. 2009;209(1):119–122.
  • Koch CG, Li L, Hixson E, et al. What are the real rates of postoperative complications: elucidating inconsistencies between administrative and clinical data sources. J Am Coll Surg. 2012;214(5):798–805.
  • Lethaby A, Vollenhoven B, Sowter M. Preoperative GnRH analogue therapy before hysterectomy or myomectomy for uterine fibroids. Cochrane Database Syst Rev. 2001;(2):CD000547.
  • National Nurses United. IHSP Calculations of federal hospital cost reports, FY 1996–1997 through 2011–2012. http://nurses.3cdn.net/e2086b18382cb8e96d_0dm6b5ad3.pdf. Accessed July 28, 2019.
  • Pritts EA. Fibroids and infertility: a systematic review of the evidence. Obstet Gynecol Surv. 2001;56(8):483.
  • Siskin GP, Shlansky-Goldberg RD, Goodwin SC, et al. A prospective multicenter comparative study between myomectomy and uterine artery embolization with polyvinyl alcohol microspheres: long-term clinical outcomes in patients with symptomatic uterine fibroids. J Vasc Interv Radiol. 2006;17(8):1287–1295.
  • Song T, Kim M, Kim ML, et al. Impact of opportunistic salpingectomy on anti-Müllerian hormone in patients undergoing laparoscopic hysterectomy: a multicentre randomised controlled trial. BJOG. 2017;124(2):314–320.
  • Wallach EE, Buttram VC, Reiter RC. Uterine leiomyomata: etiology, symptomatology, and management. Fertil Steril. 1981;36(4):433–445.

UP NEXT

Our Yearlong Look: Emergency Care Examined
By Robert Calandra

A blueprint for high-volume, high-quality lung cancer screening that is detecting cancer earlier—and helping to save lives

Clinical Brief

Multiple Sclerosis: New Perspectives on the Patient Journey–2019 Update
Summary of an Actuarial Analysis and Report