Laparoscopic supracervical hysterectomy versus laparoscopic-assisted vaginal hysterectomy

Data were derived from the Thomson Reuters MarketScan Commercial Claims and Encounter Database (Commercial Database) from the time period July 1, 2006 through March 31, 2009. The database is constructed from claims and enrollment data provided by over 130 large employer-sponsored health plans from across the US and is Health Insurance Portability and Accountability Act (HIPAA) compliant. The Commercial Database contains the healthcare experience of privately insured individuals covered under a variety of fee-for-service, fully capitated, and partially capitated health plans. There were approximately 34.6 million covered lives in 2008. Utilization, outcomes, and cost data are captured across the full continuum of care for insurance reimbursable services delivered in all settings, including physician office visits, emergency room (ER) visits, hospital stays, and outpatient pharmacy claims. The age and sex distribution of patients in MarketScan are similar to that in the Medical Expenditure Panel Survey (MEPS), which is a nationally representative database. Women with evidence of LSH or LAVH between January 1, 2007 and September 30, 2008 were selected into 2 cohorts based on hysterectomy type, with July through December 2006 as the potential preperiod and October 2008 through March 2009 as the potential follow-up period. LSH was determined by the presence of a claim with any of the following codes: International Classification of Diseases, Ninth Edition, Clinical Modification (ICD-9-CM) procedure code 68.31, Current Procedural Terminology (CPT) codes 58541 – 58544, and CPT code 58548. Codes for LAVH included ICD-9-CM procedure code 68.51 and CPT codes 58550, 58552, 58553, and 58554. The date of the first LSH or LAVH procedure code in the selection period was assigned as the index date. Where both an ICD-9-CM procedure code and a CPT code indicating the same procedure appeared for the same patient, the ICD-9-CM procedure code in the inpatient setting was selected as the index date. Patients were required to have 6 months of continuous medical and prescription coverage prior to the index date and 6 months subsequent to the index date. The postindex period included the index date. Patients were excluded if they were <18 years of age or if they had a diagnosis of cancer in the pre- or postindex periods (malignant neoplasms, ICD-9-CM diagnosis codes 140.xx through 209.xx, and carcinoma in situ and neoplasms of uncertain behavior, ICD-9-CM diagnosis codes 230.xx through 239.xx). Patients were also excluded if they had index date procedure codes for both LSH and LAVH (possibly due to coding errors) or if the length of stay (LOS) associated with the index procedure exceeded 20 days (based on the 99th percentile distribution of LOS). A number of explanatory and outcome variables were determined. (Diagnosis, procedure, and drug codes used in their definition are available from the authors.) Demographic variables were measured at index and included age, insurance plan type, geographic region, and urban versus rural residence. Clinical variables included the primary diagnoses on the index procedure claim; physician type on the index procedure claim; year of hysterectomy; emergent presentation, defined as an ER record associated with an inpatient admission for the index LSH or LAVH; a flag for the inpatient setting for the index LSH or LAVH; Charlson Comorbidity Index (CCI) score, Deyo version, separately for the preindex period and the index date, as well as counts by disease category; specific conditions of interest occurring in the preindex period and on index date based on both primary and secondary diagnosis codes and procedure codes; pelvic or abdominal surgeries occurring in the preindex period; medications of interest occurring in the preindex period; primary diagnosis on index date, and uterine weight at index. Outcome variables included perioperative outcomes and gynecologic-related (GYN-related) postoperative resource utilization and costs. The perioperative outcomes included infection, procedure-specific complications, analgesic use, inpatient mortality, and injury. The number and percentage of patients with these outcomes were determined as were costs for claims containing diagnosis, procedure, and/or drug codes consistent with their definition within the 30 days following and including the index date. GYN-related postoperative outcomes included healthcare resource utilization for inpatient, ER, and outpatient service categories for claims containing codes consistent with a GYN condition within the 30 days following the index date or index admission discharge date and separately for the 6 months following the index date or index admission discharge date (excluding the index procedure or admission itself). Diagnosis codes for GYN-related conditions can be found in Appendix 1 . The number and percentage of patients, mean number of services, and costs were reported. The costs for claims processed under a fee-for-service arrangement were the allowed charges (ie, the actual amounts paid by primary and secondary insurers plus patient cost share amounts [ie, copayments and deductibles]). The costs for claims processed under a capitated arrangement were estimated using the average cost of noncapitated claims, by procedure, geographic region, and year. All costs were adjusted to 2008 dollars using the medical service component of the Consumer Price Index (CPI). Descriptive analysis was performed on study variables, with counts and percentages reported for categorical variables and means (standard deviations [SDs]) reported for continuous variables. Statistical tests of significance for differences between the LSH and LAVH cohorts were conducted, with chi-square tests used to evaluate differences for categorical variables and t tests for continuous variables. Multivariate analysis was performed to compare outcomes between the LSH and LAVH cohorts. Logistic regression models were used to estimate the impact of hysterectomy type on clinical outcomes and utilization, with calculation of odds ratios (ORs), and generalized linear models (GLM) were used to estimate the impact on costs, with calculation of the marginal impact (ie, incremental cost difference). All multivariate analysis controlled for patients’ demographic characteristics, comorbid conditions, evidence of surgery in the preperiod, uterine weight, and diagnosis on the hysterectomy date.

A total of 13,551 and 31,232 women were selected with LSH and LAVH, respectively, from January 1, 2007 through September 30, 2008 ( Table 1 ) . Among them, 407 had claims for both LSH and LAVH, and 3 had an LOS of >20 days. They were excluded from the study. After screening for all inclusion and exclusion criteria, the final sample consisted of 6,198 LSH patients and 14,181 LAVH patients. Sample Attrition Mean age was 43.2 and 43.5 years for the LSH and LAVH cohorts, respectively ( Table 2 ) . A greater percentage of younger and older women (18 to 24, 25 to 34, and 55 to 64 years) underwent LAVH, while the reverse was true for middle-aged women (35 to 44 and 45 to 54 years) (P<.05 for all). The majority of patients were covered under health maintenance organization (HMO) and preferred provider organization (PPO) plans, approximately 80% from each cohort. Patients resided primarily in the South (57.7% of LSH and 61.2% of LAVH patients), followed by the North Central and West regions, and were in predominantly urban areas (82.7% of LSH and 75.0% of LAVH patients). Demographic Characteristics There was a low level of comorbidity in both cohorts, as indicated by the CCI score ( Table 3 ) . The only significant difference in individual comorbidities was in renal disease in the preindex period, with a higher percentage of LAVH patients having renal disease. However, sample sizes were small, <1% in either cohort. Analgesic use occurred in 46% to 48% of LSH and LAVH patients in the preindex period, immunosuppressives in 16% to 17%, and antibiotics in 11% to 12%. Endometriosis was diagnosed (in both primary and secondary positions) in approximately 12% of the LSH and LAVH cohorts in the preindex period and in about 33% of the cohorts at index. In the preindex period, adhesions were diagnosed in 2% of each cohort, with all other conditions of interest occurring in <1% of patients. At index procedure, adhesions were diagnosed in 13.6% and 12.0% of the LSH and LAVH cohorts, respectively, and pelvic inflammatory disease was found in about 2% of each cohort. All other conditions of interest at index occurred in <1% of patients. Surgery in the preindex period was uncommon, with the most frequently performed procedures being adhesiolysis and ovarian cystectomy, occurring in 1% of LSH and LAVH patients. Clinical Characteristics N/A=not applicable, ie, characteristic not measured in period. Values were reported for mean CCI score and for counts of individual diseases used in score. Individual disease counts were reported for diseases where at least 1% of patients had disease in either the preperiod or at index or where P-value was significant at <.05. A greater percentage of the LSH cohort had dysfunctional uterine bleeding (32.6%) and leiomyomas (38.0%) as their primary diagnosis compared to the LAVH cohort (27.9% and 26.3%, respectively, P<.001 for both) ( Table 3 ) . Within the LAVH cohort, more patients had endometriosis (10.4%) and prolapse (8.0%) as the primary diagnosis, versus the LSH cohort (9.3%, P=.008 and 1.5%, P250 grams (16.4%), compared to LAVH patients (10.3%, P<.001). Physician specialty at index for over 75% of each cohort was obstetrics and gynecology (data not reported in table). LAVH patients were significantly more likely to have their procedures performed in the inpatient setting than LSH patients (49.5% versus 36.6%, P<.001) (data not reported in table). Less than 1% of LSH and LAVH patients had their inpatient procedure performed following an ER visit (data not reported in table). LAVH patients had significantly higher overall infection rates compared to LSH patients (7.4% versus 6.2%, P=.002) primarily due to increased urinary tract infections (4.1% versus 3.0%, P<.001) ( Table 4 ) . Analgesic use was also higher in LAVH patients (79.4% versus 75.3%, P<.001), driven by opiate use (76.5% versus 72.0%, P<.001). While costs associated with analgesic use were greater in the LAVH patients (P=.03), no other significant cost differences were found in the indexing event or in perioperative outcomes of interest. Index and Perioperative Outcomes: Occurrence and Costs During Index Procedure and in 30-Day Follow-up Period The percentage of patients with an inpatient readmission after the index procedure was significantly lower in the LSH cohort than in the LAVH cohort (1.9% versus 3.3%, P<.001, within 30 days of follow-up and 2.4% versus 3.8%, P<.001, within 180 days of follow-up) ( Table 5 ) . While LOS for these postoperative inpatient admissions was longer in the LSH than the LAVH cohort (P<.001 for both follow-up periods), the difference was small in magnitude (0.15 to 0.16 days). The percentage of patients with an ICU stay was also lower in the LSH versus LAVH cohorts (P<.001 for both follow-up periods). LSH patients had a significantly lower mean number of GYN-related outpatient office visits in both the 30-day and 180-day follow-up periods (P<.001 for both), driven primarily by a lower mean number of primary care visits. Overall, LSH patients had significantly lower total GYN-related costs ($252 versus $385, P<.001, within 30 days of follow-up and $350 versus $569, P<.001, within 180 days of follow-up). GYN-related Resource Utilization and Costs in 30-Day and 180-Day Follow-up Periods GYN-related utilization and costs were identified using all claims with a primary or secondary non rule-out diagnosis containing codes consistent with GYN-related medical care. Re-admission rate is reported only for patients whose procedure was performed in the inpatient setting. The LSH cohort had a significantly lower risk of infection compared to the LAVH cohort (OR 0.830, P=.004) ( Table 6 ) . Significant differences in favor of LSH were also found for hematologic complications and analgesic use. For both the 30-day and 180-day follow-up periods, the LSH cohort had significantly lower risks of GYN-related inpatient readmissions, ER visits, and outpatient office visits. The total GYN-related postoperative costs were $108 and $174 less for the LSH cohort compared to the LAVH cohort in the 30-day and 180-day follow-up periods, respectively (P<.001 for both). Impact of LSH Versus LAVH on Outcomes: Multivariate Regression Results a Reference=LAVH for all models.

The current study found distinct differences between the LSH and LAVH cohorts. LSH patients were more likely to have a hysterectomy due to dysfunctional uterine bleeding and leiomyomas and less likely to undergo the procedure due to endometriosis and prolapse. Unadjusted analyses showed lower rates of overall infection and procedure-specific complications in LSH patients. Compared to LAVH patients, LSH patients had a lower number of GYN-related outpatient office visits and lower total GYN-related costs in both the 30-day and 180-day follow-up periods. Adjusted analyses found significant differences in favor of LSH with regard to the occurrence of overall infection, hematologic complication, and analgesic use despite the fact that LSH patients were also more likely than LAVH patients to have a uterus that weighed >250 grams. The marginal impact on costs was also in favor of LSH, with $108 and $174 less in incurred total GYN-related costs in the 30-day and 180-day follow-up periods, respectively. Wu et al 12 estimated that 538,722 hysterectomies for benign disease were performed in 2003, and 11.2% of them were done through the laparoscopic route, which suggests a total of 60,337 laparoscopic hysterectomies. If we assume 40% of these laparoscopic hysterectomies were performed using the LSH procedure rather than the LAVH procedure, the cost savings would be $2.6 million and $4.2 million in the 30-day and 180-day follow-up periods, respectively. These results, while consistent with much of the previous work comparing LSH and LAVH, are important in that the size of the study cohorts allowed for a robust comparison of the 2 procedures using a diverse, commercially insured population, and thus add further support to the existing evidence demonstrating improved outcomes and fewer costs with LSH. The study by Milad et al 8 was a cohort analysis of 132 patients, 27 undergoing LSH and 105 undergoing LAVH, from a university based medical center. Lalonde et al 9 performed an office and hospital chart review and conducted a postoperative questionnaire on 40 patients, equally divided between LSH and LAVH patients, from a private gynecology practice and private hospital. El-Mowafi et al 10 studied 259 patients, 123 with LSH and 136 with LAVH, from US and non-US sites. Hospital stays for LSH patients compared to LAVH patients were significantly shorter in Milad et al 8 and Lalonde et al 9 and showed no difference in El-Mowafi et al. 10 Though our study did not examine the LOS for the index procedure, LAVH patients were significantly more likely to have their procedures performed in the inpatient setting. Hysterectomy costs were significantly less for LSH versus LAVH patients in both Milad et al 8 and Lalonde et al 9 but showed no difference in our study. However, hysterectomy costs in Milad et al 8 and Lalonde et al 9 were calculated from hospital costs, while costs in our study were calculated from both inpatient and outpatient costs, depending on where the hysterectomy was performed. Further work is needed to evaluate cost differences based on setting of care. Though defined differently, the current study as well as the 3 previous studies (Milad et al, 8 Lalonde et al, 9 and El-Mowafi et al 10 ) found fewer complications in the LSH patients. The current study also found significantly lower GYN-related postoperative costs in LSH versus LAVH patients. In addition to the commonly recognized constraints of administrative claims data, 13 limitations specific to this study should be noted when interpreting the results. First, selection of LSH versus LAVH patients was based on the occurrence of specific procedure codes in the claims history and thus is dependent on the accuracy of these codes. However, because these are distinct surgeries and because we excluded patients with both types of codes, we would expect any misclassification to be small. Second, although we attempted to control for confounding variables in the multivariate regression analyses, other factors not captured in administrative claims data, such as information on level of experience of the surgeon or information on other patient factors that drive surgeon preference for LSH versus LAVH, may have impacted results. Finally, the study population comprised commercially insured patients covered with large employers, thus results may not be representative of all patients with hysterectomy in the US, especially the uninsured or those covered by Medicaid. In addition, more than half of the study population lived in the South. and the regional distribution of the study sample does not represent the regional distribution of the US.

LSH patients demonstrated fewer perioperative complications and lower GYN-related postoperative costs compared to LAVH patients. Additional comparison of these 2 laparoscopic approaches to hysterectomy is needed to further clarify the clinical and cost effectiveness of each procedure.

Of major gynecologic surgical procedures, hysterectomy is the most prevalent worldwide and is second only to cesarean delivery in the United States (US). 1 , 2 Over 600,000 women undergo hysterectomies each year in the US, and approximately one-third of women will have had the procedure performed by the age of 60 years. 2 , 3 , 4 Hysterectomies can be performed using abdominal, vaginal, or laparoscopic approaches, and given the estimated $5 billion in hospital charges resulting from this procedure annually, the outcomes and costs associated with each approach are important considerations. 5 Choice of approach can be impacted by the indication for hysterectomy, with abdominal hysterectomies often used for gynecological cancers. Vaginal hysterectomies are more commonly performed for prolapse or menstrual disorders when the uterus is of normal or slightly enlarged size. 2 , 5 , 6 Laparoscopic hysterectomies include laparoscopic supracervical hysterectomy (LSH), laparoscopic-assisted vaginal hysterectomy (LAVH), total laparoscopic hysterectomy and da Vinci hysterectomy. Both the American College of Obstetrics and Gynecology (ACOG) and the American Association of Gynecologic Laparoscopists (AAGL) support minimally invasive alternatives to abdominal hysterectomy, as these procedures are associated with lower surgical risks, shorter lengths of stay, and quicker return to normal activities than abdominal hysterectomy is. 7 Warren et al 4 also found significant cost savings associated with laparoscopic hysterectomy versus abdominal hysterectomy. Several studies have compared LSH to LAVH evaluating operating time, blood loss, length of hospital stays, and morbidity rates. Findings suggest more favorable outcomes associated with LSH, although sample sizes were small and differences were not always significant. 8 , 9 , 10 , 11 The purpose of our study was to determine the incidence of perioperative complications, postoperative healthcare utilization, and costs in LSH versus LAVH patients, using a large, commercially insured population.