Background
Complex oncologic surgeries, including those for endometrial cancer, increasingly have been concentrated to greater-volume centers, owing to previous research that has demonstrated associations between greater surgical volume and improved outcomes. There is a potential for concentration of care to have unwanted consequences, including cost burden, delayed treatment, patient dissatisfaction, and possibly worse clinical outcomes, especially for more vulnerable populations.
Objective
To describe changes in site of care for patients with endometrial cancer in New York State and to determine whether the distance women traveled for hysterectomy has changed over time.
Study Design
We used the New York Statewide Planning and Research Cooperative System to identify women with endometrial cancer who underwent hysterectomy from 2000 to 2014. Demographic and clinical data as well as hospital data were collected. Trends in travel distance (straight-line distance) were analyzed within all hospital referral regions and differences in travel distance over times and across sociodemographic characteristics analyzed.
Results
We identified 41,179 subjects. The number of hospitals and surgeons performing hysterectomy decreased across all hospital referral regions over time. The decline in the number of hospitals caring for women with endometrial cancer ranged from –16.7% in Syracuse (12 to 10 hospitals) to –76.5% in Rochester (17 to 4 hospitals). Similarly, the percentage of surgeons within a given hospital referral region operating on women declined from –45.2% in Buffalo (84–46 surgeons) to –77.8% in Albany (72 to 16 surgeons). The median distance to the index hospital for patients increased in all Hospital Referral Regions. For residents in Binghamton, median travel distance increased by 46.9 miles (95% confidence interval, 33.8–60.0) whereas distance increased in Elmira by 19.7 miles (95% confidence interval, 7.3-32.1) and by 12.4 miles (95% confidence interval, 6.4–18.4) in Albany. For residents of Binghamton and Albany, there was a greater than 100% increase in distance traveled over the 15-year time period, with increases of 551.8% (46.9 miles; 95% confidence interval, 33.8–60.0 miles) and 102.5% (12.4 miles; 95% confidence interval, 6.4–18.4 miles), respectively. Travel distance increased for all races and regardless of insurance status but was greatest for white patients and those with private insurance ( P <.0001 for both).
Conclusion
The number of surgeons and hospitals caring for women with endometrial cancer in New York State has decreased, whereas the distance that patients travel to receive care has increased over time.
Many studies have demonstrated a positive association between surgeon and hospital procedural volume and improved outcomes for oncologic surgeries. This association has led to initiatives to concentrate complex oncologic surgeries to greater-volume, tertiary centers. For endometrial cancer, the association between surgeon and hospital volume and outcomes are more modest.
Why was the study conducted?
To describe site of care for endometrial cancer in New York State and determine whether the distance women traveled for hysterectomy has changed over time.
Key findings
The number of hospitals and surgeons performing hysterectomy decreased across all hospital referral regions over time. The median distance to the index hospital for patients increased in all hospital referral regions. Travel distance increased for all races and regardless of insurance status but was greatest for white patients and those with private insurance. Increase in travel distance for endometrial cancer was most pronounced for women living in rural areas.
What does this add to what is known?
Efforts to promote referral of women with endometrial cancer to more specialized surgeons and centers have influenced patterns of care. Patients residing in rural areas seem to be adversely affected by the concentration of care for endometrial cancer, whereas the effect on minority patients and Medicaid recipients has been less substantial.
For gynecologic cancers, there is evidence that centralization of care has becoming increasingly common over time. An analysis of women with endometrial cancer in New York State found that the number of surgeons treating women with endometrial cancer decreased by more than 60% from 2000 to 2014. Although the concentration of care for oncologic surgery may improve outcomes, there is also a potential that centralization of care may have unwanted consequences. In particular, if fewer surgeons and hospitals provide care for a given condition, patients may be forced to travel greater distances to receive care, which may lead to an increased cost burden, delayed treatment, patient dissatisfaction, and possibly worse clinical outcomes. Regionalization of care may be particularly problematic for vulnerable populations, including those of low socioeconomic status and with more comorbidities.
There are few data to describe how regionalization of care has altered where women with endometrial cancer receive care. The objectives of this study were to describe patterns of care for endometrial cancer in New York State and to determine whether the distances women traveled for hysterectomy have changed over time and whether this varies by demographic and socioeconomic factors.
Materials and Methods
We used data from the Statewide Planning and Research Cooperative System (SPARCS), maintained by the New York State Department of Health, for this analysis. SPARCS captures patient characteristics, diagnoses, services, and charges for hospital inpatient admissions and outpatient visits. Encrypted physician and hospital identifiers and limited hospital information also are included. Data quality is ensured through periodic reviews and by comparing SPARCS data with data from other Department of Health databases. This study used deidentified data and was classified as non-human subject research by the institutional review board of Columbia University Medical Center.
Women diagnosed with endometrial cancer ( International Classification of Diseases , Ninth Revision (codes 179, 182.x) who underwent hysterectomy (abdominal, robot-assisted, laparoscopic, and vaginal) from 2000 to 2014 were identified. Patients were excluded if they had hysterectomy before the index hospital admission, had missing data on facility or surgeon identifiers, or had unknown health insurance status or an invalid New York State zip code for their area of residence.
Patients’ demographic characteristics and clinical factors included age (<40, 40–49, 50–59, 60–69, ≥70 years), race (white, black, Hispanic, other, unknown), primary health insurance (none, Medicare, Medicaid, private insurance, other), and Elixhauser comorbidity score (0, 1, ≥2), and year of discharge. We also noted perioperative transfusion, complications, and in-hospital mortality using a previously described coding algorithm.
Each patient’s area of residence was classified into their given Hospital Referral Region (HRR) based on the zip code classification grouping of the Dartmouth Atlas of Health Care. HRRs represent regional health care markets for tertiary medical care. Each HRR generally requires the services of a major referral center. Each patient was classified into the HRR in which they resided based on their zip code regardless of where they received care. Using hospital zip codes, we also classified each hospital into its respective HRR. Patients who were not linkable to Dartmouth Atlas data and those who were assigned to an HRR outside of New York State were excluded. There were 10 HRRs in New York State: Albany, Binghamton, Bronx, Buffalo, East Long Island, Elmira, Manhattan, Rochester, Syracuse, and White Plains ( Figure 1 ). The population density of each HRR in 2014 was calculated as the sum of population of all zip code area within an HRR divided by the land area in square miles. The population density of each HRR in 2014 was calculated as the sum of population of all zip code area within an HRR divided by the sum of land area in square miles. Population and land area of each zip code area were retrieved from the website of the US Census Bureau (American FactFinder). The HRRs with the greatest population density were Manhattan, the Bronx, and East Long Island. In contrast, the HRRs with the lowest population density were Elmira, Binghampton, and Syracuse.
Each patient’s travel distances from her home to the index hospital where she underwent hysterectomy were calculated using the straight-line distance, which is the distance from the center of each patient’s residential zip code area to the center of the hospital’s zip code area. This method for measuring distance has been validated and used in a number of studies examining geographic patterns of care.
We analyzed trends in the number of hospitals and surgeons within a given HRR over time as well as the travel distance for patients to their treating hospitals. Changes over time are expressed descriptively as percentage change from 2000 to 2014 (value difference in year 2014 and 2000 divided by the base value in 2000). To examine changes in extremes of travel distance, we analyzed the number of patients who traveled <10 miles or >50 miles from their place of residence to the treating hospital. Similarly, changes in travel distance over time were explored across the clinical and demographic characteristics (age, race/ethnicity, insurance status, comorbidity score). Fully saturated models were developed to evaluate whether percentage change across HRR and demographics were statistically significant using generalized estimation equations with normal distributions and identity linkage. Quantile regression models were used to calculate the difference and 95% confidence interval (CI) of median travel distance between the years 2000 and 2014 to test whether the median travel distance was statistically different between the years.
All analyses were conducted with SAS 9.4 (SAS Institute Inc., Cary, NC) and R 3.5.1 (Foundation for Statistical Computing, Vienna, Austria) with “ggplot” package. All statistical tests were 2-sided. A P -value of <.05 was considered statistically significant.
Results
We identified a total of 41,179 patients with endometrial cancer who underwent hysterectomy in New York State between 2000 and 2014. Abdominal hysterectomy was performed in 65.6% of the patients, whereas 21.7% underwent laparoscopic hysterectomy, 10.9% a robotic-assisted procedure, and 1.9% a vaginal hysterectomy ( Table 1 ). The majority of patients (88.4%) were 50 years of age or older. White patients accounted for 68.1% of the cohort, whereas 11.9% of the patients were black. One-half of the patients (50.6%) had commercial insurance, and more than one-half (52.3%) had a comorbidity score of ≥2.
| n (%) | |
|---|---|
| Total patients | 41,179 |
| Total hospitals | 215 |
| Annualized hospital volume at hospital level, median (IQR) | 4.2 (2.0–10.3) |
| Total surgeons | 2706 |
| Annualized surgeon volume at surgeon level, median (IQR) | 1.0 (1.0–1.6) |
| Hysterectomy type | |
| Abdominal | 27,008 (65.6) |
| Robotic | 4473 (10.9) |
| Laparoscopic | 8916 (21.7) |
| Vaginal | 782 (1.9) |
| Year of discharge | |
| 2000 | 2392 (5.8) |
| 2001 | 2617 (6.4) |
| 2002 | 2500 (6.1) |
| 2003 | 2559 (6.2) |
| 2004 | 2604 (6.3) |
| 2005 | 2703 (6.6) |
| 2006 | 2596 (6.3) |
| 2007 | 2589 (6.3) |
| 2008 | 2918 (7.1) |
| 2009 | 2767 (6.7) |
| 2010 | 2915 (7.1) |
| 2011 | 2941 (7.1) |
| 2012 | 3011 (7.3) |
| 2013 | 3115 (7.6) |
| 2014 | 2952 (7.2) |
| HRR a | |
| Albany | 3958 (9.6) |
| Binghamton | 958 (2.3) |
| Bronx | 2453 (6.0) |
| Buffalo | 3578 (8.7) |
| East Long Island | 11,369 (27.6) |
| Elmira | 732 (1.8) |
| Manhattan | 10,296 (25.0) |
| Rochester | 3102 (7.5) |
| Syracuse | 2247 (5.5) |
| White Plains | 2486 (6.0) |
| Age, y | |
| <40 | 1037 (2.5) |
| 40–49 | 3738 (9.1) |
| 50–59 | 11,252 (27.3) |
| 60–69 | 13,150 (31.9) |
| ≥70 | 12,002 (29.1) |
| Race | |
| White | 28,040 (68.1) |
| Black | 4900 (11.9) |
| Hispanic | 2673 (6.5) |
| Other | 4069 (9.9) |
| Unknown | 1497 (3.6) |
| Health insurance | |
| Uninsured/Medicaid/dual | 4336 (10.5) |
| Medicare | 15,973 (38.8) |
| Private | 20,848 (50.6) |
| Other | 22 (0.05) |
| Comorbidity | |
| 0 | 8674 (21.1) |
| 1 | 10,975 (26.7) |
| ≥2 | 21,530 (52.3) |
| Outcomes | |
| Any morbidity | 6809 (16.5) |
| Discharge mortality | 253 (0.6) |
| Transfusion | 7508 (18.2) |
| Total charges, median, IQR | $32,013 ($15,682–$35,826) |
a HRRs represent regional healthcare markets for tertiary medical care that generally requires the services of a major referral center. The regions were defined by determining where patients were referred for major cardiovascular surgical procedures and for neurosurgery.
From 2000 to 2014, the number of hospitals and surgeons performing hysterectomy for endometrial cancer in New York State decreased across all HRRs ( Figure 2 ). The decline in the number of hospitals caring for women with endometrial cancer ranged from –16.7% in Syracuse (12 to 10 hospitals) to –76.5% in Rochester (17 to 4 hospitals). Similarly, the percentage of surgeons within a given HRR operating on women declined from –45.2% in Buffalo (84 to 46 surgeons) to –77.8% in Albany (72 to 16 surgeons) across the years of study. The percentage change across HRR was statistically significant (both P <.0001).
During the same time period, the median distance to the index hospital for patients increased in all HRRs ( Table 2 ). For residents in the Binghampton HRR, median travel distance increased by 46.9 miles (95% CI, 33.8–60.0) whereas distance increased in the Elmira HRR by 19.7 miles (95% CI, 7.3–32.1) and by 12.4 miles (95% CI, 6.4–18.4) in the Albany HRR. The lowest increase in travel distance was found in Manhattan and White Plains HRRs.
| HRRs | Hospitals | Surgeons | Median travel distance (miles) | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| 2000 | 2014 | Change over time (%) a | 2000 | 2014 | Change over time (%) a | 2000 | 2014 | Difference in miles (95% CI) b | Change over time (%) a | |
| Elmira | 4 | 3 | –25 (–30.6, –19.4) | 12 | 3 | –75.0 (–80.6, –69.4) | 25.6 | 45.3 | 19.7 (7.3–32.1) | 77.0 (71.4–82.6) |
| Binghamton | 5 | 4 | –20 (–25.6, –14.4) | 22 | 6 | –72.7 (–78.3, –67.1) | 8.5 | 55.4 | 46.9 (33.8–60.0) | 551.8 (546.2–557.4) |
| Syracuse | 12 | 10 | –16.7 (–22.3, –11.1) | 36 | 16 | –55.6 (–61.2, –50.0) | 19.3 | 24.4 | 5.1 (–8.7, 18.9) | 26.4 (20.8–32.0) |
| Albany | 19 | 9 | –52.6 (–55.4, –49.8) | 72 | 16 | –77.8 (–80.6, –75) | 12.1 | 24.5 | 12.4 (6.4–18.4) | 102.5 (99.7–105.3) |
| Rochester | 17 | 4 | –76.5 (–82.1, –70.9) | 70 | 25 | –64.3 (–69.9, –58.7) | 7.8 | 9.1 | 1.3 (–1.6, 4.2) | 16.7 (11.1–22.3) |
| Buffalo | 18 | 9 | –50 (–55.6, –44.4) | 84 | 46 | –45.2 (–50.8, –39.6) | 6.6 | 8.5 | 1.9 (0.06–3.74) | 28.8 (23.2–34.4) |
| White Plains | 14 | 8 | –42.9 (–48.5, –37.3) | 57 | 20 | –64.9 (–70.5, –59.3) | 9.3 | 10.2 | 0.9 (–1.9, 3.7) | 9.7 (4.1–15.3) |
| East Long Island | 34 | 21 | –38.2 (–43.8, –32.6) | 219 | 55 | –74.9 (–80.5, –69.3) | 7.2 | 9.1 | 1.9 (0.8–3.0) | 26.4 (20.8–32.0) |
| Bronx | 11 | 5 | –54.5 (–60.1, –48.9) | 50 | 20 | –60.0 (–65.6, –54.4) | 2.0 | 3.1 | 1.1 (0.6–1.6) | 55 (49.4–60.6) |
| Manhattan | 37 | 22 | –40.5 (–46.1, –34.9) | 214 | 114 | –46.7 (–52.3, –41.1) | 3.2 | 3.7 | 0.5 (0.04–0.96) | 15.6 (10.0–21.2) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
