Elsevier

European Urology Focus

Volume 6, Issue 5, 15 September 2020, Pages 975-981
European Urology Focus

Bladder Cancer
Defining a “High Volume” Radical Cystectomy Hospital: Where Do We Draw the Line?

https://doi.org/10.1016/j.euf.2019.02.001Get rights and content

Abstract

Background

Centralization of radical cystectomy (RC) to “high volume” centers can lead to decreased morbidity but also limits access to care. In the context of centralization, there is a need to systematically define the hospital volume cutoffs for this procedure.

Objective

To systematically examine the effect of hospital volume on inpatient complications of RC for bladder cancer and to define a threshold to minimize RC morbidity.

Design, setting, and participants

This was a retrospective analysis of data for 6790 adults undergoing RC for nonmetastatic bladder cancer during 2008–2011 from the National Inpatient Sample (weighted population estimate of 33 249 RCs in the USA during this period).

Intervention

RC.

Outcome measurements and statistical analysis

Overall and major complications were defined according to International Classification of Diseases (9th revision) diagnosis and procedure codes. To define the relationship between hospital volume and morbidity, logistic regression analyses within a generalized estimating equation framework with restricted cubic splines were used.

Results and limitations

The inpatient complication rate was 4769/6790 (70.2%), of which 1572/6790 (23.2%) were major complications. Restricted cubic spline analysis revealed a significant inverse nonlinear association between hospital volume and complications. The odds of complications decreased with increasing volume, with a plateau at 50–55 cases/yr for any complications (p = 0.024) and 45–50 cases/yr for major complications (p = 0.007).

Conclusions

The relationship between hospital volume and RC morbidity is nonlinear, with a plateau for the complication rate at 50–55 cases/yr. Restricting RC to centers with such high thresholds will restrict access to care. There is a need to identify and publish best practices from high-volume centers in quality improvement initiatives to improve morbidity at low-volume centers.

Patient summary

There is a nonlinear relationship between the annual number of radical cystectomy procedures performed at a hospital and the inpatient complication rate. Complications decrease with increasing hospital volume and reach a plateau at 50–55 cases per year, beyond which the incremental benefit of increasing volume is minimal.

Introduction

Radical cystectomy (RC), one of the most technically complex procedures in urology, remains the standard of care for muscle-invasive and recurrent high-risk non–muscle-invasive bladder cancer. Up to two-thirds of patients undergoing RC experience postoperative morbidity, including a high rate of potentially life-threatening complications such as hemorrhage, bowel leakage, and sepsis [1], [2], [3], [4].

Major cancer surgery such as RC involves a team of multiple members responsible for patient care. This collaborative effort is crucial for communication, early identification of complications, and access to critical services. Increasing annual volume for such teams can potentially decrease the mortality and morbidity of RC [2], [5], [6], [7], [8]. Birkmeyer et al [9] showed a decrease in inpatient mortality with increasing hospital volume after elective major cancer surgery.

However, limiting surgical procedures to a few “high volume” centers (centralization) decreases access to care for many patients [10], [11]. This fine balance between lower morbidity and access to care makes it important that the cutoffs defining a “high volume” hospital be systematically and objectively defined. The literature on this subject has assumed a linear form for this association, dividing hospital volume into quartiles, quintiles, or deciles in analyses [2], [6], [12], [13]. As a result, a “high volume” hospital has been arbitrarily defined in the literature from as low as eight to as high as 33.5 cases/yr [13], [14], [15], [16].

There is evidence that the association between hospital volume and complications due to surgical procedures might not be a linear one [17], [18]. In this context, using the National Inpatient Sample (NIS) we sought to objectively define an annual hospital volume threshold to optimize the inpatient morbidity of elective RC for bladder cancer at the national level without prior assumption of the form for this association.

Section snippets

Data source

The US Healthcare Cost and Utilization Project (HCUP) NIS is the largest publicly available inpatient care database in the USA. The NIS comprises a 20% stratified sample of inpatient records from US community hospitals (www.hcup-us.ahrq.gov/overview.jsp). Stratified sampling allows for generation of national estimates from these data using HCUP-provided survey weights. From 2008 to 2011, there were 31 793 174 records within the NIS (weighted 156 919 107 records in the USA). This study period

Results

A total of 6790 records met the inclusion criteria (weighted 33 429 RCs performed in the USA during the same period). Descriptive statistics for patients undergoing RC overall and stratified by hospital volume quartile are shown in Table 1.

The median hospital RC volume among hospitals performing this procedure as averaged over 4 yr was 24 cases/yr. There were significant differences between the hospital volume quartiles in terms of age (p < 0.001), gender (p = 0.002), Charlson comorbidity score (p < 

Discussion

In this study, we used the NIS to systematically examine the relationship between annual hospital volume and RC morbidity to define a hospital volume cutoff to minimize the morbidity of this procedure while balancing access to care. Using logistic regression analysis with restricted cubic splines, we found an inverse nonlinear association between hospital volume and RC morbidity. However, there was a plateau for the complication rate at a high threshold of 50–55 cases/yr (45–50 cases/yr for

Conclusions

The relationship between hospital volume and RC morbidity is nonlinear. A plateau for the complication rate is seen at 50–55 cases/yr. This extends beyond the thresholds previously reported in the literature. Restricting RC to centers with such a high threshold for the annual caseload would restrict access to care for most patients. There is a need to share and publish best practices from high-volume centers in quality improvement initiatives to improve morbidity at low-volume centers.


Author

References (30)

  • A. Shabsigh et al.

    Defining early morbidity of radical cystectomy for patients with bladder cancer using a standardized reporting methodology

    Eur Urol

    (2009)
  • A.L. Merrill et al.

    Clinical effect of surgical volume

    N Engl J Med

    (2016)
  • P. Ravi et al.

    Benefit in regionalisation of care for patients treated with radical cystectomy: a nationwide inpatient sample analysis

    BJU Int

    (2014)
  • M. Sun et al.

    Is there a relationship between leapfrog volume thresholds and perioperative outcomes after radical cystectomy?

    Urol Oncol

    (2014)
  • J.D. Birkmeyer et al.

    Hospital volume and surgical mortality in the United States

    N Engl J Med

    (2002)
  • Cited by (18)

    • Disparities in access to high-volume centers and in hospital discharge status following radical cystectomy in Florida

      2023, Urologic Oncology: Seminars and Original Investigations
      Citation Excerpt :

      As might be expected for such a complex operation, centralization of RC to high-volume centers has reliably led to improvements in perioperative outcomes. However, centralization has also created unique challenges in patient access to quality RC care within the United States [5–8]. Indeed, population-based data show that only 50% of patients in the United States with muscle-invasive bladder cancer receive potentially curative therapy, and that those living > 50 miles from a center that performs RC have 40% lower odds of undergoing the operation [9].

    • Defining radical cystectomy using the ICD-10 procedure coding system

      2022, Urologic Oncology: Seminars and Original Investigations
      Citation Excerpt :

      However, despite the need to define a valid list of codes for RC, we have been unable to identify a comprehensive list of ICD-10-PCS codes to define RC within the peer-reviewed literature. Contemporary published data focus primarily on the time period up to and including 2015, and thus use ICD-9 codes to identify RC [3–7]. Given this evident gap in the literature, and acknowledging the time and effort required to compile a comprehensive list of codes to define RC using the ICD-10 system, we sought to identify and internally validate a list of ICD-10-PCS codes for RC for use by bladder cancer researchers.

    View all citing articles on Scopus
    View full text