National Trends in Prostate Biopsy and Radical Prostatectomy Volumes Following the US Preventive Services Task Force Guidelines Against Prostate-Specific Antigen Screening

Key Points

Question  What are the downstream effects of the 2012 US Preventive Services Task Force recommendation against prostate-specific antigen screening on practice patterns in prostate cancer diagnosis and treatment?

Findings  Among operative case logs from a nationally representative sample of urologists, prostate biopsy and radical prostatectomy volume decreased by 28.7% and 16.2%, respectively, following the 2012 US Preventive Services Task Force recommendation.

Meaning  These findings represent the downstream effects of the US Preventive Services Task Force recommendation.

Importance  Studies demonstrate that use of prostate-specific antigen screening decreased significantly following the US Preventive Services Task Force (USPSTF) recommendation against prostate-specific antigen screening in 2012.

Objective  To determine downstream effects on practice patterns in prostate cancer diagnosis and treatment following the 2012 USPSTF recommendation.

Design, Setting, and Participants  Procedural volumes of certifying and recertifying urologists from 2009 through 2016 were evaluated for variation in prostate biopsy and radical prostatectomy (RP) volume. Trends were confirmed using the New York Statewide Planning and Research Cooperative System and Nationwide Inpatient Sample. The study included a representative sample of urologists across practice settings and nationally representative sample of all RP discharges. We obtained operative case logs from the American Board of Urology and identified urologists performing at least 1 prostate biopsy (n = 5173) or RP (n = 3748), respectively.

Exposures  The 2012 USPSTF recommendation against routine population-wide prostate-specific antigen screening.

Main Outcomes and Measures  Change in median biopsy and RP volume per urologist and national procedural volume.

Results  Following the USPSTF recommendation, median biopsy volume per urologist decreased from 29 to 21 (interquartile range [IQR}, 12-34; P < .001). After adjusting for physician and practice characteristics, biopsy volume decreased by 28.7% following 2012 (parameter estimate, −0.25; SE, 0.03; P < .001). Similarly, following the USPSTF recommendation, median RP volume per urologist decreased from 7 (IQR, 3-15) to 6 (IQR, 2-12) (P < .001), and in adjusted analyses, RP volume decreased 16.2% (parameter estimate, −0.15; SE, 0.05; P = .003).

Conclusions and Relevance  Following the 2012 USPSTF recommendation, prostate biopsy and RP volumes decreased significantly. A panoramic vantage point is needed to evaluate the long-term consequences of the 2012 USPSTF recommendation.

Prostate cancer is the most common nondermatologic malignancy among men in the United States, with an estimated 220 000 new cases and 27 540 deaths in 2015.¹ Owing to its high incidence and the potential for cure with early detection, population-based screening programs were widely implemented in the United States during the 1990s. Quiz Ref IDHowever, following the Prostate, Lung, Colorectal, and Ovarian Cancer randomized clinical trial² showing no mortality benefit to prostate-specific antigen (PSA) screening, the US Preventive Services Task Force (USPSTF) recommended against population-based PSA screening in 2012.²

Following the USPSTF recommendation against prostate cancer screening, the use of screening PSA and and digital rectal examinations decreased substantially over the ensuing years.^3-6 These sequelae raise the undesirable possibility that the decline in screening may ultimately result in excess mortality and adverse cancer-specific outcomes.^7,8 Given these concerns, it is critical to examine the downstream effects of decreased PSA screening to inform future policy. Single-institutional studies demonstrate a reduction in prostate biopsies following the USPSTF recommendation; however, nationally representative studies are lacking.^9,10 Moreover, there is a dearth of studies evaluating national practice patterns in prostate cancer treatment following the USPSTF recommendation. We sought to determine changes in practice patterns in prostate biopsy and radical prostatectomy (RP) following the 2012 USPSTF recommendation using procedural volumes from a nationally representative sample of urologists and then validating with biopsy volumes from New York state and RP volume from the Nationwide Inpatient Sample (NIS).

We obtained operative case logs from the American Board of Urology (ABU) for urologists applying for certification between 2009 and 2016 (n = 5270). The ABU is the national agency responsible for conferring urology board certification on licensed US medical practitioners. Eligible candidates for initial board certification or recertification are required to submit operative case logs for 6 consecutive months during the 18-month period preceding the certification application date. Logs include all procedures captured by Healthcare Common Procedural Coding System code that are billed by the candidate urologist. Of note, the case log for application in a particular year documents cases that were actually performed in the prior year. For example, a recertification application for 2015 is accompanied by a case log documenting procedures performed in 2014. Recertification is mandated every 10 years for all practicing urologists with the exception of those who were initially certified prior to 1985. Thus, annual aggregate case log data represent approximately 10% of practicing urologists.

We identified all urologists who applied for ABU initial certification or recertification who performed at least 1 prostate biopsy (n = 5173) or RP (n = 3748). Prostate biopsy was identified by Healthcare Common Procedural Coding System codes 55700.¹¹ Radical prostatectomy was identified by codes 55801, 55810, 55812, 55815, 55821, 55831, 55840, 55842, 55845, and 55866.¹²

Data from 2 additional sources were used to validate trends within the ABU data. First, we identified all prostate biopsies performed within the New York Statewide Planning and Research Cooperative System between 2011 and 2014. The New York Statewide Planning and Research Cooperative System is a comprehensive, all-payer reporting system within New York state that captures both inpatient and outpatient clinical encounters. Using these data, trends in prostate biopsies were delineated according to biopsy indication that was obtained from International Classification of Diseases, Ninth Revision diagnosis codes associated with each biopsy and not available in the ABU data. Second, we queried the Healthcare Cost and Utilization Project NIS provided by the Agency for the Healthcare Research and Quality to examine national trends in annual hospitalizations for RP. The NIS is a 20% sample of discharges from all community hospitals participating in Healthcare Cost and Utilization Project, which represents more than 95% of the US population.¹³ We used International Classification of Diseases, Ninth Revision procedure code 60.5 to identify all RPs performed over time.

A waiver of Weill Cornell Medicine institutional review board approval for the study was granted by the Office of Research Integrity because no identifiable patient information was obtained for the current analysis. As such, no patient or surgeon consent was required.

Summary statistics including median and interquartile range (IQR) were used to describe the number of procedures and surgeon characteristics before and after the USPSTF recommendation. Surgeon characteristics included sex, urologic subspecialty (self-reported by each certifying physician), surgeon practice region, and population as reported by the ABU. Owing to the stepwise evolution in the USPSTF recommendation during the course of multiple publications from October 2011 through May 2012, we used 2012 as the threshold to detect change in use, consistent with prior studies.^2,14,15 Comparison of the number of procedures in the 4-year intervals prior (2009-2012) and following (2013-2016) the cutoff were performed using t test and χ² test, where applicable. Log-linear regression was performed to identify predictors of procedural volume and assess trends following the USPSTF recommendation. Sensitivity analyses were performed to evaluate trends using alternate intervals surrounding the USPSTF recommendation (eTable in the Supplement).

Quiz Ref IDFollowing the USPSTF recommendation, the median prostate biopsy volume among urologists decreased from 29 (IQR, 16-44) to 21 (IQR, 12-34) (P < .001) (Table 1). The temporal trend in total prostate biopsy volume is presented in Figure 1. Prostate biopsy volume decreased the most in practice settings of less than 100 000 people in the population (median, 29; IQR, 19-45 vs median, 19; IQR, 11-29; P < .001). Additionally, there was slight regional variation in median prostate biopsy volume with the fewest biopsies performed in the New England region (median, 27; IQR, 15-42). The temporal trend in prostate biopsy volume in the New York Statewide Planning and Research Cooperative System is presented in eFigure 1 in the Supplement. Following the USPSTF recommendation, the total number of annual biopsies decreased by 12.7%. The greatest decrease in biopsy volume was seen in those biopsies performed for an indication of abnormal PSA (26.7%), whereas biopsy volume for an indication of cancer surveillance increased by 28.8% during the study period (597 vs 769; P < .001).

Quiz Ref IDLinear regression revealed that certification year after 2012 was associated with lower biopsy volume (parameter estimate [PE] [SE], −0.25 [0.03]; P < .001) (Table 2). Male urologist sex was associated with higher biopsy volume (PE [SE], 1.00 [0.07]; P < .001), whereas oncologic subspecialty was associated with lower biopsy volume (PE [SE], −0.15 [0.07]; P = .02). After adjusting for physician and practice characteristics, there remained a significant 28.7% decrease in biopsy volume following 2012 (PE [SE], −0.25 [0.03]; P < .001).

Quiz Ref IDFollowing the USPSTF recommendation, median RP volume per urologist decreased from 7 (IQR, 3-15) to 6 (IQR, 2-12) (P < .001) (Table 3). Temporal trend in total RP volume is presented in Figure 2. Radical prostatectomy volume decreased among urologists practicing in all population densities except practice settings with populations of 100 000 to 250 000 people. There was regional variation in RP volume, with New England urologists performing the fewest RPs (median, 4; IQR, 2-11) and North Central urologists performing the most (median, 8; IQR, 3-16.5). Similarly, the number of hospital discharges following RP from the NIS decreased by 30.3% (from 83 881 to 58 430) following the USPSTF recommendation (eFigure 2 in the Supplement).

Linear regression revealed that certification year after 2012 was associated with lower RP volume (PE [SE], −0.15 [0.05]; P = .003) (Table 2). Male urologist sex and oncologic subspecialty were associated with increased RP volume (PE [SE], 0.77 [0.13]; P < .001 and 0.94 [0.09]; P < .001, respectively). Radical prostatectomy volume significantly varied by geographic region, with lower RP volume in the New England practices compared with the North Central, South Central, Southeastern, and Western practice regions (PE [SE], 0.58 [0.08], 0.38 [0.09], 0.29 [0.08], and 0.37 [0.08], respectively; all P < .001). Practice region of greater than 1 million was associated with higher RP volume (PE [SE], 0.35 [0.07]; P < .001). After adjusting for physician and practice characteristics, there remained a significant 16.2% decrease in RP volume following 2012 (PE [SE], −0.15 [0.05]; P = .003).

The 2012 USPSTF recommendation against PSA screening was associated with drastic change in US prostate cancer screening practices. Following the 2012 USPSTF recommendation, multiple studies demonstrate a substantial decrease in PSA screening and incidence of early-stage prostate cancer following the USPSTF recommendation.^3,4 However, few studies have examined the downstream effects of the effect of the USPSTF recommendation on diagnostic and therapeutic prostate cancer practice patterns.

We present the novel finding of a decrease in RP volume for prostate cancer following the USPSTF recommendation. Although the reason for this decline is likely multifactorial, stemming from a decrease in the incidence of prostate cancer as well as greater adoption of active surveillance for low-risk prostate cancer and treatment substitution with nonsurgical therapies, our finding that RP volume has declined is noteworthy given that peak RP rates occurred in the mid 1990s following the advent of PSA testing.^16,17 Additionally, we found that national variation in prostate cancer care persists, as previously demonstrated in 1993 by Lu-Yao et al,¹⁷ who found lower rates of RP in the New England and Mid-Atlantic regions of the United States.¹⁷

We also found a significant decrease in prostate biopsy volume following the USPSTF recommendation, consistent with single-institutional studies. Banerji et al¹⁰ evaluated the number of prostate biopsies performed at an academic institution during the 30-month period before and after the USPSTF recommendation and found a 31% decrease in the number of biopsies. Likewise, McGinley et al⁹ examined biopsy rates in a large urology group practice of 32 physicians and found a significant decrease following the USPSTF recommendation. Our national study is noteworthy because we demonstrate geographic variation in prostate biopsy volumes across academic and community practice settings and health plans and demonstrate an overall decrease in prostate biopsies following USPSTF recommendation.

Furthermore, we demonstrate changing trends in the indications for prostate biopsy, with fewer patients undergoing this procedure for abnormal PSA. This finding strengthens the associational relationship found in prior studies insofar as it directly links decreased PSA testing with biopsy volume. In contrast, the number of biopsies performed for active surveillance of prostate cancer has increased, consistent with studies demonstrating an uptake in observation of men with low-risk prostate cancer.¹⁸

Experts have expressed concern regarding the long-term effect of the USPSTF recommendation on prostate cancer–specific morbidity and mortality in the United States.⁷ This concern is reinforced by the findings of this study demonstrating that decreased prostate cancer screening has been accompanied by a downtrend in diagnostic and therapeutic procedures for prostate cancer. However, given that lead time associated with PSA screening is estimated at 5 to 7 years and the indolent nature of prostate cancer, a resultant increase in prostate cancer–specific mortality would likely remain undetectable for another decade.¹⁹ Thus, longer-term population-based studies will be necessary to determine the true effect of the USPSTF recommendation on prostate cancer–specific mortality.

Our findings must be interpreted within the context of the study design. First, the study population represents approximately 10% of all practicing urologists per year, which may introduce sampling bias, and our volume decrease estimates likely underestimate the true shift in procedural volume because trained urologists before 1985 do not require recertification, and there may be nonboard-certified proceduralists that we are not capturing. However, ABU data have been used widely to assess practice pattern changes across the spectrum of urologic procedures and conditions.^12,20-22 Second, ABU data consist of self-reported 6-month case logs from individual practitioners’ billing software, and the ABU reserves the right to audit all submitted case logs. Practitioners typically submit Healthcare Common Procedural Coding System codes from accounting software.²³ Therefore, there is no reason to suspect overestimation or underestimation of procedural volume. Third, the NIS does not capture outpatient RP, which may account for an increasing proportion of RP owing to favorable reimbursement in the out-of-network preferred provider organization setting. This could result in the underestimation of RP volume, which may contribute to the observed downtrend. Finally, the trends evaluated herein cannot be interpreted as the causal effects of the USPSTF recommendation alone. Many other factors may have contributed to the downtrend in procedural volume such as the uptake of active surveillance, substitution with alternative therapies, uptake of MRI, use of biomarkers to improve risk stratification in the setting of abnormal PSA, and concerns regarding biopsy-related morbidity.^16,18,24 Therefore, it is difficult to delineate the exact contribution of these additional factors.

Following the 2012 USPSTF recommendation, prostate biopsy and RP volumes decreased significantly. These findings represent the direct downstream effects of the USPSTF recommendation. Quiz Ref IDWhile the pendulum of prostate cancer screening continues to swing, a more extended vantage point is needed to evaluate the long-term consequences of the 2012 USPSTF recommendation with regard to stage at presentation, outcomes following treatment, and disease-specific mortality in prostate cancer. Because revision of the USPSTF recommendation is in progress, policy makers should weigh the downstream effects of the 2012 USPSTF recommendation and consider future unintended consequences.

Corresponding Author: Jim C. Hu, MD, MPH, Department of Urology, Weill Cornell Medicine, 525 E 68th St, Starr 900, New York, NY 10065 (jch9011@med.cornell.edu).

Accepted for Publication: July 7, 2016.

Published Online: November 2, 2016. doi:10.1001/jamasurg.2016.3987

Author Contributions: Dr Hu had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Halpern, Shoag, Artis, Ballman, Sedrakyan, Hu.

Acquisition, analysis, or interpretation of data: All Authors.

Drafting of the manuscript: Halpern, Shoag, Artis, Ballman, Hu.

Critical revision of the manuscript for important intellectual content: All Authors.

Statistical analysis: Halpern, Artis, Ballman, Wright.

Administrative, technical, or material support: Sedrakyan, Shih.

Study supervision: Shih, Hu.

No additional contributions: Shoag, Hershman.

Conflict of Interest Disclosures: Dr Hu is a member of the Speakers’ Bureau for Intuitive Surgical and Genomic Health. Dr Sedrakyan has received funding for establishing the US Food and Drug Administration’s Medical Device Epidemiology Network. No other disclosures are reported.