The Impact of Recent Screening Recommendations on Prostate Cancer Screening in a Large Health Care System Afshin Aslani, Brian J. Minnillo, Ben Johnson, Edward E. Cherullo, Lee E. Ponsky and Robert Abouassaly* From the Department of Urology, Case Western Reserve University, Urological Institute, University Hospitals Case Medical Center, Cleveland, Ohio
Purpose: The United States Preventive Services Task Force recently recommended against routine prostate cancer screening, stating that the risks of screening outweigh the benefits. We determined the impact of this recommendation on prostate cancer screening in a large health system. Materials and Methods: We obtained data on all screening prostate specific antigen tests performed at University Hospitals Case Medical Center and affiliated hospitals in northeastern Ohio from January 2008 to December 2012. We examined the total number of prostate specific antigen tests ordered with time and adjusted for patient volume by fitting a regression line. The overall trend was examined and stratified by location (urban, suburban or rural), patient age and provider type (primary care or urology). Results: A total of 43,498 screening prostate specific antigen tests were performed from January 2008 to December 2012. Most tests were ordered by specialists in internal medicine (64.9%), followed by family medicine (23.7%), urology (6.1%) and hematology/oncology (1.3%). Prostate specific antigen screening increased with time until March 2009, when initial screening trials were published. Prostate specific antigen testing then decreased significantly and continued to decrease after the task force recommendations. Similar patterns were noted in almost all subgroups. The greatest decrease in screening was observed by urologists and in patients in the intermediate age group (50 to 59 years). Conclusions: United States Preventive Services Task Force recommendations appeared to have decreased prostate cancer screening. The greatest impact was seen for urologists and patients in the intermediate age group. Further study is needed to determine the long-term effects of these recommendations on the screening, diagnosis, treatment and prognosis of this prevalent malignancy.
Abbreviations and Acronyms ACP ¼ American College of Physicians AUA ¼ American Urological Association PCP ¼ primary care physician PSA ¼ prostate specific antigen USPSTF ¼ United States Preventive Services Task Force Accepted for publication December 6, 2013. Study received institutional review board approval. Supported by National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases Award T32DK091213. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. * Correspondence: Department of Urology, Case Western Reserve University, Urological Institute, University Hospitals Case Medical Center, 11100 Euclid Ave., Mailstop LKD 5046, Office 4576, Cleveland, Ohio 44106 (telephone: 216-844-4831; e-mail: robert.abouassaly@ uhhospitals.org).
See Editorial on page 1648.
Key Words: prostate, prostatic neoplasms, prostate-specific antigen, mass screening, practice guidelines as topic
PROSTATE cancer is one of the leading causes of cancer related death in men and has a substantial burden on the health care system.1,2 In 1994 the Food and Drug Administration approved PSA testing for screening in prostate cancer in men 50 years old
or older together with digital rectal examination.3 This led to a dramatic increase in the incidence of prostate cancer in the United States.4,5 Stage migration was subsequently noted with a greater proportion of patients diagnosed with low risk, low volume
0022-5347/14/1916-1737/0 THE JOURNAL OF UROLOGY® © 2014 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION AND RESEARCH, INC.
http://dx.doi.org/10.1016/j.juro.2013.12.010 Vol. 191, 1737-1742, June 2014 Printed in U.S.A.
www.jurology.com
j
1737
1738
IMPACT OF RECOMMENDATIONS ON PROSTATE CANCER SCREENING IN HEALTH CARE SYSTEM
prostate cancer.6 Paralleling this was a decrease in mortality from this malignancy. Although a causal relationship with PSA screening was hypothesized, it has not been definitively established.7 However, PSA screening is not without controversy due to the potential over diagnosis and overtreatment of clinically insignificant prostate cancer. As a result, some organizations have advised against widespread PSA screening. For example, the United Kingdom National Health Service does not mandate or advise PSA screening, instead allowing men to decide based on the recommendation of their physician.8 Recently, large randomized trials were performed in the United States and Europe on the impact of prostate cancer screening on prostate cancer mortality.9,10 Although the 2 studies had various methodological issues, the PLCO (Prostate, Lung Colorectal and Ovarian) screening trial performed in the United States revealed no significant benefit to prostate cancer screening using PSA.9 On the other hand, the European study showed that PSA based screening decreased the rate of death from prostate cancer, although the number needed to screen was high.10 After reviewing the evidence surrounding the usefulness of PSA screening the USPSTF concluded that the potential benefit did not outweigh expected harms in patients not already diagnosed or being treated for prostate cancer.11 The Cochrane Library subsequently released an updated review and metaanalysis of PSA screening showing that screening did not significantly decrease prostate cancer specific mortality.12 In addition, others did not find that PSA screening is cost-effective in men at average to high risk of prostate cancer but it may be costeffective in those at very high risk.13 As a consequence, clinical practice guidelines for prostate cancer screening vary and are controversial due to uncertainty as to whether the benefits of screening ultimately outweigh the risks of over diagnosis and overtreatment.14e17 There are also the 2 recent clinical practice guidelines for prostate cancer screening by the ACP18 and AUA.19 The ACP only recommends that clinicians inform men between ages 50 and 69 years about the benefit and harm of PSA screening to make a shared decision considering general health status and life expectancy. It does not recommend PSA screening in men younger than 50 or older than 69 years, or with life expectancy less than 10 years. The new AUA guideline on early detection of prostate cancer only recommends PSA screening in men older than 40 years at high risk for prostate cancer or shared decision making based on patient preferences in those 55 to 69 years old.18 It does not advise screening in men younger than 40 or older than 70 years, or those with less than 10-year life expectancy.
We determined whether there was a recent change in the use of PSA screening in a large university health system after the release of new evidence from various large trials as well as the updated ACP and AUA guidelines.18,19 We also assessed whether effects differed by provider or patient characteristics.
MATERIALS AND METHODS After obtaining institutional review board approval we performed a retrospective, secondary data analysis of deidentified laboratory data. We obtained for analysis the records of all PSA screening tests performed at University Hospital Case Medical Center and affiliated hospitals from January 1, 2008 to December 31, 2012. Screening PSA tests were identified by the V76.44 screening PSA diagnostic code and confirmed by the billing code. PSA tests ordered for other diagnostic purposes or for monitoring prostate cancer were excluded from analysis. PSA screening tests were extracted from the health system central laboratory, including variables of provider location (facility), physician specialty, month and year of service, and patient age. Analysis included 7 hospitals in northeastern Ohio. University Hospitals Case Medical Center was the only urban hospital. Bedford, Richmond and Ahuja medical centers are in a suburban setting, and Conneaut, Geauga and Geneva medical centers are in rural areas. To adjust for patient volume and population changes in this large health system we used the ratio of the number of PSA screening tests to cholesterol tests done at the hospitals per month. The study period was divided into 3 segments based on the publication of the 2 mentioned screening trials in March 20099,10 and USPSTF recommendations in May 2012.11 The slope of each segment was calculated and compared using linear regression and segmented regression models (supplementary table, http://jurology.com/ and figs. 1 to 3). The segmented regression model included the effect of earlier period(s) while the linear regression model did not. Analysis was then repeated, stratified by patient age (less than 50, 50 to 59, 60 to 69, 70 to 79 and 80 years or older), physician specialty (internal medicine, family medicine, urology and medical oncology) and facility type (urban, suburban and rural). We included all physician specialties in the analysis but performed followup stratified analysis of physician specialty for those that contributed most to screening, including primary care, urology and hematology/oncology. In each subgroup we used the ratio of PSA tests performed to total cholesterol tests performed per month for stratified analysis using the equation, Yt ¼ b0 þ b1 time1 (January 2008 to March 2009) þ b2 publication of PSA trial result (March 2009) þ b3 time3 (after March 2009 and before May 2012) þ b4 USPSTF recommendation (May 2012) þ b5 time5 (after 2012 recommendation) þ b6 subgroups (using dummy variable for more than 2 groups), where b1, b3 and b5 estimate the change in the slope for each period, and b2 and b4 evaluate the change in levels (ie immediate effect) after the publication of screening trails9,10 and the USPSTF recommendation,11 respectively.
IMPACT OF RECOMMENDATIONS ON PROSTATE CANCER SCREENING IN HEALTH CARE SYSTEM
1739
Figure 1. Trend in overall screening cholesterol testing, PSA screening and adjusted PSA or PSA ratio. For cholesterol and PSA screening b represents overall change in number of tests per month. For adjusted PSA b indicates change in ratio of PSA to cholesterol tests.
In the stratified analysis and plot for location the first segment (January 2008 to March 2009) was ignored for rural hospitals since data were only available for analysis from these hospitals starting in late 2008. As a result, changes in screening in this region during the initial period were partially due to health service development.
Figures 1 to 3 show statistically significant changes with time (slopes of trend lines) in bold. The slope is represented as b ¼ beta. All statistically significant p values are p
Purpose: The United States Preventive Services Task Force recently recommended against routine prostate cancer screening, stating that the risks of screening outweigh the benefits. We determined the impact of this recommendation on prostate cancer screening in a large health system. Materials and Methods: We obtained data on all screening prostate specific antigen tests performed at University Hospitals Case Medical Center and affiliated hospitals in northeastern Ohio from January 2008 to December 2012. We examined the total number of prostate specific antigen tests ordered with time and adjusted for patient volume by fitting a regression line. The overall trend was examined and stratified by location (urban, suburban or rural), patient age and provider type (primary care or urology). Results: A total of 43,498 screening prostate specific antigen tests were performed from January 2008 to December 2012. Most tests were ordered by specialists in internal medicine (64.9%), followed by family medicine (23.7%), urology (6.1%) and hematology/oncology (1.3%). Prostate specific antigen screening increased with time until March 2009, when initial screening trials were published. Prostate specific antigen testing then decreased significantly and continued to decrease after the task force recommendations. Similar patterns were noted in almost all subgroups. The greatest decrease in screening was observed by urologists and in patients in the intermediate age group (50 to 59 years). Conclusions: United States Preventive Services Task Force recommendations appeared to have decreased prostate cancer screening. The greatest impact was seen for urologists and patients in the intermediate age group. Further study is needed to determine the long-term effects of these recommendations on the screening, diagnosis, treatment and prognosis of this prevalent malignancy.
Abbreviations and Acronyms ACP ¼ American College of Physicians AUA ¼ American Urological Association PCP ¼ primary care physician PSA ¼ prostate specific antigen USPSTF ¼ United States Preventive Services Task Force Accepted for publication December 6, 2013. Study received institutional review board approval. Supported by National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases Award T32DK091213. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. * Correspondence: Department of Urology, Case Western Reserve University, Urological Institute, University Hospitals Case Medical Center, 11100 Euclid Ave., Mailstop LKD 5046, Office 4576, Cleveland, Ohio 44106 (telephone: 216-844-4831; e-mail: robert.abouassaly@ uhhospitals.org).
See Editorial on page 1648.
Key Words: prostate, prostatic neoplasms, prostate-specific antigen, mass screening, practice guidelines as topic
PROSTATE cancer is one of the leading causes of cancer related death in men and has a substantial burden on the health care system.1,2 In 1994 the Food and Drug Administration approved PSA testing for screening in prostate cancer in men 50 years old
or older together with digital rectal examination.3 This led to a dramatic increase in the incidence of prostate cancer in the United States.4,5 Stage migration was subsequently noted with a greater proportion of patients diagnosed with low risk, low volume
0022-5347/14/1916-1737/0 THE JOURNAL OF UROLOGY® © 2014 by AMERICAN UROLOGICAL ASSOCIATION EDUCATION AND RESEARCH, INC.
http://dx.doi.org/10.1016/j.juro.2013.12.010 Vol. 191, 1737-1742, June 2014 Printed in U.S.A.
www.jurology.com
j
1737
1738
IMPACT OF RECOMMENDATIONS ON PROSTATE CANCER SCREENING IN HEALTH CARE SYSTEM
prostate cancer.6 Paralleling this was a decrease in mortality from this malignancy. Although a causal relationship with PSA screening was hypothesized, it has not been definitively established.7 However, PSA screening is not without controversy due to the potential over diagnosis and overtreatment of clinically insignificant prostate cancer. As a result, some organizations have advised against widespread PSA screening. For example, the United Kingdom National Health Service does not mandate or advise PSA screening, instead allowing men to decide based on the recommendation of their physician.8 Recently, large randomized trials were performed in the United States and Europe on the impact of prostate cancer screening on prostate cancer mortality.9,10 Although the 2 studies had various methodological issues, the PLCO (Prostate, Lung Colorectal and Ovarian) screening trial performed in the United States revealed no significant benefit to prostate cancer screening using PSA.9 On the other hand, the European study showed that PSA based screening decreased the rate of death from prostate cancer, although the number needed to screen was high.10 After reviewing the evidence surrounding the usefulness of PSA screening the USPSTF concluded that the potential benefit did not outweigh expected harms in patients not already diagnosed or being treated for prostate cancer.11 The Cochrane Library subsequently released an updated review and metaanalysis of PSA screening showing that screening did not significantly decrease prostate cancer specific mortality.12 In addition, others did not find that PSA screening is cost-effective in men at average to high risk of prostate cancer but it may be costeffective in those at very high risk.13 As a consequence, clinical practice guidelines for prostate cancer screening vary and are controversial due to uncertainty as to whether the benefits of screening ultimately outweigh the risks of over diagnosis and overtreatment.14e17 There are also the 2 recent clinical practice guidelines for prostate cancer screening by the ACP18 and AUA.19 The ACP only recommends that clinicians inform men between ages 50 and 69 years about the benefit and harm of PSA screening to make a shared decision considering general health status and life expectancy. It does not recommend PSA screening in men younger than 50 or older than 69 years, or with life expectancy less than 10 years. The new AUA guideline on early detection of prostate cancer only recommends PSA screening in men older than 40 years at high risk for prostate cancer or shared decision making based on patient preferences in those 55 to 69 years old.18 It does not advise screening in men younger than 40 or older than 70 years, or those with less than 10-year life expectancy.
We determined whether there was a recent change in the use of PSA screening in a large university health system after the release of new evidence from various large trials as well as the updated ACP and AUA guidelines.18,19 We also assessed whether effects differed by provider or patient characteristics.
MATERIALS AND METHODS After obtaining institutional review board approval we performed a retrospective, secondary data analysis of deidentified laboratory data. We obtained for analysis the records of all PSA screening tests performed at University Hospital Case Medical Center and affiliated hospitals from January 1, 2008 to December 31, 2012. Screening PSA tests were identified by the V76.44 screening PSA diagnostic code and confirmed by the billing code. PSA tests ordered for other diagnostic purposes or for monitoring prostate cancer were excluded from analysis. PSA screening tests were extracted from the health system central laboratory, including variables of provider location (facility), physician specialty, month and year of service, and patient age. Analysis included 7 hospitals in northeastern Ohio. University Hospitals Case Medical Center was the only urban hospital. Bedford, Richmond and Ahuja medical centers are in a suburban setting, and Conneaut, Geauga and Geneva medical centers are in rural areas. To adjust for patient volume and population changes in this large health system we used the ratio of the number of PSA screening tests to cholesterol tests done at the hospitals per month. The study period was divided into 3 segments based on the publication of the 2 mentioned screening trials in March 20099,10 and USPSTF recommendations in May 2012.11 The slope of each segment was calculated and compared using linear regression and segmented regression models (supplementary table, http://jurology.com/ and figs. 1 to 3). The segmented regression model included the effect of earlier period(s) while the linear regression model did not. Analysis was then repeated, stratified by patient age (less than 50, 50 to 59, 60 to 69, 70 to 79 and 80 years or older), physician specialty (internal medicine, family medicine, urology and medical oncology) and facility type (urban, suburban and rural). We included all physician specialties in the analysis but performed followup stratified analysis of physician specialty for those that contributed most to screening, including primary care, urology and hematology/oncology. In each subgroup we used the ratio of PSA tests performed to total cholesterol tests performed per month for stratified analysis using the equation, Yt ¼ b0 þ b1 time1 (January 2008 to March 2009) þ b2 publication of PSA trial result (March 2009) þ b3 time3 (after March 2009 and before May 2012) þ b4 USPSTF recommendation (May 2012) þ b5 time5 (after 2012 recommendation) þ b6 subgroups (using dummy variable for more than 2 groups), where b1, b3 and b5 estimate the change in the slope for each period, and b2 and b4 evaluate the change in levels (ie immediate effect) after the publication of screening trails9,10 and the USPSTF recommendation,11 respectively.
IMPACT OF RECOMMENDATIONS ON PROSTATE CANCER SCREENING IN HEALTH CARE SYSTEM
1739
Figure 1. Trend in overall screening cholesterol testing, PSA screening and adjusted PSA or PSA ratio. For cholesterol and PSA screening b represents overall change in number of tests per month. For adjusted PSA b indicates change in ratio of PSA to cholesterol tests.
In the stratified analysis and plot for location the first segment (January 2008 to March 2009) was ignored for rural hospitals since data were only available for analysis from these hospitals starting in late 2008. As a result, changes in screening in this region during the initial period were partially due to health service development.
Figures 1 to 3 show statistically significant changes with time (slopes of trend lines) in bold. The slope is represented as b ¼ beta. All statistically significant p values are p
