Answers are grounded in this paper and cite their sources.
Prostate cancer is the most commonly diagnosed (non-skin) cancer among men in the United States, with 179 000 new cases and 37 000 deaths expected in 1999 ( 1 ). Although $5 billion is spent annually for the care of those diagnosed with prostate cancer ( 2 ), a systematic evaluation of practice patterns and the impact of treatments on health-related quality of life (HRQOL) on a national scale has yet to be conducted. In 1994, the National Cancer Institute (NCI) initiated the Prostate Cancer Outcomes Study (PCOS) to investigate variations in the initial treatment of prostate cancer and to describe HRQOL outcomes in a large, heterogeneous cohort of newly diagnosed prostate cancer patients treated in community medical practices. The study is the most comprehensive population-based outcomes study focusing on prostate cancer ever conducted and complements the national prostate cancer research agenda. The main purpose of this commentary is to provide the rationale and objectives for the PCOS and to describe how the study fills a critical gap in the understanding of men's experiences with prostate cancer. Better knowledge of treatment outcomes is desired by thousands of patients, families, and clinicians who deal with prostate cancer every day. The PCOS was initiated to obtain more comprehensive and generalizable information about the effects of treatments on HRQOL outcomes. Information about HRQOL is also important, from a cancer surveillance perspective, for providing a broader perspective for interpreting recent trends in prostate cancer incidence and mortality. Given the uncertainty about the efficacy of selected interventions, it is essential to document the complete burden of the disease in order to characterize the extent to which progress is being made. A second purpose of this commentary is to describe the design, methods, and limitations of the PCOS in order to enhance the understanding and interpretation of numerous ongoing and planned research projects using the data collected. The incidence rate of prostate cancer has increased dramatically during the past decade, largely as a result of the rapid dissemination of prostate-specific antigen (PSA) screening, although incidence rates have been declining in more recent years ( 3 , 4 ). The use of radical prostatectomy has also increased, particularly among men 65 years old or older ( 5 , 6 ). Mortality rates have changed little over the past two decades, although a declining trend has occurred since 1991, following an increase during the mid-1980s. The failure to observe a more substantial and rapid impact of interventions on mortality may be due to several reasons. First, the majority of men initially diagnosed with the disease die with, rather than of, prostate cancer ( 7 , 8 ). This is due in part to the advanced median age at diagnosis of prostate cancer (age 71 years) as well as to the prolonged natural history of early-stage prostate cancer ( 9 ). Therefore, it may take many years to observe an impact of some interventions on population rates. Second, screening and aggressive treatments may not reduce cause-specific mortality. To date, there are no completed randomized trials that have definitively established the efficacy of screening for prostate cancer or of aggressively treating patients with early-stage disease with either external-beam radiation therapy or radical prostatectomy ( 10 , 11 ). Observational studies of large population-based cohorts demonstrate that 15-year survival of selected patients with localized tumors who are conservatively managed is quite favorable and appears similar to that of the age-matched U.S. population ( 8 , 12 ). The prolonged natural history and late age at diagnosis, in addition to the lack of definitive evidence for the efficacy of interventions, have contributed to the uncertainty and disagreement about whether to screen asymptomatic men and how to best manage patients diagnosed with nonmetastatic prostate cancer ( 13 - 16 ). While there is a lack of convincing evidence on the effects of surgery or radiotherapy on prostate cancer mortality, there is evidence that these therapies can have detrimental effects on urinary, bowel, and sexual functions in some individuals. Given the incompleteness of information regarding mortality effects, it is difficult to quantify and reconcile the tradeoffs of specific treatment strategies. Uncertainty remains about the exact levels of impairment that patients can expect after receiving different treatments because, until recently, the impact of prostate cancer on long-term HRQOL outcomes had not been thoroughly investigated in representative cohorts. Much of the early research on outcomes of prostate cancer treatment was limited to small samples from highly selected case series treated in tertiary referral centers ( 14,17-20 ). Some of these influential reports may have underestimated the long-term effects of treatment. For example, estimates of incontinence were as low as 5% and impotence rates have been estimated at lower than 50% following radical prostatectomy in major treatment centers. Studies ( 21 - 27 ) have documented a more substantial decrement in urinary, bowel, and sexual functioning. These reports have been instrumental in identifying the long-term effects of prostate cancer treatment on outcomes, but they are also limited by one or more design characteristics, including ( a ) ascertaining HRQOL cross-sectionally at a single point in time, long after the acute effects of treatment; ( b ) including only long-term survivors, who may have better outcomes; ( c ) not including all age or racial/ethnic groups; ( d ) not stratifying outcomes by important clinical and nonclinical confounding characteristics associated with treatment choice; and ( e ) selecting patients treated in a particular health plan or major tertiary referral center. The PCOS was designed to address or overcome these limitations. Information about HRQOL cannot by itself establish which treatments are desirable, since efficacy is likely to be the primary consideration for most men. Rather, the information to be collected in the PCOS can be used as part of a broader decision-making framework to assist clinicians and their patients, who desire information about HRQOL effects in order to make decisions tailored to their own preferences. Given the scientific uncertainty about efficacy, decisions about prostate cancer interventions should incorporate individual preferences. One study ( 28 ) has shown that preferences for prostate cancer screening and treatment are affected by information about medical uncertainties. A National Institutes of Health Consensus Development Conference ( 29 ) recommended radical prostatectomy or radiation therapy for treating locally confined tumors but noted that patient preferences and quality of life were important considerations when choosing among multiple therapies. Results from clinical decision models ( 30 - 32 ) suggest that both clinical and policy decisions may be based not only on the natural history of disease and treatment efficacy but also on patient preferences for outcomes of competing treatment strategies. The uncertainties surrounding treatment have given rise to variable practice styles, illustrated by considerable geographic variations in the use of therapies for clinically localized prostate cancer across the United States and in different racial groups ( 6 , 33 ). Therefore, another objective of the PCOS is to clarify the contributions of various clinical and nonclinical factors to variations in diagnostic and treatment practices. The PCOS will collect the most extensive data ever assembled for investigating patterns of prostate cancer care. The PCOS has a unique advantage over prior outcomes studies because it was designed within an existing population-based cancer registry system. Since 1973, the NCI's Surveillance, Epidemiology, and End Results (SEER) Program 1 has provided information on cancer incidence and survival for defined regions of the United States ( 34 ). The SEER Program consists of 10 central cancer registries that routinely collect and report information on all cancers diagnosed in their catchment areas, including sociodemographic information and tumor characteristics, such as final pathologic stage and histologic grade. Active and passive follow-up is also conducted to ascertain vital status and underlying cause of death. More than 11 000 men diagnosed with biopsy-proven, primary invasive carcinoma of the prostate during the period from October 1, 1994, through October 31, 1995, in six of the SEER cancer registries were eligible for the PCOS. Patients were diagnosed in the states of Connecticut, Utah, and New Mexico and in the Metropolitan areas of Atlanta (GA), Los Angeles (CA), and Seattle (WA). Men younger than age 90 years at diagnosis were eligible, except in Seattle where men younger than age 60 years were excluded because they were eligible for another ongoing study. All sampled men (with the exception of one man who had radiologic confirmation only) had microscopic confirmation of prostate cancer. The six participating registries each employed a Rapid Case Ascertainment System (RCAS) to identify eligible patients within 6 months of diagnosis. The RCAS is a widely used, labor-intensive method employed by cancer registries. It requires frequent communication with pathology laboratories servicing the registry catchment areas, including some laboratories that are outside the registry areas. The proportion of cases missed by the RCAS, as determined by comparing the number of cases identified as eligible with all of the cases reported by participating SEER registries through February 1999, was approximately 5%. Eligible patients were sampled according to a prespecified sampling design to ensure a sample representative of the population of eligible prostate cancer patients. Another objective of the sampling design was to obtain a sufficient number of minority men and younger men for assessments of HRQOL in different population subgroups. To achieve this objective, a random sample of non-Hispanic whites 60 years old or older at diagnosis was drawn from among eligible patients in each registry. Men younger than 60 years were over-sampled, with all registries other than Seattle sampling all men in this age group. Hispanic men in New Mexico and black men in Atlanta were over-sampled. Los Angeles, Connecticut, Utah, and Seattle selected all black and Hispanic men. Sample weights for the PCOS have been calculated as the inverse of the sampling proportions within each region-race-age group stratum. This method permits estimates of combined data across age-race-region strata that are appropriately weighted to the total number of eligible prostate cancer patients. The centerpiece of the PCOS data collection effort was a survey questionnaire designed to obtain self-reported HRQOL at 6, 12, and 24 months after initial diagnosis. An essential premise of the PCOS is that patients are the best source of information about their HRQOL and the presence and severity of disease and treatment-related dysfunction. There is evidence showing substantial differences in reports of HRQOL by cancer patients and their clinicians ( 35 , 36 ). In a study based on the “CaPSURE” Database ( 37 ), which contains extensive data for more than 4000 men with prostate cancer, investigators documented that physicians' and patients' ratings of symptoms following treatment did not agree. A mixed-mode survey approach was used to achieve a high response rate and to elicit accurate responses to sensitive questions because some men in focus groups indicated a preference for the self-administered versus a telephone format. After the necessary local Institutional Review Board approvals and (in some registries) physician consent were obtained, patients were sent a self-administered questionnaire along with a cover letter explaining the purpose of the study and requesting their participation. If no questionnaire was returned within 2 weeks, the men were called and a second survey was mailed, if necessary. After another 2 weeks, a second reminder call was made and a telephone interview was attempted. A second identical survey was sent at approximately 12 months after diagnosis. Men who actively refused participation at 6 months and men who had died or were lost to follow-up were not mailed a 12-month survey and are considered nonresponders. Some men who were passive refusers or not located at 6 months completed only the 12-month survey. A third survey, identical to the 12-month survey, was mailed at 24 months after diagnosis or 1 year after the 12-month survey. Those eligible to complete a 24-month survey included men who had completed a 6-month and/or a 12-month survey and who had not actively refused to participate at any previous contact. Disease-specific HRQOL was measured with the use of an index derived from three existing instruments with demonstrated validity and reliability from prior studies ( 21,24,38 ). A new index for measuring urinary, bowel, and sexual dysfunctions was from existing because of the for an tailored to the of a large, self-administered survey across patient subgroups. The newly PCOS prostate index was with the use of several focus groups and a study on prostate cancer patients. This was to ensure the of the and to the validity of the The final PCOS index contains six similar to the approach by ( ) ( For each of three and sexual the index consists of two One scale men to rate their of in each using or and the other scale men to the extent to which their of is a for using a single The survey was and among men with prostate cancer. To over due to prostate cancer, the 6-month survey to report their urinary, bowel, and sexual functions during the past as well as their functions prostate cancer diagnosis. A study among men from a sample is the of 6-month of and is comparing and of in on the survey were about symptoms of with and the presence and severity of major to be associated with treatment and questions were used to ascertain and To the index and to other health that are part of selected from the Outcomes Study health survey were used ( ). of the to burden and to the focus on excluded three of the and and health that were to be sensitive to differences treatment groups and over of the PCOS survey are from Another unique of the PCOS data collection effort extensive medical to establish characteristics and to of long-term HRQOL outcomes according to these The participating registries in medical from radiologic or of Health and physician Information difficult to obtain from patients was such as diagnostic tumor characteristics, of treatments and acute of therapies. The of physician was necessary to collect information not routinely collected by such as clinical stage and and specific therapies from multiple and radiation were on the of of treating from patients at the of the survey and from registry was conducted at 1 year after diagnosis in physicians' to ensure more complete collection of treatments given within the For comparing and age at diagnosis, stage and and of initial treatment were from the SEER Program and were measured with the use of information at the from the U.S. of the and with SEER was defined as the of in the who had a and median were in within each geographic with the the for each were used to and on all and the was used to the age of the with that of the nonresponders. All reported are More than 11 men with prostate cancer were initially identified for sampling for the PCOS. of these patients were excluded as of for the PCOS. patients those of other than or those determined to to have been diagnosed outside the study or outside the registry catchment areas, and those who died the 6-month survey patients missed by the RCAS and by SEER and February 1, 1999, were as After these a total of 11 men the eligible patient population for the PCOS. The of response for the study was as the number of men either a 6-month or a 12-month HRQOL survey, or by the total number of men sampled 1 the for no the sampled patients, were at to complete a and/or a 12-month survey. all sampled patients, men completed either a 6-month survey only a 12-month survey only or both a and a 12-month survey levels by from to The two main that men did not participate in the PCOS included or defined as when a questionnaire was returned or was not by a self-administered mailed survey with the participating by or of those men eligible to complete the 24-month follow-up survey did The of PCOS for medical were completed was the men who in the and/or the 12-month survey, all but also had completed 2 the levels of response to the PCOS according to age group and response levels with age at diagnosis, with the response among men and were for non-Hispanic whites than for non-Hispanic or The population of for the PCOS is men diagnosed with prostate cancer in the United Although the PCOS was in six defined the sampling design a representative sample of all eligible patients in those areas. may be in the survey by the high rate of survey among the sampled patients, since may from nonresponders. To the for 3 the of selected characteristics from the SEER Program among PCOS survey and nonresponders. to the PCOS were younger than more likely to be non-Hispanic and more likely to be in areas with levels of versus in and median versus in The included more men with stage and cancers than although this was not clinically The were more likely than to radical prostatectomy versus These differences that some estimates from PCOS that information across and status may not be generalizable to prostate cancer patients will be used to reduce the extent of when practice patterns and are First, all planned studies using the PCOS data to outcomes after treatment will by age and initial treatment and will for other clinical and sociodemographic characteristics with the use of In survey to the inverse of the response within each region-race-age have been These weights may be by the sampling weights to estimates for and weighted to the total eligible population of patients with prostate cancer. These may only to the extent that the all eligible case within each stratum. It is that survey in are in health than In the were older and from lower status areas, this it is also that some men more frequent and following their treatment for prostate cancer are more to in which case may have outcomes. The and of this are to health status information about the nonresponders. It has been well established that the of treatments for clinically localized prostate cancer quite by ( 6 , 33 ). The SEER regions to be areas, with more and more medical than in the of the United States ( , ). weighted estimates from PCOS of treatment may not be generalizable to the as a the factors associated with among competing and the of treatments and long-term effects are to in the six PCOS regions with the of the United The PCOS will describe variations in medical and HRQOL among men diagnosed with prostate cancer. The study will be a critical source of information for the burden of the disease in the U.S. which is for new and at progress in prostate The PCOS will also randomized which interventions in and by providing data on the impact of interventions in community practice The of an extensive population-based study of HRQOL among patients with cancer in medical on the and of cancer registry was demonstrated by the of rapid case by six registries. The study also the of men with prostate cancer and their to be part of research about the The to patient survey data with medical clinical characteristics and treatments is an important advantage for HRQOL The PCOS survey was designed to focus on urinary, and to be most to men with prostate cancer ( ). The of both and of HRQOL following prostate cancer is necessary to outcomes of since specific may be more sensitive for the and of small that are important to patients and clinicians ( ). vital status follow-up of the PCOS cohort is There will be a HRQOL survey initiated in the of 1999 to research are the of diagnostic and treatment practice of pathologic stage following radical patterns in the use of treatments for and and in HRQOL over a period of 24 months after various treatments for prostate cancer. The extensive data collected by the PCOS should provide important new information the diagnosis and of prostate cancer in the United States to patients, and policy of the following best frequent no have or at than a about a or a more than a many or if have used to with or one two three or more have had to than 2 after or not at than the about the more than the how a have had with or small small have had more than three on a single have had any or or have had not to to the have had in the have had with or every some or not at all how a have had with or small small have been in sexual or A only a not at all have in any sexual a a two to three a a not at all had any for sexual had any that were not for sexual have had an during sexual a little some a of not at all how a sexual to small small for no response in the Prostate Cancer Outcomes Study in of not to because of 29 patients for was other than or for no response in the Prostate Cancer Outcomes Study in of not to because of 29 patients for was other than or Prostate Cancer Outcomes Study survey response levels according to age at diagnosis and sampled patients. men who completed a and/or a 12-month survey. response were calculated with the use of all sampled patients in the patients with at of sampling are included in the Prostate Cancer Outcomes Study survey response levels according to age at diagnosis and sampled patients. men who completed a and/or a 12-month survey. response were calculated with the use of all sampled patients in the patients with at of sampling are included in the of Prostate Cancer Outcomes Study and For each the number of from that of for age at are weighted to the total eligible of prostate cancer patients. not to because of SEER Surveillance, Epidemiology, and End of Prostate Cancer Outcomes Study and For each the number of from that of for age at are weighted to the total eligible of prostate cancer patients. not to because of SEER Surveillance, Epidemiology, and End SEER is a of central cancer registries in the United States, by local to the National Cancer Institute data are to the on a and the the data to the for scientific the following for their in this study as the for the Prostate Cancer Outcomes for of in the design of the survey study. Cancer and Cancer on the design of the survey. in each who locally managed data and and New and Los and and and also the many other in these regions who in data also the in these six areas who in the collection of data from their patients and from medical to particularly the men by their participation in the have contributed to a better understanding of the effects of prostate cancer on men's
No takes yet. Share an insight, caveat, or question.
Potosky et al. (1999) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: