Prostate cancer radiation treatment

Dramatic changes have occurred in the U.S. healthcare system in the past two decades, with enrollment in health maintenance organizations (HMOs) increasing from 4 percent to 20 percent of the population between 1980 and 1996 (National Center for Health Statistics 1997). There has been a particularly rapid increase in HMO enrollment among the Medicare population, with 14 percent of Medicare beneficiaries now enrolled in HMOs, compared to just 6 percent in 1990. The impact of these trends on the process and outcomes of healthcare delivery for Medicare beneficiaries remains uncertain. An evaluation of the Medicare HMO program in the late 1980s found health outcomes to be similar between HMO and fee-for-service (FFS) patients, with a less intensive use of resources in the HMOs (Clement et al. 1994; Brown, Bergeron, Clement, et al. 1993). One study demonstrated similar survival but less frequent use of rehabilitation facilities among stroke patients in Medicare HMOs compared to FFS care (Retchin, Brown, Yeh, et al. 1997), while other studies reported mixed results, with some outcomes worse for Medicare HMO enrollees with chronic conditions (Miller and Luft 1994; Retchin, Clement, Rossiter, et al. 1992; Ware, Bayliss, Rogers. et al. 1996; Shaughnessy, Schlenker, and Hittle 1994; Manton, Newcomer, Lowrimore, et al. 1993).

There has been increasing concern that the rapid growth of the managed care industry may be limiting access to treatments for seriously ill patients. This concern stems from the perceived focus of managed care on cost containment through strict utilization management and limits on access to specialty care (Clancy and Brody 1995; Kassirer 1996; Brook, Kamberg, and McGlynn 1996). Numerous anecdotal accounts about restricted access and poor outcomes in managed care plans have appeared in the popular media (Mechanic 1997). Researchers, policymakers, clinicians, and the general public might benefit from objective, comparative information across diverse healthcare delivery systems on treatment practices and health outcomes for patients with prevalent chronic conditions.

Cancer remains the second-leading cause of death in the United States, behind heart disease. Although some studies have shown that HMO enrollees may receive more breast, cervical, and colorectal screening tests than FFS beneficiaries (Potosky et al. 1998; Riley et al. 1994; Retchin and Brown 1990a), only a limited number of studies compare cancer treatments and outcomes for patients in HMO versus FFS settings (Riley, Potosky, Klabunde, et al. 1999; Retchin and Brown 1990b; Greenwald and Henke 1992; Lee-Feldstein, Anton-Culver, and Feldstein 1994).

Prostate cancer is a most commonly diagnosed non-skin cancer among men, with approximately 179,000 new cases and 37,000 deaths expected in 1999 (Landis et al. 1999). More than 80 percent of new cases and 90 percent of deaths occur in men 65 years of age and older. Approximately two-thirds of the patients are initially diagnosed with tumors confined to the prostate gland (Merrill and Brawley 1997). There is considerable uncertainty and disagreement about the most effective treatment for clinically localized prostate cancer, because of lack of evidence of efficacy from randomized trials. An NIH Consensus Development Conference recommended radical prostatectomy or radiation therapy (considered "aggressive therapy") for treating locally confined tumors (National Cancer Institute 1988). However, the long-term survival of selected patients whose care is managed more conservatively is similar to that of men without prostate cancer (Albertsen, Fryback, Storer, et al. 1995; Chodak, Thisted, Gerber, et al. 1994; Johansson, Holmberg, Johansson, et al. 1997).

In a previous study, we compared treatments and survival, for Medicare-covered women diagnosed with breast cancer, between FFS care and two large prepaid staff/group practice non-profit HMOs: Group Health Cooperative of Puget Sound (GHC) and Kaiser Permanente, Northern California Region (KPNC) (Potosky, Merrill, Riley, et al. 1997). We have examined treatments and survival among colorectal cancer patients in these plans using similar methods (Merrill, Brown, Potosky, et al. 1999). In this study we extend our investigations of the association between healthcare delivery systems and cancer care and outcomes by examining whether differences exist in prostate cancer treatment and survival between the cases enrolled in these same two HMOs and cases in the FFS system.

METHODS

Data Sources and Sample Selection

Data for prostate cancer cases diagnosed between 1985 and the end of 1992 were obtained from two population-based cancer registries, one covering the five-county San Francisco-Oakland metropolitan area, and the other covering the 13-county Seattle-Puget Sound area. Both registries participate in the National Cancer Institute's (NCI) Surveillance, Epidemiology, and End Results (SEER) Program (Ries, Kosary, Hankey, et al. 1997) and collect information on cancers diagnosed in their catchment areas. Reported items used for this study include the date of diagnosis, age and stage at initial diagnosis, histological grade, and the date and causes of death. The underlying cause of death (ICD-9-CM) is obtained from death certificates through linkages with state vital records. The identification of cancer versus noncancer causes as the underlying cause of death has been shown to be highly reliable (Percy, Stanek, and Gloeckler 1981). Localized tumors are confined to the prostate gland, regional tumors are spread to contiguous organs or lymph nodes, and distant tumors are spread to remote organs. Although treatment is determined primarily by the clinical stage at diagnosis, the SEER Program (prior to 1995) collected only the final pathological stage, which is determined using all of the surgical information obtained. Clinically localized tumors may be upstaged to final regional stage after surgery. Therefore, we used both localized and regional pathologically staged cancers for our comparisons of the treatment of clinically localized cancers in order to avoid excluding men receiving surgery who may have been upstaged to regional disease. Histological grade is defined by SEER in increasing level of severity as either low grade (well-differentiated), medium grade (moderately differentiated), or high grade (poorly differentiated) carcinoma of the prostate. These categories are equivalent to the more widely used Gleason grading system as Gleason score 2-4 (low), 5-7 (medium), and 8-10 (high).

SEER registries also collect information on the initial surgery and radiation therapy administered within four months of diagnosis. Men diagnosed by death certificate or autopsy reports, or with unknown date of diagnosis or follow-up, were excluded ([less than]2 percent). Cases eligible for study were initially diagnosed with malignant prostate cancer at age 65 or over without any previous cancer diagnoses. The cases were actively followed for vital status and cause of death for a maximum of ten years through December 1994.

We used a SEER-Medicare linked database to augment the SEER registry data (Potosky, Riley, Lubitz, et al. 1993). Approximately 93 percent of all SEER prostate cancer cases age 65 and older were successfully linked with Medicare claims data. We identified all new prostate cancer cases reported to the two SEER registries from 1985 through 1992 by the two largest HMO plans in these regions, Kaiser Permanente of Northern California (KPNC) and Group Health Cooperative of Puget Sound (GHC). KPNC, the largest prepaid group plan in the United States, was established in the early 1940s, and during 1987-1990 it grew from 2.1 to 2.4 million members in Northern California, of which about half reside in the five counties covered by the SEER registry. GHC was established as the first consumer-controlled health plan in 1947. Unlike KPNC, which contracts with a distinct physician organization (The Permanente Medical Group) to provide services to plan members, GHC directly employs staff physicians. GHC covered about 385,000 persons in the Puget Sound area during the study period.