likely to have parttime therapists, compared to national averages. From these results, the authors recommended that facilities using cobalt units should upgrade their treatment equipment, give palliative treatment only, or close. While this study helped to show that cobalt units were not as good as other types, the evidence available when the care was provided did not indicate that cobalt use was inappropriate. Therefore, this study was quite valuable in showing how to improve care, but it is not evidence of poor quality. Quality has to be judged by the level of knowledge and standards in place at the time care was delivered.
Treatment for advanced prostate cancer is palliative. Data from randomized controlled trials demonstrate a survival benefit as well as relief from bone pain by treatment with androgen ablation (the elimination of sources of male hormone such as testosterone), which can include orchiectomy alone, monotherapy with a luteinizing hormone-releasing hormone (LHRH) analogue, or "maximal androgen blockade" with either orchiectomy or an LHRH analogue and anti-androgen therapy (Garnick, 1996). When prostate cancer progresses on androgen ablation therapy, treatment is less effective; however, various drugs (ketoconazole or aminoglutethimide, estramustine, suramin, mitoxantrone with prednisone or steroids) can improve pain control and quality of life (Garnick, 1996). Because patients with prostate cancer that has metastasized to the bone often suffer excruciating pain, a primary focus in the care of patients with metastatic prostate cancer is control of their pain, with either narcotics, radiation therapy, or chemotherapy. So, although advanced prostate cancer is not curable, multiple treatment options exist and there is evidence in the scientific literature that they improve quality of life and, in some cases, prolong survival. Thus, there is sufficient evidence for process-outcomes links in advanced prostate cancer that process measures could be developed to evaluate the quality of care.
Prostate cancer provides a particular challenge for quality-of-care assessment: methods for early detection are available, but there is not yet definitive information about whether early detection improves survival. There are a number of treatment modalities for early-stage disease, but there is not definitive information about the efficacy of early treatment. The results of the PLCO trial will better inform decisions about whether routine screening for prostate cancer should be performed and for whom; and the results of the PIVOT trial will provide information about the efficacy of primary treatment of localized prostate cancer by surgery.
One candidate indicator for prostate cancer quality assessment is to identify whether information about alternative treatment modalities was presented to patients, as recommended by the AUA's practice guidelines (Middleton et al., 1995). A second candidate indicator may be to assess the rates of surgical treatment among men with life expectancies less than 10 years, using age 70 to represent a proxy for 10-year remaining life expectancy for the average patient (a high rate of surgical treatment in men with a life expectancy of less than 10 years would be an indicator of poor quality).