The widespread availability of screening coupled with high risks of over diagnosis, over treatment, and a rising risk of dying from competing causes of deaths pose a major challenge for prostate cancer patients and our health care system. Data on outcomes and identification of patients who are likely to benefit from therapies are urgently needed to empower patients and health professionals and to allocate limited health resources wisely. The long-term goal of this population-based study is to improve prostate cancer care for elderly patients. In 2004, approximately thirty percent of the 230,110 new prostate cancer cases diagnosed will be in men over age 75. Elderly men have a high risk of being diagnosed with prostate cancer, but a relatively low risk of dying of this disease, and are often excluded from cancer trials. While the value of primary androgen deprivation for localized prostate cancer remains questionable, this treatment is widely used among elderly patients. Currently, there is insufficient outcomes data to guide treatment choices for elderly patients.The following are the key background considerations that provide the rationale for this proposed project:· Prostate cancer is prevalent among elderly men. · Over diagnosis and competing causes of death are important issues for elderly men with prostate cancer.· Androgen deprivation therapy (ADT) is being increasingly used as primary therapy among elderly patients with localized disease· There are insufficient outcomes data to support the use of primary androgen deprivation for non metastatic prostate cancer· Preliminary dataThe study population in this sub-project is a sub-sample of the large study (IRB # 5177 ) and the research methods are identical in both studies.

Inclusion criteria:

• Diagnosed at age 75 or older.
• Diagnosed with non-metastatic prostate cancer in 1992-1999.
• Enrolled in Medicare for the 12 months before diagnosis for an adequate assessment of baseline comorbidity.
• Enrolled in Medicare Part A and Part B through death or the end of the study period to have an adequate claim history of cancer treatment over the follow-up period.
• Alive for at least 6 months after cancer diagnosis. This criterion will remove men who are diagnosed at death or who are dying and have severe comorbidities, and are therefore poor candidates for 'aggressive' cancer interventions.

Exclusion criteria:

• Enrolled in Medicare managed care plans during the study period since no Medicare claims are available during the period of enrollment.
• Have other types of cancer. This criterion will ensure that all cancer therapies identified from the Medicare claims are for prostate cancer exclusively.

To compare overall and prostate cancer specific survival among elderly patients receiving primary androgen deprivation therapy and conservative management for non-metastatic prostate cancer diagnosed in 1992-1999