Page 16 - An Evidence Review of Active Surveillance in Men With Localized Prostate Cancer

management strategies. We also searched for studies of specific databases, including SEER
(Surveillance Epidemiology and End Results) and CaPSURE (Cancer of the Prostate Strategic
Urologic Research Endeavor). For Key Question 4, we relied on previous systematic reviews on
prostate cancer conducted for the AHRQ EPC program. Searches were supplemented with
studies recommended to us by the Technical Expert Panel, reference lists of eligible primary
studies and relevant review articles, and targeted searches for economic evaluations. We did not
include unpublished data.
Below are the study eligibility criteria we used for the first four Key Questions (no specific
literature search was performed for Key Question 5):

Key Question 1. Studies of large U.S.-based databases of patients with prostate cancer with
time-trend data (reporting changes over a range of years) between 1980 and 2011. Studies must
have had a sample size of at least 1000 patients. We also reviewed the latest version of the
1
2
American Cancer Society Cancer Statistics report , a recent SEER Survival Monograph , and
b
data available on the SEER Web site.

Key Question 2. Studies of any design that reported protocols and management strategies for
patients receiving observational management (i.e., no immediate curative treatment). We
included both studies where the goal of observation was to identify disease progression
indicative of the need for curative treatments, and studies where the goal of observation was to
determine the need for palliative treatments.

Key Question 3. Three types of studies were included. Firstly, we included studies that used
quantitative methods to analyze databases or cohorts of patients to elucidate predictors of the
offer or acceptance of or adherence to observational management strategies (including AS and
WW). We excluded studies that analyzed ADT together with observational management
strategies. We required multivariable analyses adjusting for a minimum of age and tumor stage
(if the analysis was not limited to localized cancer) or using a propensity score. Secondly, we
included studies using qualitative research methods (e.g., focus groups or surveys) to obtain
information on factors that affect the offer or acceptance of or adherence to AS or WW. Eligible
studies must have used a predefined approach to collect information. Thirdly, we also searched
for experimental studies evaluating the effect of tools, such as decision aids, on the offer or
acceptance of or adherence to AS (however, no such studies were found).

Key Question 4. We included randomized and nonrandomized, prospective or retrospective
longitudinal comparative studies performed in a multicenter setting. Nonrandomized studies
must have used multivariable or other methods to adjust for possible confounding, specifically
for age and tumor stage, to warrant inclusion. The population of interest was men with clinically
localized prostate cancer (T1-T2), without known lymph nodes (N0-X) or metastases (M0-X).
No more than 20 percent of the study sample could exhibit more advanced disease. Studies had
to compare observational management strategies (without ADT) to active treatment, including
RP, external beam RT (EBRT), or brachytherapy (BT), all with or without ADT. However, ADT
monotherapy was not considered an active treatment. Outcomes of interest included: prostate-
cancer mortality, all-cause mortality, morbidity of primary treatment, metastatic disease, quality
of life, and costs.

b Available at http://seer.cancer.gov/faststats/; last accessed September 30, 2011.

ES-6