Page 118 - An Evidence Review of Active Surveillance in Men With Localized Prostate Cancer

Figure 7. Evidence map for key question 3

Surveys

Focus groups

MV modeling

Offer Acceptance Adherence Predict WW

Studies considered for Key Question 3 are depicted as circles with size proportional to the study sample size, stratified by design
(focus groups, surveys, or studies reporting multivariable prediction models) and outcomes assessed (“offer” of, “acceptance” of,
“adherence” to observational management strategies or prediction of having received observational management). One study that
reported factors influencing the acceptance of randomization versus non-randomized choice of treatment is not shown. Hollow
circles represent studies including prostate cancer patients only; black circles represent studies of physicians only; gray circles
represent studies including both patients and physician. Placement within each box is random. MV = multivariable; WW =
watchful waiting.

Only the qualitative results of the multivariable analyses are described in this section.
Because the reviewed studies used heterogeneous coding schemes for their predictors (for
example, age was used as a continuous variable in some studies but as a nominal [discrete]
variable in others) and adjusted for varying sets of confounding variables, meaningful
comparison of effect sizes across studies is precluded. For these reasons, we do not present effect
estimates for each predictor of interest from these analyses (such as odds ratios for predicting
treatment received, or hazard ratios for WW interruption), although detailed quantitative
information is available in Appendix Tables C3.1 to C3.3.
It should also be noted that the common method for reporting “adherence to AS” in the
literature is the “interruption of AS” to seek definitive treatments and we follow this convention
in our review. A man could interrupt AS to seek curative treatments for several reasons, among
which: 1) the person meets some criteria on AS protocol indicative of disease progression that
would call for curative treatment, 2) the person does not meet criteria for curative treatment (i.e.,
continued surveillance is indicated), but due to personal preference, he decides to stop AS and
pursue curative treatment, and 3) the person decides to forgo present or future curative treatment
(e.g., because of advanced age or new comorbidities) and switches to WW. The first reason
would commonly be considered “adherent” (the person is following the protocol) and the latter
two “not adherent” (he chose to discontinue the AS protocol even though there was no indication

64