br The covariance plots showing differences between noncance
The covariance plots showing differences between noncancer patients and each treatment group are shown in Supplementary Figure 1. After matching, the noncancer and treatment groups were similar regarding several demographic characteristics.
Differences in functional status between noncancer patients and those undergoing treatment are shown in Figures 1-3. Figure 1 shows no difference in the change over time for those choosing conservative management compared with noncancer patients in ADL, PCS, or MCS score (all P > .05). Figure 2 shows that sur-gery patients experienced greater decline in PCS score (adjusted mean difference = 4.5, P < .001) and MCS score (adjusted mean difference = 3.3, P = .01), but not in ADL (P = .96). Finally, Figure 3 shows that those receiving 2-NBDG had similar changes in functional status over time as noncancer patients (all P > .05).
After adjusting for several patient characteristics, there were no differences in functional status among noncancer patients and those undergoing treatment with conservative
management or radiation. Patients undergoing surgery had a decline in their general functional status compared with their noncancer peers, although this difference may not be clinically significant.
Among the 2 measures of physical function, patients undergoing surgery experienced no differences in their ADL scores yet a greater decline in their PCS scores com-pared to matched noncancer patients. The different types of questions that comprise these measurements help explain this incongruent finding. Deficits in ADL’s (ie, bathing, dressing, toileting, getting in and out of bed or chairs, walking, and eating) represent significant declines in physical function that are uncommon among patients eligible for prostate cancer surgery. Patients who undergo surgery for prostate cancer tend to be fully functional with a life expectancy of at least 10 years.18 Patients are typically performing all their ADL’s within days after surgery, sug-gesting that this measurement of physical function may be low yield in this patient population. The PCS score, on the other hand, incorporates questions about more subtle defi-cits in physical function and inquires about activities that are more relevant for post-prostatectomy patients. For example, specific questions about walking a mile, climbing stairs, bending, difficulties with work, and pain are highly relevant, especially in the initial recovery period.19
In addition to changes in physical function, surgery patients experienced a decline in emotional well-being as measured by the MCS score. The MCS score includes ques-tions pertaining to emotional role (eg, reduce time working), vitality (eg, full of pep or tired), mental health (eg, nervous, happy, nothing can cheer you up), and social functioning (interference with social activities).19 A decline in MCS scores after surgery could certainly be attributed to emotion distress generated from complications, such as urinary incon-tinence and erectile dysfunction.20 Along these lines, a decline in emotional health after surgery may, in part, be attributed to regret. Many patients regret having their sur-gery20 and expectations about future health states can affect patient-reported quality-of-life.21 These 2 factors together highlight the importance of preoperative counseling regard-ing the potential risks and benefits of surgery.20 Several tech-niques can help align preoperative expectations with postoperative outcomes, including a careful evaluation of baseline functional status22 and the use of decision aids to enhance the shared decision making process.23
One challenge in interpreting the declining physical function and emotional well-being measurements for sur-gery patients is understanding their clinical significance. While there are varying opinions regarding what point dif-ference is clinically significant, typically a difference of half a standard deviation (5 points in the case of PCS and MCS scores) represents a meaningful change in the quality of life.24 Using campodactyly parameter, the changes observed in PCS scores (adjusted mean difference of 4.5, 95% confidence interval [CI] 2-7) and MCS scores (adjusted mean differ-ence of 3.3, 95% CI 1-6) were not clinically significant. This complements a recent randomized study that found no differences in patients’ general physical or mental health
ARTICLE IN PRESS
Table 1. Characteristics of the noncancer and cancer patients in our study population
Patients Management Surgery Radiation
Geographic region (%)
Who completed the survey, (%)