Abstract
Background
Active surveillance/watchful waiting (AS/WW) is feasible and effective for favorable‐risk prostate cancer (PCa). Understanding socioeconomic determinants of AS/WW may help determine the target population for social support and improve cancer‐related survival.
Methods
The Surveillance, Epidemiology, and End Results Prostate with Watchful Waiting Database 18 Registries identified 229,428 adult men diagnosed with primary localized PCa (clinical T1‐T2c, N0M0) during a median follow‐up of 45 months between 2010 and 2016. Socioeconomic determinants included socioeconomic status (SES) tertiles, marital status (unmarried vs married), and residency (urban vs rural). Multivariable logistic regression and Cox models determined the adjusted odds ratios (aOR) for AS/WW utilization, and adjusted hazard ratio (aHR) for cancer‐specific survival (CSS) and overall survival (OS). The extent of missing data was evaluated by multiple imputation. Sensitivity analyses were performed in multiple imputation datasets.
Results
Unmarried patients were more likely to receive AS/WW in low‐risk group (aOR, 1.20 [95%CI, 1.12–1.28]; p < 0.001) and favorable intermediate‐risk group (aOR, 1.41 [95%CI, 1.26–1.59]; p < 0.001) than married patients. Urban patients had 0.77‐fold lower likelihood of AS/WW than rural patients in low‐risk group (95% CI, 0.68–0.87; p < 0.001), but not in favorable intermediate‐risk groups. Among patients undertaking AS/WW, a significantly worse OS was observed among unmarried patients comparing to married group (aHR, 1.98 [95% CI, 1.50–2.60]; p < 0.001), and patients with high SES had better CSS than low group (aHR, 0.08 [95%CI, 0.01–0.69]; p = 0.02). No significant survival difference was found between urban and rural patients.
Conclusions and Relevance
Unmarried or urban patients had significantly higher rates of AS/WW. The utilization and efficacy of conservative management were affected by socioeconomic factors, which might serve as a barrier of treatment decision‐making and targeted a population in need of social support.
Keywords: marital status, prostatic neoplasms, rural population, watchful waiting
In this population‐based, cross‐sectional study, the utilization rates of conservative management were significantly higher in low socioeconomic status (SES), unmarried and urban patients diagnosed with low‐risk prostate cancer. Unmarried status and low SES were associated with worse overall survival among patients undertaking active surveillance/watchful waiting. The utilization of conservative management was affected by socioeconomic factors, which might serve as a barrier of treatment decision‐making and targeted a population in need of social support.
Key points.
Question: Do the socioeconomic factors influence the utilization and outcomes of active surveillance or watchful waiting (AS/WW) among prostate cancer (PCa) patients?
Findings: In this population‐based, observational cohort study, the utilization rates of conservative management were significantly higher in low socioeconomic status (SES), unmarried and urban patients diagnosed with low‐risk PCa. Unmarried status and low SES were associated with worse overall survival among patients undertaking AS/WW.
Meaning: The utilization of conservative management was affected by socioeconomic factors, which might serve as a barrier of treatment decision‐making and targeted a population in need of social support.
1. INTRODUCTION
Management of localized prostate cancer (PCa) remains controversial for both clinicians and patients when weighing the benefits and the potential treatment harms. Conservative management including active surveillance or watchful waiting (AS/WW) has been shown to be feasible and effective in clinically insignificant PCa with less adverse effects and limited survival disadvantages, comparing to radiotherapy or surgery (RT/Surgery). 1 , 2 , 3 , 4 , 5 The 2021 National Comprehensive Cancer Network (NCCN) guidelines stated AS or observation (also known as WW) was recommended for very low‐, low‐, and favorable intermediate‐risk PCa based on different life expectancies. 6 In the past 15 years, the utilization of AS/WW has significantly increased. Most recently, AS has become the dominant management for low‐risk PCa, although the absolute rates varied across different health care settings. 7 , 8 , 9 , 10
The decisions of AS/WW are often made by both clinicians and patients as resectable tumors will be reserved comparing to radical treatment. In addition, the efficacy of AS/WW depends on rigorous follow‐up, including routine prostate‐specific antigen (PSA) tests, repeated magnetic resonance imaging and prostate biopsies, etc. It highly depends on the patients' compliance. 11 , 12 A recent study suggested that socioeconomic status (SES) played important roles in the decision of AS/WW. 13 SES was derived from median household income, median house value, median rent, percent below 150% of the poverty line, education index, percent working class, and percent unemployed. 14 Besides these factors, other socioeconomic factors might be also important. For instance, in radical treatment, one of the major side effects is the reduction of sexual function, 2 which is highly concerned based on personal marital status. Some studies suggested that the patients' compliance with AS/WW might be affected by marital status. 8 , 9 Another important factor would be the accessibility of medical resources. Both treatment and outcomes of PCa differed between urban and rural patients due to the different accessibility of urologists and geographic barriers. 15 , 16 , 17
In order to better understand the influence of socioeconomic factors on AS/WW utilization and outcomes, we conducted this study based on the Surveillance, Epidemiology, and End Results (SEER) Prostate with Watchful Waiting Database. Our objective is to investigate the socioeconomic determinants (including marital status, residency locations, SES, etc.) of AS/WW tendency and disease survivals.
2. MATERIALS AND METHODS
2.1. Study cohort
The SEER Program collected AS/WW information for PCa cases since 2010. 18 Using the SEER Prostate with Watchful Waiting Database, we identified 229,428 adult men diagnosed with NCCN primary localized prostate adenocarcinoma (age ≥ 18 years, one primary only, pathology‐confirmed, clinical T1‐T2, N0M0) between 2010 and 2016. The intermediate‐risk group was divided into two subgroups (favorable and unfavorable) according to the NCCN guidelines. We extracted clinical and socioeconomic variables, such as initial management (AS/WW vs RT/Surgery), age at diagnosis, year of diagnosis, PSA (prior to diagnostic biopsy of prostate and treatment), number of positive cores, race, insurance status, residency (two‐category Rural Urban Commuting Area [RUCA] codes), marital status and SES tertiles.
Two variables were transformed before analysis. The expected survival was generated from US Actuarial Life Tables (life expectancy minus age at diagnosis). 19 The marital status included five sub‐categories: married, single, divorced, widowed, and separated. All categories other than “married” are considered as “unmarried” in subgroup analyses.
We excluded patients with NCCN high‐risk features 6 (n = 38,795), missing clinical variables for risk stratification (n = 63,105), missing socioeconomic variables (n = 26,752), or patients whose initial recommended treatments were not identified or managed (n = 11,232, e.g., actual treatment were unknown, patients refused, or physicians decided not to treat for reasons such as the presence of comorbidities). A complete case dataset was generated based on the exclusion criteria.
The study was approved by the Institutional Review Board of Shanghai Ruijin Hospital, Shanghai, China.
2.2. Multiple imputations for missing data
Due to a large amount of missing data of SEER Prostate with Watchful Waiting Database, 20 we performed multiple imputation on initial treatment (AS/WW vs RT/Surgery) using the Amelia II (version 1.8.0). 21 Besides, both the clinical variables for risk stratification (T stage, log‐transformed PSA, biopsy Grade Group, and % of positive cores) and socioeconomic variables (race, insurance status, residency, SES, marital status) were imputed. In addition, the SEER registry was set as a cross‐sectional variable and year of diagnosis as a time‐series variable. In the present study, we only focused on the favorable‐risk PCa cases, so we classified T stage as a binary variable (T1‐T2a vs. T2b‐T2c) and set “T2NOS” as a missing value in the multiple imputation models. A multiple imputation dataset was generated under 1000 maximum resampling. All the association analysis were performed in the multiple imputation dataset repeatedly as a sensitivity analysis.
2.3. Statistical analysis
In descriptive statistics, continuous variables were described as median (interquartile range) and categorical variables were described as number (%). The AS/WW utilization rates between subgroups were compared using Fisher's exact test. Ordinal variables (year of diagnosis) were compared using Cochran‐Armitage tests for trend.
To identify the predictors of initial management (AS/WW vs RT/Surgery) in each risk group, we performed univariable and multivariable logistic regression and calculated crude and adjusted odds ratios (aOR), 95% confidence intervals (95% CI) and p values for each covariate. In SEER database, the survival months were defined as the intervals between date of diagnosis and date of last contact. The overall survival (OS) was calculated from all‐cause mortality, which means death due to any cause. The cancer‐specific survival (CSS) was calculated from deaths caused by PCa. We used Cox models to determine adjusted hazard ratios (aHR) for CSS and OS in each risk group. The multicollinearity of regression models was all checked by a variance inflation factor (VIF) and a VIF of 5 or more was considered as a multicollinearity problem.
All statistical analyses were performed with R software (version 4.1.0). 22 A two‐sided p < 0.05 was considered statistically significant.
3. RESULTS
3.1. Baseline characteristics
The flowchart of NCCN risk stratification and multiple imputation was shown as Figure 1. The characteristics of the entire cohort, complete case dataset, and multiple imputation dataset are described in Table 1. The percentage of initial missing data ranged between 2.9% and 39.3%. As a result, only half (55.6%, n = 127,528) of entire cohort could be accurately classified as low‐ to intermediate‐risk PCa. On the basis of our exclusion criteria, a total of 89,544 men were included in the complete case dataset and 188,180 cases were in multiple imputation dataset (Table 1).
FIGURE 1.
Flowchart of NCCN risk stratification and multiple imputation among men with primary localized prostate adenocarcinoma in SEER Prostate with Watchful Waiting Database. NCCN, National Comprehensive Cancer Network; PSA, prostate‐specific antigen.
TABLE 1.
Baseline characteristics of entire cohort, complete case dataset, and multiple imputation dataset in men with primary localized prostate adenocarcinoma.
| Characteristic | Entire cohort (n = 229,428) | Complete case dataset (n = 89,544) | Multiple imputation dataset (n = 188,180) | |
|---|---|---|---|---|
| N (%) | Value | Value | Value | |
| Age at diagnosis (years) | 229,428 (100) | 66 (60–72) | 64 (59–69) | 65 (59–70) |
| PSA (ng/mL) | 201,801 (88.0) | 6.2 (4.7–9.3) | 5.8 (4.6–8.0) | 5.9 (4.5–8.2) |
| Race | ||||
| Non‐Hispanic White | 222,724 (97.1) | 151,400 (68.0) | 62,355 (69.6) | 129,486 (68.8) |
| Non‐Hispanic Black | 38,170 (17.1) | 14,855 (16.6) | 31,392 (16.7) | |
| Hispanic | 21,766 (9.8) | 7761 (8.7) | 18,348 (9.8) | |
| Non‐Hispanic Others | 11,388 (5.1) | 4573 (5.1) | 8954 (4.8) | |
| Insurance status | ||||
| Insured | 202,545 (88.3) | 188,548 (93.1) | 84,536 (94.4) | 174,115 (92.5) |
| Medicaid | 10,917 (5.4) | 3909 (4.4) | 10,580 (5.6) | |
| Uninsured | 3080 (1.5) | 1099 (1.2) | 3485 (1.9) | |
| Residency | ||||
| Urban | 220,438 (96.1) | 204,323 (92.7) | 82,952 (92.6) | 174,449 (92.7) |
| Rural | 16,115 (7.3) | 6592 (7.4) | 13,731 (7.3) | |
| SES index | ||||
| Low tertile | 216,591 (94.4) | 59,001 (27.2) | 22,633 (25.3) | 49,729 (26.4) |
| Middle tertile | 71,410 (33.0) | 29,536 (33.0) | 61,562 (32.7) | |
| High tertile | 86,180 (39.8) | 37,375 (41.7) | 76,889 (40.9) | |
| Marital status | ||||
| Married | 192,134 (83.7) | 144,126 (75.0) | 69,088 (77.2) | 135,640 (72.1) |
| Single | 23,738 (12.4) | 10,100 (11.3) | 25,014 (13.3) | |
| Divorced | 14,849 (7.7) | 6827 (7.6) | 16,500 (8.8) | |
| Widowed | 7583 (4.0) | 2748 (3.1) | 8245 (4.4) | |
| Separated | 1838 (1.0) | 781 (0.9) | 2781 (1.5) | |
| T stage | ||||
| T1‐T2a | 130,342 (56.8) | 54,741 (61.1) | 124,315 (66.1) | |
| T2b‐T2c | 65,538 (28.6) | 34,803 (38.9) | 63,865 (33.9) | |
| T2NOS | 33,548 (14.6) | |||
| Biopsy Grade Group | ||||
| 1 | 222,254 (96.9) | 105,186 (47.3) | 45,538 (50.9) | 105,123 (55.9) |
| 2 | 60,651 (27.3) | 30,101 (33.6) | 58,790 (31.2) | |
| 3 | 26,034 (11.7) | 13,905 (15.5) | 24,267 (12.9) | |
| 4 | 18,766 (8.4) | |||
| 5 | 11,617 (5.2) | |||
| % of positive cores | 139,169 (60.7) | 0.29 (0.16–0.50) | 0.25 (0.15–0.48) | 0.26 (0.13–0.46) |
| NCCN risk group | ||||
| Low | 127,528 (55.6) | 50,082 (39.3) | 30,547 (34.1) | 64,667 (34.4) |
| Favorable intermediate | 24,454 (19.2) | 18,030 (20.1) | 51,066 (27.1) | |
| Unfavorable intermediate | 52,992 (41.6) | 40,967 (45.8) | 72,447 (38.5) | |
| Initial management | ||||
| RT/Surgery | 185,075 (80.7) | 159,692 (86.3) | 74,730 (83.5) | 154,015 (81.8) |
| AS/WW | 25,383 (13.7) | 14,814 (16.5) | 34,165 (18.2) | |
| Follow‐up duration, months | 229,428 (100) | 41 (18–63) | 42 (19–63) | 43 (20–64) |
| Overall mortality | 229,428 (100) | 11,857 (5.2) | 2415 (2.7) | 7030 (3.7) |
| Cancer‐specific mortality | 229,428 (100) | 2615 (1.1) | 247 (0.3) | 757 (0.4) |
Note: Continuous variables were described as median (interquartile range) and categorical variables were described as number (percentage).
Abbreviation: AS/WW, Active surveillance/watchful waiting; NCCN, National Comprehensive Cancer Network; PSA, prostate‐specific antigen; SES, socioeconomic status.
3.2. Socioeconomic factors and AS/WW trends
The AS/WW rates increased over time regardless of marital status and residency in NCCN low‐ and favorable intermediate‐risk groups from 2010 to 2016 (all p trend <0.001, Figure 2). In low‐risk group, AS/WW rates were significantly lower in married patients than in unmarried patients (38.3% vs 40.8%, p < 0.001), although significant difference were only found in 2011 and 2012 in subgroups analyses by year of diagnosis (Figure 2). Notably, urban patients had significantly higher AS/WW rates than rural patients within each year (all p < 0.001). In favorable intermediate‐risk group, similarly, AS/WW rates were significantly lower in married patients than in unmarried patients (10.2% vs 14.4%, p < 0.001), and the difference were still significant except for 2010 in each year. While no significant difference was observed between urban and rural patients (p = 0.38). Similar results were found in multiple imputation dataset (Figure S1).
FIGURE 2.
Yearly rates of active surveillance/watchful waiting management among men with NCCN favorable‐risk prostate cancer from 2010 to 2016 in complete case dataset. NCCN, National Comprehensive Cancer Network. Results were shown by marital status (married vs unmarried), residency (urban vs rural) and NCCN risk stratification (low‐ and favorable intermediate‐risk groups). Error bars represent 95% confidence intervals. The percentages between subgroups were compared using Fisher's exact test (an asterisk represented a significant difference in a specific year). The yearly trends were assessed using Cochran–Armitage test (p trend).
Among patients with low‐risk PCa (n = 30,547), multivariable logistic regression analysis showed that unmarried patients had a significant increased AS/WW rates compared to married patients (aOR, 1.20 [95%CI, 1.12–1.28]; p < 0.001). Similar results were found in sub‐categories except for “widowed” (aORsingle, 1.25 [95% CI, 1.14–1.36], p < 0.001; aORdivorced, 1.23 [95% CI, 1.10–1.36], p < 0.001; aORseparated, 1.47 [95% CI, 1.08–1.99], p = 0.01, Table 2). Rural patients had ~0.77‐fold lower likelihood of AS/WW management than urban patients (95% CI, 0.68–0.87, p < 0.001). In addition, non‐Hispanic Black and Medicaid beneficiaries were considered having less preference of AS/WW management than others (all p < 0.05). Patients in middle and high SES tertile would prefer AS/WW than low group (aORmiddle: 1.29, aORhigh: 1.66, all p < 0.001). Similar results were found in multiple imputation dataset (Table S1).
TABLE 2.
Effect of baseline and socioeconomic factors on initial management (AS/WW versus RT/Surgery) among men with NCCN favorable‐risk prostate cancer in complete case dataset.
| Characteristic | Low risk (n = 30,547) | Favorable‐intermediate risk (n = 18,030) | ||||||
|---|---|---|---|---|---|---|---|---|
| Crude OR (95% CI) | p | Adjusted OR (95% CI) | p | Crude OR (95% CI) | p | Adjusted OR (95% CI) | p | |
| Expected survival (years) | ||||||||
| ≥ 10 | 1.0 (Ref.) | <0.001 | 1.0 (Ref.) | 0.001 | 1.0 (Ref.) | <0.001 | 1.0 (Ref.) | <0.001 |
| < 10 | 0.74 (0.66–0.82) | 1.28 (1.11–1.47) | 2.89 (2.45–3.40) | 2.36 (1.97–2.84) | ||||
| Year of diagnosis | 1.34 (1.32–1.35) | <0.001 | 1.40 (1.38–1.42) | <0.001 | 1.30 (1.27–1.34) | <0.001 | 1.32 (1.28–1.35) | <0.001 |
| PSA (ng/ml) | 1.40 (1.33–1.46) | <0.001 | 1.09 (1.02–1.16) | 0.007 | 6.58 (5.89–7.36) | <0.001 | 5.28 (4.70–5.92) | <0.001 |
| No. positive cores | ||||||||
| < 3 | 1.0 (Ref.) | <0.001 | 1.0 (Ref.) | <0.001 | 1.0 (Ref.) | <0.001 | 1.0 (Ref.) | <0.001 |
| ≥ 3 | 0.43 (0.41–0.46) | 0.39 (0.36–0.41) | 0.42 (0.38–0.46) | 0.43 (0.38–0.47) | ||||
| Race | ||||||||
| Non‐Hispanic White | 1.0 (Ref.) | 1.0 (Ref.) | 1.0 (Ref.) | 1.0 (Ref.) | ||||
| Non‐Hispanic Black | 0.82 (0.77–0.88) | <0.001 | 0.91 (0.84–0.99) | 0.025 | 1.05 (0.93–1.20) | 0.416 | 1.02 (0.88–1.18) | 0.823 |
| Hispanic | 0.89 (0.82–0.96) | 0.004 | 0.92 (0.83–1.02) | 0.100 | 1.00 (0.84–1.19) | 0.996 | 0.91 (0.75–1.10) | 0.316 |
| Others | 1.25 (1.13–1.39) | <0.001 | 1.06 (0.94–1.20) | 0.353 | 1.40 (1.15–1.70) | 0.001 | 1.00 (0.81–1.24) | 0.995 |
| Insurance status | ||||||||
| Insured | 1.0 (Ref.) | 1.0 (Ref.) | 1.0 (Ref.) | 1.0 (Ref.) | ||||
| Medicaid | 0.74 (0.65–0.83) | <0.001 | 0.83 (0.71–0.96) | 0.011 | 1.57 (1.27–1.95) | <0.001 | 1.15 (0.90–1.46) | 0.264 |
| Uninsured | 1.36 (1.10–1.69) | 0.004 | 1.53 (1.17–1.99) | 0.002 | 1.33 (0.89–1.99) | 0.161 | 1.08 (0.70–1.66) | 0.739 |
| Residency | ||||||||
| Urban | 1.0 (Ref.) | 1.0 (Ref.) | 1.0 (Ref.) | 1.0 (Ref.) | ||||
| Rural | 0.63 (0.57–0.69) | <0.001 | 0.77 (0.68–0.87) | <0.001 | 0.92 (0.76–1.10) | 0.355 | 0.90 (0.73–1.11) | 0.308 |
| Marital status | ||||||||
| Married | 1.0 (Ref.) | 1.0 (Ref.) | 1.0 (Ref.) | 1.0 (Ref.) | ||||
| Single | 1.19 (1.11–1.28) | <0.001 | 1.25 (1.14–1.36) | <0.001 | 1.43 (1.24–1.64) | <0.001 | 1.37 (1.18–1.60) | <0.001 |
| Divorced | 1.14 (1.05–1.25) | 0.002 | 1.23 (1.10–1.36) | <0.001 | 1.47 (1.25–1.73) | <0.001 | 1.43 (1.19–1.71) | <0.001 |
| Widowed | 0.77 (0.67–0.88) | <0.001 | 0.90 (0.75–1.07) | 0.214 | 1.54 (1.19–1.99) | 0.001 | 1.22 (0.92–1.62) | 0.165 |
| Separated | 1.30 (1.01–1.68) | 0.042 | 1.47 (1.08–1.99) | 0.014 | 2.29 (1.55–3.40) | <0.001 | 2.84 (1.83–4.40) | <0.001 |
| SES status | ||||||||
| Low tertile | 1.0 (Ref.) | 1.0 (Ref.) | 1.0 (Ref.) | 1.0 (Ref.) | ||||
| Middle tertile | 1.33 (1.25–1.42) | <0.001 | 1.29 (1.19–1.39) | <0.001 | 1.09 (0.96–1.24) | 0.166 | 1.28 (1.11–1.47) | 0.001 |
| High tertile | 1.71 (1.62–1.82) | <0.001 | 1.66 (1.53–1.79) | <0.001 | 1.15 (1.02–1.30) | 0.020 | 1.42 (1.23–1.63) | <0.001 |
Abbreviation: AS/WW, active surveillance/watchful waiting; NCCN, National Comprehensive Cancer Network; OR, odds ratio; PSA, prostate‐specific antigen; Ref, reference; RT, radiation therapy; SES, socioeconomic status.
Among patients in favorable intermediate‐risk group (n = 18,030), unmarried patients tended to receive AS/WW compared to married patients (aOR, 1.41 [95%CI, 1.26–1.59]; p < 0.001). Similar results were also found in sub‐categories except for “widowed” (p < 0.001, Table 2). Higher SES was associated with higher rates of AS/WW (aORmiddle: 1.28, aORhigh: 1.42, all p < 0.001). However, no statistical difference was found among different races, insurance statuses and residency. All the VIFs of covariables were <2.0 in collinearity analyses, which indicated that no multicollinearity problems were found in our logistic regression models. Similar results were found in multiple imputation dataset (Table S1).
3.3. Socioeconomic factors and survival outcomes
The median follow‐up was 41 months (range: 18–63 months) in entire cohort. Among men with favorable‐risk PCa (n = 48,577), 13,875 (28.6%) men received AS/WW as initial management (Table 3). We then performed survival analysis in the AS/WW subset. Interestingly, men in high SES tertile had better outcomes than low tertile group (aHR, 0.08 [95% CI, 0.01–0.69]; p = 0.02). Neither the NCCN risk group nor the other socioeconomic factors (race, insurance status, and residency) were not significantly associated with CSS. A significantly worse OS was observed among unmarried patients comparing to married group (aHR, 1.98 [95%CI, 1.50–2.60]; p < 0.001). However, “single” and “divorced” patients had significantly worse OS than married patients (aHRsingle, 1.88 [95% CI, 1.29–2.73], p = 0.001; aHRdivorced, 2.47 [95% CI, 1.72–3.55], p < 0.001, Table 3). Repeated subgroup analyses showed similar results in multiple imputation dataset (Table S2).
TABLE 3.
Multivariable adjusted hazard ratios of socioeconomic factors on cancer‐specific survival and overall survival among favorable‐risk patients receiving AS/WW in complete case dataset.
| Characteristic | No. of total patients | Cancer‐specific survival | Overall survival | ||
|---|---|---|---|---|---|
| Adjusted HR (95% CI) | p | Adjusted HR (95% CI) | p | ||
| Age at diagnosis (years) | 13,875 | 1.05 (0.98–1.13) | 0.183 | 1.09 (1.07–1.11) | <0.001 |
| Year of diagnosis | 0.99 (0.60–1.65) | 0.980 | 1.01 (0.91–1.12) | 0.819 | |
| PSA (ng/ml) | 3.25 (0.62–17.09) | 0.163 | 1.15 (0.85–1.55) | 0.356 | |
| No. positive cores | |||||
| < 3 | 9290 | 1.0 (Ref.) | 0.211 | 1.0 (Ref.) | 0.591 |
| ≥ 3 | 3078 | 2.06 (0.66–6.41) | 1.08 (0.81–1.46) | ||
| Race | |||||
| Non‐Hispanic White | 9869 | 1.0 (Ref.) | 1.0 (Ref.) | ||
| Non‐Hispanic Black | 2009 | 2.19 (0.55–8.75) | 0.267 | 0.90 (0.62–1.30) | 0.567 |
| Hispanic | 1179 | 2.44 (0.47–12.72) | 0.290 | 0.80 (0.48–1.36) | 0.415 |
| Others | 818 | 3.07 (0.35–26.91) | 0.311 | 0.44 (0.18–1.08) | 0.075 |
| Insurance status | |||||
| Insured | 13,158 | 1.0 (Ref.) | 1.0 (Ref.) | ||
| Medicaid | 529 | — | — | 0.64 (0.28–1.47) | 0.293 |
| Uninsured | 188 | — | — | 0.73 (0.18–2.95) | 0.657 |
| Residency | |||||
| Rural | 13,101 | 1.0 (Ref.) | 0.309 | 1.0 (Ref.) | 0.671 |
| Urban | 774 | 2.30 (0.46–11.41) | 1.11 (0.69–1.79) | ||
| Marital status | |||||
| Married | 10,459 | 1.0 (Ref.) | 1.0 (Ref.) | ||
| Single | 1752 | 2.09 (0.53–8.29) | 0.293 | 1.88 (1.29–2.73) | 0.001 |
| Divorced | 1150 | 1.47 (0.30–7.28) | 0.636 | 2.47 (1.72–3.55) | <0.001 |
| Widowed | 373 | — | — | 1.38 (0.76–2.52) | 0.296 |
| Separated | 141 | — | — | 1.47 (0.46–4.65) | 0.516 |
| SES status | |||||
| Low tertile | 2885 | 1.0 (Ref.) | 1.0 (Ref.) | ||
| Middle tertile | 4398 | 0.53 (0.16–1.78) | 0.304 | 0.62 (0.45–0.85) | 0.003 |
| High tertile | 6592 | 0.08 (0.01–0.69) | 0.022 | 0.33 (0.23–0.47) | <0.001 |
| Risk group | |||||
| Low | 11,873 | 1.0 (Ref.) | 1.0 (Ref.) | ||
| Favorable‐intermediate | 2002 | 1.22 (0.28–5.31) | 0.793 | 1.57 (1.13–2.17) | 0.006 |
Abbreviations: AS/WW, active surveillance/watchful waiting; HR, hazard ratio; PSA, prostate‐specific antigen; NCCN, National Comprehensive Cancer Network; Ref, reference; RT, radiation therapy; SES, socioeconomic status.
4. DISCUSSION
The current study presented a comprehensive analysis of the association between socioeconomic factors and AS/WW utilization, as well as the disease outcomes. To the best of our knowledge, this the first study presents the association between marital status, residency and AS/WW based on the US population. We found that (1) the AS/WW rates significantly increased over time from 2010 to 2016. Such trends remained significant when stratified by clinical risk, marital status or residency; (2) Unmarried patients (single, divorced and separated, except for widowed) had significantly higher AS/WW rates than married patients in both low‐ and favorable intermediate‐risk groups; (3) Rural patients had significantly lower likelihood (~0.77‐fold) of AS/WW use than urban patients only in the low‐risk group; (4) Unmarried patients had significantly worse OS than married patients after receiving AS/WW.
Several retrospective studies demonstrated that the AS/WW utilization rates for low‐ to intermediate‐risk PCa have increased significantly since 2010 in different US population‐based cohorts. 7 , 9 , 10 , 23 We found that the yearly trends were still significant after being stratified by marital status and residency, which indicated the preference for AS/WW was rising nationwide.
We found unmarried status and its sub‐categories (except for “widowed”) were significantly associated with higher rates of AS/WW within each clinical risk group. The difference might be due to that widowed status was not an active, voluntary choice, comparing to other unmarried components (single, divorced, or separated). Remarkably, in the low‐risk group, the unmarried patients had ~1.20‐fold higher likelihood of AS/WW management than married patients (95% CI: 1.12–1.28), which was similar as the results based on US Veterans database from 2005 to 2015 (aOR: 1.18, 95% CI: 1.15–1.21). 9 Similar conclusions were also reported in Sweden population. 8 A potential explanation is that the unmarried patients are likely to pay more attention to the quality of life and receive a newer form of treatment.
Mahal et al and Butler et al reported that several factors (including age, year of diagnosis, and Yost index) were predictive of AS/WW utilization in the SEER database, which were consistent with our results based on an extra follow‐up year (2010–2016). 23 , 24 Another study based on SEER database demonstrated that African American men would be more likely to undergo AS/WW than other races in low‐risk PCa patients. 25 Similar results were reported in the intermediate‐risk group by Butler et al. 23 However, in the present study, when analyses were limited in non‐Hispanic population, African American men were less likely to undergo AS/WW than Caucasian men in the low‐risk group in our adjusted model. In addition, no significant association between AS/WW rates and ethnicities (non‐Hispanic white or black patients) were observed in the favorable intermediate‐risk group. The definition of Hispanic/Latino population varies by regions in United States. Therefore, the effect of race on survival can be more precise when limited in non‐Hispanic population, due to the complexity of the Hispanic population.
A recent study by Maganty et al indicated that rural PCa patients were less likely to undergo any treatment in entire cohort and different risk groups. 15 In our study, we found that rural and low‐SES patients were less likely to undergo AS/WW, and a significant difference was found in the low‐risk group. As AS/WW requires regular follow‐up, multiple PSA tests and annual prostate biopsy, the relatively low accessibility of medical resources (such as low urologist density, long travel distance to the clinic) would be a major concern for rural or low‐SES patients. In addition, rural residency is usually associated with relatively low household income. It would be another factor that could have driven the patients from choosing AS/WW, as studies showed that despite the cost‐effective of AS/WW at the population level, AS/WW would increase costs on disease management at individual level. 26 , 27
In the perspective of clinical practice, AS/WW, as a recommended management by NCCN for low‐ and favorable intermediate‐risk PCa, would not lead to worse CSS. Such results were confirmed in our current study and other reported studies. 23 , 24 Both married status and high SES tertile served as protective factors for men receiving AS/WW, even after adjusting for NCCN risk group.
Missing data are an unavoidable issue in all medical studies, especially for the analyses based on the population‐based databases. 28 , 29 Jeong et al reported the new SEER Prostate with Watchful Waiting Database contained a large amount of missing data, including the necessary variables for NCCN risk classification. 20 As a result, simply excluding cases with missing data will bring potential bias. In the present study, the multiple imputation dataset (N = 188,180) is approximately 2.1 times larger than the complete case dataset (N = 89,544, Figure 1), which is similar to the previous study (118,821 vs. 257,060, ~2.2 times). 20 Furthermore, all the regression analyses were repeated in the multiple imputation dataset, which strengthened the original results and conclusions and reduced the bias. Recently, the multiple imputation has been widely applied in clinical studies. 30 , 31 As for an imputation approach, it is important to assess whether data were missing at random (MAR). 28 In other words, the multiple imputation process is implemented based on the MAR assumption. 32 Missing completely at random (MCAR) is the optimal but rare pattern in clinical studies. More importantly, we can consider MAR as an approximation of MCAR when including as many covariates as possible in the models of the current study. 33
Several limitations of this study should be noted. First, we did not perform analyses in very low‐risk PCa patients due to the lack of data (cancer percentage in each positive core, PSA density, or volume 6 , 18 ). However, the benefits of AS/WW for patients with very low‐risk PCa may overcome the harms compared to radical treatments. It is less important to perform a similar evaluation in these patients. Further analyses are optional if additional data can be acquired. Second, AS involves active monitoring and potential curative treatment after progression, and WW means monitoring and palliative therapy if needed. The dataset did not distinguish AS from WW, which may reduce the generalizability of our results to either AS or WW. It is worth further investigation. Third, the median follow‐up was less than 4 years, and longer observation will be needed. The time‐dependent variables (such as marital status, income, and residency) were collected at a single time of diagnosis, so the timing of assessment cannot be evaluated in the present study. Fourth, large numbers of cases lacked necessary characteristics for the NCCN risk stratification. 20 However, we performed sensitivity analysis in multiple imputation dataset and showed similar results as complete case dataset. Our conclusions from this retrospective cohort still need high‐level evidence or models for further verification.
4.1. Conclusion
In conclusion, marital status, residency and SES (including factors of income, education and employment) independently influenced the utilization of AS/WW and served as a barrier to treatment decision‐making among favorable‐risk PCa patients. Married and high SES patients undertaking AS/WW would have better OS. Personalized social supports should be provided to these patients to improve the disease outcome.
AUTHOR CONTRIBUTIONS
Da Huang: Data curation (equal); formal analysis (equal); funding acquisition (equal); methodology (equal); visualization (equal); writing – original draft (equal). Xiaohao Ruan: Data curation (equal); formal analysis (equal); visualization (equal); writing – original draft (equal). Jingyi Huang: Data curation (equal); formal analysis (equal); investigation (equal); writing – original draft (equal). Ning Zhang: Resources (equal); software (equal). Guangliang Jiang: Resources (equal); software (equal). Yi Gao: Resources (equal); software (equal). Danfeng Xu: Conceptualization (equal); supervision (equal); writing – review and editing (equal). Rong Na: Conceptualization (equal); funding acquisition (equal); supervision (equal); writing – review and editing (equal).
FUNDING INFORMATION
This work was in supported by grants from National Natural Science Foundation of China (grant nos 81772741, 1972405, 81972645, and 82173045), Shanghai Youth Talent Support Program, and the Shanghai Rising‐Star Incubator Program (22YF1440500). All the funders had no role in study design, data collection, data analysis, interpretation, and writing of the report.
CONFLICT OF INTEREST STATEMENT
The authors declare that they have no competing interests.
Supporting information
Appendix S1.
ACKNOWLEDGMENTS
We thanked SEER program to approve our protocol and provide the custom datasets.
Huang D, Ruan X, Huang J, et al. Socioeconomic determinants are associated with the utilization and outcomes of active surveillance or watchful waiting in favorable‐risk prostate cancer. Cancer Med. 2023;12:9868‐9878. doi: 10.1002/cam4.5650
Da Huang, Xiaohao Ruan and Jingyi Huang contributed equally.
Contributor Information
Da Huang, Email: huangdasjtu@gmail.com.
Danfeng Xu, Email: xdf12036@rjh.com.cn.
Rong Na, Email: narong.hs@gmail.com.
DATA AVAILABILITY STATEMENT
We thanked SEER program to approve our protocol and provide the custom datasets. All data used in this research are publicly available to qualified researchers on application to the SEER program (http://seer.cancer.gov/).
REFERENCES
- 1. Neal DE, Metcalfe C, Donovan JL, et al. Ten‐year mortality, disease progression, and treatment‐related side effects in men with localised prostate cancer from the ProtecT randomised controlled trial according to treatment received. Eur Urol. 2020;77(3):320‐330. [DOI] [PubMed] [Google Scholar]
- 2. Hoffman KE, Penson DF, Zhao Z, et al. Patient‐reported outcomes through 5 years for active surveillance, surgery, brachytherapy, or external beam radiation with or without androgen deprivation therapy for localized prostate cancer. JAMA. 2020;323(2):149‐163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Tosoian JJ, Mamawala M, Epstein JI, et al. Active Surveillance of grade group 1 prostate cancer: long‐term outcomes from a large prospective cohort. Eur Urol. 2020;77(6):675‐682. [DOI] [PubMed] [Google Scholar]
- 4. Klotz L, Vesprini D, Sethukavalan P, et al. Long‐term follow‐up of a large active surveillance cohort of patients with prostate cancer. J Clin Oncol. 2015;33(3):272‐277. [DOI] [PubMed] [Google Scholar]
- 5. Bill‐Axelson A, Holmberg L, Garmo H, et al. Radical prostatectomy or watchful waiting in prostate cancer—29‐year follow‐up. N Engl J Med. 2018;379(24):2319‐2329. [DOI] [PubMed] [Google Scholar]
- 6. Network National Comprehensive Cancer. NCCN Clinical Practice Guidelines in Oncology: Prostate Cancer (Version 2.2022) Web site. http://www.nccn.org/professionals/physician_gls/pdf/prostate.pdf. Accessed Jan 10, 2022.
- 7. Cooperberg MR, Carroll PR. Trends in management for patients with localized prostate cancer, 1990‐2013. JAMA. 2015;314(1):80‐82. [DOI] [PubMed] [Google Scholar]
- 8. Loeb S, Folkvaljon Y, Curnyn C, Robinson D, Bratt O, Stattin P. Uptake of active surveillance for very‐low‐risk prostate cancer in Sweden. JAMA Oncol. 2017;3(10):1393‐1398. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Loeb S, Byrne N, Makarov DV, Lepor H, Walter D. Use of conservative management for low‐risk prostate cancer in the Veterans Affairs Integrated Health Care System from 2005‐2015. JAMA. 2018;319(21):2231‐2233. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Mahal BA, Butler S, Franco I, et al. Use of active Surveillance or watchful waiting for low‐risk prostate cancer and management trends across risk groups in the United States, 2010‐2015. JAMA. 2019;321(7):704‐706. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Kalapara AA, Verbeek JFM, Nieboer D, et al. Adherence to active surveillance protocols for low‐risk prostate cancer: results of the Movember Foundation's global action plan prostate cancer active surveillance initiative. Eur Urol Oncol. 2020;3(1):80‐91. [DOI] [PubMed] [Google Scholar]
- 12. Detsky JS, Ghiam AF, Mamedov A, et al. Impact of biopsy compliance on outcomes for patients on active surveillance for prostate cancer. J Urol. 2020;204(5):934‐940. [DOI] [PubMed] [Google Scholar]
- 13. Butler SS, Loeb S, Cole AP, et al. United States trends in active surveillance or watchful waiting across patient socioeconomic status from 2010 to 2015. Prostate Cancer Prostatic Dis. 2020;23(1):179‐183. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Yost K, Perkins C, Cohen R, Morris C, Wright W. Socioeconomic status and breast cancer incidence in California for different race/ethnic groups. Cancer Causes Control. 2001;12(8):703‐711. [DOI] [PubMed] [Google Scholar]
- 15. Maganty A, Sabik LM, Sun Z, et al. Under treatment of prostate cancer in rural locations. J Urol. 2020;203(1):108‐114. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Colli J, Sartor O, Thomas R, Lee BR. Does urological cancer mortality increase with low population density of physicians? J Urol. 2011;186(6):2342‐2346. [DOI] [PubMed] [Google Scholar]
- 17. Odisho AY, Cooperberg MR, Fradet V, Ahmad AE, Carroll PR. Urologist density and county‐level urologic cancer mortality. J Clin Oncol. 2010;28(15):2499‐2504. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Surveillance, Epidemiology, and End Results (SEER) Program . Prostate with Watchful Waiting Database (2010. ‐2016). Web site. http://seer.cancer.gov/seerstat/databases/prostate‐ww/index.html. Accessed May 15, 2020.
- 19. The United States Social Security Administration. Actuarial Life Tables: Period Life Table , 2010. ‐2016 Web site. https://www.ssa.gov/OACT/STATS/table4c6.html. Accessed May 15, 2020.
- 20. Jeong CW, Washington SL, Herlemann A, Gomez SL, Carroll PR, Cooperberg MR. The new Surveillance, epidemiology, and end results prostate with watchful waiting database: opportunities and limitations. Eur Urol. 2020;78(3):335‐344. [DOI] [PubMed] [Google Scholar]
- 21. Honaker J, King G, Blackwell M. Amelia II: a program for missing data. J Stat Softw. 2011;45(1):1‐47. [Google Scholar]
- 22.R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R‐project.org/. Published 2020. Accessed.
- 23. Butler SS, Mahal BA, Lamba N, et al. Use and early mortality outcomes of active surveillance in patients with intermediate‐risk prostate cancer. Cancer. 2019;125(18):3164‐3171. [DOI] [PubMed] [Google Scholar]
- 24. Mahal AR, Butler S, Franco I, et al. Conservative management of low‐risk prostate cancer among young versus older men in the United States: trends and outcomes from a novel national database. Cancer. 2019;125(19):3338‐3346. [DOI] [PubMed] [Google Scholar]
- 25. Awamlh BAH, Ma X, Christos P, Hu JC, Shoag JE. Active Surveillance for black men with low‐risk prostate cancer in the United States. N Engl J Med. 2019;381(26):2581‐2582. [DOI] [PubMed] [Google Scholar]
- 26. Sathianathen NJ, Konety BR, Alarid‐Escudero F, Lawrentschuk N, Bolton DM, Kuntz KM. Cost‐effectiveness analysis of active Surveillance strategies for men with low‐risk prostate cancer. Eur Urol. 2019;75(6):910‐917. [DOI] [PubMed] [Google Scholar]
- 27. Sharma V, Wymer KM, Borah BJ, et al. Cost‐effectiveness of active Surveillance, radical prostatectomy and external beam radiotherapy for localized prostate cancer: an analysis of the ProtecT trial. J Urol. 2019;202(5):964‐972. [DOI] [PubMed] [Google Scholar]
- 28. Pedersen AB, Mikkelsen EM, Cronin‐Fenton D, et al. Missing data and multiple imputation in clinical epidemiological research. Clin Epidemiol. 2017;9:157‐166. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Luo Y. Evaluating the state of the art in missing data imputation for clinical data. Brief Bioinform. 2022;23(1):bbab489. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Lonergan PE, Jeong CW, Washington SL, Herlemann A, Carroll PR, Cooperberg MR. Active surveillance in intermediate‐risk prostate cancer with PSA 10‐20 ng/mL: pathological outcome analysis of a population‐level database. Prostate Cancer Prostatic Dis. 2021;19:e1911. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. McGurnaghan SJ, Weir A, Bishop J, et al. Risks of and risk factors for COVID‐19 disease in people with diabetes: a cohort study of the total population of Scotland. Lancet Diabetes Endocrinol. 2021;9(2):82‐93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Schafer JL, Graham JW. Missing data: our view of the state of the art. Psychol Methods. 2002;7(2):147‐177. [PubMed] [Google Scholar]
- 33. Collins LM, Schafer JL, Kam CM. A comparison of inclusive and restrictive strategies in modern missing data procedures. Psychol Methods. 2001;6(4):330‐351. [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Appendix S1.
Data Availability Statement
We thanked SEER program to approve our protocol and provide the custom datasets. All data used in this research are publicly available to qualified researchers on application to the SEER program (http://seer.cancer.gov/).