SEARCH

SEARCH BY CITATION

Keywords:

  • prostate cancer;
  • treatment;
  • demographic factors;
  • clinical factors;
  • decision-making

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. FUNDING SOURCES
  9. CONFLICT OF INTEREST
  10. REFERENCES

Study Type – Prognosis (cohort series)

Level of Evidence 2b

What's known on the subject? and What does the study add?

Men diagnosed with prostate cancer are increasingly involved in making their own treatment decisions and the current recommendations for treatment are based on informed choice. The absence of scientific evidence regarding optimum treatment choices underlines the importance of understanding which factors influence the selection of treatment by men diagnosed with prostate cancer. Previous studies have found that men diagnosed with prostate cancer were more likely to choose radiation therapy over radical prostatectomy if they were older and had a higher PSA level.

This is the first large-scale prospective study conducted outside the USA to quantify the factors associated with treatment decisions for men diagnosed with prostate cancer. It found that men who chose surgery were younger, above average physical health, and had lower grade cancers on the Gleason scale; men who had radiation therapy were older and had reduced physical health, with ADT added when men had more advanced disease. About two-thirds of the men said they primarily made the decision about treatment themselves, with the remaining men either sharing the decision-making process with their doctor or else leaving the decision more or less completely up to their doctor. These results highlight the importance of having quality up-to-date information readily available to guide these decisions.

OBJECTIVE

  • • 
    To examine demographic, clinical and quality-of-life indicators for the treatments received by men diagnosed with prostate cancer in Australia.

SUBJECTS AND METHODS

  • • 
    This prospective trial included men diagnosed with prostate cancer (n= 1064, response rate = 82%) between 2005 and 2007 in Queensland, Australia, sampled from urologists and hospital outpatient clinics.
  • • 
    Data were collected through telephone interviews and self-administered questionnaires.
  • • 
    Treatment received was categorized into five groups: radical prostatectomy; radiation therapy with neoadjuvant androgen deprivation therapy (ADT); radiation therapy alone; ADT alone; and monitoring.

RESULTS

  • • 
    Sharp contrasts in the choice between radical prostatectomy (47% of men) vs radiation therapy with ADT (30%) were evident among age at diagnosis, travel time to facilities offering radiation treatment, Gleason score, stage, body mass index and physical health.
  • • 
    Men who underwent surgery were younger and of above average physical health, and had lower grade cancers; men who underwent radiation therapy were older and less fit. ADT, in both neoadjuvant and definitive forms, was administered for high-risk and more advanced disease.
  • • 
    Two-thirds (66%) of men stated that they made the final treatment selection themselves.

CONCLUSIONS

  • • 
    These results suggest that men's baseline health and tumour characteristics influence treatment choices.
  • • 
    Distance from tertiary treatment centres also influenced the treatment received and access to specialist urologists may play a role.
  • • 
    With most men indicating high levels of decisional control, the importance of having quality up-to-date information readily available to guide their decisions cannot be overstated.

Abbreviations
ADT

androgen deprivation therapy

QCR

Queensland Cancer Registry

SF-36

36-item short-form health survey

APP

adjusted predicated probability

SES

socio-economic status

INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. FUNDING SOURCES
  9. CONFLICT OF INTEREST
  10. REFERENCES

Prostate cancer is the most common invasive cancer diagnosed among males in more developed countries, comprising nearly a quarter (22%) of incident cancers, excluding non-melanoma skin cancer [1]. Unlike other types of cancer, there is a lack of definitive evidence regarding optimum treatment once the diagnosis is confirmed [2–5]. Definitive treatment options mainly involve surgery (radical prostatectomy) or radiation therapy (either brachytherapy or external beam radiation) with or without androgen deprivation therapy (ADT). Other options include: active surveillance for low-risk disease, based on an intention to intervene curatively if a change in disease character is detected with further biopsies; watchful waiting with a view to using ADT subsequently when tumour progression is determined; and ADT alone, immediately after diagnosis [6,7].

As a result of differing side effects, risks and benefits associated with each treatment option [2], it is not surprising that men diagnosed with prostate cancer are increasingly involved in making their own treatment decisions [8,9]. Current recommendations are therefore based on informed choice [10], which typically involves input from a variety of sources, including physicians, partners, family, friends, support groups, the internet and other media [8,9]. In addition, factors such as clinical characteristics, life expectancy as a result of comorbidities, quality of life and personal preferences also affect the decision-making process [8,9,11].

The absence of scientific evidence regarding optimum treatment choices means it is important to understand which factors influence the selection of treatment by men diagnosed with prostate cancer. The aim of the present Australian study was to examine the relationship between treatment received and a range of demographic, clinical and quality-of-life indicators, using a large sample of men diagnosed with prostate cancer.

SUBJECTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. FUNDING SOURCES
  9. CONFLICT OF INTEREST
  10. REFERENCES

The Queensland University of Technology Human Research Ethics Committee and ethics committees of ten public hospitals in Queensland provided approval to conduct this study, while Queensland Health provided approval to access additional data from the Queensland Cancer Registry (QCR).

We have previously described the methodology used to collect the data for this study [12]. Eligible men, who comprised those newly diagnosed with prostate cancer in Queensland between February 2005 and July 2007 (inclusive), were referred to the study by participating urologists in outpatient clinics in selected public hospitals and urology private practices in Brisbane (and surrounding area), Townsville and Mackay. Men who had a previous diagnosis of prostate cancer, were unable to read, write or speak English or had no regular access to a telephone were excluded. In total, 1064 men were eligible and agreed to participate in the study (82% of all 1291 referrals made). To assess the representativeness of the study cohort, data on all men diagnosed with prostate cancer in Queensland during the study period were extracted from the QCR.

DATA COLLECTION

Data was collected through telephone interviews and self-administered questionnaires up to 6-month after treatment, including demographic details, medical history, self-reported diagnostic and treatment details, health-related quality of life, and decisional involvement.

Information about prostate cancer grade was obtained by extracting clinical information from pathology forms held within the QCR. Stage was recorded by the interviewing officer based on the man's description of his disease, and categorized into ‘local’ (organ-confined), ‘locally advanced’ (rising PSA level but unknown metastasis) and ‘advanced’ (known distant metastasis).

Health-related quality of life was assessed using the 36-item short-form health survey (SF-36) [13] that contains a mental health and physical health summary scale suitable for measuring patients' wellbeing.

Treatment decision involvement was assessed using a single item that ascertained who was responsible for the management decision: the patient alone; mainly the patient; both the patient and the doctor equally; mainly the doctor; or the doctor alone [14].

Treatment was categorized into five groups; radical prostatectomy, radiation therapy with neoadjuvant ADT, radiation therapy alone, ADT alone and monitoring. Monitoring included both active surveillance and watchful waiting as it was not possible with de-identified data to differentiate between these two clinical approaches. Radiation therapy included both external beam radiation therapy and brachytherapy.

Road travel time from each patient's usual residential address at diagnosis to the closest radiation facility was calculated using Geographical Information Systems software. Area-level disadvantage at diagnosis was measured using the Index of Relative Socioeconomic Disadvantage [15].

STATISTICAL METHODS

Bivariate relationships between treatment choice and each of the potential explanatory variables were initially explored using chi-squared tests. Multinomial regression models were then used to model the five-category treatment choice outcome measure. Estimates from these models are presented in terms of adjusted predicated probabilities (APPs). These represent the mean predicted probability of a man choosing a specific treatment when they have a certain value of a selected explanatory variable. The APPs were generated using the margins command in Stata v11.2 for Windows (StataCorp LP, USA).

A stepwise model building process was used, in which all variables were initially entered into the model, and then those showing no significant association with the treatment choice (which, for the model building purposes, was set to P < 0.20 [16]) were sequentially removed from the model. Once the provisional final model was obtained, each deleted variable was returned individually to the model to determine whether their inclusion improved the model fit using the likelihood ratio test. A similar method of analysis considered the self-reported involvement of the patient in the decision-making process.

Some cases were excluded since they were missing values for the explanatory variables, and the number of these missing cases was insufficient to form its own ‘unknown’ category in the analysis.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. FUNDING SOURCES
  9. CONFLICT OF INTEREST
  10. REFERENCES

COMPOSITION OF THE COHORT

Of the original 1064 eligible men in the study cohort, 108 (10%) were excluded owing to missing information, leaving 956 men for the final analysis. Information regarding the composition of the cohort can be found in Table 1. Men in the study cohort comprised 11.5% of all diagnoses of prostate cancer in Queensland during the study period. Compared with all men diagnosed with prostate cancer, members of the study cohort were slightly younger (median age at diagnosis 63 vs 67 years), more likely to have intermediate Gleason score (‘3+4’ or ‘4+3’– 59 vs 49%), more likely to be married (76 vs 72%) and more likely to live <1 h away from a hospital facility offering radiation treatment (70 vs 67%).

Table 1. Distribution of prostate cancer treatment choices (crude %) by demographic, clinical and quality-of-life indicators
Total (N= 956)RPRT and ADTRT onlyADT onlyMonitoring
47.329.612.44.76.0
  • **

    Data suppressed due to a cell count <5. DF, degrees of freedom; RP, radical prostatectomy; RT, radiation therapy.

Age group Chi-square = 261.25; DF = 8; P < 0.001
 40–59 years (n= 326)73.610.78.9 ** 5.8
 60–69 years (n= 443)44.934.514.23.62.7
 ≥70 years (n= 187)7.050.814.413.913.9
Education level Chi-square = 65.38; DF = 12; P < 0.001
 University (n= 184)63.619.67.64.34.9
 Senior/Trade/Technical (n= 425)50.125.914.14.55.4
 Junior high school (n= 220)42.734.512.33.27.3
 Primary school or lower (n= 127)22.048.014.28.77.1
Private health insurance Chi-square = 21.50; DF = 4; P < 0.001
 Yes (n= 715)51.227.012.63.95.3
 No (n= 241)35.737.312.07.17.9
Work status Chi-square = 129.84; DF = 8; P < 0.001
 Employed full-time/part-time/casual (n= 418)67.217.78.11.75.3
 Retired (n= 449)29.640.316.37.36.5
 Other/Not stated (n= 89)42.731.513.55.66.7
Income Chi-square = 113.43; DF = 20; P < 0.001
 <$20 000 (n= 115)17.451.316.57.87.0
 $20–39 999 (n= 286)36.433.915.07.37.3
 $40–59 999 (n= 134)54.525.412.7 ** 4.5
 $60–79 999 (n= 107)64.519.69.3 ** 4.7
 $80 000+ (n= 208)67.317.87.2 ** 5.8
 Not stated (n= 106)43.433.014.24.74.7
Area-level SES Chi-square = 56.43; DF = 16; P < 0.001
 Most disadvantaged (n= 122)45.137.79.8 ** 4.1
 Quintile 2 (n= 216)41.231.915.75.16.0
 Quintile 3 (n= 253)36.836.016.64.75.9
 Quintile 4 (n= 189)52.425.99.33.78.5
 Least disadvantaged (n= 176)65.915.97.46.34.5
Travel time (TRACT) Chi-square = 57.46; DF = 12; P < 0.001
 <1 h (n= 648)53.224.211.64.86.2
 1 to <2 h (n= 72)41.726.413.99.78.3
 2 to <4 h (n= 94)42.641.510.6 ** **
 ≥4 h (n= 142)26.147.916.94.24.9
Reason for initial appointment Chi-square = 52.46; DF = 16; P < 0.001
 Symptoms checked (n= 168)38.735.15.49.511.3
 Screening test (n= 187)60.423.58.63.24.3
 Doctor initiated (n= 211)44.531.815.22.85.7
 General check-up (n= 348)46.829.016.14.04.0
 Other/Not stated (n= 42)40.528.614.3 ** **
Clinical Gleason score >Chi-square = 278.80; DF = 20; P < 0.001
 ≤3+3 (n= 200)34.017.031.03.015.0
 3+4 (n= 370)59.721.611.91.65.1
 4+3 (n= 196)59.721.64.62.6 **
 4+4 (n= 59)39.052.5 ** ** **
 ≥4+5 (n= 118)28.847.5 ** 20.3 **
 Unknown (n= 13) ** 53.9 ** ** **
Stage >Chi-square = 242.80; DF = 4; P < 0.001
 Localized (n= 882)51.027.113.52.06.3
 Locally advanced/Advanced (n= 74) ** 59.5 ** 36.5 **
Paternal history of prostate cancer >Chi-square = 20.71; DF = 4; P < 0.001
 Yes (n= 136)64.722.17.4 ** 4.4
 No (n= 820)44.430.913.35.26.2
Smoking Status >Chi-square = 29.13; DF = 8; P < 0.001
 Current smoker (n= 86)34.936.016.39.3 **
 Previous smoker (n= 492)42.532.513.25.56.3
 Never smoker (n= 378)56.324.310.62.66.1
BMI >Chi-square = 18.72; DF = 8; P= 0.016
 Under/Normal weight (n= 246)49.628.011.84.16.5
 Overweight (n= 377)52.823.912.75.05.6
 Obese (n= 333)39.337.212.64.86.0
Number of comorbidities >Chi-square = 59.76; DF = 12; P < 0.001
 None (n= 203)63.117.29.93.06.9
 1 (n= 314)51.928.39.94.15.7
 2–3 (n= 353)39.935.713.65.75.1
 ≥4 (n= 86)23.338.423.37.08.1
General health >Chi-square = 52.62; DF = 12; P < 0.001
 Excellent (n= 110)62.723.66.4 ** 4.5
 Very good (n= 376)54.523.710.14.86.9
 Good (n= 349)41.334.116.03.74.9
 Fair/Poor (n= 121)28.140.514.99.17.4
Physical health (SF-36) >Chi-square = 73.31; DF = 8; P < 0.001
 Below average (n= 162)27.242.611.78.010.5
 Average (n= 302)39.135.813.66.35.3
 Above average (n= 492)58.921.512.02.64.9
Mental health (SF-36) >Chi-square = 11.13; DF = 8; P= 0.195
 Below average (n= 305)54.125.29.85.25.6
 Average (n= 417)44.830.513.94.36.5
 Above average (n= 234)42.733.813.24.75.6

FACTORS ASSOCIATED WITH TREATMENT CHOICE

All of the demographic, clinical and quality-of-life indicators considered in the present study, with the exception of mental health, had a significant bivariate relationship with the treatment received (Table 1). Of these variables, age group, work status, area-level socio-economic status (SES), travel time to a radiation facility, reason for initial appointment, Gleason score, stage at diagnosis, paternal history of prostate cancer, smoking status, body mass index and physical health all remained significantly and independently associated with treatment choice in the multinomial logistic regression model (Table 2).

Table 2. Adjusted predicted probabilities for prostate cancer treatment choices by demographic, clinical and quality-of-life indicators
 RPRT & ADTRT onlyADT onlyMonitoring
APP (95% CI) %APP (95% CI) %APP (95% CI) %APP (95% CI) %APP (95% CI) %
  • *

    APP was significantly (P < 0.05) different to the reference category. Income, Private health insurance, Mental health, Number of comorbidities and General health were not included in the final model. DF, degrees of freedom; RP, radical prostatectomy; RT, radiation therapy; REF, reference category.

Age group >Chi-square = 115.48; DF = 8; P < 0.001
 40–59 years (n= 326)REF66.5 (61–72)16.2 (12–20)10.5 (7–14)1.8 (0–4)5.0 (3–7)
 60–69 years (n= 443)46.2 (42–50)*33.9 (30–38)*13.9 (11–17)3.1 (2–5)2.9 (1–4)
 70 years and over (n= 187)10.9 (5–17)*44.6 (36–53)*13.0 (8–18)10.4 (6–15)*21.1 (12–31)*
Education level >Chi-square = 17.09; DF = 12; P= 0.146
 University (n= 184)REF51.5 (46–58)26.7 (20–33)11.4 (6–17)4.6 (1–7)6.3 (2–10)
 Senior/Trade/Technical (n= 425)47.9 (44–52)26.8 (23–30)14.9 (12–18)4.6 (3–6)5.7 (4–8)
 Junior high school (n= 220)45.8 (41–51)32.6 (27–38)11.0 (7–15)3.2 (1–5)7.5 (4–11)
 Primary school or lower (n= 127)41.6 (34–49)36.4 (29–44)10.3 (6–15)6.5 (3–10)4.8 (2–8)
Work status >Chi-square = 19.57; DF = 8; P= 0.012
 Employ FT/PT/Casual (n= 418)REF50.7 (47–55)28.2 (23–33)8.8 (6–12)4.0 (1–7)8.3 (5–12)
 Retired (n= 449)44.8 (41–49)29.9 (26–33)16.3 (13–20)*4.4 (3–6)4.6 (3–6)
 Other/Not stated (n= 89)41.0 (33–49)*31.2 (23–40)10.0 (5–15)9.7 (4–15)8.0 (2–14)
Area-level SES >Chi-square = 27.92; DF = 16; P= 0.032
 Most disadvantaged (n= 122)50.7 (44–58)30.0 (23–37)10.5 (5–16)3.6 (1–6)5.3 (0–12)
 Quintile 2 (n= 216)51.8 (46–57)26.0 (21–31)14.4 (10–19)3.0 (1–5)*4.8 (2–7)
 Quintile 3 (n= 253)40.3 (35–45)*33.8 (29–39)*14.4 (11–18)6.0 (3–9)5.6 (3–8)
 Quintile 4 (n= 189)45.6 (40–51)31.7 (26–38)10.1 (6–14)3.7 (1–6)*8.8 (5–13)
 Least disadvantaged (n= 176)REF51.3 (45–58)24.4 (17–31)9.7 (5–15)9.0 (4–14)5.5 (2–9)
Travel time (TRACT) >Chi-square = 29.16; DF = 12; P= 0.004
 <1 h (n= 648)REF50.2 (47–53)26.2 (23–29)12.5 (10–15)4.5 (3–6)6.5 (5–8)
 1–<2 h (n= 72)46.3 (37–57)24.7 (15–34)11.5 (5–18)10.0 (4–16)7.4 (2–13)
 2–<4 h (n= 94)44.8 (37–51)38.4 (30–47)*9.1 (4–14)1.9 (0–5)5.8 (1–11)
 ≥4 h (n= 142)35.0 (28–42)*41.3 (34–49)*15.5 (10–21)4.3 (1–7)3.8 (1–7)
Reason for initial appointment >Chi-square = 26.98; DF = 16; P= 0.042
 Symptoms checked (n= 168)45.7 (39–52)31.8 (26–38)6.0 (2–10)*6.0 (3–9)10.4 (6–15)*
 Screening test (n= 187)51.2 (46–57)27.8 (22–34)10.8 (6–15)4.9 (2–8)5.3 (2–9)
 Doctor initiated (n= 211)45.4 (40–51)31.6 (26–37)14.7 (10–19)2.7 (1–5)5.6 (3–8)
 General check-up (n= 348)REF47.6 (44–52)29.9 (26–34)14.1 (11–17)4.6 (3–7)3.8 (2–6)
 Other/Not stated (n= 42)41.3 (29–53)21.8 (12–32)17.8 (6–29)7.5 (1–14)11.6 (1–22)
Clinical Gleason Score >Chi-square = 143.28; DF = 12; P < 0.001
 3+3 or less (n= 200)REF32.1 (27–38)17.2 (12–22)30.0 (24–36)5.7 (3–9)15.0 (10–20)
 3+4 (n= 370)53.4 (50–57)*26.8 (23–31)*12.0 (9–15)*2.5 (1–4)5.3 (3–8)*
 4+3 (n= 196)53.4 (48–59)*37.9 (32–44)*4.6 (2–7)*2.3 (0–4)1.8 (0–4)*
 4+4 (n= 59)46.6 (37–56)*46.2 (36–57)*3.4 (1–7)*3.8 (0–8)0.0 (0–0)*
 4+5 or more (n= 118)44.5 (37–52)*42.8 (34–51)*0.1 (0–8)*9.6 (6–13)2.2 (0–5)*
 Unknown (n= 13)45.1 (39–52)29.5 (7–52)6.2 (0–18)*6.9 (0–17)12.3 (0–28)
Stage >Chi-square = 51.46; DF = 4; P < 0.001
 Localised (n= 882)REF49.3 (47–52)29.0 (26–32)13.0 (11–15)2.4 (1–3)6.4 (5–8)
 Locally adv./Advanced (n= 74)10.0 (0.0–21)*65.6 (53–79)*0.0 (0–0)*21.3 (12–30)*3.0 (0–9)
Paternal history of prostate cancer >Chi-square = 10.41; DF = 4; P= 0.034
 Yes (n= 136)REF55.7 (49–62)29.5 (22–36)7.1 (3–11)2.1 (0–5)5.5 (2–9)
 No (n= 820)45.8 (43–48)*29.7 (27–32)13.4 (11–15)*5.0 (4–6)6.1 (5–8)
Smoking Status >Chi-square = 27.18; DF = 8; P= 0.001
 Current smoker (n= 86)35.3 (27–43)*33.6 (25–43)15.6 (9–23)11.7 (5–18)*3.8 (0–8)
 Previous smoker (n= 492)44.9 (41–48)*30.1 (27–34)13.9 (11–17)*4.4 (3–6)6.6 (5–9)
 Never smoker (n= 378)REF53.1 (49–57)28.0 (24–32)9.8 (7–13)3.5 (2–5)5.6 (4–8)
BMI >Chi-square = 18.19; DF = 8; P= 0.020
 Under/Normal weight (n= 246)REF51.1 (46–56)27.1 (22–32)11.6 (8–15)4.3 (2–6)5.8 (3–8)
 Overweight (n= 377)51.3 (47–55)26.1 (22–30)12.1 (9–15)5.1 (3–7)5.5 (3–8)
 Obese (n= 333)40.2 (36–44)*34.8 (30–39)*13.6 (10–17)4.6 (3–6)6.7 (4–9)
Physical health (SF-36) >Chi-square = 17.04; DF = 8; P= 0.030
 Below average (n= 162)44.1 (37–51)33.0 (27–39)10.0 (6–14)4.3 (2–6)8.7 (5–13)
 Average (n= 302)REF41.8 (37–46)33.1 (29–38)14.6 (11–18)5.2 (3–7)5.3 (3–8)
 Above average (n= 492)51.3 (48–55)*26.2 (23–30)*12.4 (10–15)4.7 (3–7)5.4 (3–7)
Radical prostatectomy

Radical prostatectomy was the most common form of treatment (47% of men [Table 1]). After adjusting for the other variables in the model, radical prostatectomy was a more common treatment among younger men, and those who lived closer to a hospital offering radiation facilities, had a Gleason score > ‘3+3’, had localized cancer, paternal history of prostate cancer, were never smokers, were not obese or had above average physical health (Table 2).

Radiation therapy combined with ADT

Almost one third (30%) of participants received a combination of radiation therapy and ADT (Table 1). Men were independently more likely to have the combination of radiation therapy and ADT if they were older, lived a longer distance from a hospital offering radiation therapy, had more advanced disease (based on either Gleason score or stage), were obese or had average rather than above average physical health (Table 2).

Radiation therapy alone

Approximately one in eight men (12%) had radiation therapy alone for their prostate cancer treatment (Table 1). Men were comparatively more likely to choose radiation therapy alone if they were retired, went to the doctor for a general check-up rather than to have symptoms checked, had low Gleason grade (≤‘3+3’), localized prostate cancer or had a previous smoking history (Table 2).

Androgen deprivation therapy alone

Less than 5% of men underwent ADT alone as their treatment for prostate cancer (Table 1). After adjustment for the other variables, these men were more likely to be aged ≥70 years at diagnosis, live in a disadvantaged area, have locally advanced or advanced stage cancer or be a current smoker (Table 2).

Monitoring

About 6% of participants were being monitored (Table 1). Monitoring was significantly more common for men who were aged ≥70 years, had their initial appointment as a result of symptoms rather than a general check-up or had low grade prostate cancer (Gleason score of ≤‘3+3’; Table 2).

INVOLVEMENT IN DECISION-MAKING

Information on the extent of involvement in the decision-making process for treatment was available for a subset of 885 men. Two-thirds (66%) of these men stated that they made the final treatment selection mainly or completely themselves, 18% shared the decision equally with their doctor, and the remaining 16% mainly or completely left the decision to their doctor. The multivariate model indicated that men who made their own treatment decision were more likely to be aged <70 years, have a higher level of education, have a higher income, live closer to a hospital with radiation facilities, have a combined Gleason score of ≤7 and have either no comorbidities or multiple comorbidities (full results not shown).

Involvement in decision-making was also significantly associated with treatment choice. In particular, 58% of men who made their own decision had a radical prostatectomy compared with 20% of those for whom the doctor made the treatment decision (Table 3). The reverse was observed for the combination of radiation therapy and ADT, which was selected by 21% of men who made their own decision and by 57% for whom the decision was left to the doctor.

Table 3. Prostate cancer treatment choices (crude %) by the extent of the patient's involvement in the decision-making process
 RPRT and ADTRT onlyADT onlyMonitoring
  1. DF, degrees of freedom; RP, radical prostatectomy; RT, radiation therapy.

Involvement in decision-making >Chi-square = 116.80; DF = 8; P < 0.001
 Made own decision (n= 585)58.520.913.32.15.3
 Decision equally shared (n= 160)32.540.615.65.65.6
 Doctor made decision (n= 140)20.057.17.97.17.9

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. FUNDING SOURCES
  9. CONFLICT OF INTEREST
  10. REFERENCES

To the best of our knowledge, this is the first large-scale prospective study conducted outside the USA to examine the factors associated with treatment decisions for men diagnosed with prostate cancer. It complements the findings of the large scale CaPSURE studies [17] and recent retrospective studies that have used Surveillance, Epidemiology and End Results data [18].

We show that the choice of treatment by men diagnosed with prostate cancer is strongly influenced by demographics, clinical characteristics and geographical location. Sharp contrasts in the choice between radical prostatectomy vs radiation therapy with or without ADT were evident by age at diagnosis, travel time to facilities offering radiation treatment, Gleason score, stage, body mass index and physical health. Men who chose surgery were younger, had above average physical health, and had lower grade cancers on the Gleason scale; men who had radiation therapy were older and had reduced physical health, with ADT added when men had more advanced disease. Other treatment options (ADT alone or monitoring) were chosen by only ∼10% of the cohort. Men >70 years were more often treated with ADT alone than were younger men. The patterns of these results are generally consistent with the Australian clinical practice guidelines [19], in that men with localized cancer, no significant surgical risk factors, and who prefer surgery, are most likely to benefit from radical prostatectomy.

Two previous studies have been conducted in the USA, based on separate, large cohorts diagnosed during the mid-1990s [20,21] around the time that PSA testing was becoming more prevalent. While direct comparisons are difficult owing to the significant advances in management, these studies had similar results to ours, in that men diagnosed with prostate cancer were more likely to choose radiation therapy over radical prostatectomy if they were older and had a higher PSA level suggestive of more advanced cancer [20,21].

Similarly to other studies [17,22,23], we found that younger men were significantly more likely to choose radical prostatectomy than older men, and these men were also more likely to have played a greater role in the decision-making process. It has been shown previously that a man aged 50 diagnosed with prostate cancer has a greater likelihood of dying prematurely from prostate cancer than a man diagnosed when age 70 [24]. While having no definitive evidence regarding the most effective treatment [2,3], the increased involvement among younger men in the decision to have surgery may suggest a preference to have the cancer physically removed rather than manage it with other methods, including active surveillance.

A previous Queensland study on breast cancer [25] reported mastectomy was a more common treatment choice for women living in rural areas compared with breast-conserving surgery (typically followed by radiotherapy). By contrast we found that men living further from radiation facilities, which are in the major urban centres, were more likely to have radiation therapy and ADT but were far less likely to choose radical prostatectomy, independently of age, stage and other related factors. We found this pattern consistently held when stratifying by Gleason categories (results not shown). Consistent with our results, a recent US study [26] indicated that patients with prostate cancer from rural areas in Georgia were more likely to receive external radiation and less likely to undergo radical prostatectomy than those living in an urban locality. This consistent finding from both studies is somewhat counterintuitive considering that radiation therapy requires ongoing treatment for a longer period of time, for which increased travel has been seen as a deterrent. The reasons behind this decision require further investigation, but could relate to the specialist urology services that may be more accessible at tertiary cancer treatment centres.

We found fairly limited evidence of any association between SES and treatment choices for men in this cohort, with the effect limited to a significant association with work status and a significant, but inconsistent association with area-level SES. This contrasts with other studies that have found that disparities in prostate cancer treatment patterns have been traced to SES differences [17,27]. Our inability to find a socioeconomic differential may have been influenced by the potentially higher representation of men who were more affluent than the general population. For example about three quarters of the men in our cohort had private health insurance, substantially higher than the corresponding population average of 48% [28]; however, Australia, like the UK, has a publicly funded hospital system, which means, in theory, the choice of treatment should not be influenced by their ability to pay for that treatment.

About two-thirds of the patients in the present study said they primarily made the decision about treatment themselves, with the remaining men either sharing the decision-making process with their doctor or else leaving the decision more or less completely up to their doctor. The proportion of men taking an active role is consistent with that reported by Gwede et al.[22] who showed, in a small (n= 119) study of men in the USA, that 63% of patients wanted an active role in treatment decision-making. There is evidence that there has been a transition from a passive to a more active role in recent years [9], and this may be attributable, at least in part, to the increased accessibility of information through the internet [8]. This highlights the importance of having quality and updated information available for men newly diagnosed with prostate cancer so they can understand the benefits and risks associated with each treatment choice.

All men in the present cohort were initially referred to the study by urologists, as with the large CaPSURE cohort in the USA [17]. This gave the present cohort the advantage of earlier recruitment (before treatment) into the study than would have otherwise been possible. Specifically, patients recruited through cancer registries in Australia are typically recruited 4–6 months after diagnosis, limiting the ability to assess pre-treatment quality of life and increasing recall bias. The design of the present study increased the likelihood, therefore, of recruiting men who were typically in the early stages of their treatment decision-making process. This may have increased the proportion of men receiving surgical treatment, because separate studies [18,22,23]have noted that patients enrolled by urologists are more likely to choose radical prostatectomy; however, we found that men who either shared or left the decision to their doctor were relatively more likely to be treated by radiation therapy than by having a radical prostatectomy.

In addition, data on stage at diagnosis are based on self-report. We were unable to validate this information, but when we compared self-reported Gleason score with the Gleason score extracted from the QCR, there was 95% concordance between the Gleason categories used in this paper. We were also unable to assess the impact of PSA levels at diagnosis, or the extent of biopsy involvement, on the treatment choices.

In conclusion, the present results show that men's baseline health and tumour characteristics influence treatment choices, suggesting a consistency with Australian clinical practice guidelines. With the majority of men indicating high levels of decisional control, the importance of having quality up-to-date information readily available to guide these decisions cannot be overstated.

ACKNOWLEDGEMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. FUNDING SOURCES
  9. CONFLICT OF INTEREST
  10. REFERENCES

The research team would like to acknowledge the Northern Section of the Urological Society of Australia and New Zealand and Urologists in Queensland for their support; Rob McDowall for registry data collection; Jan Howell and Lorraine Caesar for patient data collection; and those men who took part in the study.

FUNDING SOURCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. FUNDING SOURCES
  9. CONFLICT OF INTEREST
  10. REFERENCES

This project was awarded funding by the Australian National Health and Medical Research Council (NHMRC 442301) and Cancer Council Queensland Research Grant. Peter Baade and Suzanne Chambers are supported by NHMRC Career Development Fellowships (1005334 and 496003, respectively).

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. SUBJECTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. FUNDING SOURCES
  9. CONFLICT OF INTEREST
  10. REFERENCES
  • 1
    Ferlay J, Shin H, Bray F, Forman D, Mathers C, Parkin D. GLOBOCAN 2008: Cancer Incidence and Mortality Worldwide. Lyon: IARC, 2010. Available at: http://globocan.iarc.fr/. Accessed 25 May 2011
  • 2
    Wilt TJ, MacDonald R, Rutks I, Shamliyan TA, Taylor BC, Kane RL. Systematic review: comparative effectiveness and harms of treatments for clinically localized prostate cancer. Ann Intern Med 2008; 148: 43548
  • 3
    Heidenreich A, Bellmunt J, Bolla M et al. EAU guidelines on prostate cancer. Part 1: screening, diagnosis, and treatment of clinically localised disease. Eur Urol 2011; 59: 6171
  • 4
    Carter HB. Management of low (favourable)-risk prostate cancer. BJU Int 2011; 108: 168495
  • 5
    Huang GJ, Sadetsky N, Penson DF. Health related quality of life for men treated for localized prostate cancer with long-term followup. J Urol 2010; 183: 220612
  • 6
    Hegarty J, Beirne PV, Walsh E, Comber H, Fitzgerald T, Wallace Kazer M. Radical prostatectomy versus watchful waiting for prostate cancer. Cochrane Database Syst Rev 2010; (11): CD006590
  • 7
    Albertsen PC. When is active surveillance the appropriate treatment for prostate cancer? Acta Oncol 2011; 50 (Suppl. 1): 1206
  • 8
    Cox J, Amling CL. Current decision-making in prostate cancer therapy. Curr Opin Urol 2008; 18: 2758
  • 9
    Zeliadt SB, Ramsey SD, Penson DF et al. Why do men choose one treatment over another? A review of patient decision making for localized prostate cancer. Cancer 2006; 106: 186574
  • 10
    Branney P, White A, Jain S, Hiley C, Flowers P. Choosing health, choosing treatment: patient choice after diagnosis of localized prostate cancer. Urology 2009; 74: 96871
  • 11
    Singh J, Trabulsi EJ, Gomella LG. The quality-of-life impact of prostate cancer treatments. Curr Urol Rep 2010; 11: 13946
  • 12
    Baade PD, Aitken JF, Ferguson M, Gardiner RA, Chambers SK. Diagnostic and treatment pathways for men with prostate cancer in Queensland: investigating spatial and demographic inequalities. BMC Cancer 2010; 10: 452
  • 13
    Medical Outcomes Trust and Quality Metric Incorporated. SF-36: SF-36v2TM Health Survey. (IQOLA SF36v2 Standard, English (Australia), 7/03). Medical Outcomes Trust and Quality Metric Incorporated, Health Assessment Las: Lincoln, RI, 1992, 2003
  • 14
    Thompson SC, Pitts JS, Schwankovsky L. Preferences for involvement in medical decision-making: situational and demographic influences. Patient Educ Couns 1993; 22: 13340
  • 15
    Australian Bureau of Statistics. Census of Population and Housing: Socio-Economic Indexes for Areas (SEIFA), Australia. 2006. Canberra: ABS2008. Report No.: Cat. No. 2033.0.55.001
  • 16
    Sun GW, Shook TL, Kay GL. Inappropriate use of bivariable analysis to screen risk factors for use in multivariable analysis. J Clin Epidemiol 1996; 49: 90716
  • 17
    Cooperberg MR, Broering JM, Carroll PR. Time trends and local variation in primary treatment of localized prostate cancer. J Clin Oncol 2010; 28: 111723
  • 18
    Jang TL, Bekelman JE, Liu Y et al. Physician visits prior to treatment for clinically localized prostate cancer. Arch Intern Med 2010; 170: 44050
  • 19
    National Health and Medical Research Council and Australian Cancer Network. Clinical Practice Guidelines: evidence-based information and recommendations for the management of localised prostate cancer. A report of the Australian Cancer Network Working Party on Management of Localised Prostate Cancer. Canberra, 2003
  • 20
    Harlan LC, Potosky A, Gilliland FD et al. Factors associated with initial therapy for clinically localized prostate cancer: prostate cancer outcomes study. J Natl Cancer Inst 2001; 93: 186471
  • 21
    Schymura MJ, Kahn AR, German RR et al. Factors associated with initial treatment and survival for clinically localized prostate cancer: results from the CDC-NPCR Patterns of Care Study (PoC1). BMC Cancer 2010; 10: 152
  • 22
    Gwede CK, Pow-Sang J, Seigne J et al. Treatment decision-making strategies and influences in patients with localized prostate carcinoma. Cancer 2005; 104: 138190
  • 23
    Sommers BD, Beard CJ, D'Amico AV, Kaplan I, Richie JP, Zeckhauser RJ. Predictors of patient preferences and treatment choices for localized prostate cancer. Cancer 2008; 113: 205867
  • 24
    Baade PD, Steginga SK, Pinnock CB, Aitken JF. Communicating prostate cancer risk: what should we be telling our patients? Med J Aust 2005; 182: 4725
  • 25
    Thompson B, Baade P, Coory M, Carriere P, Fritschi L. Patterns of surgical treatment for women diagnosed with early breast cancer in Queensland. Ann Surg Oncol 2008; 15: 44351
  • 26
    Steenland K, Goodman M, Liff J et al. The effect of race and rural residence on prostate cancer treatment choice among men in Georgia. Urology 2011; 77: 5817
  • 27
    Chu DI, Freedland SJ. Prostate cancer. Socioeconomic status and disparities in treatment patterns. Nat Rev Urol 2010; 7: 4801
  • 28
    PHIAC. Quarterly Statistics: Hospital Treatment, General Treatment. Private Health Insurance Administration Council ed., Barton: Australian Government, 2011. Available at: http://www.phiac.gov.au/resources/file/membershipdata/MC%20Jun11.pdf. Accessed 25 May 2012