SEARCH

SEARCH BY CITATION

Keywords:

  • prostatic neoplasms;
  • quality of life;
  • urination disorders;
  • decision making

Abstract

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

OBJECTIVE

To investigate whether men’s self-reported health-related quality of life and lower urinary tract symptoms (LUTS) are associated with acceptance of prostate-specific antigen (PSA) testing and subsequent prostate biopsy.

PATIENTS AND METHODS

In a prospective questionnaire study of men aged 50–69 years, nested within the primary-care-based Prostate testing for cancer and Treatment study in nine UK areas, the Hospital Anxiety and Depression Scale (HADS), 12-item Short Form Health Survey (SF-12) and a self-reported LUTS measure (ICSmaleSF) were completed immediately before having a PSA test or prostate biopsy, or after not responding to an invitation for a PSA test or refusing a biopsy. Analyses compared 348 men accepting or 232 not responding to invitations for PSA testing and 318 accepting or 48 refusing a prostate biopsy.

RESULTS

Men accepting or not responding to the invitation for a PSA test had similar HADS, SF-12 and LUTS scores. Men accepting a biopsy had similar HADS and SF-12 scores to those refusing biopsy, but significantly more LUTS (P < 0.01 for hesitancy, reduced stream, intermittency, incomplete emptying, frequency during the day).

CONCLUSION

Depressed or anxious mood, comorbidity and LUTS were not associated with the decision to respond to invitations for a PSA test. Men agreeing to a biopsy were more likely to have LUTS than those refusing, suggesting that men believe that LUTS are a symptom of prostate cancer. Men needing a prostate biopsy require more information about LUTS so that they can make informed choices about testing for prostate cancer.


Abbreviations
HRQoL

health-related quality of life

ProtecT

Prostate testing for cancer and Treatment (study)

HADS

Hospital Anxiety and Depression Scale

SF-12

12-item Short Form Health Survey

ERSPC

European Randomised Study of Screening for Prostate Cancer.

INTRODUCTION

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

Prostate cancer is now the most common cancer in men in the Western world, and an important public health problem [1]. Early detection and radical treatment can cure prostate cancer, but there is a high risk of serious side-effects, and many untreated men would not develop advanced disease. Because the balance of benefits and harms is so unclear, formal screening programmes are rare, except in the context of randomized controlled trials or other prospective evaluation studies [2]. However, PSA testing is widespread in several countries in the absence of an organized programme, notably in the USA [3] and Australia [4].

Whether PSA testing is available through a formal screening programme or on request, men must decide whether to take the test and they might also need to decide whether to have a prostate biopsy if their PSA level is raised. In the UK, where men can request a PSA test through their GP, an information leaflet is available in which men are advised to consider the benefits and risks before having the test (http://www.cancerscreening.nhs.uk/prostate/index.html). However, such advice is likely to be just one factor amongst many contributing to a man’s decision about whether to have his PSA level measured. Other influences include how the advice is presented [2], media reports [5], presence of LUTS [5,6], and the emotional impact of anticipating the screening procedure [7].

We studied the health-related quality of life (HRQoL) of men who decided in favour of or against PSA testing and biopsy in the context of the Prostate testing for cancer and Treatment (ProtecT) study (International Standard Randomized Controlled Trial Number 20141297). ProtecT is the largest primary care-based randomized trial of PSA testing and treatment in the world, with men aged 50–69 years at general practices in nine UK centres invited (via a letter from their GP) for PSA testing [8]. In the HRQoL study, men who refused PSA testing, or who refused a biopsy having been found to have a high PSA level, were invited to complete the same questionnaires as men undergoing these tests as part of the ProtecT study. This allowed us to investigate the association between LUTS, measures of HRQoL, and decisions made about undergoing PSA testing and prostate biopsy.

PATIENTS AND METHODS

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

At participating general practices, men aged 50–69 years were sent an information sheet detailing the study aims, process, information about prostate cancer (including that it might be asymptomatic), treatment alternatives and possible side-effects, and an appointment to attend a nurse-led ‘prostate check clinic’. At the clinic, men received further information about the implications of PSA testing, uncertainties about testing and treatments, and the need for a treatment trial. Men giving written informed consent to enter the ProtecT study had their PSA level tested, and subsequently participants with a PSA level of ≥3.0 ng/mL were invited to attend the centre’s urology department to have a DRE and TRUS-guided biopsy. Men with clinically localized disease were eligible to participate in the treatment trial, while men with a normal PSA result (<3.0 ng/mL) exited the study; 11% of men had an abnormal PSA result, and 25% with a raised PSA level had prostate cancer at biopsy. Of men with a high PSA level, ≈12% refused a biopsy.

The HRQoL study included men who: (a) accepted a PSA test; (b) did not respond to a PSA test invitation (‘nonresponders’); (c) accepted a prostate biopsy or; (d) refused biopsy. Men who accepted a PSA test completed the questionnaires at the prostate clinic immediately before having the test. These men all subsequently received a negative PSA result and were randomly sampled over the data collection period. Men who did not respond to the PSA test invitation within a mean (range) of 43 (25–79) days were posted a separate letter asking them to consent to this study and complete the questionnaires; no reminder letter was sent to these men. Non-responders were recruited from seven general practices in one centre.

Men accepting a biopsy completed the questionnaires at the prostate biopsy clinic immediately before having the procedure. Men who were not found to have cancer at biopsy were included in this study. Men refusing a biopsy were posted a questionnaire within a mean (range) of 9 (0–70) days after refusing the procedure. A reminder letter and questionnaires were sent to men not responding within 21 days.

A target sample size of 250 men in each group allowed a 1-point difference in means to be detected with 80% power at the 5% significance level when comparing two groups on a component of the Hospital Anxiety and Depression Scale (HADS). A sd of 3.5 in each group was assumed [9] and a design effect of 1.25 accommodated to allow for clustering at the practice level. In the event, over-recruitment was allowed in some groups to allow for shortfalls in others. Approval was obtained from Trent Multi-centre Research Ethics Committee.

At each assessment point men completed the HADS [10,11], and the 12-item Short Form Health Survey [12] (SF-12). Selected questions from a validated self-reported LUTS measure (ICSmaleSF) [13] were also administered at the time of PSA testing, with the full set of questions being administered at the time of biopsy. The appendix gives further details of these measures.

HRQoL and LUTS scores were compared between: (i) men having accepted and not responding to an invitation for PSA testing; and (ii) men with a raised PSA level accepting or refusing biopsy. As most outcome measures were not normally distributed, differences in mean scores were presented with bias-corrected and accelerated bootstrap CIs [14], the latter being based upon 4999 re-samples stratified by group. P values were calculated using the usual methods but were always in broad agreement with the CIs.

A data-based adjustment of sems to allow for clustering within practices was not used, as the few practices represented by men not responding to the PSA invitation would provide unreliable robust sems [14], and bootstrap estimates based on re-sampling of practices would include different numbers of men in each re-sampled dataset. Hence the analysis was mostly done with no allowance for clustering. To allow the effect of clustering to be gauged, P values for key results are recalculated with the sem inflated by 1.27. This is the square root of a design effect of 1.61 that, once the variable number of men per practice is taken into account [15], accommodates an intraclass correlation of 0.023 for comparison (i) and 0.067 for comparison (ii) above. This design effect is larger than that used in the sample size calculation, but accommodates intraclass correlations of a magnitude since reported for similar studies [16]. Differences in means, CIs and P values were adjusted for age, with a significance criterion of 1% used throughout.

RESULTS

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

Response rates for this study are shown in Table 1; the participants’ characteristics were similar in men accepting or not responding to the PSA invitation. Men accepting a biopsy were more likely to be in paid employment (= 0.003) and married or living as married (= 0.001) than men refusing.

Table 1.  Baseline data for the four groups of men
 Invited to prostate clinic and PSA testPSA ≥3 ng/mL, invited for biopsy
Accept (A)No response (B)PAccept (A)Refuse (B)P
  1. N/A, not available; *Wilcoxon rank sum tests, other P values are t-tests.

Mean (sd) age, years 59.5 (5.8)  59.6 (6.7)0.81 61.5 (5.7) 62.8 (4.6)0.130
Mean (range) PSA, ng/mL  1.0 (0.1–2.9)N/A  4.8 (1.5–24)  4.6 (3.0–14)0.360*
Paid employment, n (%)204 (58.1)N/A186 (53.5) 16 (31.4)0.003
Married or living as married, n (%)286 (81.5)N/A292 (83.9) 33 (64.7)0.001
Questionnaires posted, n4971268380102
Response rate, n (%)348 (70) 232 (18)318 (84) 48 (47)

Men who accepted the PSA test had similar scores for anxiety, mental and physical health and LUTS to those who did not respond. They had less depression than nonresponders but the evidence supporting this difference was weak (= 0.065; Table 2). Inflating the sems to allow for the clustered design did not alter the conclusions supported by these results.

Table 2.  HRQoL and LUTS scores reported by men who accepted a PSA test compared to men who did not respond to the test invitation
VariableMean (sd)Age-adjusted difference (A–B) (95% CI)P
Accept PSA test (A)No response (B)
HRQoL
 HADS – Anxiety 5.36 (3.55) 5.49 (4.35)−0.17 (−0.84, 0.52)0.610
 HADS – Depression 3.34 (2.95) 3.81 (3.36)−0.50 (−1.04, 0.04)0.065
 SF-12 Mental52.48 (8.76)52.18 (9.44) 0.74 (−0.88, 2.43)0.370
 SF-12 Physical48.60 (9.75)48.08 (10.2) 0.51 (−1.29, 2.38)0.590
Urinary symptoms
 Hesitancy 0.62 (0.77) 0.48 (0.73) 0.10 (−0.02, 0.23)0.100
 Frequency during day 0.95 (0.83) 1.00 (0.89)−0.06 (−0.20, 0.08)0.430
 Frequency during night 0.99 (0.86) 1.00 (0.91)−0.04 (−0.19, 0.10)0.600
 LUTS interfere with life 0.27 (0.52) 0.25 (0.59) 0.03 (−0.07, 0.11)0.590

Men who accepted a prostate biopsy had similar anxiety, depression, mental, and physical health scores, but significantly more LUTS, than men who refused (Table 3). Allowing for the clustered design, there remained strong evidence of higher rates of reduced stream (= 0.002) and intermittency (= 0.005) in men who chose to have a biopsy, and weak evidence of higher rates of hesitancy (= 0.026), incomplete emptying (= 0.017) and greater frequency of urination during the day (= 0.033).

Table 3.  HRQoL and LUTS scores reported by men who accepted a prostate biopsy compared to men who refused a biopsy
VariableMean (sd)Age-adjusted difference(C–D) (95% CI)P
Accept biopsy (C)Refuse biopsy (D)
HRQoL
 HADS – Anxiety4.89 (3.45)4.23 (3.55)0.65 (−0.44, 1.71)0.230
 HADS – Depression2.59 (2.66)3.13 (3.38)−0.55 (−1.72, 0.32)0.200
 SF-12 – Mental53.86 (7.63)53.74 (8.10)0.18 (−2.16, 3.38)0.890
 SF-12 – Physical50.70 (8.82)49.66 (10.1)1.07 (−1.86, 4.89)0.480
Urinary symptoms
 Hesitancy0.74 (0.75)0.42 (0.65)0.33 (0.10, 0.51)0.005
 Straining0.50 (0.75)0.42 (0.65)0.08 (−0.15, 0.26)0.480
 Reduced stream0.92 (1.06)0.35 (0.60)0.60 (0.38, 0.79)<0.001
 Intermittency0.99 (0.89)0.53 (0.65)0.49 (0.25, 0.68)<0.001
 Incomplete emptying0.86 (0.96)0.42 (0.77)0.45 (0.17, 0.66)0.003
 Frequency during day1.30 (0.88)0.94 (0.82)0.37 (0.11, 0.62)0.007
 Frequency during night1.18 (0.90)1.02 (0.85)0.18 (−0.10, 0.43)0.190
 LUTS interfere with life0.43 (0.68)0.31 (0.62)0.12 (−0.11, 0.28)0.250

DISCUSSION

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

This prospective study showed that men accepting PSA testing had similar HRQoL and self-reported LUTS scores as men not responding to an invitation. Those accepting a prostate biopsy had similar HRQoL scores to men refusing the invitation but, notably, those who refused a biopsy had significantly fewer self-reported LUTS. These results are important because it suggests that men might not understand the relationship between LUTS and prostate cancer, and those at risk of prostate cancer could subsequently fail to attend for biopsy because they have no LUTS.

While several studies have assessed the impact of screening on anxiety and depression, few have assessed the association between these two factors and the decision-making process. There was no evidence in the present study that anxious or depressed mood was associated with the decisions about PSA testing or biopsy. This is consistent with a similar investigation in the European Randomised Study of Screening for Prostate Cancer (ERSPC) Dutch centre, where men waiting to have their PSA tested reported similar levels of trait anxiety to men contacted 3–10 months after declining to take part in the study [17] (although in the ERSPC men decide whether or not to take part in the study, with subsequent random allocation to PSA testing or usual management). It might be that men do not ruminate over this decision-making sufficiently for it to affect their mood; a small group of men from the present study also took part in qualitative interviews, and several of the responses to those interviews indicated rapid decision making [18].

Findings from the ERSPC suggest that men who refuse PSA testing do so because of a perceived low risk of prostate cancer in the absence of LUTS [6]. In that study 34% of men who refused to participate agreed with the statement ‘A man may have prostate cancer, even without having symptoms’, compared to 48% of men who participated (P < 0.001) [6]. However, while those results suggest that a man’s beliefs influence his decision making, in the present study there is no evidence of an underlying association between a man reporting LUTS and his decision about having his PSA level tested. In the qualitative part of the present study several men reported being influenced by their LUTS when asked directly about their decision-making [18]. This suggests that studies might overestimate the association between the presence of LUTS and the decision to undergo testing, if questions are asked in a way that makes it more likely that men who undergo testing will report LUTS.

The qualitative interviews also suggest that at least some men weigh a raised PSA level against the presence or absence of LUTS when deciding whether to proceed to biopsy [18]. This is consistent with the present study, with men reporting LUTS being more likely to proceed with biopsy. However, it is likely that men are misjudging their risk of having prostate cancer by taking LUTS into account, as prostate cancer rarely causes symptoms while localized, with LUTS more often being a symptom of benign prostate conditions [19].

The major strengths of this study are the population-based design, allowing the participation of any man in the study population to be invited, and the detailed assessments of LUTS. Furthermore, in ProtecT decisions about whether to undergo PSA testing and biopsy are separated from and precede the decision about whether to participate in the treatment trial. This makes the results of this study easier to interpret than those from the ERSPC, where men were invited to participate in a study within which PSA testing was randomly allocated [6]. The present study also has some limitations. Men not responding to PSA test invitations were only sampled from one UK area (albeit from seven general practices within that area), because contacting this group required the researcher (K.A.) to visit the GP practice to access men’s details. The response rate in men not having the PSA test was low; these participants had already not responded to the invitation for a test within the context of the ProtecT study and were difficult to reach. There were relatively few men recruited to this study and refusing biopsy, due to limitations in approaching these men if they had withdrawn from the main ProtecT study and the low response rate in this group. It was inappropriate to contact men who had actively refused participation in the ProtecT study, but this is unlikely to have caused bias in our results, as these men represented only 5% of those invited to undergo PSA testing.

In part these limitations are inherent in the nature of our study. While other studies have explored the impact of HRQoL in men undergoing testing for prostate cancer [17,20], this study focused on the HRQoL of men who are difficult to reach, i.e. those not-responding to PSA testing and those refusing prostate biopsy.

In conclusion, men agreeing to biopsy were more likely to have self-reported LUTS than those who refused, suggesting that men perceive LUTS as a symptom of prostate cancer. Men require more information about LUTS and the likelihood that they are related to benign rather than malignant disease, so that they can make informed choices about their need for a prostate biopsy. This study did not identify HRQoL or symptomatic differences between men responding or not responding to PSA test invitations. However, it might be more practical to provide information about LUTS at the time of PSA testing, so those men who subsequently need to consider biopsy are well prepared. In addition, the results of the present study may not be generally applicable to contexts where men must approach their GP to request a PSA test, and it remains possible that earlier provision of information on LUTS in those contexts might assist shared decision-making about the test.

FUNDING

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

This research was funded by Cancer Research UK. The ProtecT study is funded by the Department of Health NHS Health Technology Assessment Programme (projects 96/20/06, 96/20/99). The views and opinions expressed in this report are those of the authors and do not necessarily reflect those of the Department of Health or Cancer Research UK. The funders had no involvement in this work.

ACKNOWLEDGEMENTS

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

The authors acknowledge the tremendous contribution of all members of the ProtecT study research group, and especially the following who were involved in this research: Prasad Bollina, Sue Bonnington, Debbie Cooper, Andrew Doble, Alan Doherty, Emma Elliott, David Gillatt, Pippa Herbert, Peter Holding, Joanne Howson, Mandy Jones, Roger Kockelbergh, Howard Kynaston, Teresa Lennon, Norma Lyons, Hilary Moody, Philip Powell, Stephen Prescott, Pauline Thompson. Pete Shiarly is thanked for his contribution to the design and management of the study database.

REFERENCES

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

Appendix

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING
  8. ACKNOWLEDGEMENTS
  9. CONFLICT OF INTEREST
  10. REFERENCES
  11. Appendix
INTERPRETATION OF QUESTIONNAIRE SCORES

HADS: Two seven-item scales assess anxiety and depression separately. Scores range from 0 to 21, with higher scores indicating higher levels of anxiety and depression [11]. Scores of 0–7 are considered normal, with 8–10 mild, 11–15 moderate and 16 or above severe cases [10].

SF-12: The 12 items contribute to two factors which assess mental and physical health. Mean score ranges from 0 to 100, with scores below 50 indicating poorer than average health [12].

ICSmaleSF: Two scales assess voiding LUTS (five items), including hesitancy, straining, reduced stream, intermittency and incomplete emptying (item score range 0–4, subscale score range 0–20), and urinary incontinence (six items), including urgency, urge incontinence, stress incontinence, unpredictable/miscellaneous incontinence, nocturnal incontinence and postmicturition dribble (item score range 0–4, subscale score range 0–24). Three separate items assess frequency during the night (score range 0–4), frequency during the day and overall interference of urinary symptoms with life (score range 0–3). For all items and scales, a higher score indicates greater symptoms [13].