- Top of page
- Patients and methods
Objective To determine the level of psychopathology, traumatic distress and quality of life in men with newly diagnosed clinically localized prostate cancer, the effect on these of a consultation in a combined-specialist early-prostate clinic, and predictors of psychopathology.
Patients and methods Eighty-eight patients were recruited from the combined clinic; they completed a battery of questionnaires including the Hospital Anxiety and Depression Scale (HADS), the revised Impact of Event Scale (IES) and the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire, before their first appointment. Two weeks later they completed the HADS, IES and a patient-satisfaction survey.
Results The overall level of psychopathology varied among the questionnaires used, from 0% on the HADS depression scale, 8% on the HADS anxiety scale and 14% on the IES. Anxiety and traumatic stress symptoms were commoner than depressive symptoms. The quality-of-life scores showed a relatively good level of functioning. Pre-morbid factors and disease status did not predict psychological distress. Younger age was mildly predictive of poorer psychological functioning. Anxiety symptoms reduced slightly after a joint clinic appointment, whereas depressive symptoms showed a slight increase.
Conclusion This study suggests that men with early localized prostate cancer have low levels of psychopathology overall. However, some men experience distressing psychological symptoms and it is important that future research is conducted to help develop clear guidelines on the optimal methods of detecting and managing men with prostate cancer who have mental health difficulties.
- Top of page
- Patients and methods
Psychological distress after a diagnosis of cancer is well recognized; one study found that 34–46% of newly diagnosed patients with breast cancer had a psychiatric disorder . In 117 newly diagnosed outpatients aged 21–75 years with various forms of cancer, just over a quarter had significant psychological distress, anxiety (26%) being commoner than depression (7%) . Similar results were found in a larger study where 187 (23%) of 809 inpatients and outpatients with various types of cancer scored above the standard threshold of 11 on the Hospital Anxiety and Depression Scale (HADS), suggesting psychological difficulty . Other studies have confirmed relatively high levels of psychopathology in a wide range of patients with cancer [4,5].
Few studies have considered psychopathology in men with prostate cancer, and in a literature review we found only one study that considered psychopathology in a group of patients with early localized disease. Of 163 men with prostate cancer, 29% complained of worry and 21% complained of depression . The two-thirds who had advanced disease described significantly more psychopathology than the third with early-stage disease. In another study the HADS was administered to 92 men with advanced-stage prostate cancer. Thirty (32.6%) scored at or above a threshold of 7 on the anxiety subscale and 14 (15.2%) did so on the depression scale .
More studies have focused on measures of quality of life (QoL); some have suggested that QoL is an independent prognostic factor for survival in various types of cancer  and the routine measurement of QoL has been advocated in the research and clinical management of men with prostate cancer. In 1072 patients who had undergone radical prostatectomy for clinically localized prostate cancer QoL scores were comparatively high and the same as in patients with BPH who had undergone TURP . Several studies of the treatment of prostate cancer have found better physical outcome to be associated with better QoL scores [10–12]. One study found no significant difference in QoL scores between patients with localized prostate cancer treated with surgery or radiotherapy .
In an attempt to improve knowledge of psychological distress and QoL in patients with prostate cancer, we studied individuals with newly diagnosed clinically localized prostate cancer. The objectives of the study were to determine: (i) the level of psychopathology, traumatic distress and QoL in the sample soon after a diagnosis of early prostate cancer; (ii) the effect of a consultation in a combined, specialist early-prostate cancer clinic on the level of psychopathology; and (iii) predictors of psychopathology.
Patients and methods
- Top of page
- Patients and methods
Patients were consecutive referrals with a new diagnosis of clinically localized prostate cancer to a clinic jointly run by the departments of urology and oncology. The aim of the clinic was to facilitate individualized decision-making among the three possible management approaches (radical surgery, radiotherapy or ‘watchful waiting’). Recruitment was between May 1998 and December 1999; the only exclusion criterion was cognitive deficits which would make completing the questionnaires impossible. Ethical committee approval was obtained.
A letter of invitation to take part in the study, explaining the nature of the study, was sent to each prospective clinic attendee, with the clinic appointment letter. On arrival at the clinic, the specialist nurse discussed the study with patients, answered any questions and obtained informed consent. Questionnaires (see below) were administered to consenting patients immediately before the joint clinic appointment (Assessment 1). After the appointment, the lead clinician completed a data sheet on each patient, including clinical information on the severity of disease, including PSA level, TNM classification, Gleason grade and ASA score. Two weeks later, a postal patient-satisfaction survey about the clinic was sent to each participant, with further questionnaires (Assessment 2).
Five questionnaires were used: (i) The 30-Item General Health Questionnaire (GHQ30) , which comprises 30 questions relating to the individual's mood and feeling of well-being over the previous few weeks, each question requiring one of four responses, two of which would receive a score of 0 and two scoring 1. The maximum possible score is therefore 30; a score of >4 has been suggested as being indicative of psychological distress, although in a population with physical symptoms, a threshold of 10/11 has been advocated. The GHQ30 has been validated for use in cancer populations . (ii) The HADS , a 14-item scale specifically designed for patients with medical illnesses, as it excludes somatic items and relies only on emotional symptoms of depression and anxiety. It has been validated in cancer populations [17,18]. Seven questions relate to symptoms of depression and seven to anxiety over the previous week, again using a Likert scale. The maximum score is 21 on the anxiety and depression subscales, with a threshold of 10/11 being used to detect likely psychiatric ‘caseness’ (a term used to describe a level of psychological distress that is believed to be clinically significant). (iii) the Impact of Event Scale-Revised (IESR) [19,20], a well validated questionnaire of 22 questions with Likert scales covering the three characteristic groups of symptoms found in post-traumatic stress disorder (intrusion, avoidance and hyperarousal) over the preceding week. The diagnosis of prostate cancer was considered the traumatic event in this study. (iv) the EORTC Quality of Life Questionnaire-C30 version 2.0 (EORTC-30) , a 30-item questionnaire which considers physical, psychological and social aspects of QoL using visual analogue scales (VAS). It has been shown to have good validity when used in patients with cancer . (v) a 0–100 VAS to measure pain, included because of the association of higher levels of pain with poorer outcome in previous studies . All five questionnaires were administered at Assessment 1, and the HADS and IESR at Assessment 2.
Basic demographic data, the information recorded by the lead clinician and the questionnaire data were analysed statistically; the mean (sd) was obtained for continuous data, with frequencies of individuals who scored above thresholds on the questionnaires. Frequencies were obtained for discrete data. Differences in scores between Assessments 1 and 2 were evaluated using Student's t-test; odds ratios were calculated for categorical data. To consider the possible predictive value of variables on psychological distress, forward stepwise linear regression analysis was applied using the total IESR score at Assessment 1 as the dependent variable. Age, physical functioning score on the EORTC-30 scale, medical history, psychiatric history, PSA score, pain VAS and marital status were entered as independent variables.
- Top of page
- Patients and methods
Of 104 patients attending the joint clinic during the recruitment period, 88 agreed to take part in the study, three refused and 13 were excluded (10 were found to have metastatic disease and three felt unable to complete the questionnaires after consenting). Eighty-eight men completed Assessment 1 and 61 (69%) also completed Assessment 2. The mean (sd, range) age of the sample was 64.5 (6.7, 48–78) years; 77 (89%) were married and only 4% had any significant physical comorbidity (ASA 3 or 4). Thirteen (15%) had no medical history, 30 (34%) had previous currently inactive and unrelated medical disorders or operations, 22 (25%) had active disorders which were unrelated to prostate disease and 23 (26%) had disorders in both these categories. Seventy-nine (92%) had no psychiatric history, three (3.5%) had some psychiatric history requiring management by a GP, two (2.3%) had required assessment by a psychiatrist as an outpatient, and two (2.3%) had required admission as an inpatient for psychiatric disorder.
The mean (sd, range) PSA level was 12.93 (10.66, 3–57) ng/mL. Table 1 shows the Gleason score frequencies and clinical stage distribution. The mean score on a 0–100 VAS for pain was 15.62 (21.05, 0–85).
Table 1. Gleason scores and stage distribution
|Gleason score/stage||N (%) |
The levels of psychological distress and caseness at Assessment 1 are shown in Table 2; Table 3 compares the mean values of the HADS and IESR scores at the two Assessments. The HADS anxiety subscale score was significantly lower at Assessment 2 and the HADS depression subscale score was significantly higher at Assessment 2. There was no significant difference between IESR scores at the two Assessments. The QoL scores at Assessment 1 showed a relatively good level of functioning across the domains measured (Table 4).
Table 2. The scale for Questionnaires at Assessment 1
|Questionnaire||No. completing questionnaire||Mean (sd) [range]||Maximum possible score||Caseness threshold||No. (%) of cases|
|GHQ-30||79||3.24 (4.95) [0–25]||30|| 4/5||20 (25)|
|GHQ-30||79||3.24 (4.95) [0–25]||30||10/11|| 7 (9)|
|HADS-A||83||4.96 (3.82) [0–16]||21|| 7/8||18 (22)|
|HADS-A||83||4.96 (3.82) [0–16]||21||10/11|| 8 (10)|
|HADS-D||83||1.95 (2.5) [0–10]||21|| 7/8|| 4 (5)|
|HADS-D||83||1.95 (2.5) [0–10]||21||10/11|| 0|
|IESR-AV||85||9.06 (8.73) [0–34]||40||NA||NA|
|IESR-INT||85||8.6 (7.4) [0–35]||35||NA||NA|
|IESR-Hyp||85||4.06 (5.48) [0–21]||35||NA||NA|
|IESR I+A||85||17.54 (14.45) [0–57]||75||34/35||12 (14)|
Table 3. The mean differences in psychometric measures at Assessments 1 and 2 for the 61 pairs who completed the questionnaires at both times
|Psychometric measure||Mean (sd) ||Mean (95% CI) difference||Two-tail P|
|Assessment 1||Assessment 2|
|HADS-A||5.11 (3.89)||4.38 (3.90)||0.74 (0.033, 1.44)||0.04|
|HADS-D||1.79 (2.37)||2.46 (2.57)|| − 0.67 (− 1.14, −0.20)||0.006|
|IESR-A||8.66 (7.67)||9.49 (9.03)|| − 0.84 (− 2.76, 1.09)||0.39|
|IESR-I||8.64 (6.87)||8.05 (7.07)||0.59 (− 0.96, 2.14)||0.45|
|IESR-H||3.80 (4.89)||4.41 (5.76)|| − 0.61 (− 1.79, 0.58)||0.31|
|IESR-A+I||17.30 (13.40)||17.54 (14.61)|| − 0.25 (− 3.32, 2.83)||0.87|
|IESR total||21.21 (17.41)||21.80 (19.37)|| − 0.59 (− 4.43, 3.25)||0.76|
Table 4. The EORTC-30 results at Assessment 1
|Variable||No. of men||Mean (sd) raw score||Scoring range|
|Physical functioning||83||1.32 (0.5)||1–2|
|Role functioning||86||1.30 (0.63)||1–2|
|Emotional functioning||86||1.54 (0.69)||1–4|
|Cognitive functioning||84||1.40 (0.59)||1–4|
|Social functioning||85||1.39 (0.87)||1–4|
|Overall health status||85||5.50 (1.9)||1.33–7|
|Nausea and vomiting||86||1.04 (0.14)||1–4|
|Appetite loss||86||1.07 (0.30)||1–4|
|Financial difficulties||86||1.14 (0.46)||1–4|
Fifty-three of 60 responders (88%) gave positive responses to the confidential patient-satisfaction survey of the joint clinic at Assessment 2. There were three (5%) negative responses; two expressed concerns about the administration of the clinic and one that his prostate cancer had not been fully discussed (in fact, a renal tumour had been found coincidentally and the consultation involved prolonged discussion of this).
The forward stepwise linear regression analysis showed that the only independent variable entered that significantly affected the total IESR score was age (T=− 2.97, P=0.004), with younger men faring worse than older men, although the effect of this was not large and only accounted for 10% of the total variance in the IESR score (adjusted R2=0.096).
- Top of page
- Patients and methods
The overall level of psychopathology in this study varied among the questionnaires used, from 0% on the HADS depression scale, 8% on the HADS anxiety scale, 14% on the IESR and 20% on the GHQ-30 (using a threshold of 4/5) and 9% using a threshold of 10/11. Anxiety and traumatic stress symptoms were commoner than depressive symptoms. Pre-morbid factors and disease status did not predict psychological distress. Younger age was mildly predictive of poorer psychological functioning. Anxiety symptoms reduced slightly after a joint clinic appointment, whereas depressive symptoms showed a slight increase.
A key strength of this study is that it considered consecutive attendees to a joint clinic who had a confirmed diagnosis of clinically localized prostate cancer. To the best of our knowledge this is the first study to focus on psychopathology in a sample of men with early localized prostate cancer. Previous studies have considered different cancers together and individuals at different stages of disease and treatment. The recruitment rate of 84.6% is high and should maximize the likelihood that the results are generally applicable to this population group. The sample size of 88 participants is reasonable, although not large. We were disappointed that only 61 (69%) returned the questionnaire at 2 weeks, despite reminders, and do not know how this might have biased the results. A few patients failed to complete every questionnaire adequately at the first assessment, but there were sufficiently few that is was unlikely to have caused any significant bias in the results. The questionnaires used have all been well validated and cover a wide range of symptoms, but ideally all patients would have also received a structured psychiatric interview. The findings of the linear regression analysis should be interpreted with caution because some of the continuous variables had a nonparametric distribution.
The mean scores on the questionnaires are not high and are lower than in most studies of patients with cancer in general [2–5] and in those that have considered men with prostate cancer and more advanced disease [6,7]. Indeed, the mean values are little higher than those in the normal population, suggesting that the diagnosis of early prostate cancer did not precipitate a significant psychiatric reaction in most participants. The manner and timing of assessment, assessment tools used, stage of disease and other factors are possible explanations for the lower levels of psychological distress in this study. Particularly important is that the present participants were highly selected, having already been seen and counselled by the urologist who referred them to the specialist clinic.
That men with prostate cancer are generally older may be important; it has previously been suggested that older men are less likely to discuss their emotions , which may result in particularly low levels of psychopathology. Perhaps at such an early stage, particularly in the absence of physical disability, denial is used as a defence mechanism by many men, with psychological distress only emerging with disease progression. This gains some support by the finding of much higher levels of psychological distress (37% HADS cases) in patients with advanced lung cancer . In addition, the present participants may have already been advised that the prognosis of early, localized prostate cancer is relatively good compared with other common cancers.
Females have been consistently shown to experience prevalence rates of about double those of males in large community studies of depression, anxiety and post-traumatic stress disorder . This may account, at least partly, for the higher levels of reported psychopathology in women with breast cancer than in men with prostate cancer. Another possible explanation for the relatively low levels of psychopathology in the present study is that men referred to a joint clinic are in some way specially selected and less likely to be psychologically distressed than those who are not, although this is speculation.
The reduction in anxiety levels and increase in depression levels between Assessments 1 and 2 is interesting, although probably of limited clinical significance, given the low levels of psychopathology at both times. Anxiety levels were possibly reduced by an in-depth discussion of the diagnosis, treatment options and prognosis. However, it may be that this exploration resulted in a fuller recognition of the implications of the diagnosis, with some increase in depressive symptoms.
The results of the linear regression analysis was not promising in terms of being able to predict who will develop psychological difficulties after diagnosis, suggesting that younger men should be considered as being at slightly higher risk. However, with only 10% of the total variance explained, age is, at best, a weak predictor. Other studies have found other factors to be predictive of outcome, including advanced stage of disease, surgical intervention, greater pain, fatigue and urinary problems . Previous studies have found independent predictors of psychological distress in those with cancer to be female gender, experience of disturbance in family and social life caused by illness, nausea and vomiting, and perception of poor physical health . We were unable to support these findings in men with early prostate cancer, but several of the predictors would not have been applicable to this study. Studies of psychopathology after other traumatic events have usually found that distress soon after the event is highly predictive of later outcome, highlighting the importance of considering psychological symptoms at this time .
The relatively low level of psychopathology suggests that for most men with a diagnosis of early prostate cancer in this selected population there is no need to involve mental health professionals. However, a few do have difficulties and could benefit from mental health support. There is also a need to provide adequate support and information to all patients through other than mental health professionals as part of routine care.
The ability of brief questionnaires to detect psychopathology and the apparent willingness of individuals to complete them suggests that it is appropriate to screen for mental health difficulties using questionnaires, and then to offer those who score above predetermined thresholds further assessment and management of specific difficulties, as indicated. This diverges from calls to offer preventive counselling to all those with cancer. There is no good evidence that an ‘intervention for all’ approach works; it is costly and difficult to justify, not least because it would result in many individuals with no significant psychological distress receiving an intervention. A ‘screen and intervene as indicated’ approach would probably be more cost-effective and allow a staged delivery of care.
In conclusion, the present study suggests that men with early localized prostate cancer have low levels of psychopathology overall. Future research should include more men and be prospective, following them for as long as possible from diagnosis to determine the natural course of any psychological response. It would also be useful to determine the effectiveness of interventions (e.g. brief psychological and pharmacological interventions) for men with early distressing psychological symptoms. Such research should result in the development of clear guidelines on the optimal methods of detecting and managing men with prostate cancer and mental health difficulties.