The development and validation of a quality-of-life measure to assess partner morbidity in benign prostatic enlargement
Mr Henry Sells, Department of Urology, Treliske Hospital, Truro TR1 3LJ, UK.
Objective To investigate morbidity in the partners of patients with benign prostatic enlargement (BPE) by developing and validating a disease-specific questionnaire.
Patients and methods Phase 1 of the study comprised preliminary interviews with 15 patients who had newly diagnosed BPE, and with their partners, to determine the relevant issues for the partners. In phase 2, using these issues, a questionnaire was produced and tested on the 15 partners. In phase 3 the questionnaire was completed by 90 further partners, 50 at interview and the next 40 by post. As part of the validation process, the partners were also asked to complete the Short-Form-36 questionnaire, and the patients the Internation Prostate Symptom Score (IPSS) and the ICSmale questionnaires.
Results In phase 1 all 15 of the partners were affected by the patients’ disease and nine issues were identified. In phase 2, of the 90 partners, only one had no morbidity from the patient’s symptoms; 71% were worried that the patient may have cancer and 69% concerned that the patient may require an operation. Only six partners were present at the patients’ urological consultation. The partners’ questionnaire scores were related significantly to the Mental Health and Vitality domains of the SF36 and with the patients’ IPSS.
Conclusions The study confirmed the presence of significant morbidity in the partners of patients with BPE. The degree of partner morbidity was related to the severity of the patients’ symptoms. Many of the questionnaire issues can be addressed in the consulting room by open discussion with patient and partner.
The outcome of treatment for BPE has traditionally been measured using physical variables and symptom scores. Recently, studies have assessed the effect of BPH and its treatments on quality of life [1–7]. However, no previous work has focused on the partners’ attitudes towards this condition, although it seems reasonable to suppose that many of the classic symptoms of BPE may have some effect on the patients’ partner. For example, there is published evidence that other conditions such as multiple sclerosis, spinal injuries and prostate cancer adversely affect the partner and that in many cases partners suffered greater psychological stress than patients [8–10].
Partner morbidity can be measured by using modified patient questionnaires [9,11]. This approach is limited by focusing on factors likely to be important to patients and ignoring other issues that have an impact on partners’ lives. The aim of the present study was to develop a quality-of-life (QoL) questionnaire specifically for partners of patients with BPE, including an assessment of its psychometric performance, and to explore partner morbidity in this condition.
Patients and methods
The construction of QoL measures is a structured and complex process and demands strict, well-defined psychometric assessment . It is important to determine what health-related QoL issues are relevant to the population being studied by carrying out semi-structured qualitative interviews in association with a thorough literature search and advice from appropriate healthcare professionals (Phase 1). Once established, the issues can then be ‘operationalized’ to produce an initial questionnaire, which is further refined to ensure that the questions are clinically relevant, clear, unambiguous and inoffensive (Phase 2). This questionnaire should then be field-tested in a larger population to establish its psychometric characteristics, in particular validity and reliability (Phase 3).
The issues in Phase 1 were established in three ways. The literature was searched to determine which issues were relevant to partners of patients with other chronic diseases. Urologists and allied health professionals were interviewed to determine the BPE symptoms that were perceived to be relevant to the partners. Seventeen patients with symptomatic BPE, having seen a urologist and requiring medical or surgical treatment, were asked whether they and their partners would be willing to take part in preliminary interviews about the effect of prostatic symptoms; 15 couples agreed. The interviews were semi-structured and designed to determine the degree to which the partners were affected by the patients’ symptoms. The interviews were transcribed, analysed by two urologists, and a list of partner issues produced.
In Phase 2, using these issues, two provisional questionnaires were devised; one with nominal answer scales (Yes/No) and one with ordinal answer scales (ordered categories). The ordinal scale allowed each question to be scored (No = 1, A little = 2, Somewhat = 3, Quite a lot = 4, A great deal = 5) and for a questionnaire total score to be calculated. The two provisional questionnaires were submitted to each of the original 15 partners to determine which questionnaire was easier to complete and to ensure that the questions were clinically relevant, clear, unambiguous, inoffensive and not duplicated. As well as the questions related to these issues, an overall impact question was included, similar to that used in the IPSS  and ICSmale[14,15] instruments. Partners were also asked whether they had been present during the patients’ consultation.
In Phase 3, the refined questionnaire (Appendix 1) was then completed by a further 90 partners. The first 50 were completed during interviews with the couples conducted by one of the authors (H.S.). Each partner was asked to voice any apparent difficulties or faults in the questions. Illiterate partners (two) had the questions and answer options read to them. When it was clear that the questionnaire could be self-completed by partners, 40 further couples were sent the questionnaire by post for completion.
The questionnaire was tested psychometrically as follows. Content validity tests whether the questionnaire measures what it is supposed to measure, ensures that questions address relevant issues and that they are clear and unambiguous. These aspects were addressed by conducting interviews to establish the issues before devising the questionnaire and then by requesting feedback from the first 50 partners who completed the questionnaire.
Construct validity tests the questionnaire against expected or reasonable hypotheses. We hypothesized that partner morbidity would correlate with the severity of patients’ symptoms. To allow the present questionnaire to be compared against this hypothesis, the IPSS of the patients when seen at clinic were recorded and the ICSmale[14,15] and ICSsex questionnaire were given to patients during Phase 3. Areas of expected correlation between patient and partner questions assessing similar issues were targeted in the statistical analysis.
Criterion validity compares the questionnaire against a similar, previously validated instrument or ‘gold standard’. In this study, as partner morbidity has not previously been addressed, there were no specific measures available and hence in Phase 3 the new questionnaire was compared against the Short-Form 36 (SF36), a generic QoL questionnaire containing eight broad health dimensions (bodily pain, general health, mental health, physical function, role emotional, role physical, social function, vitality). This measure has been used previously to assess patients with BPE [17,18]. In response to the SF36 results, the final 20 patients completed the Hospital Anxiety and Depression (HAD) questionnaire. Pearson correlations were used with two-tailed P-values. The values should be seen as indicative, as no correction was made for multiple testing.
Internal consistency is a measure of how the individual questions correlate with one another and ensures that none of the individual question scores differ enough from the others that they skew the total questionnaire scores. This is assessed by computerized correlation calculations resulting in a ‘coefficient α’, which approaches unity when the questions correlate satisfactorily.
In Phase 1, from the 15 couples who took part in the preliminary interviews, all of the partners were affected by at least one aspect of the patient’s prostatic disease. The 15 interviews yielded nine different issues (Table 1). The most common problem for the partners was fatigue during the day after being woken up at night by the patients going to the toilet.
Table 1. The partner issues identified in Phase 1 (n = 15), the percentage of partners affected by the issues (Phase 2, 90 couples) and the mean score for each question
| Phase 1 |
|Worry that patient needs an operation|| 1|| |
|Sex life adversely affected|| 1|| |
|Depressed|| 2|| |
|Difficulty going out of house|| 3|| |
|Pity for patient|| 3|| |
|Woken during the night by patient|| 4|| |
|Social life affected|| 4|| |
|Worry about patient having cancer|| 6|| |
|Fatigue through being woken by patient at night|| 8|| |
| Phase 2 and 3 |
|1 How many times does your husband wake you up at night?||68 (76)||2.87|
|2 Do you get tired because of being woken at night?||38 (42)||1.81|
|3 Is your social life affected by your husband’s symptoms?||42 (47)||1.81|
|4 Do your husband’s symptoms make it difficult for you to do essential tasks?||24 (27)||1.46|
|5 Do you get upset by the distress that your husband suffers because of his symptoms?||59 (66)||2.11|
|6 Have you noticed a worsening in your sex life?||59 (66)||2.04|
|7 Do you worry that your husband’s symptoms may be caused by cancer?||64 (71)||2.37|
|8 Are you worried about the possibility of your husband needing an operation?||59 (66)||2.22|
In Phase 2, the nine-item questionnaires consisted of eight items relating to issues established in Phase 1. The remaining issue (partners feeling depressed by the patients’ symptoms) was addressed by the overall impact question. Of the two questionnaires devised using these issues, partners preferred the ordinal to the binary answer scale (Appendix 1), which was therefore adopted for use in Phase 3. Question 1 had different answer categories and was therefore scored differently (Not at all = 1, Less than once = 2, Once = 3, 2–3 times = 4, 4 times = 5). None of the questions required alteration during the study.
In Phase 3, of the 90 partners who completed the questionnaire, only one was not adversely affected at all by the patient’s symptoms. Most partners were woken up more than once a night and nearly half were tired the next day because of it (Table 1). Nearly three-quarters of partners were worried that the patient may have prostate cancer. Two-thirds of partners were worried about the patient needing an operation or had noticed a worsening in their sex life.
Table 1 also shows the mean scores for each question. The mean score for question 1 indicates that the partners were only woken up, on average, less than once a night. Of the other issues, worry about the patient having cancer (question 7) was the highest scoring and thus the most serious issue for the partners. Fear of the patient needing an operation for his prostate problems (question 8) also scored highly. Of the 90 couples in phase 3, 53% of the patients were awaiting either TURP or laser prostate treatment, the rest being treated pharmacologically. However, the proposed mode of treatment of the patient was not shown to have a significant effect on the scoring of question 8. The mean score for question 9 (overall impact) was 3.98, approximating to the answer ‘mixed equally satisfied and dissatisfied’. Of the 90 patients in phase 3, only six (7%) were accompanied by their partner during the consultation with the urologist.
Content validity. Partners experienced no problems completing the questionnaire and none of the questions was deemed ambiguous or difficult to understand. Four partners who received questionnaires by post did not answer question 6 (sex).
Construct validity. The total partner questionnaire score was related significantly with the patients’ IPSS (Table 2) as predicted by the study hypothesis. Similarly, questions 1 (number of times woken) and 2 (fatigue) of the partner questionnaire correlated significantly with ICSmale question 2 (nocturia; correlation coefficients, CCs, 0.424 and 0.294, P = 0.001 and P = 0.005, respectively). As expected, partner question 6 (sex) was related significantly with ICSsex questions 24 (sex; CC 0.629, P = 0.001), 25 (erection rigidity; CC 0.555, P = 0.001), 26 (ejaculation; CC 0.399, P = 0.001) and 27 (discomfort during ejaculation; CC 0.242, P = 0.021). Partner question 9 (overall impact) correlated significantly with question 33 of the ICSmale (overall symptom impact; CC 0.467, P = 0.001).
Table 2. Correlation of the total partner questionnaire score with the patient IPSS and SF-36 domains
|Total patient IPSS||0.297||0.004|
| Physical function||0.021||0.861|
| Role physical||− 0.105||0.389|
| Bodily pain||− 0.030||0.804|
| General health||− 0.025||0.836|
| Vitality||− 0.263||0.028|
| Social function||− 0.184||0.128|
| Role emotional||− 0.172||0.155|
| Mental health||− 0.410||0.001|
Criterion validity. Seventy partners completed the SF-36 questionnaires. As expected, question 2 (fatigue), related to General Health (CC − 0.243, P = 0.043) and Vitality (CC − 0.466, P = 0.001) domains of the SF-36. Question 3 (social life) related to Social function (CC − 0.264, P = 0.027). Questions 5 (pity) and 7 (worry about cancer) related to Mental Health (CC − 0.462, P = 0.001, and − 0.254, P = 0.034, respectively).
The total partner questionnaire score related to Mental Health and Vitality (Table 2). In view of the significant correlation of the Mental Health domain with total and individual partner questions, the next 20 partners were asked to complete the HAD score. None of the individual questions or total partner scores correlated significantly with the Anxiety or Depression domains. For reliability, the eight questions produced a coefficient α of 0.807.
The questionnaire developed in the present study was tested psychometrically, and was acceptable and relevant to the partners of patients with BPE. Using this measure, we showed for the first time that partners have morbidity as a consequence of the patients’ urinary symptoms and that this morbidity is related to the severity of the patient’s symptoms. Unlike most measures used previously to assess partner morbidity in other diseases, the present questionnaire benefits from being developed and validated specifically for partners.
Almost all partners experienced some morbidity as a consequence of the patients’ condition, with the most common issues being sleep disturbance, fear of cancer and surgery, social disruption and deterioration in sex life. Some of the issues raised by the partners were as expected. Partner fatigue, disrupted social life and difficulty leaving the house can all be related to BPE symptoms such as nocturia, frequency and urgency. However, it was surprising that partners still expressed considerable concerns about cancer and surgery, because all patients had been consulted by a urologist who had diagnosed a benign condition and planned treatment. It is possible that the urologist failed to discuss the issues of cancer and surgery adequately or, in the light of poor partner attendance at the consultation (< 10%), that the patient subsequently failed to relate the information to the partner. Partners may be reluctant to accompany patients at the consultation simply because the patients do not want the partners to be present or there is no specific invitation to partners, leading to an assumption that they are not allowed or not encouraged to attend. We now routinely add a sentence to the outpatient letter explaining that partners are welcome to attend. Further studies will assess whether this has any effect on partner morbidity.
The high frequency of partners whose social life (question 3) was affected was, from the interview, ascribed to the reluctance of patients to go to social events or day trips where frequency, urgency and incontinence cause embarrassment. Furthermore, many partners revealed they were unwilling to attend social activities alone. The many partners noticing a deterioration in their sex lives was consistent with recent findings that BPE symptoms are related to deterioration in all aspects of patient sexual function [16,19]. However, at interview many partners expressed relief that answering the questionnaire had forced the couple to address an issue that had not previously been discussed. This may represent another focus of discussion to be raised actively by the clinician at consultation.
Of the partners, 57% scored 4 or 5 on at least one of the questions, implying that they suffered significant morbidity from at least one of the issues. However, when analysing the absolute scores for each question it appeared that the morbidity relating to most issues was moderate. Only 66% of partners were woken at least once a night and similarly, of the many partners worried about cancer, most were worried either ‘A little’ or ‘Somewhat’, with only 23% worried ‘Quite a lot’ or ‘A great deal’. This illustrates the importance of determining, individually, the relevant issues for each partner, to enable the specific problems to be targeted. The partner questionnaire allowed these issues to be identified easily in an outpatient setting.
Psychometrically, the questionnaire performed well. Content validity was maximized both by undertaking preliminary interviews and using continuous partner feedback during the study. The statistically significant correlation of overall partner scores with the patient IPSS supports our hypothesis that the degree of partner morbidity depends on the severity of patient symptoms. Assuming the hypothesis is valid, this provides construct validity for the questionnaire. This is further established by the correlation of individual partner questions with appropriate patient questions of the ICSmale questionnaire. The relationship between partner question scores and the corresponding domains of the SF-36 suggests further criterion validity for the questionnaire. The absence of the expected statistically significant relationships between the partner and HAD questionnaires may be because there were relatively few patients (n = 20). The satisfactory coefficient α indicates that the measure is consistent but test-retest studies are needed before the questionnaire is used to assess outcome. The performance of the present partner questionnaire in terms of content, construct and criterion validity justifies its use in further clinical studies as a QoL measure for partners of patients with BPH.
The partner questionnaire has only been developed and validated for the partners of patients who required treatment for BPE. It is not yet known how it will perform in the population of partners of men with BPE symptoms who do not warrant treatment or who have not presented to a doctor. If the hypothesis is confirmed in this population, it would provide a greater test of the questionnaire’s sensitivity.
Having detected morbidity in the partners of patients with BPE, we are evaluating methods to improve the partners’ quality of life. As well as inviting the partners to attend the consultation, specific issues, e.g. treatment options, sexual function and cancer, can be targeted for discussion. Using the partner questionnaire before the consultation will allow these relevant questions to be identified and to determine any improvement in the partner’s QoL using these methods.
In summary, the study questionnaire is acceptable to partners, easy to complete unaided and satisfied suitable psychometric criteria. Using this measure, we have identified significant partner morbidity in BPE. However, it remains to be determined whether improving the partners’ QoL reduces the impact of the condition on the patient.
H. Sells, FRCS, Specialist Registrar.
J. Donovan, PhD, Lecturer in Health and Health Care.
P. Ewings, PhD.
R.P. MacDonagh, FRCS, Consultant Urologist, Senior Clinical Lecturer.
Questionnaire for wives of patients with prostate enlargement
This questionnaire is designed to see how much your husband’s symptoms from his enlarged prostate affect you. Please read each question carefully and answer by ticking the appropriate answer.
Please answer according to your situation at the moment.
1. How many times on average does your husband wake you up each night when he gets up to go to the toilet?
Not at all
Less than once
4 times or more
2. Do you get tired during the day because of being woken up at night?
Quite a lot
A great deal
3. Your social life may include seeing friends, going on day trips and your hobbies. Are you limited in these sorts of activities by your husband’s urinary symptoms?
Quite a lot
A great deal
4. Do your husband’s urinary symptoms make it difficult for you to do essential tasks, for example doing the shopping?
Quite a lot
A great deal
5. Do you get upset by the distress that your husband suffers because of his urinary symptoms?
Quite a lot
A great deal
6. Have you noticed a worsening in your sex life since your husband started having his urinary symptoms?
Quite a lot
A great deal
7. Do you worry that your husband’s urinary symptoms may be caused by cancer?
Quite a lot
A great deal
8. Are you worried about the possibility of your husband needing an operation to make him better?
Quite a lot
A great deal
9. How would you feel if you were to spend the rest of your life with your husband’s symptoms the way they are now?
Mixed (equally satisfied and dissatisfied)