Evaluation of an instrument to assess the needs of men diagnosed with prostate carcinoma

An assessment of the validity and reliability of a self-administered questionnaire developed to measure the needs experienced by men diagnosed with prostate carcinoma

Authors

  • Janine M. Duke Ph.D.,

    Corresponding author
    1. Centre for Clinical Epidemiology and Biostatistics, The University of Newcastle, Royal Newcastle Hospital, Newcastle, Australia
    • Centre for Clinical Epidemiology and Biostatistics, The University of Newcastle, 3rd Floor DMB, Royal Newcastle Hospital, Newcastle, NSW, 2300 Australia
    Search for more papers by this author
    • Fax: 011-02-49236148

  • Carla J. Treloar Ph.D.,

    1. National Centre for HIV Social Research, University of New South Wales, Sydney, Australia
    Search for more papers by this author
  • Julie E. Byles Ph.D.

    1. Centre for Clinical Epidemiology and Biostatistics, The University of Newcastle, Royal Newcastle Hospital, Newcastle, Australia
    Search for more papers by this author

Abstract

BACKGROUND

This study was conducted to evaluate a self-administered questionnaire developed to measure the needs experienced by men diagnosed with prostate carcinoma (the Prostate Cancer Needs Questionnaire [PCNQv1.1]). The PCNQv1.1 was constructed in two parts. Part 1 measures the needs at diagnosis and initial treatment and Part 2 measures current needs.

METHODS

A random sample of 650 men diagnosed with prostate carcinoma who were being treated by a urologist of the Hunter Urology Group in Newcastle, New South Wales, Australia, were invited to participate in the study. They were sent by post the self-administered PCNQv1.1. Completed questionnaires were received from 385 men. Information provided on 100 additional men was considered ineligible.

RESULTS

The principal components method of factor analysis with varimax orthogonal rotation identified eight factors with eigenvalues greater than 1, which accounted for 61.7% of the variance in Part 1 of the PCNQv1.1. Likewise, six factors were identified in Part 2, which accounted for 63.5% of the variance. Internal reliability coefficients (Cronbach alpha) were adequate for identified factors with values ranging from 0.70 to 0.88 for Part 1 and from 0.71 to 0.87 for Part 2. Test-retest reliability was acceptable with intraclass correlation coefficients ranging from 0.61 to 0.78 for Part 1 and from 0.60 to 0.82 for Part 2.

CONCLUSIONS

These results suggest that the PCNQv1.1 is a valid and reliable instrument to assess the needs experienced by men diagnosed with prostate carcinoma. Cancer 2003;97:993–1001. © 2003 American Cancer Society.

DOI 10.1002/cncr.11156

Prostate carcinoma is characteristically a disease of older men and is a growing concern in Western societies as longevity increases. According to 1996 Australian national statistics, prostate carcinoma was the most common registrable cancer: 10,055 new cases were reported and and 2644 deaths were recorded, making prostate carcinoma the second leading cause of cancer death among older men.1 Although prostate carcinoma is found rarely among men younger than 50 years old, it is the most common form of cancer among men 55 years and older.2 Due to the rising incidence of prostate carcinoma, the experience and burden of living with this disease are increasingly prevalent. Men with either localized or advanced prostate carcinoma face a range of treatment options associated with troublesome side effects and functional impairments.

Three primary strategies have been used to study the morbidity outcomes for cancer patients: patient satisfaction, quality of life, and needs assessments. Quality of life surveys have been valuable in identifying consequences of cancer diagnosis and treatment for different domains of life experiences,3–6 whereas patient satisfaction surveys have focused more closely on perceived quality of care issues.7 These measures do not directly assess or indicate the patient's need for help, services, or resources.8–14

Needs assessment strategies offer a range of utilities. First, they provide a direct assessment of patient's perceived need for help, allowing a more direct indication of resources required. Second, they identify the magnitude of help needed, allowing priority direction of services or resources. Third, they identify patients or subgroups of patients most in need of help.11–13, 15

The Cancer Patient Needs Questionnaire11 and the Supportive Care Needs Survey13, 16 have provided useful insights into the global needs experienced by both male and female cancer patients in New South Wales, Australia, across the domains of psychological needs, information needs, physical and daily living, patient care and support needs, interpersonal communication needs, and sexuality. The findings of these studies are consistent with other research that have addressed the needs of cancer patients.9, 14

A MEDLINE search of the literature published between 1990 and 1998 failed to reveal any published needs assessment surveys specific for men diagnosed with prostate carcinoma. A needs assessment survey (the Prostate Cancer Needs Questionnaire [PCNQv1.0]) was developed based on results of a qualitative study, conducted with men diagnosed with prostate carcinoma. Items related to issues or concerns associated with initial diagnosis, treatment decisions, and initial and current treatment experiences. Items also covered issues relating to information, choices and decisions, preparation for treatment, recovery and general well-being, impotence and relationships, incontinence, emotional well-being, communication, and support.

To determine the usefulness of any developed instrument, the dual issues of validity and reliability have to be established.17–19 Validity refers to the degree to which an instrument measures the intended outcome.20–22 Measures of of validity include face validity, content validity, and construct validity. Face validity refers to whether the selected items measure the intended outcome.22 Content validity relates to the extent to which an instrument samples items from the full breadth of the content desired. This measure can be enhanced by incorporating the perspectives of those who have experienced similar problems.22, 23 Construct validity refers to the extent to which an instrument measures the abstract concepts or constructs underlying the instrument's factors or subscales.22–24

The reliability of an instrument refers to the extent to which it measures an outcome in a reproducible and consistent way. Reliability also takes into account random error.20–22 A commonly used method of assessing reliability is an index of internal consistency. We used Cronbach alpha as the index of internal consistency.25 We also used the intraclass correlation coefficient (ICC), an index of test-retest reliability, which refers to the reliability or consistency of the instrument administered at two different times.26

This article describes the development and psychometric evaluation of a self-administered instrument designed to measure the needs experienced by men diagnosed with prostate carcinoma. Specifically, the goals of this study were to 1) identify items that could be deleted from the instrument, 2) assess the construct validity of the factors or subscales, and 3) establish the reliability (internal consistency and test-retest reliability) of the subscales.

MATERIALS AND METHODS

Questionnaire Development and Pretesting

Information obtained from a qualitative study previously conducted to explore concerns and needs experienced by men diagnosed with prostate carcinoma was used to inform the development of a questionnaire to measure needs experienced by this specific group of men. Items were developed to address the themes and issues that emerged from the qualitative study using language appropriate to the target group and, finally, arranged in the questionnaire based on a time sequence from diagnosis to current experiences. Pretesting was performed according to the suggested methods of Converse and Presser.27 On the results of preliminary pretesting with members of the Newcastle Prostate Cancer Support Group (who had participated in the qualitative study) and expert opinion, the questionnaire was refined to remove repetitive items. Reducing the number of items enhanced questionnaire completion rates. Final pretesting was conducted with eight men who had participated in the qualitative study and were asked to comment on the relevance, comprehensiveness, and acceptability of the items in the final version of the PCNQv1.0. A small group of health professionals involved in the care of men diagnosed with prostate carcinoma, including a urology clinic nurse, two surgical nurses, two general practitioners, and a urologist, were asked to comment on whether the instrument was an accurate measure that covered the issues confronting this group of patients.

The PCNQv1.0 was presented in two parts. Part 1 related to past issues and experiences and Part 2 related to current issues and experiences: Part 1a (Past Issues…Time at diagnosis) included 20 items; Part 1b (Past Issues…Time of treatment decisions) included 19 items; Part 1c (Past Issues…Initial treatment experiences) comprised 10 items; and Part 2 contained items related to current issues (How do you feel now?). The PCNQv1.0 was presented as Part 1 (Part 1a, Part 1b, Part 1c) and Part 2 of the questionnaire, “A Survey for Men About Prostate Cancer.” Items were framed as statements of issues or problems requiring two responses. For example, the first response was indicated on the alternatives “Strongly Disagree” to “Strongly Agree” using a five-point Likert scale, enabling an indication of whether the item presented a problem or not. The second response for each item, which allowed the participant to indicate a level of help that would be/have been desirable with this problem (“None at all,” “A Little,” “Some,” “Quite a bit,” “A lot”), provided a direct index of perceived need. Part 3 of “A Survey for Men About Prostate Cancer” contained the validated quality of life instrument EORTC QLQ-C30 (v2).24 The EORTC QLQ-C30 (v2) was developed for use with cancer patients by the European Organization for Research and Treatment of Cancer and was included to assess the construct validity of the PCNQv1.0.

Sample and Sampling Procedure

The Hunter Urology Group comprises seven urologists who care for patients in the Newcastle and Hunter Valley areas of New South Wales, Australia, in both the private and public health systems. All members of the Hunter Urology Group participated in this study. The sampling frame for this study consisted of men who had been diagnosed with prostate carcinoma and were patients being treated by a participating urologist in the Hunter Urology Group. Men who had been diagnosed with prostate carcinoma and had visited their urologist during the period January 1996 to July 1998 were eligible to be included in the sampling frame. Men excluded from this sampling frame included those who had been diagnosed for less than 3 months or were considered by the urologist to be too ill to participate.

For each urologist, a patient was assigned a computer-generated number. A list of patients for each urologist was ordered using these random numbers. This sampling frame consisted of 1633 eligible men. A random sample of 650 was selected. They were received a package from the urologists on behalf of the research team that contained the questionnaire, “A Survey for Men About Prostate Cancer,” an information letter informing the men of the purpose and confidentiality of the study, an invitation to participate, and a reply-paid envelope for return of the completed questionnaire by consenting participants.

Sample Size

The sample size of 650 was based on an expected response rate of 70%, which would allow a case-to-item ratio of approximately 10:1 for both Part 1(Parts 1a, 1b, 1c) and Part 2 of the questionnaire for factor analysis procedures. This sample size was sufficient to test the reliability of the instrument to within ±5% (ICC = 0.80) and to test the internal consistency to within ±5% (Cronbach alpha = 0.70).

Sample sizes were also given, which enabled exploratory analyses to be conducted. This sample size allowed detection of moderate differences in study factors. We used the chi-square test for goodness-of-fit at the 5% level of significance with at least 80% power for all categories of study factors considered.28

Statistical Analyses

Statistical analyses were applied to Part 1 (Past Issues) and Part 2 (Current Issues). A combination of item analysis procedures was applied to the data to identify items that would be considered for deletion from both Part 1 and Part 2 of the PCNQv1.0. These item analysis procedures included assessment of the frequency of endorsement, factor analysis, homogeneity of the items, and reliability testing. Principal components factor analysis with orthogonal rotation was used to identify underlying subscales.29 Subscales produced by factor analysis were assessed for internal consistency using the Cronbach alpha coefficient.25 The ICC was the most appropriate method to assess test-retest reliability.26, 30, 31 Mean factor scores derived from selected subscales of the PCNQv1.0 were correlated with mean factor scores derived from the subscales of the validated quality of life instrument, EORTC QLQ-C30 (v2).24 A hypothetical association was expected. The correlation between the mean factor scores of the PCNQv1.0 and the EORTC QLQ-C30 (v2) was used to assess construct validity. The Spearman rank correlation coefficient was calculated as the measure of association.32 All item and factor analysis procedures were performed using STATA statistical software.33

RESULTS

Pretesting

Members of the Newcastle Prostate Cancer Support Group and the group of health professionals involved in the final pretesting phase agreed that the PCNQv1.0 measured the intended outcome and covered the issues confronting this group of men. During the pretesting phases of the PCNQv1.0, preliminary evidence for face and content validity was confirmed, i.e., the questionnaire measured the perceived needs experienced by men diagnosed with prostate carcinoma and included a range of items to appropriately cover domains of concern, respectively. The PCNQv1.0 was well understood by the target population and was suitable for self-administration.

Sample

Of the original sample of 650, completed questionnaires were received from 385 men who consented to participate in the study (59%). Information was provided to researchers on another 100 men who were sent questionnaires. These men were ineligible to participate due to death (n = 33), stroke/heart problems (n = 5), they were “too ill” (n = 8), wrong address (n = 27), dementia (n = 5), confused (n = 8), they were “not diagnosed with prostate cancer” (n = 13), and overseas holiday (n = 1). Discounting these ineligible participants, a response rate of 70% was obtained. Although ethics approval was granted for use of nonidentifiable group data on disease and demographic data from urologists, this information was not provided. Table 1 displays the demographic and disease characteristics of participants.

Table 1. Sample Demographics and Treatment Characteristics
CharacteristicsStudy sample
No. (%)
  • CAE: College of Advanced Education; SD: standard deviation.

  • a

    Differences in sample number are due to missing data.

Time since diagnosis (n = 374)a 
 Less than 2 yrs ago82 (22)
 2–3 yrs ago68 (18)
 3–4 yrs ago66 (18)
 4–5 yrs ago68 (18)
 More than 5 yrs ago90 (24)
Stage of cancer at diagnosis (n = 373)a 
 Localized217 (58)
 Advanced64 (17)
 Do not know92 (25)
Symptoms at diagnosis (n = 377)a 
 Yes205 (54)
 No172 (46)
Cancer spread since diagnosis (n = 377)a 
 Yes39 (10)
 No338 (90)
Treatment for prostate carcinoma (n = 331)a 
 Watchful waiting only39 (12)
 Radiotherapy only108 (33)
 Hormone only60 (18)
 Surgery only69 (21)
 Radiotherapy and hormone34 (10)
 Surgery and hormone21 (6)
Marital status (n = 382)a 
 Married321 (84)
 Not married61 (16)
Education (n = 345)a 
 < 3 yrs high school109 (32)
 3–4 yrs high school102 (30)
 5–6 yrs high school26 (8)
 Technical college76 (22)
 C.A.E./university32 (9)
Employment status now (n = 321)a 
 Employed28 (9)
 Not employed293 (91)
Location (n = 354)a 
 Urban275 (78)
 Rural79 (22)
Age (n = 381)a 
 Mean: 72 yrs (SD 7.8)b 
 maximum–minimum: 43–93 yrs 

Missing data

The data were checked for missing values and it was found that 230 respondents had no missing data. The means for overall response to each item were calculated from the available data and imputed for missing values before analysis34, 35 when a participant had missing responses to five or fewer items of the 95 items, including Part (a) and (b) of each response (≤ 2.6% [5 of 190] missing data). This resulted in a sample size of 320 complete observations available for factor analysis procedures. This sample size of 320 was adequate to conduct factor analytic procedures36, 37 and offered a case-to-item ratio of approximately 6:1 for Part 1 and a case-to-item ratio of approximately 7:1 for Part 2 of the questionnaire. As an added check, the factor analyses were performed on the smaller sample of 230 with no imputed values for missing data. The factor structures for Part 1 and Part 2, using both methods, was similar with only minor differences in factor loadings.

Item Deletion

Four items relating to family support and family needs and eight items that cross-loaded significantly (cross-loading of over 0.30)38 were deleted from Part 1 of the PCNQv1.0. Fourteen items were deleted from Part 2 of the PCNQv1.0 due to significant cross-loading as were three items that related to sexual problems and usefulness of impotence clinics. Two additional items relating to spiritual life, for which internal consistency could not be established, were also deleted. As a result of item and factor analysis procedures, 31 items were deleted from the PCNQv1.0. Therefore, the refined PCNQv1.1 contained 64 items. Items in Parts 1 and 2 of the PCNQv1.1 are displayed in Tables 2 and 3, respectively.

Table 2. PCNQv1.1 Questionnaire Items in Part I (Past Issues)
Item
  1. PCNQv1.1: Prostate Cancer Needs Questionnaire; GP: general practitioner.

Part 1a At the time of diagnosis…
 I knew very little about the prostate and what the prostate did in my body.
 I knew very little about prostate cancer.
 The GP was very supportive.
 The specialist was very supportive.
 I was given a clear explanation of what my diagnosis meant.
 I felt that I could call my GP at any time to discuss my diagnosis.
 I felt that I could call my specialist at any time to discuss my diagnosis.
 I had difficulty understanding what all the tests and results meant.
 When I was told that I had prostate cancer, I just wanted the cancer out of my body as soon as possible.
 When I was told that I had prostate cancer, I was very distressed and anxious.
 Waiting to have further tests and to be given the test results was an extremely anxious time.
 When I was told that I had prostate cancer, I felt very alone and needed someone to talk to.
 I needed to discuss my feelings with other health professionals.
 I needed to be involved with a support group.
Part 1b Treatment decisions for prostate cancer…
 I was given a clear explanation about available treatment choices.
 After I spoke with my specialist(s), I understood all the treatment options.
 I understood the risks involved with each treatment.
 I knew what I could expect if I did not have the treatment for prostate cancer.
 I felt very concerned at the time to “treat and cure” the cancer as fast as possible.
 I understood the real chance of suffering incontinence after surgical treatment or radiotherapy.
 I understood the high likelihood of being impotent after treatment for prostate cancer.
 I felt very comfortable to discuss any problems or questions that I had with my specialist.
 I would have liked to be able to discuss my concerns with men who had been in a similar situation.
 I felt very anxious during the time my treatment decision was being made.
 I felt my GP understood how I was feeling.
 I felt my specialist understood how I was feeling.
 I felt I could ask as many questions as I felt necessary.
 I needed to get information from other sources other than my doctor.
 I feel that I was well prepared for the actual procedure of the treatment I was to have.
Part 1c Initial treatment experiences…
 I felt I was well prepared for what to expect immediately after the treatment procedure.
 I was prepared well for what to expect when I was discharged from hospital.
 When I was discharged from hospital, I felt very confident that I had reliable professional backup if I needed it.
 My partner/family was well prepared for what to expect when I was discharged from hospital.
 I felt certain that I would make a full recovery from the cancer.
 I felt certain that I would eventually make a full recovery from incontinence.
 I felt certain that I would make a full recovery from impotence.
 I understood what I could expect in my future after the completion of my treatment.
Table 3. PCNQv1.1 Questionnaire Items in Part 2 (Current Issues)
Item
  1. PCNQv1.1: Prostate Cancer Needs Questionnaire; GP: general practitioner.

Part 2 How do you feel now?
 I have time when I feel very alone and need someone to talk to.
 It is not easy for me to talk with friends and family about how I am feeling.
 My GP and specialists communicate well with each other regarding my health.
 My GP is very supportive.
 My GP communicates well with me about my health.
 My specialist(s) are very supportive.
 My specialist(s) communicate well with me.
 My GP is a useful source of information.
 My GP is helpful in managing my impotence.
 My GP is helpful in managing my incontinence.
 Sometimes I feel there is no let up from the consequences of the treatment for prostate cancer.
 I worry about whether the cancer is going to spread.
 I expected that I would feel a lot better than I do.
 I feel embarrassed about being incontinent.
 I find it difficult to discuss my incontinence with my friends.
 There is a lack of spontaneity in my daily life because of incontinence.
 I feel that impotence has affected the closeness that I can share with my partner.
 I feel that my impotence has put a strain on the relationship with my partner.
 I feel that my relationship with my partner has changed.
 I find myself unable to enjoy my previous level of social life.
 I feel generally fatigued and lack motivation to do anything.
 Working has become difficult for me.
 I feel that I have lost some self-esteem.
 At times I feel very down.
 I feel that my life has been altered beyond my control.
 I feel that I am living a full and active life.
 I feel positive about the future.

Validity of the PCNQv1.1

Factor analysis

Factor analysis is used in construct validation to indicate the association among several measurements. Items measuring the same topic would be expected to be grouped together by factor analysis onto the same factor, i.e., a test of convergent validity.18, 19 Principal components factor analysis with varimax orthogonal rotation on the first or “a” response to each item identified eight factors that accounted for 61.7% of the variance in Part 1 of the PCNQv1.1 and six factors that accounted for 63.5% of the variance in Part 2 of the PCNQv1.1, with eigenvalues greater than 1. The factors in Part 1 of the PCNQv1.1 (and the proportion of the total variance explained by each) were identified as follows: 1) specialist support for diagnosis and treatment (24.2%) related to the support and understanding provided by the specialist; 2) anxiety and urgency (11.1%) related to the anxiety and distress associated with the initial diagnosis and concern for fast treatment decisions; 3) confidence in treatment preparation (5.5%) related to treatment/procedure preparation and immediate expectations after; 4) knowledge of prostate carcinoma, tests (5%) related to a lack of knowledge and understanding; 5) confidence in treatment recovery (4.1%) related to the perceived confidence in treatment recovery; 6) general practitioner (GP) support for diagnosis and treatment (4.1%) related to GP support and understanding at the time of diagnosis; 7) informed choice (3.6%) related to the provision and explanation of information regarding treatment choices, risks, and expectations of the future after completion of treatment; and 8) other support for diagnosis and treatment (3.5%) related to the support and information from sources other than the GP and specialist at the time of diagnosis and treatment. Factors 3 and 5 represented themes relating to different aspects of confidence in treatment. For the purposes of this questionnaire, these two factors were combined. The internal consistency of the combined factor was assessed and found to be acceptable.

The factors identified in Part 2 of the PCNQv1.1 (and the proportion of the total variance explained by each) were identified as follows: 1) role limitations (28.3%) related to the general fatigue and impact of cancer on social, daily, and work life; 2) GP ongoing support (12.6%) related to GP support, information, and communication; 3) impotence (7.7%) related to the impact on relationships; 4) incontinence (5.5%) related to the embarrassment and change in daily life; 5) personal integration and control (4.9%) related to feeling alone, depression, pervasiveness of cancer in one's life, and lack of control; and 6) specialist ongoing support (2.6%) related to the support and communication offered by the specialist in ongoing care.

Correlations with the EORTC QLQ-C30 (v2)

After evaluating the content and meaning of the subscales of the EORTC QLQ-C30 (v2) and the PCNQv1.1, Spearman rank correlation coefficients of greater than or equal to 0.60 (or −0.60) were expected for hypothesized subscale correlations (Table 4). Despite only one correlation coefficient exceeding 0.60 (or −0.60), all Spearman rank correlation coefficients were statistically significantly at the 5% level of significance.

Table 4. Correlations between the EORTCQLQ-C30 (v2) and PCNQv1.1 PCNQv1.1 Subscales
FactorEORTC Global health statusEORTC Role functioningEORTC Emotional functioningEORTC Social functioningEORTC FatigueEORTC Physical functioning
  • EORTC: European Organization for Research and Treatment of Cancer; PCNQv1.1: Prostate Cancer Needs Questionnaire;

  • a

    Number of observations.

Part1 F2  0.31 (n = 362)a   
 Anxiety and urgency      
Part2 F1−0.63 (n = 361)a0.55 (n = 360)a 0.52 (n = 361)a0.52 (n = 361)a0.52 (n = 361)a
 Role limitations      
Part2 F3   0.34 (n = 235)a  
 Impotence      
Part2 F4  0.30 (n = 160)a   
 Incontinence      
Part2 F5−0.50 (n = 359)a 0.55 (n = 359)a   
 Personal and integration control      

Reliability

Internal consistency

The Cronbach alpha measure was used to assess the internal consistency of the items comprising each factor. Most psychometricians agree that the value of Cronbach alpha should be greater than 0.8018, 19 or that it should be in the range 0.70–0.80.39 For this study, a value greater than 0.80 was desirable and a score greater than 0.70 was adequate. As shown in Tables 5 and 6, the reliability coefficients ranged from 0.70 to 0.88 for factors in Part 1 of the PCNQv1.1 (Past Experiences) and from 0.71 to 0.87 for factors in Part 2 of the PCNQv1.1 (Current Experiences).

Table 5. Factors in Part I of the PCNQv1.1
FactorDomain (no. of items)Variance (%)Cronbach alpha coefficient
  1. PCNQv1.1: Prostate Cancer Needs Questionnaire; GP: general practitioner.

1Specialist support for diagnosis and treatment (6)24.20.88
2Anxiety and urgency (6)11.10.80
3 and 5Confidence in treatment preparation and recovery (8)9.60.76
 (combined)  0.69
4Knowledge of prostate, cancer tests (3)5.00.61
6GP support for diagnosis and treatment (3)4.10.86
7Informed choice (7)3.60.70
8Other support for diagnosis and treatment (4)3.50.76
Table 6. Factors in Part 2 of the PCNQv1.1
FactorDomain (no. of items)Variance (%)Cronbach alpha coefficient
  1. PCNQv1.1: Prostate Cancer Needs Questionnaire; GP: general practitioner.

1Role limitations (5)28.30.84
2GP ongoing support (3)12.60.87
3Impotence (4)7.70.75
4Incontinence (4)5.50.71
5Personal integration and control (8)4.90.85
6Specialist ongoing support (3)2.60.83

Test-retest reliability

Analyses were conducted using completed data for 101 of 150 respondents (a response rate of 67.3%) who had been posted a second copy of the questionnaire 2 weeks after completion and return of the first. The ICCs were calculated for both the problem and the corresponding need scores for each subscale. As shown in Tables 7 and 8, the ICCs were acceptable. Values ranged from 0.61 to 0.78 for Part 1 and from 0.60 to 0.82 for Part 2.

Table 7. Intraclass Correlation Coefficients for Subscales for Part 1 of the PCNQv1.1 for Area of Concern and Corresponding Need Scores
SubscalesFactor 1 Specialist supportFactor 2 Anxiety and urgencyFactor 3 Combined 3 and 5 Confidence, treatment, and recoveryFactor 4 Knowledge, prostate cancer testsFactor 6 GP supportFactor 7 Informed choiceFactor 8 Other support
  1. PCNQv1.1: Prostate Cancer Needs Questionnaire; GP: general practitioner.

Area of concern0.760.760.680.610.720.800.71
Need0.710.740.770.660.620.630.78
Table 8. Intraclass Correlation Coefficients for Subscales for Part 2 of the PCNQv1.1 for Area of Concern and Corresponding Need Scores
SubscalesFactor 1 Role limitationsFactor 2 GP ongoing supportFactor 3 ImpotenceFactor 4 IncontinenceFactor 5 Personal integration and controlFactor 6 Specialist ongoing support
  1. PCNQv1.1: Prostate Cancer Needs Questionnaire; GP: general practitioner.

Area of concern0.820.710.600.640.770.78
Need0.760.630.600.650.760.67

DISCUSSION

Results of this study provide empiric evidence of the reliability, validity, and acceptability of the PCNQv1.1 to assess the needs of men diagnosed with prostate carcinoma. The self-administered questionnaire, which takes approximately 30 minutes to complete (15 minutes for each part), provided a means for men to indicate directly their need requirements using a categorical response scale.

We used factor analysis procedures to demonstrate statistically the different domains of need measured in the two parts of the questionnaire, i.e., Past Issues (which relate to the diagnosis and initial treatments) and Current Issues. The demonstrated validity and reliability of both Parts 1 (Past Issues) and 2 (Current Issues) of the PCNQv1.1 highlight the versatility of the instrument to be used either together to collect both historic and current need information or individually. Exploratory factor analysis identified seven factors in Part 1 (specialist support, anxiety and urgency, knowledge of prostate carcinoma and tests, GP support, confidence in treatment, support from other sources) and six factors in Part 2 (role limitations, GP ongoing support, impotence, incontinence, personal integration and control, and specialist ongoing support). These factors are consistent with past research that identified the importance to cancer patients of emotional well-being, physical and daily living, communication, information, and preparation for treatments.9, 11, 13, 40–43 Likewise, other factors identified correspond with published research, which suggests that impotence and the effect on relationships and incontinence are of major concern to men diagnosed with prostate carcinoma.44, 45

Throughout the questionnaire development phase, care was taken to incorporate language used by the men in their communication of concerns and needs and to construct the questionnaire on a time sequence of events from diagnosis, treatment decisions, and initial experiences to current issues. The format used in the PCNQv1.1 was an important dimension of the need questionnaire. Definitions of perceived or felt needs have been expressed as “…problems as viewed or understood by the people who experience them” (Windsor et al.,22 p 64) and as “what people in your community say they want or what they think are problems that need addressing” (Hawe et al.,46 p19). Therefore, the PCNQv1.1 provided an opportunity for participants to identify issues that were not necessarily perceived to be problems to them, for which a need for help might exist.

Although 17% of the study sample had known advanced prostate carcinoma, men who were too sick to participate were not included in this study. Likewise, the concerns and needs of men in the terminal stage of their disease were not measured. Further research is required to explore the concerns and needs of these men.

The acceptability of the PCNQv1.1 as a self-administered instrument offsets the need for more time-consuming and expensive personal interviews7, 47 and provides a more practical means of assessing needs on an ongoing basis. The PCNQv1.1 has been used to identify areas in which a high prevalence of need exists48 and to evaluate the change in need in a 12-month follow-up study. Similarly, using group data for men diagnosed with prostate carcinoma, with respect to stage, treatment, and sociodemographics, the PCNQv1.1 has been used to identify subgroups at risk of developing specific needs.48 Not all men diagnosed with prostate carcinoma may need or want referral for further assistance or specific interventions. However, results indicate that many men would like to have an opportunity for more open communication and acknowledgment of their health concerns with their practicing clinician. The PCNQv1.1 allows clinicians and health service providers a better opportunity of understanding and meeting the needs of men diagnosed with prostate carcinoma.

Ancillary