Abstract

Objective

This study was conducted to examine the psychometric properties of the Turkish version of the Expanded Prostate Cancer index for Clinical Practice (EPIC-CP).

Materials and Methods

The sample of this study consisted of 80 patients diagnosed with prostate cancer who applied to urology and oncology outpatient clinics. Data were collected between February 2021 and July 2021. The content validity of the scale was evaluated by consulting 11 experts from the field of surgical nursing. The Content Validity index, Explanatory and Confirmatory Factor Analysis were used for validity, while Pearson Correlation Analysis and Cronbach Alpha Coefficient were used for reliability.

Results

Overall prostate cancer quality of life score (minimum 0-maximum 60). The score was calculated as 26.26±9.6. For the validity of the scale, it was determined that the Content Validity index was 1.0 for each item of the scale, the factors used in the Explanatory Factor Analysis explained 74.403% of the total variance, the factor loads of the items were over 0.40 in the Confirmatory Factor Analysis, and all correlation relationships were significant. The total Cronbach’s α value, which shows the reliability and internal consistency of the scale, was determined to be 0.83.

Conclusion

As a result of the statistical evaluations, the Turkish validity and reliability of the EPIC-CP was found to be high. Considering these results, this developed scale can be used successfully for research to be conducted in Turkiye.

Introduction

Prostate cancer is the second most common type of cancer among men worldwide, after lung cancer, and its incidence is increasing (1). According to GLOBACAN 2020 data: 1,414,259 men were diagnosed with prostate cancer and 375,304 men died from prostate cancer. 3.8% of all deaths from cancer in men are due to prostate cancer (2).

Today, due to the increase in population, the more frequent screening of serum prostate-specific antigen (PSA) values in men, the expansion of prostate biopsy indications, and advances in surgical techniques, more patients are diagnosed with prostate cancer. With the increase in early detection of prostate cancer and advances in treatment, the survival rates of patients are also increasing (2, 3).

Although prostate cancer is asymptomatic at an early stage, patients may experience urinary retention, nocturia, hematuria, stool thinning due to rectal compression, painful defecation, weakness, anorexia, and weight loss in the later stages as the cancer progresses. According to the clinical stage of the disease, hormonal therapy, radiotherapy, chemotherapy, cryotherapy, or radical prostatectomy are among the treatment options (4, 5).

Depending on the prostate cancer treatment method and complications, patients are faced with many conditions that will negatively affect their quality of life, such as urinary incontinence, changes in bowel habits, and sexual dysfunction. Each of these situations has negative effects on the perception of quality of life (6). Health-related quality of life is an important parameter in cancer management that enables clinicians to evaluate how treatment side effects affect patients. Validated and reliable measurement tools should be used to assess quality of life (6, 7).

The University of California-Los Angeles Prostate Cancer index (UCLA-PCI) was developed by Litwin et al. (8) to assess the quality of life of patients with prostate cancer worldwide. The Expanded Prostate Cancer index Composite (EPIC), consisting of 50 items, EPIC-26, consisting of 26 items, were developed by Wei et al. (9), inspired by UCLA-PCI. Because these forms of EPIC are difficult to use, as the number of items is high and the application is impractical, the “Expanded Prostate Cancer index Composite for Clinical Practice (EPIC-CP)” consisting of 10 questions and 16 items was created. EPIC-CP is used in many studies evaluating the quality of life of patients with prostate cancer around the world (10).

Recently, a study to examine the Turkish psychometric properties of EPIC-CP, which is used by clinicians and researchers in the evaluation of the quality of life of patients with prostate cancer, has not been conducted. For this reason, this study, which examines the Turkish psychometric properties of EPIC-CP, will be useful in evaluating the quality of life of patients with prostate cancer and in planning care and interventions to improve it.

Materials and Methods

This research was conducted as a methodological study to examine the psychometric properties of the Turkish version of EPIC-CP. The sample of the study consisted of 80 patients who applied to the urology and oncology outpatient clinic of a university hospital between February and July 2021, and were diagnosed with prostate cancer. In the literature, it is recommended to calculate the sample size by taking at least 5 times the number of items in the measurement tool (11). EPIC-CP consists of 10 questions and 16 items. Therefore, 80 patients were included in the study, five times the total number of items.

Data were collected using the face-to-face interview method. Inclusion criteria: All patients diagnosed with prostate cancer, who are under treatment or under follow-up, who speak and understand Turkish, and who voluntarily agree to participate in the study. Exclusion criteria: patients with relapsed prostate cancer and patients with psychiatric problems using antipsychotic, antidepressant, and anxiolytic drugs.

Ethical Aspect of Research

Permission was obtained by e-mail from lead author Peter Chang, who developed the EPIC‐CP scale. Ethics committee approval was obtained from the Non-Interventional Clinical Research Ethics Committee of a Dokuz Eylül University Hospital (date: 10.04.2019 and decision no: 2019/09-05). Then, permission was obtained from the institution where the study was conducted.

Data Collection Tools

Data were collected using the Patient Descriptive Characteristics Form and EPIC-CP prepared by the researchers. Patient Descriptive Characteristics Form: consists of 10 questions in total, in which socio-demographic characteristics of patients such as age, marital status, education level, and income level, and clinical characteristics such as the patient’s PSA value, Gleason Score, tumor extent, and treatment methods are assessed. Information about the PSA value, Gleason score, and tumor spread of the patients was obtained from the medical records.

Expanded Prostate Cancer Index Composite for Clinical Practice (EPIC-CP)

This index, developed by Chang et al. (10), aims to evaluate the effect of treatment on the quality of life of prostate cancer patients in a short period of time. EPIC-CP is inspired by the first form of the index, EPIC, and the second, abbreviated form, EPIC-26. EPIC-CP is a Likert-type measurement tool consisting of 16 items, structured into 10 questions, with clinical ease of use. The index consists of five sub-dimensions that determine the quality of life of patients with prostate cancer. These: urinary incontinence, urinary irritation/retention, bowel functions, sexuality, and hormonal symptoms. The value given to each question is between zero and four points. Each sub-dimension ranges from zero to 12 points. Higher scores indicate worse symptom severity and worse quality of life. With EPIC-CP, both [the sub-dimensions affecting the quality of life of the prostate cancer patient can be evaluated (minimum 0-maximum 12)], as well as the general evaluation of the quality of life (minimum 0-maximum 60), can be conducted. The first item in the index is an independent general urinary disorder item with no scoring (10).

Statistical Analysis

In the analysis of the data using the IBM SPSS Statistics 21.0 program, number, percentage, and mean were used for descriptive statistics. The Content Validity index and Explanatory Factor Analysis were calculated for validity, while the Pearson Correlation Analysis and Cronbach’s Alpha Coefficient were calculated for reliability. The AMOS program was used in the Confirmatory Factor Analysis, for validity. The conformity of the data to the normal distribution was evaluated with the Shapiro-Wilk test. p<0.05 was considered significant.

Results

Demographic and Clinical Characteristics

The mean age of the patients participating in the study was 72.96±8.44. Sixty-five percent were married, 52% were primary-secondary school graduates, 41% had income equal to expenditure, 49% had a Gleason score of 8 to 10, and 61% had metastatic spread of the tumor. The mean PSA value of the patients was found to be 28.39±4.31 ng/mL. The overall prostate cancer quality of life score was 26.26±9.6. When the sub-dimensions were examined, it was determined that the patients had the highest sexual symptom scores (8.62±2.39), which were significantly associated with a reduction quality of life (Table 1).

Validity Analysis

Language Validity

For the EPIC-CP, which was translated into Turkish, opinions were obtained from 11 experts. To test the comprehensibility of the items in the language-valid scale, the scale was applied to a small group with the characteristics of the sample group. In this study, considering the number of scale items and the sample size, the instrument was applied to 8 prostate cancer patients with characteristics similar to those in the study. After the application, the comprehensibility of the questions was tested. Since no negative or positive feedback was received, the study continued. Patients participating in the pilot study were excluded from the sample.

Content Validity

Expert opinions were evaluated according to the Scope Validity index Davis technique. The experts were given the original form of EPIC-CP and the Turkish form together and were asked to assess and make a selection regarding the suitability of the scale items (A: item is appropriate, B: item should be reviewed, C: item should be seriously reviewed, D: item is not suitable). KGI is calculated as follows: The number of experts who assigned A and B to each item of the scale is divided by the total number of experts. If the KGI is greater than 0.80, the substance is considered sufficient in terms of scope validity (12). In this study, KGI was calculated as 1.0 for each item of the scale. According to expert opinions, the draft form of the scale was rearranged finalized.

Construct Validity

In the factor analysis, the Kaiser-Meyer-Olkin (KMO) test was applied to test whether the sample size was suitable for factor analysis. As a result of the analysis, the KMO value was 0.821. In addition, when the results of the Bartlett’s Sphericity test were examined, it was seen that the obtained chi square value was acceptable [χ²(78)=517.372; p<0.05].

Descriptive and Confirmatory Factor Analysis

To reveal the factor pattern of the scale, principal components analysis was chosen as the factorization method and varimax was selected as the rotation method. In the explanatory factor analysis conducted to reveal the factor pattern of the scale, 3 items were removed from the scale due to low factor load (EPICCP7, EPICCP8, and EPICCP9) and the remaining 13 items were collected in 4 sub-dimensions. These factors explain 74.403% of the total variance (Table 2).

Confirmatory Factor Analysis was conducted to determine whether the items and their sub-dimensions explained the original structure of the scale. In the analysis, the factor loads of the substances were above 0.40 and all correlations were significant (Figure 1).

Reliability Analysis

Internal Consistency (Cronbach Alpha Coefficient)

When the reliability of the scale and sub-dimensions of the Turkish version of EPIC-CP was evaluated separately, the reliability coefficients were found to be 0.841 for the first dimension, 0.834 for the second dimension, 0.833 for the third dimension, 0.781 for the fourth dimension, and 0.874 for the overall scale, indicating a good degree of reliability. The fact that Cronbach’s Alpha values are greater than 0.60 shows that the scales used are reliable (13). This meant that the scale used in the study was determined to have a good degree of reliability.

Item Analysis and Total Score Correlation

The relationship between the scores from the scale items, and the total score of the scale, is explained by item analysis and total score correlation. When the correlations between the variables are examined, the factor loadings of the substances are over 0.40 and all correlation relationships are significant (Table 3).

According to confirmatory factor analysis, the analysis determined that the 13 items and 4 sub-dimensions that make up the scale were related to the scale structure. The values ​​accepted for the fit index calculations are shown in Table 4.

Discussion

The KGI of the EPIC-CP scale was calculated as 1.0 for each item. In the literature, it is emphasized that these rates should be above 0.80 (12). The results of this study showed that experts reached a consensus on the content of the scale.

The adequacy of the data and sample size for factor analysis was evaluated using the Bartlett’s test of sphericity and the KMO test. The Bartlett sphericity test value is expected to be statistically significant and the KMO value is expected to be higher than 0.60 (14). In this study, Exploratory Factor Analysis, KMO, and Bartlett’s tests, met the necessary conditions, and the number of samples and data was suitable for factor analysis.

In the explanatory factor analysis performed to reveal the factor pattern of the scale, 3 items were removed from the scale (EPICCP7, EPICCP8, and EPICCP9) due to their low factor loading, and the remaining 13 items, were collected in 4 sub-dimensions. These factors explain 74.403% of the total variance. It is considered sufficient if the variance explained in multi-factor designs is above 50% (11).

EFA factor loads ranged from 0.68 to 0.88. The fact that the factor loadings measured from each sub-dimension are more than 0.30 indicates a strong factor structure (14). The results of this analysis showed that the factor loads in the EFA were at the desired level. In the original scale, the results could not be compared because EFA was not performed (10).

According to the CFA analysis, the fit indices were RMSEA ≤0.08, and GFI, CFI, IFI, RFI, NFI, NNFI were ≥0.80. These values ​​indicate that there is a significant relationship between the scale and its sub-dimensions. According to the CFA results, the data were found to be compatible with the model, and the sub-dimensions were related to the scale. In the CFA analysis, the factor loads of the items varied between 0.67 and 0.89. The factor loads are above 0.40 and all correlation relationships are significant. In the original form of the scale, the results could not be compared because CFA was not performed (10).

Cronbach’s alpha coefficient measures the average correlation between items in the scale. This value is expected to be close to 1 (15, 16). In this study, the Cronbach’s alpha coefficient for both the total scale and its sub-dimensions was greater than 0.70. This finding shows that the scale and its sub-dimensions are reliable. The items were sufficient to measure the relevant subject and the scale had good reliability. The Cronbach’s alpha coefficient of the scale and sub-dimensions was found to be 0.64-0.84 in the original scale (10). Lourenço et al. (17), in 2020, found the Cronbach’s alpha coefficient to be 0.35-0.82 for the sub-dimensions of the scale. The reliability coefficient results from this study were found to be consistent with those of the original scale.

To show that each item can measure at the expected level, an item-total score analysis was conducted to explain the relationship between the score of each item and the total score of the scale. This value is expected to be >0.40 (15). The total score correlation of the Turkish version of the scale was 0.67 to 0.89. These results show that the sub-dimensions of the scale are correlated with the total score and each item is reliable. The item total score analysis in the original scale was found to be ranging from 0.31-0.83 (10). However, since the section on sexual problems, which reduced the item total score, was included in its original format in the analysis, the item-total score correlation was higher in this study.

As a result, the Turkish version of EPIC-CP consists of four sub-dimensions and 13 items. Each sub-dimension is scored as 0-12 points. The total score of the scale is the sum of all sub-dimensions, totaling 48.

Study Limitations

Some patients in the study had undergone open radical prostatectomy, and patients who had other surgical treatment options (laparoscopic or robotic radical prostatectomy) were not included in the sample. This situation may affect the generalizability of the scale. In addition, while the scale originally had five sub-dimensions, it was reduced to four in the Turkish version, with the sexuality sub-dimension being removed from the scale. The absence of a sexuality sub-dimension in the Turkish version creates a limitation.

Conclusion

The Turkish validity and reliability of EPIC‐CP are high. It is thought that its use in Turkish society will be useful in evaluating the quality of life of patients with prostate cancer and in planning care and interventions to improve it.