M.J. Leskinen, Department of Urology, Seinäjoki Central Hospital, Hanneksenrinne 7, FIN-60220 Seinäjoki, Finland. e-mail: email@example.com
Authors from Finland have assessed a version in their language of the National Institutes of Health-Chronic Prostatitis Symptom Index; they found that their translated version was valid and easily understandable in the management of the symptoms of chronic pelvic pain syndrome. They felt it should be used as a primary outcome measure in studies with these patients.
There are three papers in this issue relating to the effect of drugs on LUTS; the first of these is a pooled analysis of three double-blind placebo-controlled studies into the safety and efficacy of the 10 mg dose of alfuzosin. The second evaluates the effect on quality-of-life issues of treatment with dutasteride. Finally, authors from Australia compare the effect of a Serenoa repens extract with placebo for LUTS.
To provide a fluent and easily comprehensible Finnish version of the National Institutes of Health Chronic Prostatitis Symptom Index (NIH-CPSI) and to study its linguistic validity and correlation with a visual pain scale (VAS).
PATIENTS AND METHODS
The double-back translation method with two interim modifications was used to produce the Finnish version of the NIH-CPSI. The validity was tested by presenting the questionnaire to 155 men with clinically confirmed chronic pelvic pain syndrome (CPPS) and 12 controls with no previous urological symptoms. Convergent validity of the NIH-CPSI was tested by determining the correlation between the Finnish NIH-CPSI and VAS. Patients’ and urologists’ opinions about the utility of the Finnish NIH-CPSI were also reviewed.
The total Finnish NIH-CPSI scores and the pain domain and voiding symptom domain scores differed significantly (P < 0.001) between the groups, suggesting good discriminant validity of the symptom index. The NIH-CPSI scores correlated well with the VAS (Pearson's correlation 0.76). The preciseness and comprehensibility of the questionnaire were consistently evaluated to be ‘good’ or ‘excellent’ both by patients and urologists.
The Finnish version of the NIH-CPSI is valid and easily comprehensible for measuring CPPS symptoms. In addition, it provides good discriminant and convergent validity in distinguishing CPPS symptoms and should be used as primary outcome measure in CPPS studies.
Until recently, the reliable evaluation of the symptoms of chronic prostatitis (CP) or chronic pelvic pain syndrome (CPPS) was very difficult. The lack of an acceptable outcome measure made interpreting the results obtained in CP/CPPS studies vague. A solution was provided by the Chronic Prostatitis Collaborative Research Network (CPCRN) when it introduced the National Institutes of Health Chronic Prostatitis Symptom Index (NIH-CPSI) in 1999 . This index, comprised of pain, voiding and quality-of-life (QoL) domains, was confirmed to be valid and easily administered both in studies and in clinical practice . The NIH-CPSI was primarily developed to evaluate the treatment responses and natural history of CP/CPPS, but since its launch it has also been useful in distinguishing prostatitis-like symptoms in the general population, therefore having additional value in epidemiological studies [3,4]. The translation and linguistic validation of the NIH-CPSI into Spanish, Japanese and German was described previously [5–7].
Although a visual pain scale (VAS) and Subjective Global Assessment (SGA) are widely accepted by medicolegal authorities as valid outcome measures in treatment studies, the NIH-CPSI lacks this status. Thus our first aim was to provide a fluent and easily comprehensible version of the NIH-CPSI in Finnish, and study its linguistic validity. Second, by using the modified validation study and comparing its correlation with the VAS we aimed to explore its discriminant and convergent validity.
PATIENTS AND METHODS
The original NIH-CPSI questionnaire was translated into Finnish by two Finnish urologists (M.J.L. and A.M.) independently. The translators then compared the two versions with each other, and several phrases and words were revised to enhance the comprehensibility. The reconciled version was then back-translated by a native English, bilingual (English and Finnish) professional translator. Another bilingual professional translator and the authors (M.J.L. and A.M.) then compared the translated version with the original NIH-CPSI. No inconsistencies were found between the versions, and the idiomatic content of the versions was practically identical. After revision for Finnish grammar and spelling by a native Finn with an MA degree in Finnish, the questionnaire was taken for trial use in the Finnish urological community. During 3 months of trial use in clinical practice, one adjustment was made to the questionnaire. The word ‘delighted’ in question 9 first translated into Finnish as ‘mielissään’, was revised to ‘erittäin tyytyväinen’ (very happy) to adapt better into Finnish phraseology (Appendix).
For the validation study, 155 men with clinical CPPS and 12 controls were included in the study. The patients with CPPS were consecutive referrals with a long history of symptoms and had failed to respond to conventional treatments, including antibiotics, analgesics and α-blockers, offered by GPs. The control group comprised 12 men from a general surgery waiting list for a haemorrhoidectomy, and who had no previous urological history. Written informed consent was obtained from all subjects before enrolment; the baseline characteristics of the study groups are shown in Table 1. The validation process included a clinical examination to verify clinical CPPS and symptom evaluation with the Finnish NIH-CPSI. At the visit, all patients self-completed the Finnish NIH-CPSI and this was followed by a clinical examination to confirm clinical CPPS. The patients with CPPS were also asked to scale their subjective opinion about the preciseness (scale 1–5; 1 = poor to 5 = excellent) of the questionnaire for their symptoms. The investigator made the same judgement independently using the same scale. The patients were re-evaluated, including a clinical examination by another investigator unaware of the subjects’ previous history or NIH-CPSI score, < 1 week later, when the subjects also self-completed the VAS (scale 0–10), which was used as an acknowledged ‘gold standard’ pain measure.
Table 1. Baseline characteristics: age, duration of symptoms, NIH-CPSI scores and VAS in the study groups
A nonparametric Mann–Whitney U-test was used to compare the groups. The correlations between the Finnish NIH-CPSI domains and the VAS were calculated using a nonparametric Pearson's test, and consistency between patients’ and investigator's perceptions by Cronbach's coefficient α.
The mean age of the patients with CPPS was 53 years and the mean duration of CPPS symptoms 6.6 years. The men in the control group were younger, with a mean age of 44 years. The Finnish NIH-CPSI scores differed significantly between the groups (P < 0.001), the total mean (sd) score in the CPPS group being 20.2 (8.6) and in the control group 5.8 (4.5). There were also significant differences between the groups in pain and voiding symptom domains, but not in the QoL domain (Table 1).
The correlation between total CPSI score and the VAS was relatively high, at 0.76. Cross-tabulations of correlations between different domains, total NIH-CPSI score and the VAS are shown in Table 2 and Fig. 1.
Table 2. Correlations (Pearson’s) between domains, NIH-CPSI and visual pain scale (VAS) in 155 patients with CPPS
The mean (sd) score of the patients’ subjective opinion about the comprehensibility and precision of the questionnaire for their symptoms was 4.2 (0.8), and that of the investigators’ 4.2 (0.7) on a scale of 1–5, thus resulting in a consistency of 0.65 calculated by Cronbach's α.
The discriminant validity of the NIH-CPSI among Finnish men was tested by comparing the NIH-CPSI scores of men with a history of CPPS symptoms with those with no previous urological history. The correlation between the NIH-CPSI and a previously acknowledged outcome measure, the VAS, was used to determine its convergent validity, with comprehensibility and precision evaluated by responses from patients and their physicians.
The double-back translation method is widely accepted as the best method in translating various symptom scores and QoL indexes into foreign languages [5–7]. Although Finnish is the first language of only ≈ 5.5 million people (4.8 million in Finland and 0.7 million in Sweden and other countries), the need for a readily understandable and valid symptom measure is obvious because of the high prevalence of CPPS in Finland . As a language, Finnish does not relate to any Indo-European languages, thus leading to certain difficulties in translation . The way of expressing some symptoms and QoL issues in Finnish differs from that of major Indo-European languages, therefore precluding exact word-to-word translation, as was also the case for the Japanese NIH-CPSI . Despite these problems the first translated version was relatively sound in clinical practice and only one phrase needed revision after trial use, to adapt better into Finnish phraseology. Before the validation study this revised Finnish NIH-CPSI was used to select patients for clinical studies in Finland and its applicability in those was promising .
The mean score of the present patients with CPPS (20.2) was consistent with the score of 19.5 obtained in the original NIH-CPSI validation study , and with the previously published linguistic validation studies in Spanish, German and Japanese [5–7]. The scores in different domains and the duration of CPPS symptoms (Table 1) were also very similar to those reported previously. As the differences in pain, urinary symptoms and total NIH-CPSI scores were statistically significantly different between the study groups the Finnish NIH-CPSI clearly has discriminant and content validity to distinguish the patients with CPPS symptoms from those without. Even though the control patients had relatively severe symptoms from haemorrhoids, often associated with local pain, their pain domain scores were very low, which further supports the high power of the Finnish NIH-CPSI in distinguishing different types and locations of pain. However, the QoL domain was not statistically different between the groups, probably reflecting the slight impairment of QoL by haemorrhoid symptoms.
The present correlation between the pain domain scores and total NIH-CPSI scores (0.89, Table 2) was lower than that with the Spanish NIH-CPSI (0.97) . This difference might be a result of the different sample sizes, as the Spanish NIH-CPSI study included only 37 subjects and thus had higher random variation. The original NIH-CPSI study with 151 patients contains no information about the correlation between the pain domain and total NIH-CPSI, but that between the pain domain and the QoL domain was 0.77, which is consistent with the correlation of 0.71 obtained in the present study . These values suggest that the construct and content validity of the Finnish NIH-CPSI is comparable with results obtained in previous validation studies [1,5,6].
The VAS is a generally recognized and validated measure of acute and chronic pain, and its correlation with QoL and functional performance disability has been well established [11,12]. This VAS is officially accepted as an outcome measure in clinical studies of pain-related disorders, including CPPS. The NIH-CPSI, although widely used in clinical studies and promoted by the CPCRN , still lacks the medico-legal status of an official outcome measure in CPPS studies. To assess the convergent validity of the Finnish NIH-CPSI, the correlation between the VAS and the Finnish NIH-CPSI was determined; the VAS correlates well not only with the pain domain (Pearson's correlation 0.87), but also with total NIH-CPSI score (Pearson's correlation 0.76; Table 2 and Fig. 1). These results support the assumption that the Finnish NIH-CPSI also has good convergent validity.
The subjective comprehensibility and precision of the NIH-CPSI for patients’ symptoms has not been evaluated in previous studies. To assess the practicality of the Finnish NIH-CPSI this was evaluated subjectively by patients and their urologists. The questionnaire was graded 1–5 for accuracy and comprehensibility, with evaluations varying from good to excellent, with a mean response of 4.2 from both patients and urologists. The evaluations were also fairly consistent, the Cronbach's α being 0.65, also supporting the usefulness and comprehensibility of the questionnaire.
The internal validity of the NIH-CPSI was confirmed to be high in previous studies; the test-retest reliability was 0.83–0.93 in the original  and 0.63–0.91 in the Japanese study , while the internal consistency (Cronbach's α) of the original NIH-CPSI was 0.86 , and 0.74–0.94 in previous linguistic validation studies [5–7]. Although being unable to calculate the test-retest reliability formally, using the VAS as a modified re-test method gave an opportunity to evaluate the convergent validity. The finding that the NIH-CPSI and VAS correlated well is new and indicates good convergent validity, which might be important especially when the medicolegal aspects of outcome measures in CPPS studies are considered.
The present study has some limitations; the modified re-test method does not fulfil the criteria of a formal validation study but must be regarded as a modified linguistic validation. In addition, the control group was relatively small, comprising patients awaiting surgery and therefore not completely asymptomatic. This probably resulted in an impaired QoL, with no statistical difference detected between the groups. However, the pain domain, urinary symptom domain and total NIH-CPSI score differentiated these men well from patients with CPPS, and therefore we are confident that the Finnish NIH-CPSI has good construct validity and that it functions well in distinguishing such patients from those with other symptoms impairing their QoL.
In conclusion, the Finnish version of the NIH-CPSI has good construct and convergent validity, and is easily comprehensible in evaluating CPPS symptoms. Although the NIH-CPSI and VAS correlate well with each other, the NIH-CPSI should be used as a primary outcome measure in CPPS studies and clinical practice, because of its good discriminant validity.
chronic pelvic pain syndrome
Chronic Prostatitis Collaborative Research Network
National Institutes of Health Chronic Prostatitis Symptom Index