[The validity of the IPSS questionnaire in a sample of 262 patients with benign prostatic hyperplasia]

The IPSS is a well-known method for symptom evaluation in patients with prostate disease. The present study analyzes the validity of this self-administered questionnaire in a group of patients from an industrial area in Spain.

513 questionnaires (test 1) were sent out to patients with benign prostatic hyperplasia (BPH) who had undergone TUR or adenomectomy. After an interval of 30-60 days, the same questionnaire (test 2) was sent out to all those patients who had correctly answered test 1. The feasibility and reliability of this test in our hands, as well as its ability to discriminate, were determined. Moreover, the relationship between symptoms (S) and quality of life (L) was analyzed.

Patient ages ranged from 50 to 86 years (mean 67.6). A) Feasibility (percentage of patients who answered the questionnaire correctly) was 71%. B) Reliability: internal consistency (Cronbach's alpha) was 0.74 and the retest reliability (correlation between the scores of tests 1 and 2) was 0.76 (Pearson's r) and 0.5 (Kendall's T). C) Correlation between symptoms (S) and quality of life (L) was 0.53 (Pearson) and 0.33 (Kendall). D) Discriminating validity (ability of the test alone to diagnose patients with BPH correctly). The patients in the study were compared with a group of 420 "healthy" males from a prostate cancer screening program; the area below the ROC curve was 0.98 (scantily evaluable: healthy patients vs surgical BPH cases).

In our setting, the self-administered IPSS questionnaire had a moderate feasibility (29% of the patients were unable to understand the questionnaire). Our results, in terms of internal consistency and retest reliability, are similar to those of other validation studies conducted in our country (and appreciably worse than those of the original validation studies in the English language). Nevertheless, some items of the questionnaire have an unacceptable retest reliability, particularly those relating to frequency or urgency. Finally, the total symptom (S) score showed a scanty correlation with quality of life (L). We can conclude that although the IPSS questionnaire is a useful method in the assessment of patients with prostate disease, its reliability remains far from optimal.