Bias in Evaluating Erectile Function in Lifelong Premature Ejaculation Patients with the International Index of Erectile Function-5.

INTRODUCTION

Lifelong premature ejaculation (LPE) is the most important ejaculating dysfunction. Relatively little is known about erectile function in this population.

AIMS: We attempted to evaluate the erectile function of patients with LPE using the International Index of Erectile Function-5 (IIEF-5) to determine if it is sufficiently reliable and accurate to make such an assessment.

METHODS

A total of 406 patients with LPE were enrolled in our study. The participants voluntarily answered the Premature Ejaculation Diagnostic Tool (PEDT) and IIEF-5 questionnaires and underwent a full history evaluation and clinical examination by doctors. Their answers were converted into data analyzed by a statistic software.

MAIN OUTCOME MEASURES: The patients were diagnosed with LPE based on the diagnostic criteria and PEDT scores. The intravaginal ejaculation latency time was recorded according to patient self-reports. The IIEF-5 was used to evaluate their erectile function. Thorough history and clinical examination helped doctors make more correct diagnoses of erectile dysfunction (ED).

RESULTS

Of the 406 patients, 70 (17.24%) patients had ED, as confirmed by doctors. IIEF-5 was accurate for the assessment of the erectile function of LPE patients when the cutoff was decreased to 15.5. Question 5 (1.34 ± 0.53) was the main reason for the drop in the total IIEF-5 score. Questions 1 and 5 shared low consistency with the other three IIEF-5 items, thus they lowered the reliability of the IIEF-5 scores. These questions created a confounding bias that decreased the diagnostic threshold of IIEF-5. However, they could not be removed from the IIEF-5 because they did not reduce its diagnostic accuracy in patients with LPE.

CONCLUSIONS

Bias from questions 1 and 5 lowered the reliability of the IIEF-5 scores; however, it did not reduce the diagnostic accuracy of the IIEF-5. The recommendation is to edit questions 1 and 5 when they are applied to populations with LPE.