Introduction
Currently, there is no clear standard assessment tool for the diagnosis and daily monitoring of hypoactive sexual desire disorder (HSDD) in postmenopausal women.
Aim
The aim of the study was to validate (i) the Women’s Sexual Interest Diagnostic Interview—Short Form (WSID-SF) which is a structured tool to identify HSDD; and (ii) the Daily Log of Sexual Activities (DLSA) which is a diary to monitor daily HSDD status in postmenopausal women. Both assessments were collected as self-reports by an interactive voice response system (IVRS).
Methods
At the initial study visit, 629 postmenopausal women, age 39–66, were evaluated for HSDD by the WSID-SF. In addition, in a subgroup of 175 subjects at five study sites, HSDD was assessed by physicians who were blinded to the WSID-SF diagnosis. During the 58-day study period, patients completed the DLSA daily through self-reported IVRS. All women also completed the Female Sexual Function Index (FSFI), Menopause Sexual Interest Questionnaire (MSIQ), Female Sexual Distress Scale (FSDS), Dyadic Adjustment Scale (DAS), and Kellner Symptom Questionnaire (KSQ) at both the initial study visit and at the end of the study. The WSID-SF-based identifications were compared with clinical diagnoses made based on physicians’ clinical experience. Construct validity of the WSID-SF and DLSA were assessed based on comparisons with questionnaire results. Internal consistency and test–retest reliability of the DLSA were also evaluated.
Main Outcome Measures
Main outcome measures include the agreement between WSID-SF diagnosis and clinician diagnosis, convergent and divergent validity of the WSID-SF and DLSA, and reliability of the DLSA.
Results
Enrolled subjects were classified into HSDD (N = 468) and non-HSDD (N = 161) groups by WSID-SF. When compared with physician’s diagnosis, WSID-SF-based diagnosis had a specificity of 89% and a sensitivity of 70% (kappa = 0.46, P < 0.001). WSID-SF showed significant correlation with each domain of the FSFI, MSIQ, and FSDS (all P < 0.001). As anticipated, WSID-SF had low or nonsignificant correlations with all domains of the DAS and the KSQ. Four different algorithms were piloted to calculate DLSA scores. Data on the detailed analysis conducted to evaluate the four scoring strategies is on file (not presented in this article). Ultimately, the weekly DLSA total score calculated by algorithm #4 was selected to validate the DLSA. In the test–retest reliability evaluation, the intraclass correlation coefficient was 0.80 for women with HSDD and 0.84 for women without HSDD (P < 0.001 for all analysis). In the known-group validity comparison, the weekly DLSA total score was significantly different (P < 0.001) among the HSDD and the no HSDD groups, with an effect size of 0.89–0.92 based on Cohen’s d. The DLSA also showed convergent validity with moderate to high correlations with each domain of the FSFI, MSIQ, and FSDS (P < 0.001 for all correlations). As anticipated, the DLSA had weak correlations with the DAS and KSQ demonstrating divergent validity.
Conclusions
The WSID-SF had good specificity and sensitivity (i.e., discriminative validity) in identifying HSDD in postmenopausal women. In addition, the DLSA is a reliable and valid patient-reported outcomes tool that can be utilized to assess effectiveness of treatments in postmenopausal women with HSDD. Further, the WSID-SF and DLSA both demonstrated good convergent and divergent validity.