LONGITUDINAL VALIDATION OF THE PREMENSTRUAL SYMPTOMS IMPACT SURVEY (PMSIS)
Author(s)
Min Yang, MD, PhD, Senior Scientist1, Aaron S Yarlas, PhD, Senior Outcomes Analyst1, Geoffrey C Hammond, PhD, Associate Scientist1, Joseph A. Gricar, MS, Health Economics Consultant21QualityMetric Incorporated, Lincoln, RI, USA; 2 Independent Health Care Consultant, New York, NY, USA
OBJECTIVES To validate the Premenstrual Symptom Impact Survey (PMSIS), a six-item self-assessment patient-reported outcomes instrument designed to evaluate impact of premenstrual symptoms on a woman's health-related quality of life. METHODS The PMSIS and SF-12v2 Health Survey were administered in two waves to females (18-45 Yrs) with premenstrual complaints (e.g., irritability, depression, headache and abdominal bloating). Retrospective criteria from the American College of Obstetricians and Gynecologists and the DSM-IV-TR were used to categorize women into either Premenstrual Syndrome (PMS) or Premenstrual Dysphoric Disorder (PMDD) groups respectively. PMSIS responses were analyzed for internal consistency, reliability, convergent and discriminant/known-group validity, at each time point. Responses across times were analyzed for test-retest reliability. RESULTS At Wave1 (N=1100) and Wave2 (N=770), 34.2% and 32.6% of participants were identified in the PMS group while 14.9% and 14.2% were identified in the PMDD group, respectively. The mean PMSIS scores at each wave were as follows: PMS groups (54.3, 52.0), Non-PMS groups (32.4, 28.3), PMDD groups (59.0, 55.0) and Non-PMDD groups (36.5, 32.8). The PMSIS showed good internal consistency at each time (α ≥ 0.88), and adequate test-retest reliability across times (intra-class correlation: 0.74). PMSIS scores correlated significantly (p <0.001) with SF-12 Physical Component Summary (PCS) and Mental Component Summary (MCS) scores. At each time, PMSIS scores discriminated well across presence/absence of PMS and PMDD (all F's > 100, all ps <0.001), and between low/medium/high PCS and MCS groups (all F's > 24, ps <0.001), indicating known-group discriminant validity. Receiver operating characteristics analyses showed satisfactory values for areas under the curve (≥ 0.78) in detecting women with PMS and/or PMDD at each time. CONCLUSIONS This study demonstrates that the PMSIS has good internal consistency, test-retest reliability, and convergent and discriminant validity, making the PMSIS a viable option for identifying and assessing premenstrual problems.
Conference/Value in Health Info
2009-05, ISPOR 2009, Orlando, FL, USA
Value in Health, Vol. 12, No. 3 (May 2009)
Code
PIH33
Disease
Pediatrics, Reproductive and Sexual Health