Psychometric Properties of Patients' Health Questionnaire Assessed on Obstetric Fistula Patients in Northwestern, Nigeria

Abstract

Nigeria is among the highest contributor to global prevalence of obstetrics fistula (OF) and weak health care policies and services have left the victims reeling in untold psycho-social and physical morbidities. This study was aimed at assessing the psychometric properties of a globally used Patients’ Health Questionnaire (PHQ-9) among these patients. A descriptive cross-sectional study design was employed to study the fertility desires, mental health and patients’ quality of life among women with obstetric fistula in three selected OF centers in Northwestern region of Nigeria. Using two-stage sampling technique four hundred twenty (420) OF patients were recruited into the study. Maximum likelihood (ML) factor analysis was deployed to produce latent factors from the responses to the PHQ-9 scale, internal reliability and construct validity were determined by Cronbach alpha values while comparative mean analysis was used to validate standardized scores of the derived PHQ-9 scale factors against socio-demographic and obstetric characteristics. Mean age and standard deviation of respondents were 26.4±8.4 years as 60% aged between 20-25 years. The married, divorced and those in polygamous were 45%, 47.1% and 73.6% respectively. More than half (55.5%) were traders while majority (89.8%) earned NGN4500 or less monthly. Three factors of somatic (r= 0.72-0.82; α=0.86), non-somatic (r= 0.64-0.76; α=0.77) and affective (r= 0.64-0.96; α=0.74) constructs emerged from the ML factor analysis. Patients of ages 20-25 years, divorced/single and widows, the formally educated and farmers significantly differed with high scores of the affective construct of the PHQ-9 scale. Those who had more than four pregnancies and greater than a single stillbirth significantly differed with affective symptoms (0.29±0.90; 0.69±0.9 The 15–17-year-olds at first birth and those who waited more than ten years for OF surgical repairs differed with non-somatic symptoms’ construct of the PHQ-9 scale (0.21±1.04; 0.31±0.87) significantly differed with affective factors. The PHQ-9 instrument exhibited possibility of being used to assess the quality of OF patients’ mental health. Future studies could expand the sample size to accommodate advanced inferential psychometric methods.