Yvonne Kuipers, Julie Jomeen, Tinne Dilles and Bart Van Rompaey
The purpose of this paper is to measure reliability, validity and accuracy of the 12-item General Health Questionnaire (GHQ-12) as a measure of emotional wellbeing in pregnant…
Abstract
Purpose
The purpose of this paper is to measure reliability, validity and accuracy of the 12-item General Health Questionnaire (GHQ-12) as a measure of emotional wellbeing in pregnant women; utility and threshold in particular.
Design/methodology/approach
The authors measured self-reported emotional wellbeing responses of 164 low-risk pregnant Dutch women with the GHQ-12 and a dichotomous case-finding item (Gold standard). The authors established internal consistency of the 12 GHQ-items (Cronbach’s coefficient α); construct validity: factor analysis using Oblimin rotation; convergent validity (Pearson’s correlation) and discriminatory ability (area under the receiver operating characteristics curve and index of union); and external validity of the dichotomous criterion standard against the GHQ-12 responses (sensitivity, specificity, likelihood ratios and predictive values), applying a cut-off value of ⩾ 12 and ⩾ 17, respectively.
Findings
A coefficient of 0.85 showed construct reliability. The GHQ-12 items in the pattern matrix showed a three-dimensional factorial model: factor 1, anxiety and depression; factor 2, coping; and factor 3, significance/effect on life, with a total variance of 59 per cent. The GHQ-12 showed good accuracy (0.84; p=<0.001) and external validity (r=0.57; p=<0.001) when the cut-off value was set at the ⩾ 17 value. Using a cut-off value of ⩾ 17 demonstrated higher sensitivity (72.32 vs 41.07 per cent) but lower specificity (32.69 vs 55.77 per cent) compared to the commonly used cut-off value of ⩾ 12.
Research limitations/implications
Findings generally support the reliability, validity and accuracy of the Dutch version of the GHQ-12. Further evaluation of the measure, at more than one timepoint during pregnancy, is recommended.
Practical implications
The GHQ-12 holds the potential to measure antenatal emotional wellbeing and women’s emotional responses and coping mechanisms with reduced antenatal emotional wellbeing.
Social implications
Adapting the GHQ-12 cut-off value enables effective identification of reduced emotional wellbeing to provide adequate care and allows potential reduction of anxiety among healthy pregnant women who are incorrectly screened as positive.
Originality/value
A novel aspect is adapting the threshold of the GHQ-12 to ⩾ 17 in antenatal care.