Assessment of Periodic Blood Loss: Interindividual and Intraindividual Variations of Pictorial Blood Loss Assessment Chart Registrations

Abstract

Study ObjectiveTo explore the interindividual and intraindividual variation of Pictorial Blood Loss Assessment Chart (PBAC) registrations of subjective perception of minimal, normal, and heavy menstrual bleeding in women using their usual sanitary protection. DesignRetrospective study (Canadian Task Force classification II-3). SettingUniversity tertiary hospital. PatientsWomen who had participated in 4 previously published prospective studies using PBAC as the outcome measure. InterventionsPatients underwent hysteroscopic removal of polyps, laparoscopic occlusion of uterine vessels, or uterine artery embolization, and control subjects received no treatment. Measurements and Main ResultsPBAC scores, hemoglobin concentration, and subjective assessment of periodic blood loss were recorded in 429 women during 1049 menstrual cycles. The median PBAC values in groups of women who assessed their bleeding as light, normal, and heavy were 45.0, 116.0, and 254.5, respectively (p < .001). Sensitivity and specificity for specific PBAC cutoff points were calculated using a receiver operating characteristic curve. The maximum sensitivity for assessment of heavy periodic blood loss (78.5) was reached at a PBAC score of 160. The corresponding specificity was 75.8. At PBAC values below 130, 90% of the women reported normal blood loss, and 91% had hemoglobin values >12.0 g/dL. The reliability of repeated PBAC scores for individual measures was assessed via calculation of the intraclass correlation coefficient, which for repeated PBAC measurements was 0.86 (95% confidence interval, 0.80–0.90) for average measures. ConclusionThe interindividual variation was high, and the intraindividual variation was low. A low PBAC score may be used to define treatment end points for clinical use or in trials.