To develop and validate a polycystic ovary syndrome (PCOS) case definition using administrative health data sources.

A validation study.

Secondary care centre outpatient gynaecology clinic in Calgary, Alberta, Canada.

3951 electronic health records of women aged 18–45 years who presented to a gynaecology clinic in Calgary, Canada, between January 2014 and December 2019 were reviewed. We identified 180 patients with PCOS using the Rotterdam criteria. Participants were excluded if they were biologically male, pregnant at the time of the consultation, did not meet the date criteria or if their consultation note was missing. The chart data were connected to the Practitioner Claims and the Discharge Abstract Database by personal health number.

Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of 68 case definitions for PCOS were estimated. Case definition performance was graded.

Of the 68 case definitions tested, none had high validity. The best performing case definitions were: (1) ≥3 instances of International Classification of Diseases-9 code 256.4 (polycystic ovaries) with exclusion codes (sensitivity 23.89%, specificity 99.59%, PPV 74.14%, NPV 96.35%) and (2) 626.X (irregular menstruation), 704.1 (hirsutism) and ≥3 instances of code 256.4 with exclusion codes (sensitivity 2.78%, specificity 99.97%, PPV 83.33%, NPV 95.40%).

We identified several case definitions for PCOS of moderate validity with high PPV (>70%) for case ascertainment in PCOS research in jurisdictions with similar administrative health data. These case definitions are limited by low sensitivity, which should be considered when interpreting research findings.

Gestational diabetes is a common metabolic disorder in pregnancy which identifies a substantial increased risk of future diabetes. Despite this risk, many individuals are not screened for dysglycaemia in the postpartum period. Continuous glucose monitoring (CGM) is an evolving technology that provides details of an individual’s glucose levels throughout the day; however, it has not yet been evaluated as a screening tool for postpartum dysglycaemia. To address this gap, this prospective cohort study will examine the use of CGM in the early postpartum period to predict the risk of maternal dysglycaemia after delivery.

The Predicting Dysglycaemia in Individuals with Gestational Diabetes Immediately Postpartum using CGM (PREDISPOSE) study is a prospective cohort study designed to assess the ability of a CGM device (Freestyle Libre 2) worn in the postpartum period to detect persistent dysglycaemia in individuals with gestational diabetes. The study will recruit 240 individuals with gestational diabetes. Each participant will wear the CGM immediately postpartum and before attending routine postpartum diabetes screening, consisting of a 75-gram oral glucose tolerance test (OGTT) and related blood work (haemoglobin A1c (HbA1c), complete blood count and lipid profile). The primary outcome is the accuracy of the area under the curve for all glucose measurements from the first CGM wear to detect postpartum dysglycaemia. We will perform sensitivity and specificity analyses to determine optimal CGM cut-offs to diagnose diabetes or prediabetes. Secondary outcomes include the incidence of postpartum dysglycaemia (based on 75-gram OGTT and/or HbA1c), incidence of postpartum dyslipidaemia, patient acceptability of CGM testing, data variability from CGM and cardiometabolic health outcomes diagnosed in years one, two and five after delivery.

All participating sites have received ethics approval of the current protocol and have started recruitment of participants to the study. The ethics boards that approved this study are the Biomedical Research Ethics Board at the University of Manitoba, the Conjoint Health Research Ethics Board at the University of Calgary, the Mount Sinai Hospital Research Ethics Board at Mount Sinai Hospital and the Comité d'éthique de la Recherche at Université Laval. Study results will be disseminated through conference presentations and publication in a peer-reviewed journal, regardless of study findings.

NCT04972955. Registration date: 28 June 2021.