Validation of a diabetes risk score in identifying patients at risk of progression to abnormal glucose tolerance post partum

FINDRISC (Finnish Diabetes Risk Score) is a risk assessment tool widely used for the prediction of the development of type 2 diabetes (T2DM), combining a questionnaire with simple anthropometric measurements to identify patients at risk of developing diabetes, with increasing score (0-26) signifying increased risk. A cut off score of 9 has previously been proposed (with drug-treated DM as the endpoint) with a positive predictive value (PPV) of 0.12 and negative predictive value (NPV) of 0.99, area under receiver operating characteristic curve (AuROC)=0.80. It has been well validated in the general population.

We examined its use in predicting progression to pre-diabetes/diabetes in a cohort of Caucasian patients with a history of gestational diabetes mellitus (GDM). 116 women with a history of GDM underwent screening 1-5 years post-index pregnancy. Those with a history of persistent post-partum dysglycaemia had fasting glucose levels taken, while others underwent a 75g OGTT.

Of the 116 women with a history of GDM, 83 showed normal glucose tolerance (NGT) post-partum (71.6%). 22 patients had abnormal OGTT at 12 weeks (18.9%). A further 11 patients (9.5%) had pre-diabetes/diabetes at rescreening. The FINDRISC score was higher in patients with pre-diabetes/diabetes than those with NGT post partum (mean score 13.6+/- 4.1 vs. 11.0+/- 3.7, p<0.01). For a cut-off score of 9, PPV was 0.30, NPV was 0.79, AuROC=0.69, comparable with published data in the general population.

This study shows the validity of an inexpensive, convenient risk score in helping to determine which patients may need more frequent screening post GDM.