Laboratory diagnosis of gestational diabetes: An in silico investigation into the effects of pre-analytical processing on the diagnostic sensitivity and specificity of the oral glucose tolerance test

- The impact of pre-analytical glucose decay on diabetes diagnosis was assessed.
- A virtual cohort of pregnant women was used in a Monte-Carlo analysis.
- Overall, diagnostic sensitivity was reduced by pre-analytical glucose decay.
- Batched pre-analytical processing of samples yielded very low sensitivity.
- Fast tracking pre-analytical processing mitigates the low sensitivity.

IntroductionDelayed separation of red cells from plasma causes pre analytical glucose loss, which in turn results in an under-diagnosis of GDM (gestational diabetes) based on the OGTT (oral glucose tolerance test). In silico investigations may help laboratory decision making, when exploring pragmatic improvements to sample processing.MethodsLate pregnancy 0, 1 and 2 h 75 g OGTT values were obtained from two distinct populations of pregnant women: 1. Values derived from the HAPO (Hyperglycemia and Adverse Pregnancy Outcome) Study and 2. New Zealand women identified as at higher risk of GDM by their caregivers, undergoing OGTT during routine antenatal care. In both populations studied, in silico modelling focussed on the effects of pre-analytical delays in plasma separation, when using fluoride collection tubes.ResultsUsing a model that 'batched' samples from the three OGTT collection times, diagnostic sensitivity was estimated as follows: 66.1% for HAPO research population and 48.4% for the 1305 women receiving routine antenatal care. If samples were not batched, but processed shortly after each blood sample was collected, then sensitivity increased to 81%.ConclusionExploration of a range of clinical and laboratory scenarios using in silico modelling, showed that delaying the processing of pregnancy OGTT samples, using batched sample collection into fluoride tubes, causes unacceptable loss of GDM diagnostic sensitivity across two distinct population groups. This modelling approach will hopefully provide information that helps with final decision making around improved laboratory processing techniques.