November 24, 2020
3 min read
Random plasma glucose tests at 12 weeks’ gestation and fasting plasma glucose or HbA1c at 28 weeks’ gestation can safely identify women with hyperglycemia at risk for suboptimal pregnancy outcomes during COVID-19, study data show.
In a retrospective analysis published in Diabetic Medicine, researchers found that the laboratory tests “perform adequately” for risk stratification when oral glucose tolerance testing is not available.
“Due to the COVID-19-related requirement for social distancing, normal testing processes for diabetes which arises in pregnancy, such as the oral glucose tolerance test for gestational diabetes, have had to change,” Claire Meek, PhD, MBChB, senior clinical research associate at the University of Cambridge Institute of Metabolic Science and consultant physician in diabetes in pregnancy at Addenbrooke’s Hospital, United Kingdom, told Healio. “Under normal circumstances, the oral glucose tolerance test is used internationally to diagnose gestational diabetes, but this requires 2 hours in a waiting room and contravenes social distancing requirements. Several new abbreviated strategies for diagnosing gestational diabetes have been recommended by different national committees, which use two to three different blood tests, including a random or fasting blood glucose and HbA1c. It is very unusual for new diagnostic processes to be introduced with no testing. Although there are some benefits to a faster, clearer diagnostic process, we were concerned that the new criteria may miss some women with gestational diabetes.”
Best tests, cutoff points
As Healio previously reported, the routine use of an OGTT to screen for gestational diabetes must be “carefully considered” in the context of local pandemic impact, including community transmission rates, researchers wrote in an article published in the European Journal of Endocrinology; however, current evidence does not support a single, alternative test.
Meek and colleagues assessed whether the new criteria would effectively diagnose gestational diabetes, and which cutoff points would be best to use. Researchers analyzed data from 17,736 women with singleton pregnancies who underwent a random plasma glucose measurement followed by a 50 g OGTT at 24 weeks’ gestation between 2004 and 2008, as well as 826 women with gestational diabetes between 2014 and 2019 who received standard clinical management. Researchers also assessed data from 361 women with at least one gestational diabetes risk factor, recruited from the OPHELIA study, an ongoing, prospective observational study of hyperglycemia in pregnancy. For OPHELIA, pregnant women with at least one gestational diabetes risk factor underwent a 75 g OGTT with measurement of HbA1c (8.3% of women had gestational diabetes).
Pregnancy outcomes included gestational diabetes using U.K. National Institute for Health and Care Excellence (NICE) criteria, diabetes in pregnancy using WHO criteria, cesarean section, large for gestational age infant, neonatal hypoglycemia and neonatal ICU admission. Researchers used receiver-operating characteristic (ROC) curves and unadjusted logistic regression analyses to compare random plasma glucose, FPG and HbA1c performance.
Gestational diabetes diagnosis was associated with random plasma glucose at 12 weeks, with an ROC of 0.81 for both diagnostic criteria (95% CI, 0.79-0.83). FPG was also associated with gestational diabetes diagnosis, with an ROC curve of 0.75 for U.K. guidelines (95% CI, 0.65-0.85) and 0.92 for WHO guidelines (95% CI, 0.85-0.98).
HbA1c at 28 weeks’ gestation was similarly associated with gestational diabetes diagnosis, with an ROC of 0.83 for U.K. guidelines (95%, CI 0.75-0.9) and 0.84 for WHO guidelines (95% CI, 0.77-0.91).
“Each measure predicts some, but not all, pregnancy outcomes studied,” the researchers wrote.
In assessing sensitivity and specificity of thresholds for random plasma glucose, HbA1c and FPG to predict gestational diabetes, researchers found that to identify a similar proportion of women as detected by the NICE criteria (5%) would require a 12-week random plasma glucose of at least 8.5 mmol/L (42% sensitivity; 96% specificity), a 28-week HbA1c of at least 39 mmol/mol (HbA1c 5.7%; 26% sensitivity; 96% specificity); or an FPG of at least 5.2 mmol/L to 5.4 mmol/L (sensitivity 18%-41%; specificity 97%-98%).
“We found that each of the three tests studied were associated with gestational diabetes diagnosis and outcomes,” Meek said. “The tests each performed well at excluding the diagnosis, with high test specificity; however, each individual test would miss around 60% to 80% of women with gestational diabetes, or poor test sensitivity.”
Meek said the findings demonstrate that this diagnostic pathway is therefore acceptable when there is no alternative, but OGTTs offer much better diagnostic performance.
“We were surprised that the new tests were associated with outcomes, but we were disappointed so many women with gestational diabetes would be missed,” Meek said.
“The new diagnostic criteria are suitable for short-term use during the COVID-19 pandemic, but are not an acceptable long-term alternative to the oral glucose tolerance test,” she said. “The new diagnostic criteria will fail to identify substantial numbers of women with gestational diabetes. Women who are concerned about this should contact their midwives for further testing if available.”
For more information:
Claire Meek, PhD, MBChB, can be reached at [email protected].