À compter du 1er décembre 2023, l’accès à POEMs et à Essential Evidence Plus ne fera plus partie des avantages offerts aux membres de l’AMC.
Question clinique
Can increased rates of gestational diabetes be linked to changes in laboratory protocol?
L’Essentiel
A seemingly innocuous change in the laboratory processing of tubes for glucose determination of women undergoing oral glucose tolerance testing was associated with an almost doubling of the number of women who received a diagnosis of gestational diabetes mellitus. If rates of diabetes seem to be going up in your practice, consider a change in protocol as the cause of spurious increases. 4
Référence
Plan de l'etude: Descriptive
Financement: Self-funded or unfunded
Cadre: Other
Sommaire
This study is a before-after, natural experiment to determine the effect of immediate versus delayed testing of glucose samples in pregnant women undergoing oral glucose tolerance tests. A total of 7509 women had blood collected and sent to a central laboratory for blood glucose analysis over a 27-month period. Then, the protocol for analysis was changed and glucose testing was performed immediately on collected samples and results were reported for 4808 women. As compared with the delayed testing protocol, mean glucose concentrations for the fasting, 1-hour, and 2-hour samples were 4.3 mg/dL (0.24 mmol/L; 5.4%), 6.1 mg/dL (0.34 mmol/L; 4.9%), and 2.8 mg/dL (0.16 mmol/L; 2.3%) higher using the immediate processing. Although a seemingly small increase (maximum 5.4%), there were almost twice as many diagnoses of gestational diabetes mellitus following the change in laboratory protocol, from 11.6% to 20.6%. Another recent analysis showed how variability in laboratory assays can affect clinical decisions.
Reviewer
Allen F. Shaughnessy, PharmD, MMedEd
Professor of Family Medicine
Tufts University
Boston, MA
Commentaires
The "Bottom Line" implies the wrong message
The original study (https://pubmed.ncbi.nlm.nih.gov/32350020/) compared two different protocols:
1) delayed centrifugation of blood samples (lab's original protocol)
2) immediate centrifugation of blood samples (changed after 6/2017)
As per the original article: "Preanalytical processing of blood samples can markedly affect the results of plasma glucose readings because ongoing glycolysis by erythrocytes and leukocytes prior to centrifugation utilizes glucose, lowering its concentration (4–9)." Which means the lab was correcting a problem with their protocol.
As per the "Bottom Line" summary of this POEM:
"A seemingly innocuous change in the laboratory processing of tubes for glucose determination of women undergoing oral glucose tolerance testing was associated with an almost doubling of the number of women who received a diagnosis of gestational diabetes mellitus. If rates of diabetes seem to be going up in your practice, consider a change in protocol as the cause of spurious increases. 4"
This statement's use of the word "spurious" implies that the increase in GDM diagnoses is the problem. In fact the delayed centrifugation protocol was the problem, resulting in half of the cases of GDM being missed. The delay was artificially causing the glucose concentrations in half the samples to drop below the threshold for diagnosing GDM.
The "Bottom Line" message is the opposite of the original article.
Upon further reflection, a…
Upon further reflection, a more fitting"Bottom Line" message:
"If you perform venipuncture in your office, you should remember that specimen handling can have a significant impact on results.
"A delay in centrifugation likely resulted in half the GDM cases being missed. If you take blood but don't have a centrifuge, you are probably under-diagnosing DM."
Sorry, that last one was…
Sorry, that last one was still me
Excellent POEM
Excellent POEM
Gestational diabetes
This is not a useful endpoint
We would need to know whether this excess number were also at risk for complications of pregnancy