Written by Clay Smith
Had the Pregnancy-Adapted YEARS (PAY) algorithm been applied in the DiPEP population, it would have missed a significant number of PEs. But there are several limitations to this study.
Why does this matter?
DiPEP looked at the diagnostic utility of a number of biomarkers for the workup of PE during pregnancy. None was useful alone. Recently, the PAY algorithm was used to effectively work up suspected PE in pregnancy. In brief, the YEARS criteria were applied and varying D-dimer cutoffs used based on risk. MDCalc has a brilliant calculator for YEARS, for which you can select pregnant or not, if you need a refresh on this. In the PAY study with 498 patients, just one DVT was later diagnosed within 3 months. How would PAY fare if applied in another patient population?
Why was DiPEP so different?
This was a secondary, retrospective analysis of the DiPEP cohort. They found far worse diagnostic accuracy. When the PAY algorithm was applied to the DiPEP cohort, the sensitivity was a paltry 58.3%. Using a Geneva + D-dimer strategy, sensitivity improved to only 75%. Why was DiPEP so low compared to the original PAY cohort? DiPEP only included patients who had PE imaging ordered. So, these were higher pretest probability patients by definition. Also, 73% in DiPEP had anticoagulation prior to obtaining a D-dimer, which may have falsely lowered the levels. In addition, “the five ‘missed’ PE [using PAY] were reported as small or segmental.” Most importantly DiPEP was a retrospective look at data that was initially collected for another purpose. How accurate was the DiPEP case report form for clinical signs of DVT, hemoptysis, or whether PE was the most likely diagnosis? For now, Pregnancy-Adapted YEARS is one of the largest and best studies we have to guide us in evaluating these patients. The poor performance in the DiPEP study gives me pause, but I don’t plan to abandon the PAY algorithm yet.
Accuracy of PE rule-out strategies in pregnancy: secondary analysis of the DiPEP study prospective cohort. Emerg Med J. 2020 Apr 9. pii: emermed-2019-209213. doi: 10.1136/emermed-2019-209213. [Epub ahead of print]
Open in Read by QxMD