February 2022

Clinical Question

Does automatic age-adjustment of D-dimer cut-off on an EHR reduce performance of imaging tests?

Title of Paper

Multi-centre implementation of automated age-adjusted D-Dimer results reduced unnecessary PE imaging

Journal and Year

American Journal of Emergency Medicine. 2021.

Lead Author

Jeffrey Dubin

Background

– The concept of age adjusted d-dimer – that is, increasing the cut-off for a ‘negative’ value based on the age of your patient – has been around for a while now – I think it was first suggested in 2013. Since then it has been slowly making its way into practice, but my experience is that it isn’t being widely applied.

Study Design

– Retrospective cohort study in 6 hospital EDs (large urban and small community hospitals)
– Positive DD cut-off increased from 0.5 to age x 0.01
– Matched data from 1.5 years before change to 1.5 years post
– D-Dimers ordered at clinician discretion (with assistance of decision aids on the EHR – geneva and PERC for PE and Wells for DVT) – in place a year before the study started

Outcomes

– Primary: proportion of patients who underwent D-dimer imaging (CTPA or V/Q) as follow up to D-dimer testing
– Secondary:
– Diagnostic accuracy
– Missed PE on index ED visit – chart review of radiology records
– PE related 30 day mortality

Summary of Results

– D-dimers performed: **Pre**: 22,302 | **Post**: 25,082
– Positive: **Pre**: 48.6% | **Post**: 43.2% (p<0.001)
– 5.4% reduction in positives, p<0.05
– No difference in age, or mean d-dimer level (1.3)
– Imaging: **Pre**: 32% (7218/22302) | **Post**: 28% (7017/25082)
– 4.4% decrease, p <0.05
– = 1104 fewer imaging studies (in 1.5 years = 2 scans a day – ***across 6 hospitals***)
– 5.2% pre and 4.2% post had imaging despite neg D-dimer
– 4/1202 were positive (3 pre: 1x intermed prob V/Q, 1x subseg CTPA, 1x lobar | 1 post: subsegmental on CTPA)
– 1 missed PE in the “new negative” group – sub-segmental PE on CTPA, 65 yr old with d-dimer 0.55
– No PE related deaths in the ‘new negative’ group

– Test Characteristics for D-dimer:
– Pre: Sn 96.7% | Sp 52.6%
– Post: Sn 96.1% | Sp 58.2%
– Difference 5.4%, p<0.05

Discussion Points

Absolute CT imaging reduction on the lower end of previously reported range (5-18%). Probably due to increased use of d-dimer, particularly in younger cohort, in American practice compared to European.

Authors Conclusion

Overall, we found that age adjusted D-dimers resulted in a significant systemwide decrease of PE imaging of 4.4% without adversly affecting our safety endpoints. We believe that automated age-adjusted D-dimer can be implemented safely in other systems to reduce imaging usage.

Clinical Bottom Line

Age-adjusted D-dimer is far more likely to get traction if its included in the EHR reporting like this. This also adds to the body of evidence that AAD-d significantly improves the accuracy and specificity of D-dimer testing without too much impact on the sensitivity.

Not measured but likely to have additional benefits on ED throughput and LOS

– This could be even more of benefit in systems like the UK where these patients would be sent to the medics for their subsequent CTPA which would (at least in my days there) have occurred the next day with tx dose clexane until then.
– In Australia slightly less of an impact as we just get on with the scan from the ED, generally withholding the anticoagulation until we’ve seen a positive result, unless we think its a big PE clinically.

Authors:
Dr Becky Maxwell
Dr Chris Connolly

Authors:
Andy Neill
Dave McCreary

Clinical Question
– what should we image post OOHCA

Paper Title
– Early headtopelvis computed tomography in outofhospital circulatory arrest without obvious etiology https://pubmed.ncbi.nlm.nih.gov/33606342/

Author
– Branch, AEM

Background
– cardiac arrest is associated with bad outcome no surprise
– we have good pathways for STEMIs and know what to do with them
– everything else gets dumped in a big vague bucket
– knowing the cause of the arrest is probably imprioatn and after the ECG the question is what other investigation shold we be doing
– CT brains are common with an unclear yield and i have seen a lot of CTPAs ordered in the context of non obvious arrests
– this paper hopes to answer if we should look further

Methods
– prospective observational in a single centre
– total body CT within 6 hrs
– CT was brain, gated CTCA and abdo/pelvis
– patients “with no obvious cause of arrest”
– they had blinded radiologists and outcome asssessors looking at cause of arrest which is a nice touch mehthods wise
– they had some time critical pre defined causes they wanted to examine

Results
– 100 pts
– CT identified cause in 40% and in >90% of those where a CT might be expected to be useful.
– 15% cause identified by CT that wouldn’t have been picked up if the CT wasn’t done.
– While this seems small I think this is huge given how sick the cohort was. Within those causes most were PE and abdo disasters
– interestingly coronary occlusions in that group were small (though they have been almost systematically excluded by this stage)

Thoughts
– this is hardly definitive data and you will scan a lot of people and not find very much
– but I don’t think cardiac arrest is the place where you want to be skimpy on your diagnostic tests
– this is not low risk chest pain where testing unnecesarily is a big issue
– these guys are all going to the ICU at 3 grand a day, getting them a CT is not the issue
– and i can tell you that if they dont’ get that CT in the ED, they will liekly be back down tomorrow from the ICU to get it done.
– with the ICU hat on, knowing the body has been imaged fully is a great reassurance to me – it means i can focus on all the supportive care and not be worrying for example if it’s safe to anticoagulate as i don’t know if there’s a brain bleed or a splenic injury from the LUCAS
– this goes against myb usual bias in EM which is less testing is probably the way to go but when it comes to the critically ill

Authors:
Dr Michelle Tipping
Dr Mark Winstanley

References:

RCEMLearning: burnout
GMC: doctors-burnout-worsens-as-gmc-report-reveals-pandemic-impact
Physician burnout uk: mental-health-resources
Civility Saves Lives

NICE Hyponatraemia
litfl.com
emergencymedicinecases.com