You Are Called to Admit Pt from ED, Diagnosed Today with Moderate Sized PE on CTPA

EBM Conference

TJ O’Neill

6/25/2009

Clinical Question:

You are called to admit pt from ED, diagnosed today with moderate sized PE on CTPA.

Pt has dyspnea only w/ exertion, not at rest VS HR 102RR 24 Sat 93% RA BP 124/82

It is 445PM, the very nice but overworked Cardiology fellow will only do stat TTE if it will change management tonight and this patient does not appear to be candidate for thrombolytics.

How can pt be stratified to floor vs stepdown w/o echocardiography?

Prior studies have demonstrated poor outcomes in patients with RV strain on echocardiogram and EKG findings have been shown to be predictive of pulmonary artery hemodynamics . EKG is of unclear prognostic significance in moderate risk patients

Study type: Prospective outcome trial based on diagnostic test

Study population: 386 pts eventually included who presented to single center ED in Italy with complaints that raised suspicion for PE, which later was confirmed with imaging. (Recruitment began prior to diagnosis of PE)

Inclusion criteria

• Age >18
• Suspicion for PE, confirmed w/ VQ, CT, or angiography

Exclusion criteria:

• Persistent SBP <100
• Hx of prior PE
• Hx of “severe COPD”
• TTE findings of long-term RV overload (RV hypertrophy)

Study protocol:

• All patients underwent EKG analysis after recruitment
• EKGs were interpreted in blinded fashion with up to 3 expert readers
• Incomplete or complete RBBB
• TWI in V1-3
• S1Q3T3 (really just S1Q3)
• All patients also underwent TTE in the ED within 1 hour of diagnosis of PE
• All patients treated with heparin and generally transitioned to Coumadin (3% received IVC filters)

Desired study outcomes:

• Primary: Death or deterioration prior to discharge
• Progression to shock
• Need for mechanical vent or CPR
• Need for catecholamines other than Dopa <5.
• Secondary: Death and deterioration separately
• Patients were followed until endpoint occurred or they were discharged

Statistical Analysis:

• Multivariant Cox proportional hazard regression model was applied based on variables with probability values of less than 0.1
• Adjustments were made for thrombolytic therapy

Results:

• RV strain on EKG was associated with highly significant risk of progression to primary endpoint (as well as secondary endpoints) Table 2
• This was maintained in a multivariable analysis (whereas tachycardia, hypoxia, diaphoresis were not)
• When TTE findings and EKG findings were considered together, a very high risk group of patients with both EKG and echo findings of RV overload were detected.

Are the results of this prognosis study valid?

Was a defined, representative sample of patients assembled at a common (usually early) point in the course of their disease?

Yes , with the important exception of exclusion of patients who may have confounded EKG analysis.

Was patient follow-up sufficiently long and complete?

Until discharge or death, reasonable although could have looked longer.

Were objective outcome criteria applied in a "blind" fashion?

Unclear, all but shock appear to be fairly clear cut and shock does not appear to be the primary

However, patients may have been managed differently given that TTE results were available up front.

For example if pt with PE drops BP knowledge of RVS or RVD might affect choice of fluids vs pressors which would count as an endpoint

If subgroups with different prognoses are identified, was there adjustment for important prognostic factors?

Multivariant analysis was conducted, with exception that age did not appear to be included

Was there validation in an independent group ("test set") of patients? No

Are the valid results of this prognosis study important?

How likely are the outcomes over time?

Composite endpoint 6%, death 3%. Probably similar now

How precise are the prognostic estimates?

HR 3.75 (1.58-8.89) with P 0.038 in univariant analysis but HR 2.58 (1.05-6.36) P= 0.38 in multivariant

Can we apply this valid, important evidence about prognosis to our patient?

Are the study patients similar to your own?

Relatively lower use of CTA, and no use of Lovenox, although these would not be expected to have large effect. European population There was also a higher rate of thrombolysis ( supposedly controlled for.)

Will this evidence make a clinically important impact on our conclusions about what to offer or tell our patient?

Two main criticisms- Exclusion of pts w/ RVH and lack of blinding of results to treating physicians in the setting of soft outcome measures

This makes it difficult to apply results to our patient. Additionally, as an independent variable it changes only incrementally the absolute risk of bad outcome (even if significant).

A better done study would have provided at least some prognostic data which could be used to determine which patients require a higher level of care, but this prognostic information is probably available without EKG.

In summary, a good negative predictor does not help manage a patient who would be admitted anyway and a good positive predictor does not help if pt will not have thrombolysis.