Zachary Messina, DO | EM/IM PGY3
The Article:
Daley et al. Increased Sensitivity of FOCUS for PE in ED Patients with Abnormal Vitals. Academic Emergency Medicine. November 2019. Vol 26 (11). 1212-1220.
The Idea:
To determine the sensitivity and specificity of the FOCUS exam as well as each component of the focused cardiac ultrasound (FOCUS) in two subsets of patients with abnormal vital signs.
The Study:
This study is a prospective, blinded, observational multicenter cohort study of two subsets of patients. The study was performed at six academic emergency rooms and involved seven ultrasound-fellowship-trained emergency attendings, three PGY-3 emergency room residents, as well as three third year medical students. The enrollment of the study was determined by a power calculation in place of a timeframe. Patients were placed in one of two groups, either the primary grouping which had a requirement of heart rate of greater than 100 and/or hypotension with systolic blood pressure less than 90 mmHg, or a subgroup which required only heart rate to be greater than 110 beats per minute. All patients were 18 years of age or older and all patients underwent computed tomography with angiography (CTA) of the chest. Prisoners, wards of the state, non-English speaking patients, and those where investigators were not able to obtain any echocardiographic data due to technical challenges were excluded. The goal was to have sonographers perform FOCUS prior to CTA; however, if FOCUS was performed after CTA the sonographers were blinded to the results of the CTA. All personnel in the study underwent standardized training consisting of a brief video and 1-hour didactic session. Two of the three residents also underwent an additional didactic session; the third already had prior experience in collecting ultrasounds. The three medical students underwent a 1-hour didactic session and 1-hour hands-on training session by the third resident as well. The medical students were required to perform 20 FOCUS exams prior to the study initiation to ensure competency. The FOCUS exam consisted of the parasternal long axis, parasternal short axis, apical four chamber, and subxyphoid views with a measurement of TAPSE, visual inspection of right ventricular enlargement, septal flattening, presence of tricuspid regurgitation on color Doppler, or monitoring for McConnell’s sign. If any of these components were present, the FOCUS exam was positive, if all components were absent the FOCUS was negative. TAPSE in this study was defined as <2.0cm instead of the conventional <1.7cm to increase sensitivity at the loss of specificity. Finally, inter-rater reliability was also tested as the site principal investigator by their own review in 104 of 136 patients in the right ventricular components in comparison to the sonographer’s interpretation.
The Findings:
First, 143 patient’s underwent CTA during the study, but seven were excluded. Three were due to technically difficult FOCUS exams, and only four were due to non-English speaking language barrier. There were 136 patients enrolled during the study, 37 had a pulmonary embolism on CTA. Of these 37 patients, six were hypotensive, 28 were normotensive with at least one right ventricular component identified, and three were normotensive without any component identified. In the comparison of FOCUS versus CTA in patients who had a PE, the following data apply. In the primary group of patients (n = 136), the sensitivity of FOCUS was 92% (95% CI = 78-98%). Of the right ventricular components, TAPSE of <2.0cm was the most sensitive at 88% (95% CI = 72-97%), and in comparison the TAPSE value of <1.7cm was at 67% (95% CI = 48-82%). The other components ranged from a sensitivity of 35% to 51%. The overall specificity of FOCUS was 64% (95% CI = 53-73%), with the McConnell’s sign component being most specific at 99% (95% CI = 94-100%). The other components ranged from 64% to 93% in specificity.
In the subgroup of patients with a heart rate of >110 (n = 98), the sensitivity of FOCUS for PE was 100% (95% CI = 88-100%), with the most sensitive component being TAPSE of <2.0cm, which was 93% sensitive (95% CI = 75-99%). The sensitivity of TAPSE at <1.7cm was 77% (95% CI = 56-91%). The other components ranged in sensitivity from 36% to 57%. FOCUS specificity in this subgroup was 63% overall (95% CI = 51-74%), the most specific being the McConnell’s sign at 100% specificity (95% CI = 95-100%). The other components ranged from 63% to 93% in specificity.
Finally, inter-rater reliability for whether the FOCUS exam was positive or negative by two separate sonographers was 1.0 (95% CI = 0.31 to 1.0). Of the individual components, TAPSE had a kappa statistic of 0.61 (95% CI = 0.31 to 1.0), septal flattening had 0.88 (95% CI = 0.69 to 1.0), right ventricular enlargement had 0.89 (95% CI = 0.7 to 1.0), McConnell’s sign had 0.89 (95% CI = 0.7 to 1.0), and tricuspid regurgitation was 0.81 (95% CI = 0.64 to 1.0).
Limitations of this study:
From the study set-up, one of the three residents performing ultrasound was more experienced. To offset this, additional training was provided to the other residents in hopes this would bridge any gap in experience. In addition, the three medical students were all third year and inexperienced in ultrasound. While their data was also compiled into the resident data, there was concern for inaccuracy given inexperience. This was attempted to be compensated for in additional training and the performance of 20 FOCUS exams prior to the study initiation. These limitations may lead to the acceptance of incorrect data, however, this was demonstrated in terms of Kappa statistics.
In terms of data collection, nine patients had missing data for TAPSE, two for right ventricular enlargement, two for septal flattening, two for McConnell’s sign, and 60 for tricuspid regurgitation. There was also unintentional unblinding of two patients in the study to the sonographer. These limitations may primarily cause the unintentional underestimation of sensitivity or overestimation of specificity in the data collected.
Finally, the study was created using a statistical power, which may cause bias, however, the power was meant to limit the width of sensitivity to no greater than 20%. Selection bias may also confound these data in terms of the exclusion of patients; however, the number of patients excluded was kept to a minimum.
The Takeaway:
The FOCUS exam in patients with abnormal vital signs may help to significantly lower the likelihood of pulmonary embolism in most patients who are suspected of having a pulmonary embolism. This was especially true in patients with a heart rate of >110 beats per minute. Further study with an unpowered cohort is needed to determine of FOCUS can exclude PE.
