News in brief: ECG model to improve detection of structural heart disease; Cardiac investigations surge after COVID-19 vaccination; Reducing unnecessary imaging is one step to decarbonisation

11 May 2022

ECG model to improve underdetection of structural heart disease 

A novel ECG-based machine learning approach can identify patients at high-risk for structural heart disease endpoints including moderate or severe valvular disease, aortic regurgitation, mitral stenosis, mitral regurgitation, tricuspid regurgitation, reduced left ventricular ejection fraction, and increased interventricular septal thickness.

A US team developed the model using data from 2,232,130 ECGs from 484,765 adult patients who met criteria for at least one positive or negative individual disease label. The final model was validated and tested in multiple independent sets of patients.

“The rECHOmmend model using age, sex, and ECG traces for prediction of the composite disease label yielded an AUROC of 0.91 [95% CI 0.90, 0.91] and a PPV of 42% at 90% sensitivity with 18% disease prevalence,” the study said.

“Furthermore, we demonstrated that the combination of these distinct endpoints into a single platform tied to a recommendation for a singular, practical clinical response—follow-up echocardiography—resulted in an overall PPV of 42% for clinically meaningful disease while maintaining high sensitivity (90%) and specificity (73%).”

The researchers said the model enables targeted echocardiographic screening to help detect unrecognised and underdiagnosed structural heart disease.

Read more in Circulation

Cardiac investigations surge after COVID-19 vaccination

Emergency departments have seen a surge in presentations by young people requiring cardiac investigations for symptoms such as chest pain and palpitations following receipt of COVID-19 vaccination.

A review of almost 1,000 patients who attended the ED at St Vincent’s Hospital Melbourne between August and October 2021 with post-vaccine symptoms found that they accounted for 6% of the total ED workload, but almost all of them were discharged home after initial investigations.

Most of the patients (83%) were under 45 years of age and 20% arrived by ambulance. Chest pain was the most common presenting complaint (44%), followed by headache (10%) and palpitations (8.2%).

The most common investigations were a full blood examination (74%), an ECG (64%) and serum troponin (49%). Two thirds (65%) of patients were directly discharged home and 22.1% were sent home after a short stay admission. Only 2.2% of patients were admitted to the hospital.

The study investigators said the surge in presentations was likely driven by public concerns about prothrombotic complications of vaccines, but only 0.3% of the study population received a diagnosis of pulmonary embolism.

Department of Health guidelines had recommended that patients with concerns should have initial assessment and investigation in a general practice or an ambulatory outpatient cardiology setting, with EDs reserved for unwell and high-risk patients.

However lack of access to GPs and cardiology outpatient consultations resulted in a huge increase in ED workload and occupancy and increased diagnostic investigations

The findings highlighted an urgent need for public education initiatives and alternate treatment pathways such as nurse assessment helplines to reduce anxiety-related presentation to emergency departments, the authors said in the Emergency Medicine Journal.

Reducing unnecessary imaging is one step to decarbonisation

Clinicians can help reduce the healthcare carbon footprint by adopting a three-pronged strategy to make more efficient use of diagnostic imaging, according to Australian researchers.

Diagnostic imaging and pathology testing account for almost 10% of the hospital carbon footprint, with MRI and CT scans accounting for a high proportion of it, a Melbourne University study found.

The carbon emissions from an MRI are equivalent to driving a car for 145 km, while a CT scan carbon emission is equal to driving 76 km, the findings in Lancet Regional Health showed.

Much of the large carbon footprint was due to electricity use by scanners, and in particular, their standby power use, said the researchers, who recommended that clinicians and administrators make efforts to reduce unnecessary imaging and/or switch imaging to a lower carbon modality such as X-rays.

Other carbon footprint reduction tips include turning scanners off when they are not required rather than leaving them on standby and ensuring existing scanners have high utilisation rates, they suggested.

Already a member?

Login to keep reading.

Email me a login link