Author + information
- Catherine S. Bennet, MD∗ (, )
- Fardous C. Abeya, MD,
- Ari Hoffman, MD,
- Joselyn Rwebembera, MD,
- Michael H. Picard, MD,
- Malissa J. Wood, MD and
- Samson Okello, MD
- ↵∗Division of Cardiology, Department of Medicine, Johns Hopkins School of Medicine, 1800 Orleans Street, Zayed 7125, Baltimore, Maryland 21287
Greater than 75% of the burden of cardiovascular disease is concentrated in low- and middle-income countries that have few trained cardiologists (1). Given the constraints of formal cardiology training in these regions and the increasing availability of diagnostic technologies there, it is prudent to determine how noncardiologists can efficaciously use these technologies to improve cardiac care. To this end, we sought to compare the transthoracic echocardiography (TTE) interpretations of internists based in Uganda who underwent abbreviated TTE training with those of a U.S.-based, board-certified cardiologist and echocardiographer to inform the ability of this training approach to allow for the accurate application of TTE technology.
Two Uganda-based internists without prior TTE training (A.H., J.R.) underwent a 3-day in-person TTE course led by cardiologists from the Ugandan Heart Institute on knobology, image acquisition, and performing and interpreting TTE measurements. They also used Catherine Otto’s Textbook of Clinical Echocardiography (2) for reference and accessed online resources. With this training, they performed and interpreted 685 TTEs for adult patients referred for TTE based on local practice at Mbarara Regional Referral Hospital from February 2013 to February 2014. TTEs were interpreted for left ventricular ejection fraction (EF) and valvular stenosis and regurgitation using an echo machine with full imaging and color, pulse-wave, and continuous-wave Doppler capabilities. Patients younger than 18 years of age and TTE reports without available images were excluded. A cardiologist (M.J.W.) with formal fellowship training and certification by the American Boards of Cardiology and Echocardiography reread 303 TTE studies meeting inclusion criteria while blinded to the Ugandan interpretations. We used Cohen’s kappa statistics to determine inter-reader agreement between assessments of clinically relevant TTE parameters—categorized ejection fraction (EF) (<35%, 35% to 55%, >55%) and the severity (none, trace, or mild vs. moderate or severe) of valvular regurgitation and stenosis.
Based on the interpretation of the U.S. echocardiographer, there were 53 TTEs with EF <35%, 57 with EF 35% to 55%, and 193 with a normal EF of >55%. There was moderate inter-reader agreement for categorized EF (Table 1). A Bland-Altman analysis of EF revealed a mean difference of 5.63% between the Ugandan interpretation and that of the U.S.-based cardiologist.
The prevalence of moderate or severe valvular lesions in our population was relatively low, with fewer stenotic than regurgitant lesions. There was 1 case of aortic stenosis, 11 cases of aortic regurgitation, 7 cases of mitral stenosis, 60 cases of mitral regurgitation, and 63 cases of tricuspid regurgitation. There was moderate agreement for the degree of aortic stenosis and aortic regurgitation. There was good inter-reader agreement for the degree of mitral stenosis, mitral regurgitation, and tricuspid regurgitation (Table 1). Tricuspid stenosis and pulmonary valve parameters were not assessed.
Our results show that the interpretations of TTEs by providers in Uganda with abbreviated TTE training had moderate to good agreement with those of a cardiologist with formal fellowship training and board certification. Correct assessment of EF had greatest agreement at the clinically relevant extremes. These results are similar to other studies looking at focused TTE training of internal medicine residents, nurses, and emergency medicine physicians (3).
Our study is limited by the low prevalence of valvular stenosis and regurgitation in our population. Furthermore, there is possible selection bias given that 51% of TTE reports did not have corresponding images and thus were not reread by the U.S. echocardiographer. However, there are no statistical differences between the characteristics of analyzed and excluded studies in terms of patient demographics or reason for TTE referral. It is also important to note that the reads of both Ugandan providers were combined in our analysis and only 1 board-certified cardiologist and echocardiographer reread the TTE images. Therefore, we were unable to quantify the degree of inter-reader variability that could be expected between TTE interpretations regardless of level of training.
In conclusion, focused TTE training may be a reasonable alternative to subspecialty cardiovascular echocardiography training in resource-limited settings. Abbreviated training allows for the reliable assessment of cardiac structure and function with the goal of answering clinically relevant questions that inform appropriate cardiac care. Importantly, our results also show that there is room for improvement in this training model and should serve as a call to action. There is a great opportunity for formally trained cardiologists, echocardiographers, and sonographers to work with clinical providers in resource-limited settings to improve goal-oriented TTE training and robustly improve cardiac care in these regions.
Please note: Dr. Okello was supported by the Bernard Lown Scholars in Cardiovascular Health Program (Boston, Massachusetts). Dr. Wood has served as a consultant for Boehringer Ingelheim; and received research support from Abbott. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- 2017 American College of Cardiology Foundation