Echocardiographic screening for subclinical rheumatic heart disease: Improving screening through simplification of the diagnostic criteria
Thesis (PhD)--Stellenbosch University, 2020.
ENGLISH ABSTRACT: Rheumatic heart disease (RHD) remains one of the leading causes of cardiovascular morbidity and mortality in developing countries withSub-Saharan Africa being identified as an endemic area. The early detection and initiation of secondary prophylaxis in children with ‘latent’ RHD remain attractive primary health care interventions, particularly in endemic regions with no or limited access to specialist cardiac services. However, the current consensus-derived screening criteria endorsed by the World Heart Federation (WHF criteria) are overly complex, require the use of expensive echocardiographic equipment with Doppler functionality and identify a large borderline diagnostic group that demonstrates a predominantly benign outcome in longitudinal study. This raises concerns regarding the feasibility of large-scale screening in resource-poor regions and questions the utility of early echocardiographic case-detection of RHD. The primary purpose of this thesis was to critically appraise the performance of the WHF criteria and todetermine whether a set of screening criteria based on a novel, focused morphological and mechanistic evaluation would simplify the current WHF guideline and reduce the number of cases ‘misclassified’ with borderline RHD whilst maintaining a similar degree of sensitivity. A literature review was undertaken that critically appraised the performance of the current WHF criteria and its impact in African RHD screening programs. This highlighted important logistical and methodological shortcomings that have curtailed the implementation of large-scale RHD screening in RHD endemic regions. The five-year experience of a large-scale, high-risk RHD screening program (Echo in Africa [EIA] project) was analysed. The results from this project highlightRHD as an ongoing, significant healthcare challenge amongst underserved communities within the Western Cape, South Africa.The estimated prevalence of WHF ‘definite-’ and ‘borderline-RHD’ of 9.1 cases/1000 and 19.5 cases/1000 reported by EIA is significantly higher thanthatpreviously described in this region. Furthermore, a critical appraisal of the WHF criteria’s performance in the EIA cohort highlighted various redundant and ambiguous criteria that require revision.Inter-scallop separations (ISS) of the posterior mitral valve leaflet (PMVL) were described in both our high-and very low-risk populations. They were a common finding and the principal cause of WHF ‘pathological’ mitral regurgitation (MR) in the ‘borderline RHD’ group in both cohorts. This supported theirstatus as a normal and importantly, non-rheumatic variant.The reliability of the current WHF anterior mitral valve leaflet (AMVL) thickness assessment was evaluated and was demonstrated to be poor amongst readers despite controlling for systematic bias. This raised the possibility of introducing a non-measurement-based AMVL thickness evaluation. A novel screening definition of AMVL restriction was introduced, enabling the description of a variable spectrum of AMVL restriction amongst children.This definition reliably identified two subtypes of leaflet restriction: a normal, ‘gradual bowing’ variant that localised predominantly to the medial portion of the leaflet and a ‘distal tip’ variant seen to affect at least the central portion of the leaflet in all cases of WHF ‘definite RHD’ in this cohort. Finally, this thesis culminated in the development and evaluation of a novel set of morpho-mechanistic (MM) echocardiographic screening criteria for RHD. Together with an abbreviated ‘rule-out’ screening test, the MM criteria were assessed alongside the current WHF criteria in a gold standard RHD-negative cohort and a gold standard RHD-positivecohort. The MM criteria significantly reduced the false-positive rate of a borderline diagnosis inthe gold standard RHD-negative cohort (2.7/1000 vs 41.8/1000) whilst maintaining a similar screening sensitivity (99.7%) compared to the WHF criteria (95.9%) within thegold standard RHD-positivecohort. Similarly, the MM RHD ‘rule-out’ test performed well by excluding the majority of cases (98%) within the gold standard RHD-negative cohort while including all cases within the gold standard RHD-positive cohort. The work presented in this thesis addresses key research needs and gaps in our current understanding of ‘screen-detected’ latent RHD. It representsa significant contribution that will impact on policy, practice and further research in the field. The discovery that ISS of the PMVL are a normal finding and the principal cause of isolated ‘pathological’ MR in the borderline group represents a key element in solving the ‘borderline conundrum’. This discovery supported the adoption of a morpho-mechanistic screening approach over a predominantly functional MV assessment. Centred around a novel definition of AMVL restriction, the MM criteria significantly improve the specificity of RHD detection by markedly reducing the size of the borderline group. Importantly, this was achieved without a reduction in the sensitivity of the criteria when compared to the current WHF criteria. Together with a simple ‘rule-out’ test, the MM criteria bring us closer to the objective of implementing large-scale screening programs that identify children with latent RHD who will benefit from secondary prophylaxis.