Rheumatic heart disease: infectious disease origin, chronic care approach

Primordial prevention of RHD is predicated on reducing exposure to GAS and GAS infections (Fig. 2). The strongest evidence for a causal association between a primordial (socioeconomic) determinant and RHD risk is for household or bedroom crowding. There have been at least 50 studies over several decades examining the association between crowding and either GAS infection, ARF or RHD. These are mostly of limited quality but collectively, supporting an association between crowding and ARF risk. Examples include the following. In an ecological study undertaken in New Zealand, Jaine et al. identified that the rate ratio of ARF cases was significantly associated with crowding quintile, according to population-level data [44]. In Wannamaker’s pivotal military barracks investigations published in 1954, decreasing distance of one’s bed from a case of GAS infection was identified as a risk factor for GAS acquisition, and rate of acquisition of GAS was also associated with the number of GAS carriers in a barrack group [45]. In a case-control study of 148 ARF cases and 444 controls from Serbia investigating the influence of socioeconomic factors on the occurrence of ARF, smaller available living space (<5 m2), having ≥2 persons per room or have ≥2 persons per bed were all significantly more common among ARF cases than controls [46]. In a more recently Australian study, the number of cases of pyoderma (mostly due to streptococci) per household was directly proportional to the number of people per bedroom in one Aboriginal community [47].

There are limited examples of primordial-level interventions in practice, since strategies to alleviate the elements of poverty that drive RHD are a global challenge. In NZ children with ARF and their families are prioritised for higher quality or larger public housing. NZ Guidelines [48, 49] and draft guidelines under review in Australia by the Communicable Disease Network of Australia, recommend children are educated about respiratory hygiene to reduce GAS spread. In Australia, GAS skin infections (impetigo) also appear to be an antecedent of ARF, hence research activities in some jurisdictions have focused on provision of ‘health hardware’, including running water and soap to reduce skin sores and ARF [50] and other initiatives incorporating community participation in household clean-ups to improve hygiene [51]. Research projects co-designed by researchers and community residents are underway in Northern Australia, aiming to help Aboriginal community members to develop durable, locally-applicable strategies to control GAS transmission and hence, RHD risk [52]. As yet no conclusive evidence is available about how to use primordial interventions as a mechanism for developing a control strategy.

Primary prevention of RHD comprises antibiotic therapy for GAS infections to prevent an initial attack of ARF [55], where administration within 9 days of pharyngitis onset can reduce ARF occurrence by up to 80% [56]. Evidence from a large-scale program to improve access to sore throat treatment in NZ [57] suggests that ARF incidence can be reduced through a multi-pronged approach has the potential to be cost-effective in high risk populations [58]. Sore throat treatment was a key component of broader, comprehensive ARF/RHD control strategies in Cuba [59] and Costa Rica [60]. Access to either laboratory culture facilities or rapid antigen detection tests to confirm GAS pharyngitis remains challenging in many LMICs, leading some to recommend either imperfect clinical prediction rules or alternatively a ‘treat-all’ approach [61]. In high resource settings with a low incidence of ARF, concerns about antibiotic overuse have appropriately curtailed routine antibiotic treatment of pharyngitis. Advances in GAS vaccine development have placed vaccination on the agenda as a future opportunity for primary prevention [62‐65].

The cornerstone of RHD control internationally is secondary prevention with long-term depot administration of penicillin, or daily oral antibiotic in individuals who are penicillin allergic [69]. In the absence of a vaccine for GAS, secondary prevention has been determined to be the most cost-effective RHD control strategy [70]. Secondary prevention comprises administration of intramuscular, long-acting penicillin every 28 days for individuals with a history of ARF (or documented RHD) to prevent ARF recurrences [55] thereby reducing cumulative valve damage. International guidelines recommend benzathine penicillin G (BPG) as the first line therapy, injected intramuscularly every four weeks [71]. Secondary prophylaxis continues from the first ARF diagnosis until the period of highest clinical risk has passed. Australian [49], NZ [72] and WHO guidelines [71] recommend prophylaxis for a minimum of ten years following last ARF episode or until 21 years of age.

Patients with severe RHD may need lifelong prophylaxis, implying a long-term burden of disability and treatment. The challenge of delivering onerous secondary prophylaxis prompted the development of RHD registers [73]. Although now rare in high income countries, the register-based model has been adapted in low resource settings with increasing standardisation of terms and clinical data collection. New register-based programs are expanding, including examples from Uganda [74], Fiji [75] and East Timor [76]. Register-based secondary prophylaxis remains a backbone of broad RHD control programs, generally considered cost effective [32, 77] although not fully evaluated in very low resource settings.

Echocardiographic screening to identify asymptomatic RHD has provided new impetus for secondary prophylaxis. Echocardiography has replaced auscultation as a screening tool, providing a more sensitive and specific test [78]. The development of echocardiographic diagnostic guidelines for screening has standardised interpretation and reporting of results [79]. Although echocardiographic screening is broadly accepted as an epidemiologic research tool to better quantify the RHD burden [80], the role of echo to identify asymptomatic valve lesions and deliver early intervention is less clear in the absence of good knowledge of the natural history of disease progression [78]. In theory, the use of echocardiographic screening for clinical benefit would be an augmentation of secondary prevention; also, detection in women of childbearing age would permit appropriate planning for pregnancy and childbirth. This is potentially feasible in RHD-endemic populations of HICs where the majority of symptomatic cases have already been detected. In LMICs where RHD is generally more advanced when diagnosed [4], echocardiographic screening detects a higher proportion of symptomatic cases. The disease-altering benefit of secondary prophylaxis in advanced valvular RHD is smaller than in asymptomatic cases as interventions shift from secondary prophylaxis to management of heart failure and surgical options. Nevertheless, screening in these settings identifies cases at a less progressed stage than self-presentation and provides the only means to identify cases at a mild stage when SP has the greatest likelihood of improving outcomes, thus still providing benefit. In reality, the utility of screening programs for clinical benefit is subject to ongoing debate but may be considered as ways of identifying cases for both secondary and tertiary interventions.

Tertiary interventions for RHD include medical management of heart failure, operative management of valve lesions and treatment for the consequences of RHD, including stroke, infective endocarditis and arrhythmia. Opportunities for operative intervention span catheter-based balloon valvotomy, open commissurotomy, valve repair or replacement. Access to these services is severely curtailed by the gross economic and geographic disadvantage in endemic RHD settings with minimal tertiary cardiac services. A number of programs aim to improve access to advanced cardiac services in RHD endemic, low-resource settings. These initiatives often involve reciprocal training of surgeons, visiting surgical teams and the development of regional centres of excellence [90].

The advantages of vertical programs– their specific disease focus, use of specialist staff, dedicated resources and focused objectives –generally do not extend to NCD services which are usually part of (horizontal) mainstream services. Where both infectious diseases and NCD impart high burden in a single population, the need for co-ordinating resources and systems is now being recognised [3, 97, 99].

The RHD sector intersects with many specialities and patient groups, providing opportunities for increased integration with both vertical programs (like TB and HIV/AIDS) as well as mainstream primary health services (like maternal and child health, NCD services and adolescent health). A stark example of the need for integration can be seen in the REMEDY study of RHD hospital patients in 12 LMICs, where only 4% of the women with RHD of childbearing age were receiving contraception [4].

A ‘diagonal’ approach is increasingly promoted as a way to strengthen health systems, harnessing aspects of successful and well-resourced vertical programs to support other health services [100, 101]. This calls for disciplinary interaction and alignment [99] by looking for opportunities to combine prevention, screening, and treatment programmes.

Diagonal approaches often start with vertical programs [101‐103], with incremental integration of subgroups of patients with overlapping conditions, expanding to other patient groups as services consolidate and service partnerships develop [104]. There are now abundant reports in the literature showing how donor funding, infrastructure, systems and services for HIV/AIDS have been leveraged to improve other generic services like maternal and child health, and NCDs [105‐107]. Collaborative frameworks for the care and control of TB and diabetes [97] and TB and HIV [96] grew out of the recognition of the interaction between these diseases, suggesting how overlaps between conditions can lead to clinical integration. An important starting point is screening for other diseases in vertical programs as shown in Malawi where TB case-finding was enhanced in HIV patients and TB programs were an entry point for HIV care [96]. In Mexico, MCH services were enhanced by building onto vertical programs in a step-wise fashion [103]. Despite these encouraging case studies, there is a need for more evidence of what works and not [96, 101, 107], while recognising the contextual complexity and that no ‘one-size fits all’ exists [102].

RHD illustrates some of the challenges in operationalising diagonal or integrative models of care, even when strong conceptual links and potential benefits are evident. There are few published examples of integration between vertical programs: a pilot study to explore the utility of screening HIV infected children for RHD has been conducted in Uganda [108] and in Eritrea a pilot program screened pregnant women for RHD [109]. A shift towards integrated, comprehensive care for RHD and other conditions is expected to accelerate – spurred by global efforts to improve universal health coverage and integrated, people-centred health services [110]. Within RHD this shift necessities a common nomenclature for describing integrative activities. A descriptive framework is presented in Table 1, differentiating between adding RHD elements to existing programs or adding other disease elements to RHD programs.

Using this framework, it is possible to describe a range of early stage activities as integrative approaches. For example, in Australia, completion of training modules on RHD in pregnancy is now a mandated component of midwifery training, demonstrating the addition of RHD elements to existing programs [111]. Nascent programs are also beginning in continental Africa to embed RHD activities into existing maternal health infrastructure. Conversely, training RHD care providers to improve management of non-RHD conditions is integrative – illustrated by recent efforts to improve care of anaphylaxis by primary care providers who administer benzathine penicillin G injections in Zambia [112]. Integration with primary care delivery may also be evident if targeted echocardiographic screening for RHD, currently being tested with regards to sensitivity and specificity when undertaken by non-specialist primary care providers [113‐116], becomes more mainstream over time.

Thus inputs from vertical programs can be harnessed for RHD control efforts at various points along the life course (Fig. 3), with the primordial factors required to be addressed at all life stages through synergies with community development and environmental health programs.

In the mid-1990s, following a thorough review of the evidence [117], the MacColl Institute for Healthcare Innovation developed the Chronic Care Model (CCM) as a conceptual model for quality improvement in chronic disease management. The CCM is an approach to chronic illness management which aims to improve service delivery by creating an ‘informed, activated patient and a prepared, proactive practice team’. The model outlines themes relevant to the health system (Health Systems, Delivery System Design, Decision Support, Clinical Information Systems) and community (Self-Management Support and Community Linkages) to consider when providing chronic care management in primary care settings. The CCM therefore provides a relevant framework for improving the delivery of secondary prevention of RHD.

This model has been successfully applied in many health care organisational settings in the United States, with research evidence accumulating that system changes consistent with CCM components/principles improve outcomes in chronic diseases [118, 119], ranging from diabetes [120‐123] to chronic heart failure [121] and COPD [124, 125]. The model has also been applied to cancer [126], depression [127] and neurology [128], and translated into practical application in the Australian Indigenous primary care setting [129].

The CCM was used as a reference point to develop the Innovative Care for Chronic Conditions (ICCC) Framework of the WHO [8, 130]. The ICCC was developed to ensure applicability in low and medium income countries through greater linkage of the policy environment and community context to health care organisation. Further, an enhanced version of the CCM, the Expanded CCM [131], was proposed to incorporate public health/health promotion concepts and strategies, in particular the social determinants of health [36, 37] which influence not only secondary and tertiary levels of prevention, but impact profoundly on primordial and primary prevention of both infectious disease and NCD. The three members of the CCM-family each have a different emphasis, however all recognise the centrality of good organisational systems for managing long-term conditions.

We propose that the CCM models can be operationalised for RHD by greater integration of secondary and tertiary prevention into existing services [132]. Secondary prophylaxis is the most important investment to make [70] yet adherence to secondary prophylaxis tends to be low internationally. Secondary prevention clearly fulfils the definition of ‘chronic care’, even though it applies to ‘acute’ rheumatic fever. After ARF, chronic care management entails 4-weekly contact with a healthcare provider for administration of secondary prophylaxis (a depot penicillin injection) for a minimum 10-year period, as well as regular medical evaluation. Adherence is challenging since the treatment is a painful injection; healthcare staff often have poor knowledge of ARF and RHD; patient populations are young, often asymptomatic and in some settings, mobile; and health literacy pertaining to the value of injections may be low [133]. Models that have the potential to improve adherence could have important value in this context. In a before-after study implementing a continuous quality improvement (CQI) process using participatory action research, we showed improved delivery of aspects of ARF/RHD care to Aboriginal patients, but a failure to improve adherence to acceptable levels [134]. CCM-based CQI methods applied to NCDs in the same settings (clinics in remote Australian Aboriginal communities) have successfully improved patient outcomes [135]. A randomised community trial is currently underway in a high-burden Australian jurisdiction to determine the impact of systems interventions incorporating the components of the CCM on uptake of regular BPG for the secondary prevention of ARF [133]. Tailored use of the CCM has been developed as the basis for the trial as shown in Fig. 4, using the six CCM themes as the scaffold for primary health care activities to implement within the study intervention. This intervention exemplifies a diagonal approach where primary health care services are augmented through ‘vertical’ support to improve outcomes. The mainstream services provide the infrastructure and context for RHD care to be integrated into other essential services, providing better opportunity for all-of-person management.

The CCM also has a particularly good fit with tertiary prevention of RHD. In well-resourced programs, other aspects of chronic care after ARF include regular (e.g. yearly) echocardiography and dental care [49]. Thus the life-long requirement for cardiac review, monitoring of medications (including anticoagulation) and the potential need for surgery means that patient-centred case management approaches can have a significant impact on quality of life and survival. The need for a model of long-term care of RHD patients in LMIC is illustrated by a recent global registry study [136] showing 17% two-year mortality among 3343 children and adults with RHD from LMIC. The financial expenditure of tertiary-level interventions is wasted if continuity of care for these patients is not provided in the community. In many low resource settings, existing primary care services are very weak and/or geared towards acute, rather than ongoing care. Adoption of the CCM for RHD patients can support and promote strengthening of the primary care system, and better integration with tertiary services.

CHD, particularly its association with diabetes, is a common focus of the CCM [121, 122]. The requirement for life-long adherence to cardiac medications and lifestyle modifications as well as the need for treatment of comorbidities means that the patient-centred CCM provides a useful framework for long-term management of CHD. In the management of TB (and other infections), some components of the CCM may be utilised during the antibiotic treatment phase, but not beyond, despite the potential benefits after microbiological cure. Well-functional TB control programs have specific public health mechanisms to support care and reduce patient default, provide patient education, provide decision support to practitioners through written national TB treatment policies, and good data collection to provide feedback nationally and to the WHO. While these structures are nominally in place internationally, there is always scope for improvement in the quality of implementation, to ensure the desired outcome of ‘informed, activated patients and communities’ and ‘prepared, pro-active practice teams and community partners’ as described in the chronic care model [131]. Two themes of the CCM emphasise the importance of patient and community engagement – the ‘self-management support; and ‘community engagement’ themes. New community-driven approaches to TB control, mentioned above [92, 93] provide informative strategies that could be used to by RHD control programs to address these themes of the CCM.

The burden of RHD is grounded in social and economic determinants of health, necessitating a political and government response. In some places, RHD has become an explicitly political issue. For example, in New Zealand, rheumatic fever rates have been a feature of two election cycles and major government investment has reduced the rate of ARF by 23 % [137, 138]. In 2015, the Addis Ababa Communique on ARF and RHD control was adopted by Heads of States of African Union member countries [139]. However, in the majority of LMIC endemic settings government and governance responses are thinly spread and often overwhelmed by competing demands. Thus there may be scope to frame RHD as a sentinel disease of inequality, facilitating strategic investment in primary care and equity advancing interventions [140]. Political interest may be further channelled into global initiatives, including growing efforts towards universal health care (UHC) [141] and a pending discussion of RHD control at the World Health Assembly 2018 [6].

Financing for RHD interventions is often skewed towards investments in costly tertiary intervention. In particular, a large number of RHD endemic LMIC send patients overseas for expensive valve surgery [142], sometimes with poor outcomes. A health systems strengthening approach would see government funds disbursed earlier in the aetiologic pathway with greater potential for prevention and primary care settings where multiple other diseases can also be addressed. However, in some settings early interventions (primary/secondary prevention) incur out-of-pocket cost, reducing access and thus efficacy of the interventions [143, 144]. RHD is a good example of how investing in UHC has the strong potential to be cost saving. Spanning the infectious disease – chronic disease continuum RHD can help illustrate to policy makers that providing appropriate antibiotics in primary care can save expensive international heart surgery. Illustrative narratives are influential for decision makers [145] and RHD offers compelling vignettes to address financing issues.

Diagnosis and management of RHD is human-resource intensive at all levels, including for early interventions of pharyngitis management, ARF diagnosis and secondary prophylaxis delivery. Increased human resources for health are needed globally. In the interim, task shifting [115] and task sharing may be opportunities to ameliorate shortages. In particular, given the shortage of expert operators for cardiac ultrasound, training programmes are under development for non-expert operators to undertake limited echo views for screening purposes [113, 116]. In addition to this kind of technological solution ongoing work is needed to train and support nurses and community health workers to diagnose and manage strep pharyngitis, ARF and RHD. Simple, standardised clinical guidelines are a critical component of this support and are urgently needed for RHD control.

Access to essential medicines and essential technologies for RHD care continue to be major global barriers. In particular, the critical injectable antibiotic, benzathine penicillin G (BPG) has been subject to global shortages over a number of years [146]. An old, off-patent, affordable antibiotic which has been on the WHO Essential Medicines List since inception, BPG should be readily available. The difficulties in procurement and use of BPG exemplify access to medicines issues globally and are an appealing target for government and multi-state intervention. The WHO Package of Essential NCD Interventions for Primary Health Care (PEN) identifies essential technologies relevant to RHD, particularly stethoscopes, scales, and blood pressure monitoring. [147] ECG capacity is identified in PEN as an essential technology ‘where resources permit’. Supporting access and supported use of the basic clinical resources in primary care should be a shared priority. Although expanding access to echocardiography for diagnosis and clinical review offers considerable scope for improve quality of care for RHD and other cardiac conditions, the role of population level echocardiographic screening remains under investigation [148]. The cost and training required to take advantage of portable and hand-held imaging technology remain significant barriers to health systems, but could revolutionise diagnosis, triaging and management of ARF/RHD in the future.

Many LMICs have insufficient contemporary data on ARF/RHD occurrence to inform health system responses [28, 149]– both at the policy and programmatic levels. Concerted efforts are needed to fill information gaps, in particular recording of longitudinal data to inform ARF/RHD care. Classically, RHD registers have been used to record clinical features and adherence with secondary prophylaxis. As electronic health information systems increase in endemic settings, there is increased scope to integrate register-based care [150]. In addition to improving case management, this may offer new opportunities to understand disease progression and mortality outcomes of RHD. Documented efforts to integrate RHD registers into routine care information systems are needed to identify best practice. This exemplifies the need for ongoing implementation research science in RHD control.

The unanswered challenge in RHD control is identifying service delivery models which are acceptable to populations at risk, to deliver effective interventions that are sustainable in the face of competing priorities. The incremental integration, diagonal approaches and health system strengthening recommended earlier in this paper are likely to be components of identifying feasible models. These models need to be evaluated to provide evidence-based recommendations for policy and program design.