The incidence of Rheumatic Heart Disease in Remote Australia appears to have increased significantly or even doubled according to some surveys in the last 10 years or so. (1)
What is happening? Are living conditions in Remote Communities getting worse still? Are we finding previously undiagnosed RHD?
We know that Acute Rheumatic Fever(ARF) and Rheumatic Heart Disease(RHD) are diseases of poverty and overcrowding. They are largely unknown in modern urban Australia but still are common in Remote Communities, particularly in NT.
While living conditions in many communities are still “third world” standard, I can find no evidence for further worsening in recent years and my own anecdotal experience over 20 years or so would suggest that things are no worse than they have been in the past. “Closing the Gap” reports show little improvement, but they do not suggest worsening of living conditions and life expectancy.
Are we detecting previously undiagnosed disease?
In my anecdotal experience virtually every person in a community presents to the clinic, often frequently. It seems intuitively unlikely that significant symptomatic heart failure as a result of valve dysfunction would not have been picked up on presentation. Echocardiogram and cardiology review is available for acute symptomatic disease. Heart failure due to acute carditis or deterioration of RHD is uncommon, but can be confused with more common conditions such as pneumonia. Adverse outcomes as a result are likely to occur (https://tjilpidoc.com/2022/03/09/poor-administration-a-health-hazard/)
There has been an understandable promotion of screening by Echocardiogram with programs such as the “Deadly Heart Trek” which have found asymptomatic RHD in some clients. It is generally accepted that prophylactic Penicillin reduces recurrence of ARF and deterioration (though there are no prospective trials to prove this). Thus finding asymptomatic clients and treating them with prophylactic Penicillin would seem intuitively a Good Idea. But like all screening processes it can be difficult to show benefit – whether this screening will result in improved outcomes is yet to be established.
Overdiagnosis
There is one other possible explanation for the apparent increased incidence – that we are over-diagnosing ARF. In recent years there has been promotion of the idea that clinicians have been missing cases of ARF and should be on the lookout for it to reduce the incidence of serious RHD with prophylaxis. Once a diagnosis of ARF is established, even provisionally, that client is subject to a regime of monthly injections and reviews for anything up to 10 years. Many clients are discharged from hospital at their initial presentation without an expert assessment and classification – this is relegated to a later date. But elective Echocardiography and Cardiology review are difficult to access for Remote clients for various reasons. It may be months or even years before these are performed. By this time the relevant clinical signs and data may be lost or otherwise unavailable. As a result even senior clinicians are reluctant to reverse a provisional diagnosis and it can be difficult if not impossible to remove the Rheumatic Fever “label” once it is applied.
ASOT AntiDNAse and streptococcal serology – what is the normal?
Acute Rheumatic Fever increases Antistreptolysin Titre (ASOT). The upper laboratory limit of normal in Australia is 200 IU. But the majority (65%) of asymptomatic subjects had a level >200 IU with some being as high as 800 IU. The level increased with age>10 and season (winter) in an Egyptian study (2). In Australia it is likely to be high in Remote Community subjects because of living conditions and frequent exposure to Group A Streptococcus, but I could find no research on this question for Remote Australia. However, it seems likely that normal levels are much higher than the accepted laboratory range. This makes it a poor positive discriminator for ARF, though it may be helpful in ruling out the disease if it is negative. Similar issues apply in the case of AntiDNAse.(4)
The Clinical Criteria for ARF
Skin infection is common and is the likely source of streptococcal infection in most cases of ARF, at least in the Top End and tropical Australia. (McDonald et al). In spite of this, conventional teaching still sees pharyngitis and tonsillitis as the primary source. Acute Rheumatic Fever remains a clinical diagnosis – there is no independent lab test or other indicator which can reliably discriminate it from other diagnoses. The diagnosis is made on the Jones Criteria, which were first introduced in 1944. They have been modified several times since to increase their sensitivity in low risk populations. This has the effect however of reducing specificity. In reading the references there still seems to be ambiguity, particularly with regard to arthritis/arthralgia. In the most strict version of the criteria, only polyarthritis was allowed as a major criterion – ie several joints involved with objective signs such as effusion and redness. In the more recent versions monoarthritis or even polyarthralgia are allowed as major criteria in high risk areas. Chorea is probably pathognomonic in young people as other causes of acute Chorea are uncommon. “Carditis” can be difficult to define in a Remote setting where echocardiography is not generally available on the spot. A small group of patients present in heart failure due to carditis – these are challenging to diagnose and manage and errors are frequent in this group.
ARF – typical presentations
A common presentation is joint pain or arthritis with or without fever and raised ESR/CRP. Chorea is less common, with acute carditis or other presentations the least common. In my 20 years experience in Remote Health I have not seen the classically described erythema marginatum or subcutaneous nodules. Because of the increased awareness of ARF as a diagnosis, the classical criteria have been relaxed – I have seen a provisional diagnosis of ARF made on a presentation of monoarthritis or even polyarthralgia and raised CRP but without other criteria. ASOT appears to be used as a de facto criterion when it is not a positive discriminator (see above) Enthusiasts argue that any potential harm from overdiagnosis is outweighed by the benefit to a client with true ARF in reducing long term disability with prophylaxis. I would argue that the imposition on clients of an unnecessary diagnosis is not trivial, with monthly painful injections and frequent reviews for up to 10 years or more. There is a workload burden on the Remote Clinic involved and an opportunity cost as a result. The results of misdiagnosis at presentation can be significant – I have personally seen a case of knee pain in a child diagnosed as ARF when in fact it was tibial osteomyelitis and definitive treatment was delayed. On another occasion knee pain was considered to be ARF when in fact the diagnosis was Slipped Capital Femoral Epiphysis. In both these cases the misdiagnosis could have resulted in significant disability. Indeed a study at Royal Darwin Hospital showed that many of the cases admitted with presumed ARF had an alternative diagnosis at discharge. (3)
RHD presentations
The majority of clients with severe RHD requiring surgical intervention or documented valve changes on ECHO either have longstanding RHD with the details of the presentation lost in the mists of time, have presented with heart failure or have been found on Echocardiography screening. No clients who presented with joint symptoms in my case reviews showed evidence of RHD on Echocardiogram. Chorea seems a more reliable criterion with at least some of these subsequently developing RHD changes
Clinical standards
As a practitioner near retirement of course I think things were better in the old days
Our medical clinical training was rigorous with an emphasis on clinical method. This emphasis appears to have been lost in recent years – many clinicians do not take a detailed relevant history of the presentation or refer to previous attendances or past history. Examination is cursory if at all.
We have come to rely on lab testing and imaging for diagnosis when a rigorous clinical method in the hands of an expert clinician remains the most effective diagnostic tool. Many clinicians are nonmedical – they have not undergone the clinical training that doctors go through. There is a heavy reliance on telemedicine, which means that examination is limited. General Practitioners have been largely relegated to administrative tasks and navigating complex chronic disease. Their role in the assessment of acute presentations has been reduced and their opinion is often not respected. They are no longer seen as “expert generalists” at the centre of the clinical process. The assessment of an acute presentation is the classic scenario where masquerades and alternative diagnoses must be considered as well as the “probability diagnosis”. (Murtagh 6) ARF has now become a “probability diagnosis” due to its promotion as a condition which must not be missed. Unsophisticated clinicians often do not consider the alternatives. ARF is a clinical diagnosis. I have noted a tendency in unsophisticated Remote Staff to overreport clinical diagnoses (Otitis media, pharyngitis, bronchiolitis for example). Is this happening with ARF also?
Workforce issues in Remote Australia
The Remote workforce is heavily “casualized” and there is massive staff turnover in most Remote Clinics. Health encounters have become “commoditized” and anonymous – client and clinician often do not know each other. (see previous post) Many Remote Services are struggling to maintain their workforce numbers. These factors further reduce the quality and safety of clinical assessments.
ECHO – how reliable is it?
In file reviews I have noted on some occasions that an echocardiogram was reported as abnormal with Rheumatic changes but subsequent echocardiograms were reported as normal. In one case there was a normal report with abnormal reports before and after. We have always been taught that Rheumatic valve changes do not resolve with time. If this is the case then the quality of echocardiograms must be brought into question. Ultrasonography is a difficult skill, with cardiac ultrasound even more so. Where there is doubt, there is a tendency to overreport changes to avoid missing significant lesions.
Conclusions
The apparent increase in Rheumatic fever and RHD in the last decade can be explained in part by screening and finding asymptomatic patients. But it is likely that the increase in ARF diagnosis is in part due to overdiagnosis, as a result of casualization of the workforce, reduction in clinical standards, promotion of the diagnosis and reduction in the role of expert clinicians such as doctors.
This overdiagnosis has significant consequences for patients and Remote Clinics and it can be difficult to reverse the “label” once it is applied. Any patient admitted with a provisional diagnosis of ARF should undergo careful assessment by a senior clinician before discharge and classification as ARF. While it is important not to miss cases of ARF, we should be aiming to improve our diagnostic precision so that we do not impose an unnecessary burden of treatment on clients and the health service, and do not miss other potentially serious conditions.
An answer should be sought to the question – Do RHD changes resolve with time? Echocardiogram is a difficult skill – there is a need for review of some results and rigorous standards.
References
(1) AIHW Acute Rheumatic Fever and Rheumatic Heart Disease in Australia 2022
https://www.aihw.gov.au/reports/indigenous-australians/arf-rhd-2022/contents/arf
(2) Antistreptolysin O titer in health and disease: levels and significance
Alyaa Amal Kotby, Nevin Mamdouh Habeeb, and Sahar Ezz El Elarab
Pediatr Rep. 2012 Jan 2; 4(1): e8.
(3) The challenge of acute rheumatic fever diagnosis in a high-incidence population: a prospective study and proposed guidelines for diagnosis in Australia’s Northern Territory
Anna Ralph 1, Susan Jacups, Kay McGough, Malcolm McDonald, Bart J Currie
Heart Lung Circ . 2006 Apr;15(2):113-8.
(4) Detection of upper limit of normal values of anti-DNase B antibody in children’s age groups who were admitted to hospital with noninfectious reasons
Servet Delice,1 Riza Adaleti,2 Simin Cevan,3 Pinar Alagoz,4 Aynur Bedel,5 Cagatay Nuhoglu,5 and Sebahat Aksaray2
North Clin Istanb. 2015; 2(2): 136–141.
(5) Low rates of streptococcal pharyngitis and high rates of pyoderma in Australian aboriginal communities where acute rheumatic fever is hyperendemic
Malcolm I McDonald 1, Rebecca J Towers, Ross M Andrews, Norma Benger, Bart J Currie, Jonathan R Carapetis
Clin Infect Dis . 2006 Sep 15;43(6):683-9.
(6) General Practice 8th edition
John Murtagh
TjilpiDoc 