In the developed countries of the world, the incidence of rheumatic fever has fallen markedly during the last century. For example, in the USA the incidence per 100 000 was 100 at the start of the last century, 45–65 between 1935 and 1960, and is currently estimated to be approximately 10 cases per 100000.⁴ This decrease in rheumatic fever incidence preceded the introduction of antibiotics and is a reflection of improved socio-economic standards, less overcrowded housing and improved access to medical care. The current prevalence of rheumatic heart disease in the USA and Japan, 0.6–0.7 per 1000 population, contrasts sharply with that in the developing countries of Africa and Asia where rates as high as 30 per 1000 have been reported.⁵ As GAS pharyngitis and rheumatic fever are causally related, both diseases share similar epidemiologic features.⁴ The age of first infection is commonly between 6 and 15 years. Also, the risk for developing rheumatic fever is highest in situations where GAS is more common, for example where people live in crowded conditions.⁶

Clinical, epidemiologic, and immunologic observations tend to strongly support the causative role of untreated GAS pharyngitis in rheumatic fever.^6,7 Beyond this, however, the pathogenesis of acute rheumatic fever (ARF) and clinical heart disease remains unclear and several important and unexplained observations render the management of this important disease extremely difficult. These are:

• individual variability of susceptibility to GAS pharyngitis
  
• individual variability of development of symptomatic GAS pharyngitis
  
• individual variability of development of acute rheumatic fever after an episode of GAS pharyngitis
  
• individual variation in the development of carditis and chronic rheumatic heart disease after an attack of acute rheumatic fever
  
• the development of chronic rheumatic heart disease (RHD) in patients who have no definite history of acute rheumatic fever.

Streptococcal skin infection (impetigo) is considered not to cause rheumatic fever. However, a report of acute rheumatic fever following streptococcal wound infection⁸ and the high prevalence of pyoderma with relative paucity of streptococcal pharyngitis in some communities with a high incidence of rheumatic fever has raised questions about the link between streptococcal skin infection and rheumatic fever.⁹ While effective antibiotic treatment virtually abolishes the risk of rheumatic fever, in situations of untreated epidemic GAS pharyngitis up to 3% of patients develop it.^1,7 Worryingly, as many as a third of patients who develop rheumatic fever do so after virtually asymptomatic GAS and in more recent outbreaks, 58% denied preceding symptoms.¹⁰ This does not augur well for the primary prevention of rheumatic fever where prompt diagnosis of GAS pharyngitis and treatment are essential.

The virulence of the streptococcal infection is dependent on the organisms’ M protein serotype which determines the antigenic epitopes which are shared with human heart tissue, especially sarcolemmal membrane proteins and cardiac myosin.¹¹ It is these variations in virulence, as a result of M protein variation, which are thought to explain the occasional outbreaks of rheumatic fever in areas of previously low incidence.¹² Other factors influencing the risk for rheumatic fever are the magnitude of the immune response and the persistence of the organism during the convalescent phase of the illness.⁷

Evidence suggests that host factors play a role in the risk for rheumatic fever. In patients who have suffered an attack of rheumatic fever, the incidence of a repeat attack is approximately 50%. A specific B-cell alloantigen has been found to be present in 99% of patients with rheumatic fever versus 14% of controls.¹³ Certain HLA antigens appear to be associated with increased risk for rheumatic fever. Approximately 60–70% of patients worldwide are positive for HLA-DR3, DR4, DR7, DRW53 or DQW2.¹⁴ Such genetic markers for rheumatic fever risk may be useful to identify those in need of GAS prophylaxis. However, in view of the frequency with which these markers occur, they are unlikely to be of practical benefit in the short term.

While there is no specific clinical, laboratory or other test to confirm conclusively a diagnosis of rheumatic fever, the diagnosis is usually made using the clinical criteria first formulated in 1944 by Duckett Jones¹⁵ and recently modified by the World Health Organization (WHO).¹⁶ The diagnosis is suggested if, in the presence of preceding GAS infection, two major criteria (carditis, chorea, polyarthritis, erythema marginatum, and subcutaneous nodules) or one major and two minor criteria (fever, arthralgia, elevated erythrocyte sedimentation rate, elevated C-reactive protein or a prolonged PR interval on ECG) are present. Evidence of preceding GAS infection, essential for the diagnosis, may be obtained from throat swab culture (only positive in approximately 11% of patients at the time of diagnosis of acute rheumatic fever³) or by demonstrating a rising titer of antistreptococcal antibodies, either antistreptolysin O (ASO) or antideoxyribonuclease B (anti-DNase B).

There has been concern that strict application of the Jones criteria may result in underdiagnosis of rheumatic fever in endemic areas. particularly in the case of recurrent episodes.¹⁷ During recurrence of rheumatic activity in a patient with pre-existing RHD, the carditis may precipitate heart failure but may not be possible to diagnose because of lack of information on previous cardiac findings or because valve replacement surgery has been performed. The new WHO criteria thus recommend that a diagnosis of a recurrence of ARF in a patient with pre-existing RHD is possible on the basis of minor manifestations and evidence of recent streptococcal infection.¹⁶ Doctors should recognize that the published criteria are guidelines, and are particularly useful in epidemiologic investigations where diagnostic rigor is essential. It is appropriate for clinical judgment to be applied and to supersede guidelines particularly in parts of the world where ARF remains common.¹⁷

The most recent recommendations on the prevention of rheumatic fever have been published by the WHO.¹⁶

Prevention of rheumatic fever may be considered to be either prevention of the initial attack (primary prevention) or prevention of recurrent attacks (secondary prevention). True primary prevention of rheumatic fever depends more on socio-economic than medical factors. Upgrading housing and other aspects of poverty eradication will do more to eradicate the disease than antibiotic prophylaxis.

Primary prevention

Prevention of the initial attack of rheumatic fever depends on the prompt recognition of GAS pharyngitis and its effective treatment. While it has been demonstrated that therapy initiated as long as 9 days after the onset of GAS pharyngitis can prevent an attack of rheumatic fever,¹⁸ early treatment reduces both the morbidity and the period of infectivity.

The first report of the use of penicillin for the treatment of GAS pharyngitis and prevention of most attacks of rheumatic fever was published in 1950.¹⁸ Over the following 40 years, attention focused on accurate diagnosis and treatment of GAS pharyngitis. A single dose of intramuscular benzathine penicillin G became the most common mode of treatment and avoided problems of non-compliance. Subsequently, as a result of the pain and possibility of allergic reaction associated with benzathine penicillin G, oral penicillin became the treatment of choice¹⁹ and remains so today²⁰ (Class I, Level A). In situations where compliance with a 10-day course of oral penicillin would be unreliable, a single dose of intramuscular benzathine penicillin G would be preferred (dosage 1.2 million u if >27kg, otherwise 600000u).²¹