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.

Accurate diagnosis of acute rheumatic fever (ARF) remains problematic in high-incidence settings and especially in the Aboriginal population of Australia's Northern Territory. Previous investigators have demonstrated that strict application of the 1992 Updated Jones Criteria results in under-diagnosis. This study's objectives were to review use of the Jones Criteria (1992 Update) in diagnosing ARF in Australian Aboriginal patients presenting with suspected rheumatic fever, and formulate a locally relevant algorithm to improve diagnosis. Patients presenting to Royal Darwin Hospital with suspected ARF were prospectively assessed during a 15-month period. Demographic information, clinical history, examination, laboratory and echocardiographic findings were documented in order to determine whether the Jones Criteria were fulfilled, and to identify alternative diagnoses. The hospital discharge diagnosis was recorded and patients were followed up 18-33 months later. Out of 35 patients with suspected ARF, all were Aboriginal Australians, 17 (49%) had a discharge diagnosis of definite ARF, 7 (20%) had definite non-rheumatic fever diagnoses (disseminated gonococcal infection, systemic lupus erythematosis, buttock abscess and other febrile illnesses in children with cardiac murmur due to previously undiagnosed RHD). The remaining 11 (31%) posed diagnostic difficulties because of mild symptoms that failed to fulfil Jones Criteria (attracting diagnoses such as 'unexplained arthralgia') or atypical features such as older age. Two patients whose illness initially failed to fulfil the Jones Criteria, who were neither diagnosed with ARF nor commenced on secondary benzathine penicillin prophylaxis, were found on follow-up to have definite and probable ARF, respectively. At least 29% (8/28) of patients without prior recognised ARF/RHD had echocardiographic evidence of established RHD, indicating that previous episodes were missed. Individual mild episodes of ARF may be overlooked, with patients missing out on the timely institution of secondary prophylaxis. The Jones Criteria should be supplemented by active exclusion of differential diagnoses and vigilant follow-up including echocardiography. 'Probable' and 'possible ARF' should be recognised as diagnostic categories applying to patients not fulfilling the Jones Criteria but who nevertheless should be offered prophylactic penicillin at least until further follow-up. A set of diagnostic guidelines is proposed