Jones criteria and underdiagnosis of rheumatic fever
http://www.100md.com
《美国医学杂志》
Projeto PRONUCLEAR, Brazilian Society of Rheumatology and Hospital das Clínicas, Federal University of Goiás, Brazil
Objective. The authors attempt to determine whether typical clinical and laboratory manifestations of acute rheumatic fever (ARF) are in accordance to what has been traditionally described and how useful the Jones criteria are for diagnosis. Methods. Data from 81 cases of ARF were retrospectively collected. Inclusion criteria: 5 to 15 years of age and diagnosis of ARF confirmed by 2 or more rheumatologists, sustained for at least 6 months and two or more visits. Results. Girls had more chorea (23/50.0% vs. 5/14.3%)(p<0.0001). Cardiovascular (65/80.2%) and joint involvements (63 / 77.8%) were the most frequent manifestations. Fever was noted in roughly half of the patients. Arthritis was more frequent than arthralgia (47/58.0% vs. 16/19.8%, respectively) (p<0.0001); however, no specific pattern of joint involvement was found to be more prevalent. Mitral insufficiency was the most frequently detected echocardiographic sign (53 / 93.0%) and its association with aortic insufficiency was noted in 27 / 47.4% patients. Only 24 / 29.6% patients fulfilled Jones criteria for ARF requiring an evidence of previous group-A streptococcal infection (GASI). When compulsory GASI was disregarded, this number rose to 71/87.7% patients (p<0.0001). Conclusion. Girls were more affected by chorea; heart valves and joints were equally affected and represented the major clinical features; no specific pattern of joint involvement (eg.: migratory arthritis) could be labeled as typical; and strict adherence to Jones criteria, with compulsory documentation of a previous GASI, may lead to underdiagnosis of ARF.
Keywords: Rheumatic fever; Diagnosis; Diagnostic errors; Pediatrics; Rheumatology
Rheumatic fever is a public health problem in developing countries with incidence rates that reach epidemic levels. It is intimately linked to poor social and economic conditions, frequently affecting individuals that depend heavily on their physical strength and well being to earn their living. Therefore, especially due to its cardiac sequels, rheumatic fever represents a great burden on afflicted patients and on society as a whole.[1],[2],[3],[4]
There are still many controversial points related to the clinical manifestations and diagnosis of rheumatic fever. Physicians who frequently deal with new cases of the disease often feel uncomfortable in establishing a definite diagnosis, specially if they strictly follow Jones criteria. This set of criteria was created by a North-American physician in the mid-40's, when incidence of acute rheumatic fever (ARF) in the USA was high.[5] The Jones criteria have been revised a few times since then - the last revision was performed by the WHO Expert Consultation panel on Rheumatic Fever and Rheumatic Heart Disease.[6] Information on chosen methods and statistical analyses used in such reviews has not yet been disclosed.[6],[7]
A significant confusing factor in diagnosing ARF comes from the fact that recent studies have shown unusual clinical manifestations of the disease. In many instances, there are contradictory findings between different studies and they are frequently in contrast to what has been described in textbooks. Therefore, physicians seldom can rely on a "typical" clinical picture of ARF and diagnosis based on clinical judgement remains a difficult task.
The present study attempts to determine whether clinical and laboratory manifestations of acute rheumatic fever (ARF) are in accordance to what has been traditionally described as typical and how useful the Jones criteria are for diagnosis.
Materials and Methods
Data on clinical and laboratory features were collected from hospital records of 81 children and adolescents who were consecutively seen at the Pediatric Rheumatology Unit of the Hospital das Clνnicas/Federal University of Goiαs, Brazil, between January 1996 and August 2001. This Unit is an academic, tertiary, referral center located in Goiβnia, a city of 1 million inhabitants in Brazil's Central region. Prevalence of ARF in this area is high.[8],[9],[10] and the staff rheumatologists are nationally board-certified and well experienced in managing this disease.
Since one of our goals was to assess the reliability of Jones criteria and since no alternative diagnostic system has been formally proposed yet, the gold standard for ARF in this study was based on the authors' clinical judgement. Two of the authors (Pereira BAF and Silva NAS) are staff pediatric rheumatologists and reviewed all cases. Subjects were included in the study if: a) their ages ranged from 5 to 15 years; b) they had a clinical diagnosis of a first episode of ARF agreed by at least 2 board-certified rheumatologists; and c) this diagnosis was sustained for 6 months and 2 or more visits. Selected cases followed the Unit's standard protocol for ARF and data on their laboratory and clinical manifestations were recorded. Subjects were excluded from the study if diagnosis was pending, not sustained or attested by only one rheumatologist or other health professional. Other exclusion criteria were only one visit to our Unit and/or follow-up shorter than 6 months. All cases were tested against the Jones criteria in order to assess its reliability as a diagnostic tool.
All patients were tested for anti-streptolysin O antibodies (ASLO), erythrocyte sedimentation rate (ESR), C reactive protein (CRP), and mucoprotein levels, within 4 weeks after onset of symptoms. ASLO tests were performed by plate agglutination technique using 50 mL. of patient's sera and one drop of reacting solution containing ASLO sensitized latex particles (Biolab, Rio de Janeiro, BR). Results were expressed in IU/mL.
Evidence of previous group A streptococcal infection (GASI) was considered positive if history revealed an episode of upper airways streptococcal infection within 5 weeks preceding the onset of symptoms or if ASLO levels were greater than 320 IU/mL. Inflammatory activity was determined by ESR, CRP, and/or mucoprotein levels. First and/or second degree relatives were considered for family history of ARF. Echocardiographic exams were performed whenever a cardiac murmur was detected on physical exam. Joint features were divided into 5 patterns of presentation: (1) classical migratory arthritis, traditionally described as the most typical in ARF[6],[7] ;(2) non steroidal anti-inflammatory drugs (NSAIDs) responsive arthritis, when joints were not affected in a migratory fashion and inflammation remitted within 72 hours after initiation of NSAID; (3) atypical arthritis, when there was joint inflammation that did not fit into any of the above; (4) migratory arthralgia, as described for classical migratory arthritis but with pain as the sole symptom of joint involvement; and (5) other arthralgias, when the patient complained of joint pain that did not show a migratory pattern of presentation.
Parents of enrolled patients agreed to sign an informed consent. The study was approved by the Bioethics Committee of the attributable institution (CEP-HC/UFG).
Data were stored in a MS-Excel 2000 database system. STATA was used to calculate Pearson's chi-square, Student's t test (for comparison of means and difference of proportions) and analysis of 95% confidence intervals.
Results
Girls were more affected by ARF than boys (46 /56.8% vs. 35/48.2%, respectively) (p<0.001). Gender significantly influenced the presence of chorea, which was clearly more frequent in girls (23/50.0% vs. 5/14.3%) (p<0.001). The mean age did not differ between patients with (10.4±3 y.o.) and without (9.9±2 y.o.) chorea (p=0.353).
The frequency of clinical features is illustrated in [Figure - 1]. Heart and joints were the most frequent sites of involvement (65/80.2% and 63/77.8%, respectively) (p>0.05 - NS ). Chorea was noted in 28/34.6% patients. Arthritis occurred more frequently than isolated arthralgia (47/58% vs. 16/19.8%, respectively) (p<0.001), but no particular pattern of joint involvement was found to be significantly more frequent when stratified into subtypes of presentation [Table - 1]. Apical systolic murmur was by far the most common clinical cardiac sign (57/96.6%). Echocardiographic abnormalities were present in 57/70.4% patients, among which 53/93% had mitral insufficiency (MI). The association of mitral and aortic insufficiency (AI) accounted for about half of these findings (27/47.4%). Carditis was present in 19/28 (67.8%) patients with chorea.
Only 24/29.6% patients fulfilled the revised Jones criteria for diagnosis of first episodes of ARF,[6],[7] i.e. evidence of a previous GASI plus 2 major signs or 1 major and 2 minor signs. Since chorea allows for a diagnosis of ARF even in the absence of other major or minor signs[6],[7] the authors added to these afore mentioned patients a group of 25 other subjects that had chorea with no evidence of previous GASI and noticed 60.5% of fulfillment (49 patients). When compulsory evidence of a previous GASI was disregarded, this figure rose to 71/87.7% [Figure - 2].
Discussion
Acute attacks of first episodes of ARF were described in 81 distinct patients. The significantly greater overall proportion of female patients had also been noted in other studies.[11],[12] The high number of patients with chorea in the present study (28/34.6%) might have had some influence in the gender distribution of the total number of cases. Indeed, Carapetis and Currrie have postulated that, with a few exceptions, reports from certain geographical areas (Africa, Asia, the Pacific, the Caribbean and Arab countries) usually show lower relative numbers of patients with chorea, whereas those from other regions and/or ethnic groups (USA, Pakistan, Turkey, and Australian aborigines) present higher proportions.[13] According to that account, this study area may belong to the latter group, with figures comparable only to Australia's and the USA's.[13],[14],[15],[16] Interestingly, in this study patients with chorea were not older than other ARF patients, as has been traditionally described. In fact, they tended to be slightly younger when compared to other patients from the present study (p=0.353, NS ) as well as to patients with chorea from other series.[1],[13],[17],[18] The fact that cases were included only if they were younger than 15 years of age might have excluded some teenagers and lowered the mean age of the sample.
Articular and cardiac manifestations have been described as the most frequent features of rheumatic fever.[1],[2],[3],[4],[19],[20] Around 80% of patients in this series had joint complaints. Migratory arthritis has been said to be the typical articular manifestation of ARF,[4],[6],[19] but in our series, although arthritis occurred more frequently than arthralgia, no particular pattern of joint involvement was found to be more frequent when stratified into clinical subtypes of presentation [Table - 1]. Other studies have already indicated the fairly frequent occurrence of atypical joint manifestations.[21] This finding represents an additional complication, since diagnosis of ARF is highly dependent on clinical information obtained by history.
Carditis, in the form of valvulitis, usually occurs in more than three-quarters of patients with ARF.[2],[3],[20] Some studies have found higher rates due to the systematic use of echo-Doppler.[2],[3] In this study patients were submitted to an echo-Doppler exam only if they presented clinical evidence of carditis, in accordance to the American Heart Association and the WHO guidelines.[6],[7] However, there has been a flood of recent works showing a much greater sensitivity for the detection of valvulitis accomplished by the systematic use of echo-Doppler in patients with other features of ARF and with no clinical signs of carditis.[3],[17],[22],[23] Although any layer of the heart can be affected, mitral valve insufficiency was present in more than 90% of patients who underwent echo-Doppler, as has been reported by other authors.[1],[2],[3],[20] In this series, 19/28 (67.8%) patients with chorea had carditis as well, a figure similar to those described elsewhere.[1],[13] In fact, it has been shown that screening of choreic patients with echo-Doppler may further increment the detection of "silent" valvulitis.[17],[24],[25], [26]
Evidence of a prior streptococcal infection was considered essential for diagnosis of ARF in the 1965 revision of the Jones criteria and has been kept as such since then.[27] In the recently published revision from WHO, there is a note reminding of the existence of "probable cases", but even for these patients the need for an evidence of previous GASI remained.[6] In the present study, only about 60% of patients with a first attack of ARF had either chorea or a positive history of an upper airways GASI and/or high titers of ASLO. Highly suggestive clinical manifestations presented by the other 40% of patients and the long time of follow-up set as part of the gold standard indicate that obligatory evidence of GASI may leave almost half of ARF patients undiagnosed. This subgroup of undiagnosed ARF patients will not receive adequate early prophylaxis and will remain susceptible to recurrences of ARF that might lead to chronic rheumatic heart disease. One reason for such low sensitivity of Jones criteria in this series might come from the fact that patients could have been seen late on the course of the disease (when ASLO titers were already low) or during recurrences of subclinical first attacks of ARF that went unnoticed. Nevertheless, criteria for diagnosis of ARF must contemplate such possibilities since it is a disease that has a devastating effect over developing nations where such problems are common. Future case-control studies shall determine some epidemiological parameters ( e.g., specificity, accuracy) and may help to better evaluate this set of diagnostic criteria.
This study is a further indication of how difficult and complex diagnosis of ARF still is. Lack of laboratory markers; atypical clinical presentations of joint involvement; low availability of echo-Doppler in areas where it is most needed; and low sensitivity of Jones criteria (with compulsory evidence of a previous streptococcal infection) undoubtedly contribute to underdiagnosis. Criteria for diagnosis of ARF shall take into account all these situations in order to be considered as a relevant tool for physicians' daily practice.
References
1.Shet A, Kaplan E. Addressing the burden of group A streptococcal disease in India. Indian J Pediatr 2004; 71 : 41-48.
2.Bittar FF, Hayek P, Obeid M, Gharzeddine W, Mikati M, Dbaibo GS. Rheumatic fever in children: a 15-year experience in a developing country. Pediatr Cardiol 2000; 21 : 119-122.
3.Mishra TK, Routray SN, Behera M, Pattniak UK, Satpathy C. Has the prevalence of rheumatic fever/rheumatic heart disease really changed A hospital-based study. Indian Heart J 2003; 55 : 152-157.
4.Terreri MT, Ferraz MB, Goldenberg J, Len C, Hilario MO. Resource utilization and cost of rheumatic fever. J Rheumatol 2001; 28: 1394-1397. [PUBMED]
5.Jones TD. Diagnosis of rheumatic fever. JAMA 1944; 126 : 481-4.
6.Rheumatic fever and rheumatic heart disease. Report of a WHO Expert Consultation. World Health Organization, Geneva, 2004 (Technical Report Series No. 923)
7.Dajani AS, Ayoub E, Bierman FZ et al. Guidelines for diagnosis of rheumatic fever: Jones criteria, Updated 1992. Circulation 1993; 87 : 302-307.
8.Haddad N, Silva MB. Mortality due to cardiovascular disease in women during the reproductive age (15 to 49 years), in the state of Sao Paulo, Brazil, from 1991 to 1995. Arq Bras Cardiol 2000; 75 : 375-379. [PUBMED] [FULLTEXT]
9.Meira, ZMA, Castilho SRT, Barros MVL et al. Prevalence of rheumatic fever in children from a public high school in Belo Horizonte. Arq Bras Cardiol 1995; 65: 331-334.
10.Gus I, Zaslavsky C, Seger JM, Machado RS. Epidemiology of rheumatic fever. A local study. Arq Bras Cardiol 1995; 65 : 321-325.
11.Risvi SF, Khan MA, Kundi A, Marsh DR, Samad A, Pasha O. Status of rheumatic heart disease in rural Pakistan. Heart 2004; 90 : 394-399.
12.Melka A. Rheumatic heart disease in Gondar College of Medical Sciences Teaching Hospital: socio-demographic and clinical profile. Clin Exp Rheumatol 1996; 14 : 567-569.
13.Carapetis JR, Curie BJ. Rheumatic chorea in northern Australia: a clinical and epidemiological study. Arch Dis Child 1999; 80 : 353-358.
14.Choong CY, Abascal VM, Weyman J et al. Prevalence of valvular regurgitation by Doppler echocardiography in patients with structurally normal hearts by two-dimensional echocardiography. Am Heart J 1989; 117 : 636-642.
15.Denny FW. A 45-year perspective on the streptococcus and rheumatic fever: The Edward H. Kass lecture in infectious disease history. Clin Infect Dis 1994; 19 : 1110-1122.
16.Al Eissa YA. Acute rheumatic fever during childhood in Saudi Arabia. Indian J Pediatr 1990; 57 : 771-773.
17.Elevli M, Celebi A, Tombul T, Gokalp AS. Cardiac involvement in Sydenham's chorea: clinical and Doppler echocardiographic findings. Acta Paediatr 1999; 88: 1074-1077. [PUBMED]
18.Ghram N, Alani C, Oudali B, Fitouri Z, Ben Becher S. Sydenham's chorea in children. Arch Pediatr 1999; 6: 1048-1052.
19.da Silva NA, Pereira BAF. Acute rheumatic fever. Still a challenge. Rheum Dis Clin North Am 1997; 23 : 545-568.
20.Kafetzis DA, Chantzi FM, Grigoriadou G, Vougiouka O, Liapi G. Incidence and clinical profile of rheumatic fever in Greece. Eur J Clin Microbiol Infect Dis 2005; 24: 68-70
21.Hilario MOE, Len C, Goldenberg J, Fonseca AS, Ferraz MB, Naspitz CK. Febre reumαtica. Manifestaηυes articulares atνpicas. Rev Assoc Med Brasil 1992; 38 : 214-216.
22.Ozkutlu S, Hallioglu O, Ayabacan C. Evaluation of subclinical valvar disease in patients with rheumatic fever. Cardiol Young 2003; 13 : 495-499.
23.Hilαrio MO, Andrade JL, Gasparian AB, Carvalho AC, Andrade CT, Len CA. The value of echocardiography in the diagnosis and follow-up of rheumatic carditis in children and adolescents: a 2 year prospective study. J Rheumatol 2000; 27: 1082-1086.
24.Lanna CC, Tonelli E, Barros MV, Goulart EM, Mota CC. Subclinical rheumatic valvitis: a long-term follow-up. Cardiol Young 2003; 13 : 431-438.
25.Saxena A. Diagnosis of rheumatic fever: current status of Jones criteria and role of echocardiography. Indian J Pediatr 2000; 67(suppl 3): 11-41.
26.Narula J, Chandrasekhar Y, Rahimtoola S. Diagnosis of active rheumatic carditis. The echoes of change. Circulation 1999; 100: 1576-1581.
27.Stollerman GH, Markowitz M, Taranta A, Wannamaker LW, Whittemore R. Report of the Adhoc Committee on Rheumatic Fever and Congenital Heart Disease of the American Heart Association: Jones Criteria (Revised) for guidance in the diagnosis of rheumatic fever. Circulation 1965; 32 : 664-668.(Pereira Breno AF, da Silva Nilzio Antoni)
Objective. The authors attempt to determine whether typical clinical and laboratory manifestations of acute rheumatic fever (ARF) are in accordance to what has been traditionally described and how useful the Jones criteria are for diagnosis. Methods. Data from 81 cases of ARF were retrospectively collected. Inclusion criteria: 5 to 15 years of age and diagnosis of ARF confirmed by 2 or more rheumatologists, sustained for at least 6 months and two or more visits. Results. Girls had more chorea (23/50.0% vs. 5/14.3%)(p<0.0001). Cardiovascular (65/80.2%) and joint involvements (63 / 77.8%) were the most frequent manifestations. Fever was noted in roughly half of the patients. Arthritis was more frequent than arthralgia (47/58.0% vs. 16/19.8%, respectively) (p<0.0001); however, no specific pattern of joint involvement was found to be more prevalent. Mitral insufficiency was the most frequently detected echocardiographic sign (53 / 93.0%) and its association with aortic insufficiency was noted in 27 / 47.4% patients. Only 24 / 29.6% patients fulfilled Jones criteria for ARF requiring an evidence of previous group-A streptococcal infection (GASI). When compulsory GASI was disregarded, this number rose to 71/87.7% patients (p<0.0001). Conclusion. Girls were more affected by chorea; heart valves and joints were equally affected and represented the major clinical features; no specific pattern of joint involvement (eg.: migratory arthritis) could be labeled as typical; and strict adherence to Jones criteria, with compulsory documentation of a previous GASI, may lead to underdiagnosis of ARF.
Keywords: Rheumatic fever; Diagnosis; Diagnostic errors; Pediatrics; Rheumatology
Rheumatic fever is a public health problem in developing countries with incidence rates that reach epidemic levels. It is intimately linked to poor social and economic conditions, frequently affecting individuals that depend heavily on their physical strength and well being to earn their living. Therefore, especially due to its cardiac sequels, rheumatic fever represents a great burden on afflicted patients and on society as a whole.[1],[2],[3],[4]
There are still many controversial points related to the clinical manifestations and diagnosis of rheumatic fever. Physicians who frequently deal with new cases of the disease often feel uncomfortable in establishing a definite diagnosis, specially if they strictly follow Jones criteria. This set of criteria was created by a North-American physician in the mid-40's, when incidence of acute rheumatic fever (ARF) in the USA was high.[5] The Jones criteria have been revised a few times since then - the last revision was performed by the WHO Expert Consultation panel on Rheumatic Fever and Rheumatic Heart Disease.[6] Information on chosen methods and statistical analyses used in such reviews has not yet been disclosed.[6],[7]
A significant confusing factor in diagnosing ARF comes from the fact that recent studies have shown unusual clinical manifestations of the disease. In many instances, there are contradictory findings between different studies and they are frequently in contrast to what has been described in textbooks. Therefore, physicians seldom can rely on a "typical" clinical picture of ARF and diagnosis based on clinical judgement remains a difficult task.
The present study attempts to determine whether clinical and laboratory manifestations of acute rheumatic fever (ARF) are in accordance to what has been traditionally described as typical and how useful the Jones criteria are for diagnosis.
Materials and Methods
Data on clinical and laboratory features were collected from hospital records of 81 children and adolescents who were consecutively seen at the Pediatric Rheumatology Unit of the Hospital das Clνnicas/Federal University of Goiαs, Brazil, between January 1996 and August 2001. This Unit is an academic, tertiary, referral center located in Goiβnia, a city of 1 million inhabitants in Brazil's Central region. Prevalence of ARF in this area is high.[8],[9],[10] and the staff rheumatologists are nationally board-certified and well experienced in managing this disease.
Since one of our goals was to assess the reliability of Jones criteria and since no alternative diagnostic system has been formally proposed yet, the gold standard for ARF in this study was based on the authors' clinical judgement. Two of the authors (Pereira BAF and Silva NAS) are staff pediatric rheumatologists and reviewed all cases. Subjects were included in the study if: a) their ages ranged from 5 to 15 years; b) they had a clinical diagnosis of a first episode of ARF agreed by at least 2 board-certified rheumatologists; and c) this diagnosis was sustained for 6 months and 2 or more visits. Selected cases followed the Unit's standard protocol for ARF and data on their laboratory and clinical manifestations were recorded. Subjects were excluded from the study if diagnosis was pending, not sustained or attested by only one rheumatologist or other health professional. Other exclusion criteria were only one visit to our Unit and/or follow-up shorter than 6 months. All cases were tested against the Jones criteria in order to assess its reliability as a diagnostic tool.
All patients were tested for anti-streptolysin O antibodies (ASLO), erythrocyte sedimentation rate (ESR), C reactive protein (CRP), and mucoprotein levels, within 4 weeks after onset of symptoms. ASLO tests were performed by plate agglutination technique using 50 mL. of patient's sera and one drop of reacting solution containing ASLO sensitized latex particles (Biolab, Rio de Janeiro, BR). Results were expressed in IU/mL.
Evidence of previous group A streptococcal infection (GASI) was considered positive if history revealed an episode of upper airways streptococcal infection within 5 weeks preceding the onset of symptoms or if ASLO levels were greater than 320 IU/mL. Inflammatory activity was determined by ESR, CRP, and/or mucoprotein levels. First and/or second degree relatives were considered for family history of ARF. Echocardiographic exams were performed whenever a cardiac murmur was detected on physical exam. Joint features were divided into 5 patterns of presentation: (1) classical migratory arthritis, traditionally described as the most typical in ARF[6],[7] ;(2) non steroidal anti-inflammatory drugs (NSAIDs) responsive arthritis, when joints were not affected in a migratory fashion and inflammation remitted within 72 hours after initiation of NSAID; (3) atypical arthritis, when there was joint inflammation that did not fit into any of the above; (4) migratory arthralgia, as described for classical migratory arthritis but with pain as the sole symptom of joint involvement; and (5) other arthralgias, when the patient complained of joint pain that did not show a migratory pattern of presentation.
Parents of enrolled patients agreed to sign an informed consent. The study was approved by the Bioethics Committee of the attributable institution (CEP-HC/UFG).
Data were stored in a MS-Excel 2000 database system. STATA was used to calculate Pearson's chi-square, Student's t test (for comparison of means and difference of proportions) and analysis of 95% confidence intervals.
Results
Girls were more affected by ARF than boys (46 /56.8% vs. 35/48.2%, respectively) (p<0.001). Gender significantly influenced the presence of chorea, which was clearly more frequent in girls (23/50.0% vs. 5/14.3%) (p<0.001). The mean age did not differ between patients with (10.4±3 y.o.) and without (9.9±2 y.o.) chorea (p=0.353).
The frequency of clinical features is illustrated in [Figure - 1]. Heart and joints were the most frequent sites of involvement (65/80.2% and 63/77.8%, respectively) (p>0.05 - NS ). Chorea was noted in 28/34.6% patients. Arthritis occurred more frequently than isolated arthralgia (47/58% vs. 16/19.8%, respectively) (p<0.001), but no particular pattern of joint involvement was found to be significantly more frequent when stratified into subtypes of presentation [Table - 1]. Apical systolic murmur was by far the most common clinical cardiac sign (57/96.6%). Echocardiographic abnormalities were present in 57/70.4% patients, among which 53/93% had mitral insufficiency (MI). The association of mitral and aortic insufficiency (AI) accounted for about half of these findings (27/47.4%). Carditis was present in 19/28 (67.8%) patients with chorea.
Only 24/29.6% patients fulfilled the revised Jones criteria for diagnosis of first episodes of ARF,[6],[7] i.e. evidence of a previous GASI plus 2 major signs or 1 major and 2 minor signs. Since chorea allows for a diagnosis of ARF even in the absence of other major or minor signs[6],[7] the authors added to these afore mentioned patients a group of 25 other subjects that had chorea with no evidence of previous GASI and noticed 60.5% of fulfillment (49 patients). When compulsory evidence of a previous GASI was disregarded, this figure rose to 71/87.7% [Figure - 2].
Discussion
Acute attacks of first episodes of ARF were described in 81 distinct patients. The significantly greater overall proportion of female patients had also been noted in other studies.[11],[12] The high number of patients with chorea in the present study (28/34.6%) might have had some influence in the gender distribution of the total number of cases. Indeed, Carapetis and Currrie have postulated that, with a few exceptions, reports from certain geographical areas (Africa, Asia, the Pacific, the Caribbean and Arab countries) usually show lower relative numbers of patients with chorea, whereas those from other regions and/or ethnic groups (USA, Pakistan, Turkey, and Australian aborigines) present higher proportions.[13] According to that account, this study area may belong to the latter group, with figures comparable only to Australia's and the USA's.[13],[14],[15],[16] Interestingly, in this study patients with chorea were not older than other ARF patients, as has been traditionally described. In fact, they tended to be slightly younger when compared to other patients from the present study (p=0.353, NS ) as well as to patients with chorea from other series.[1],[13],[17],[18] The fact that cases were included only if they were younger than 15 years of age might have excluded some teenagers and lowered the mean age of the sample.
Articular and cardiac manifestations have been described as the most frequent features of rheumatic fever.[1],[2],[3],[4],[19],[20] Around 80% of patients in this series had joint complaints. Migratory arthritis has been said to be the typical articular manifestation of ARF,[4],[6],[19] but in our series, although arthritis occurred more frequently than arthralgia, no particular pattern of joint involvement was found to be more frequent when stratified into clinical subtypes of presentation [Table - 1]. Other studies have already indicated the fairly frequent occurrence of atypical joint manifestations.[21] This finding represents an additional complication, since diagnosis of ARF is highly dependent on clinical information obtained by history.
Carditis, in the form of valvulitis, usually occurs in more than three-quarters of patients with ARF.[2],[3],[20] Some studies have found higher rates due to the systematic use of echo-Doppler.[2],[3] In this study patients were submitted to an echo-Doppler exam only if they presented clinical evidence of carditis, in accordance to the American Heart Association and the WHO guidelines.[6],[7] However, there has been a flood of recent works showing a much greater sensitivity for the detection of valvulitis accomplished by the systematic use of echo-Doppler in patients with other features of ARF and with no clinical signs of carditis.[3],[17],[22],[23] Although any layer of the heart can be affected, mitral valve insufficiency was present in more than 90% of patients who underwent echo-Doppler, as has been reported by other authors.[1],[2],[3],[20] In this series, 19/28 (67.8%) patients with chorea had carditis as well, a figure similar to those described elsewhere.[1],[13] In fact, it has been shown that screening of choreic patients with echo-Doppler may further increment the detection of "silent" valvulitis.[17],[24],[25], [26]
Evidence of a prior streptococcal infection was considered essential for diagnosis of ARF in the 1965 revision of the Jones criteria and has been kept as such since then.[27] In the recently published revision from WHO, there is a note reminding of the existence of "probable cases", but even for these patients the need for an evidence of previous GASI remained.[6] In the present study, only about 60% of patients with a first attack of ARF had either chorea or a positive history of an upper airways GASI and/or high titers of ASLO. Highly suggestive clinical manifestations presented by the other 40% of patients and the long time of follow-up set as part of the gold standard indicate that obligatory evidence of GASI may leave almost half of ARF patients undiagnosed. This subgroup of undiagnosed ARF patients will not receive adequate early prophylaxis and will remain susceptible to recurrences of ARF that might lead to chronic rheumatic heart disease. One reason for such low sensitivity of Jones criteria in this series might come from the fact that patients could have been seen late on the course of the disease (when ASLO titers were already low) or during recurrences of subclinical first attacks of ARF that went unnoticed. Nevertheless, criteria for diagnosis of ARF must contemplate such possibilities since it is a disease that has a devastating effect over developing nations where such problems are common. Future case-control studies shall determine some epidemiological parameters ( e.g., specificity, accuracy) and may help to better evaluate this set of diagnostic criteria.
This study is a further indication of how difficult and complex diagnosis of ARF still is. Lack of laboratory markers; atypical clinical presentations of joint involvement; low availability of echo-Doppler in areas where it is most needed; and low sensitivity of Jones criteria (with compulsory evidence of a previous streptococcal infection) undoubtedly contribute to underdiagnosis. Criteria for diagnosis of ARF shall take into account all these situations in order to be considered as a relevant tool for physicians' daily practice.
References
1.Shet A, Kaplan E. Addressing the burden of group A streptococcal disease in India. Indian J Pediatr 2004; 71 : 41-48.
2.Bittar FF, Hayek P, Obeid M, Gharzeddine W, Mikati M, Dbaibo GS. Rheumatic fever in children: a 15-year experience in a developing country. Pediatr Cardiol 2000; 21 : 119-122.
3.Mishra TK, Routray SN, Behera M, Pattniak UK, Satpathy C. Has the prevalence of rheumatic fever/rheumatic heart disease really changed A hospital-based study. Indian Heart J 2003; 55 : 152-157.
4.Terreri MT, Ferraz MB, Goldenberg J, Len C, Hilario MO. Resource utilization and cost of rheumatic fever. J Rheumatol 2001; 28: 1394-1397. [PUBMED]
5.Jones TD. Diagnosis of rheumatic fever. JAMA 1944; 126 : 481-4.
6.Rheumatic fever and rheumatic heart disease. Report of a WHO Expert Consultation. World Health Organization, Geneva, 2004 (Technical Report Series No. 923)
7.Dajani AS, Ayoub E, Bierman FZ et al. Guidelines for diagnosis of rheumatic fever: Jones criteria, Updated 1992. Circulation 1993; 87 : 302-307.
8.Haddad N, Silva MB. Mortality due to cardiovascular disease in women during the reproductive age (15 to 49 years), in the state of Sao Paulo, Brazil, from 1991 to 1995. Arq Bras Cardiol 2000; 75 : 375-379. [PUBMED] [FULLTEXT]
9.Meira, ZMA, Castilho SRT, Barros MVL et al. Prevalence of rheumatic fever in children from a public high school in Belo Horizonte. Arq Bras Cardiol 1995; 65: 331-334.
10.Gus I, Zaslavsky C, Seger JM, Machado RS. Epidemiology of rheumatic fever. A local study. Arq Bras Cardiol 1995; 65 : 321-325.
11.Risvi SF, Khan MA, Kundi A, Marsh DR, Samad A, Pasha O. Status of rheumatic heart disease in rural Pakistan. Heart 2004; 90 : 394-399.
12.Melka A. Rheumatic heart disease in Gondar College of Medical Sciences Teaching Hospital: socio-demographic and clinical profile. Clin Exp Rheumatol 1996; 14 : 567-569.
13.Carapetis JR, Curie BJ. Rheumatic chorea in northern Australia: a clinical and epidemiological study. Arch Dis Child 1999; 80 : 353-358.
14.Choong CY, Abascal VM, Weyman J et al. Prevalence of valvular regurgitation by Doppler echocardiography in patients with structurally normal hearts by two-dimensional echocardiography. Am Heart J 1989; 117 : 636-642.
15.Denny FW. A 45-year perspective on the streptococcus and rheumatic fever: The Edward H. Kass lecture in infectious disease history. Clin Infect Dis 1994; 19 : 1110-1122.
16.Al Eissa YA. Acute rheumatic fever during childhood in Saudi Arabia. Indian J Pediatr 1990; 57 : 771-773.
17.Elevli M, Celebi A, Tombul T, Gokalp AS. Cardiac involvement in Sydenham's chorea: clinical and Doppler echocardiographic findings. Acta Paediatr 1999; 88: 1074-1077. [PUBMED]
18.Ghram N, Alani C, Oudali B, Fitouri Z, Ben Becher S. Sydenham's chorea in children. Arch Pediatr 1999; 6: 1048-1052.
19.da Silva NA, Pereira BAF. Acute rheumatic fever. Still a challenge. Rheum Dis Clin North Am 1997; 23 : 545-568.
20.Kafetzis DA, Chantzi FM, Grigoriadou G, Vougiouka O, Liapi G. Incidence and clinical profile of rheumatic fever in Greece. Eur J Clin Microbiol Infect Dis 2005; 24: 68-70
21.Hilario MOE, Len C, Goldenberg J, Fonseca AS, Ferraz MB, Naspitz CK. Febre reumαtica. Manifestaηυes articulares atνpicas. Rev Assoc Med Brasil 1992; 38 : 214-216.
22.Ozkutlu S, Hallioglu O, Ayabacan C. Evaluation of subclinical valvar disease in patients with rheumatic fever. Cardiol Young 2003; 13 : 495-499.
23.Hilαrio MO, Andrade JL, Gasparian AB, Carvalho AC, Andrade CT, Len CA. The value of echocardiography in the diagnosis and follow-up of rheumatic carditis in children and adolescents: a 2 year prospective study. J Rheumatol 2000; 27: 1082-1086.
24.Lanna CC, Tonelli E, Barros MV, Goulart EM, Mota CC. Subclinical rheumatic valvitis: a long-term follow-up. Cardiol Young 2003; 13 : 431-438.
25.Saxena A. Diagnosis of rheumatic fever: current status of Jones criteria and role of echocardiography. Indian J Pediatr 2000; 67(suppl 3): 11-41.
26.Narula J, Chandrasekhar Y, Rahimtoola S. Diagnosis of active rheumatic carditis. The echoes of change. Circulation 1999; 100: 1576-1581.
27.Stollerman GH, Markowitz M, Taranta A, Wannamaker LW, Whittemore R. Report of the Adhoc Committee on Rheumatic Fever and Congenital Heart Disease of the American Heart Association: Jones Criteria (Revised) for guidance in the diagnosis of rheumatic fever. Circulation 1965; 32 : 664-668.(Pereira Breno AF, da Silva Nilzio Antoni)