Material and methods

Karachi is the largest cosmopolitan city of Pakistan and lies within the Tropics. The climate is hot and humid in summer, and mild in winter. There is scant rainfall. It has a population of over 12 million, representing people from all four provinces of Pakistan.

For our study, we registered 161 children, 57 suffering their first episode, and 104 with recurrent episodes of acute rheumatic fever. The paediatric departments of three major teaching hospitals of Karachi participated in the study, namely the National Institute of Cardiovascular Diseases, a purely cardiac facility, the National Institute of Child Health, a Paediatric Hospital, and the Civil Hospital of Karachi, a general hospital. These hospitals serve as primary and tertiary centers for the care of children. The period of study was from March 1991 to March 1994. Both authors independently confirmed the clinical findings. The incidence of new cases registered varied slightly during the year, being higher in the months of winter and spring. A mean of 28.5 new cases were registered in a year for those suffering their first episode, and 52 having recurrent episodes. The mean age of patients was almost similar, at 9 ± 1.95 years for those having their first episode, and 10 ± 1.6 years for those with recurrent episodes. The overall gender was also equal in the two groups, and patients in both groups were under weight for their age, at 22.6 ± 6.2 kg for those suffering the first episode, and 24 ± 5.7 kg for those having recurrent episodes.

Prior to referral to tertiary care, 68% of the patients had been treated by their general practitioner, 14% by homeopaths, and 10% by faith healers, but were not appropriately diagnosed or treated. The remaining 8% of children had not been seen by any of these. The duration of symptoms from the onset to the time of registration ranged between 8 and 75 days, with mean of 36 ± 18 days.

Two-thirds of the children were referred from the urban slums of Karachi, one-fifth from crowded urban areas, and one-sixth from rural areas, so that five-sixths of children came from urban areas. At the time of registration, four-fifths of children suffering their initial attack were attending school, compared to half of those having recurrent episodes.

The average income per month from all sources for the family was estimated to be US$55 for those suffering the first episode, compared to US$44.25 in those having recurrent episodes. The number of members living together in the family were 10 ± 6 in those suffering an initial attack, compared to 9 ± 5 for those having recurrent episodes, with most living in 2 room houses when suffering a first episode, but in 1 roomed houses for those having recurrent episodes. On average, 4 members of the family of those having a first episode shared a single room, compared to 7 members in those having recurrent episodes. A history of recurrent episodes amongst the family was elicited for 5% of patients suffering their first episode, compared to 10% of those having recurrent episodes.

Management

We advised complete bed rest for all patients, with gradual ambulation after the erythrocytic sedimentation rate had started to return towards normal. The mean duration of the acute attack, indicated by amelioration of symptoms and normalization of the erythrocytic sedimentation rate, was shorter in those suffering their initial attack, at 53 ± 24 days, compared to 67 ± 32 days in those having recurrent episodes. It was shorter in patients with arthritis, with an average of 22 days, intermediate in those with chorea, and longest in those with carditis, the latter ranging from 45 to 60 days.

In order to eradicate streptococcal throat infections, we gave injections of Benzathine penicillin to all patients, after a test dose, in the dose of 0.6 million units for children weighing less than 27 kg, and 1.2 million units for those heavier than 27 kg. The first injection was given at the time of admission, and repeated after 2 weeks, and then at 4 weekly intervals. Inflammation was controlled with Aspirin, which was given in the dose of 100 mg/kg/day in 4 divided doses, being reduced to 75 mg/kg/day after 2 weeks, and continued in the same dose until the erythrocytic sedimentation rate had returned to normal. Steroids were administered to one-tenth of patients suffering their first episode, and to a similar proportion of those having recurrent episodes, in the setting of severe toxemia, abdominal pain, vomiting, pericarditis and dominant myocarditis. All patients with carditis were treated with Digoxin, while anti-failure medication was added over the first one or two weeks for one-tenth of patients suffering an initial attack, but was needed for seventeen-twentieths of those having recurrent episodes. Children with severe carditis complicating recurrent episodes needed more aggressive vasodilator therapy, such as inhibitors of angiotensin converting enzyme, which were used in 15% of patients, intravenous inotropic support in 3% of patients, and sodium nitroprusside as dermal patches in 5%. Surgery was advised for uncontrolled congestive cardiac failure. Patients with chorea were free from symptoms within 2 to 3 months of being treated with oral Phenobarbitone.

Results: clinical profile according to the modified Jones criteria

Major manifestations (Fig. 1)

Figure 1. Comparative clinical profile of acute rheumatic fever in those seen after a first episode versus those suffering recurrent episodes diagnosed by the Jones criterions.

Carditis. The mean heart rate calculated by electrocardiographic means in those suffering their first episode was 112 ± 20 beats per minute, compared to 122 ± 20 in those having recurrent episodes. The heart rate was greater than 101 beats per minute in 88% of patients.

Carditis was significantly less frequent in those suffering their first episode, at 64% compared to 96% in those having recurrent episodes (p < 0.05). Congestive cardiac failure was also significantly less frequent, being seen in only 21.1% of those having their initial attack compared to 65.4% in those having recurrent episodes (p < 0.0001). Pericarditis was observed in 5.3% of those having a first episode, compared to 9.6% for those with recurrent episodes. Oedema was also much less frequent in those having an initial episode, at 7% compared to 32.6% of those having recurrent episodes (p < 0.05). The carditis was graded at mild in 29.8% of those having a first episode, compared to 10.6% in those with recurrent episodes, moderate in 26.3% compared to 53.8%, and severe in 5.3% of those with a first episode compared to 31.8% of those having recurrent episodes (p < 0.05 – Tables 1 & 2). Pulmonary arterial hypertension was significantly less frequent in those suffering a first episode, at 29.8% compared to 86.5% for those with recurrent episodes (p < 0.05). Atrial fibrillation was absent in those having their initial attack, but seen in 3.8% of those having recurrent episodes.

Isolated mitral valvar regurgitation was a frequent finding in those having a first episode, being noted in 80%, but was found in only 13% of those with recurrent episodes. Aortic regurgitation combined with mitral regurgitation was detected in 14% of patients having an initial attack, but in 43% of those having recurrent episodes. Tricuspid, mitral, and aortic regurgitation were detected in 6% of the patients having an initial attack, but in 11% of those having recurrent episodes. Stenotic lesions were found only in those suffering recurrent episodes, with mitral stenosis and regurgitation being present in 15% of cases. Mitral regurgitation and stenosis with tricuspid regurgitation were present in 17%, and aortic stenosis with mitral stenosis and regurgitation and tricuspid regurgitation found in 1%. All the findings were significantly more severe in those having recurrent episodes.

Arthritis. This feature was more frequent during the first episode, at 61.4%, compared to 36.5% in those suffering recurrent episodes (p < 0.05).

Chorea. This was seen more frequently in those suffering the initial attack, 15.8% compared to 3.8% in those with recurrent episodes (p < 0.05). It was isolated in 14%, and associated with carditis in 2% of those having their first episode whereas, amongst those having recurrent episodes, 3% had isolated chorea and 1% had chorea associated with carditis.

Skin lesions. These were always associated with carditis. Sub-cutaneous nodules were more frequent in those having a first episode (7%), as opposed to 5.8% of those with recurrent episodes (p > 0.05). Erythema marginatum was also observed more frequently in the setting of the first episode, at 3.5%, compared to only 0.9% of those having recurrent episodes (p < 0.05).

Minor manifestations (Fig. 1)

Fever. Fever was present in all children except those who had isolated chorea as a manifestation. Thus, 84% of children suffering their first episode complained of fever, compared to 94% of those having recurrent episodes (p > 0.05).

A toxic state was observed in 31% patients suffering the first episode, compared to 41% of those having recurrent episodes, while 47% children with a first attack were anaemic, compared to 63% of those having recurrent episodes.

Arthralgia. These problems with the joints were found in 21.1% of those having a first attack, compared to 32.7% of those with recurrent episodes (p < 0.05). Arthralgia was always seen associated with carditis in both groups.

Streptococcal throat infection. A history of a sore throat was obtainable from 54.4% of children having a first episode, compared to 21.2% of those having recurrent episodes (p < 0.05).

Laboratory work. The throat swab and cultures for β-hemolytic streptococci were positive in 21.1% of those suffering an initial attack, compared to 16.3% of those with recurrent episodes. The mean total leukocytic count was comparable, at 11,907 ± 3762/cu.mm for those having a first episode, and 13,409 ± 9159 in those with recurrent episodes. All patients had erythrocytic sedimentation rates of more than 12 mm in 1st hour. The mean rate was 78 ± 33 mm in the first hour of those suffering a first episode, compared to a higher mean of 97 ± 33 mm in those with recurrent episodes. The anti Streptolysin O Titer was more than 400 i.u. in 87.7% of those suffering a first episode, and 95.2% in those with recurrent episodes, and was less than 200 i.u. in 12% of those having a first episode, and in 5% of those with recurrent episodes, all of whom had isolated chorea.

Follow-up. Amongst those having their initial attack, 2 of 57 children (3.6%), compared to 12 of 104 (13.3%) of those with recurrent episodes, were lost to follow-up. Thus 147 of 161 (81.3%) were followed to the end of the study, 55 having had a first attack, and 92 suffering recurrent episodes. These children together were followed for a mean period of 13,494 patient years. Patients suffering their first episode were followed for a mean period of 18 ± 7.8 months, as opposed to a period of 17.3 ± 8.3 months for those suffering recurrent episodes. The mean numbers of follow-up visits were 5 ± 2.5 visits/patient for those having their first attack, and 5.6 ± 2.9 per patient in those suffering recurrent episodes.

During the period of follow-up, 3 children (5.3%) first seen after their initial attack developed a recurrence of acute rheumatic fever, giving a recurrence rate of 0.036/patient year, compared to 11 children (10.8%) in those who had already suffered recurrent episodes, with their recurrence rate being 0.085/patient year (p < 0.05). Secondary prophylaxis with long acting penicillin was achieved in 94.5% of those first seen after an initial attack, and in 85.5% of the patients who had already suffered recurrent episodes.

In those seen after their first attack, 38.6% had no carditis at their initial examination. In those with carditis, it was mild in 29.8%, moderate in 26.3%, and severe in 5.3%. At the conclusion of the follow up, 55 remained of the children seen after their first episode, and of them, 32 (56.1%) had no carditis, 13 (22.8%) had mild carditis, and 10 (17.5%) had moderate carditis. Thus, of those seen after their first attack, 30.3% patients were cured of carditis, an improvement in severity was marked in 39.4%, and 30.3% of patients showed some improvement. A statistically significant improvement in the severity of carditis, therefore, was noted in all grades (p < 0.09 – Table 1 and Fig. 2).

Table 1. Outcome of carditis in patients seen after the first episode of acute rheumatic fever*.

Figure 2. Comparative outcome of carditis at the end of follow-up between those seen after the first episode as opposed to recurrent episodes of acute rheumatic fever.

When comparing the data for carditis for those seen already having had recurrent episodes, at the start of study 3.8% patients had no carditis, it was mild in 10.6%, moderate in 53.8% and severe in 31.8%. At the conclusion of follow-up, 90 patients remained in the study, as 2 of the initial 104 had died, and 12 were lost to follow-up.

The outcome for these patients at the time of last follow-up visit showed that 11 (10.6%) had no carditis, 13 (12.6%) had mild carditis, 62 (59.6%) had moderate carditis, and only 4 of the 90 (3.8%) had severe carditis (Table 2 and Fig. 2). There was significant improvement in 44 of the 90 patients (48.9%), no improvement in 42 (46.7%), and deterioration in 4 patients (4.4%). Two patients had died (1.9%). Statistical comparison between the first and last assessments of the grading of carditis amongst these 90 patients showed significant improvement at the time of the last examination. (p < 0.000001).

Table 2. Outcome of carditis in patients seen with recurrent episodes of acute rheumatic fever*.

One patient seen after having had recurrent episodes developed endocarditis on the aortic valve during a recurrence, and required replacement of the valve. Mitral valvar commissurotomy was performed in 2 children, while a symptomatic child at Grade IV of the classification of the New York Heart Association with severe mitral regurgitation underwent replacement of the mitral valve.

When we compared the outcome of carditis between those seen after their first episode and those who had suffered recurrent episodes, we found that 69.7% of the children seen after their initial attack improved, compared to 48.9% of those who had suffered recurrent episodes, while moderate to severe carditis was statistically more significant in those who had suffered recurrent episodes (Fig. 2).

Discussion

In our prospective study, we registered approximately twice as many patients suffering recurrent episodes of acute rheumatic fever compared to those seen after their first attack. The clinical profile was significantly milder for those seen after their first attack compared to those with recurrent episodes, this being similar to the trend reported in the past in developed countries. It is of interest to note that the recently reported clinical profile of resurgent acute rheumatic fever in seven communities in the United States of America was in large measure similar to that for our patients seen after their first attack. The mean incidence of carditis of 57% was comparable to 61.4% in our study. The points to note were that the epidemics occurred in suburban settings, and at least one of the communities was middle class. The virulent strains of streptococci re-emerged after a long period of dormancy in the suburban communities in the United States of America.¹⁰

Our results confirm the reports of Sanyal et al. from Delhi,⁹ and Majeed and colleagues from Kuwait.⁴ They also noted that the incidence and severity of carditis was significantly less in those seen after their first episode compared to those who had already suffered recurrent episodes. Severe carditis was largely confined to those having recurrent episodes. Thus, if no distinction was made between those suffering an initial attack compared to those having recurrent episodes, then patients diagnosed with acute rheumatic fever according to the modified Jones criteria would appear to be manifesting severer disease, with the dominant lesion as carditis, and would resemble the reports from most of the developing countries.^{6, 9, 11–16} Our data, therefore, suggests that those reporting severe and fulminant acute rheumatic fever in developing countries might have included patients suffering recurrent episodes along with those seen after the initial attack.

Subcutaneous nodules, and Erythema marginatum, are reported to be seen less frequently as manifestations of acute rheumatic fever in the developing countries.^17–21 Our study does not support this observation. We found subcutaneous nodules, as well as erythema marginatum, with similar frequency to the reports from India⁹ and Kuwait,⁴ all of which studied patients seen after their first episode of infection.

Follow-up of our patients, as with previous studies, showed that chronic rheumatic valvar disease was more severe in those suffering recurrent episodes^{6, 7, 10, 11} compared to those seen after their first infection,^{2, 4, 22, 23} and that the symptoms improved in both groups, but significantly more in those seen subsequent to their first episode (p < 0.05). If the chronic rheumatic valvar disease was moderate or severe in those seen at the first visit with recurrent episodes, then the severity did not improve significantly, and deteriorated in some cases. Indeed, at the conclusion of our study, there were no children seen after their initial attack who still had severe carditis, and both patients who died had already had recurrent episodes when first seen. Thus, the chances of cure or improvement in the severity of carditis were significantly greater for those children seen after an initial attack, a finding reported by several other investigators.^{2, 4, 9, 22, 23}

At the time of registration, stenotic valvar lesions were present only in those seen after recurrent episodes. At the conclusion of the follow up, no patients seen after their first attack had developed mitral stenosis. This is at variance to the report of Sanyal and colleagues.²⁴ The period of follow-up for our children, however, was probably too short for any to have developed mitral stenosis.²⁵

Follow-up of all our patients showed that the adherence to secondary prophylaxis was much less in those seen after already having recurrent episodes compared to those seen after their first attack. This undoubtedly influenced the significantly higher recurrence rate in those seen after already suffering recurrent episodes. The rate of recurrence during the period of follow-up for those seen after their first attack was 0.036/patient year, this being half that seen for those already having had recurrent episodes, with 10.8% of those already having had recurrences developed still further recurrences, compared to 5.3% of those seen after their first episode of acute rheumatic fever. This is also in keeping with previous reports.^24–30

The comparison of the socioeconomic conditions of our patients showed that all the children came from the lower socioeconomic communities in Pakistan. The children came from the urban slums, where sanitation and hygiene is poor.^{8, 19, 31–35} The huts and houses are huddled together across narrow passages. Over-crowding was rife, with 9 to 10 persons living in one house, which comprised of one or two rooms.^{17, 35–38} The registers for school attendance showed that 58 of 161 (36%) children with acute rheumatic fever did not attend school, which suggests that the assessments of the incidence of acute rheumatic fever based on school attendance in the developing countries do not truly reflect its prevalence.^38–40 When we compared the socioeconomic parameters in those suffering recurrent episodes with those seen after their first attack, we found that those suffering recurrent episodes came from poorer families. This may in part explain the higher recurrences and higher non-compliance to secondary prophylaxis amongst these children.^{20, 26, 30, 41}

Recurrences due to non-compliance to the secondary prophylaxis resulted in a significant number of those seen after their first episode joining those who had suffered recurrent episodes at the conclusion of the study.^{24, 26, 29, 42} Thus, the process of conversion to recurrent episodes is a continuous phenomenon in children with acute rheumatic fever, and is expected to continue. It also explains why we registered more children who had already suffered recurrent episodes than those seen after their first attack. This suggests that the pool of those with recurrent acute rheumatic fever will continue to increase due to the persistent recurrences, and will add to the burden of severe cardiac sequels.^{24, 26–30}

This phenomenon of an ever-increasing pool of severe cardiac disease also explains the reports from the developing countries showing the prevalence of established rheumatic valvar heart disease, which is often severe and fulminant.^{6, 7, 11, 12} Our data showed that complications such as bacterial endocarditis and chronic cardiac failure were seen only in those suffering recurrent episodes with moderate, or more often severe, established chronic rheumatic valvar disease.

We found that the majority of our patients came from urban rather than rural areas, and none of the children belonged to the middle or high-income groups.^{8, 11, 31–35} It can, of course, be argued that only poor children come to the public hospitals. One of the hospitals included, nonetheless, was the National Institute of Cardiovascular Diseases, where large sections of patients are from the middle or lower middle class.^{14, 41} Our study, therefore, shows that acute rheumatic fever is endemic in the urban slums, where the causative factors of adverse socioeconomic and environmental factors are aggregated. Preventive strategies should be targeted to these communities. The local providers of health care need to be trained to apply preventive measures. Sore throat should be treated in these communities with Penicillin, with or without the facility for throat culture, particularly if the examination shows pus or other features of streptococcal infection.^{5, 40}

Local non-medical providers of health care initially treated the majority of our children, and there was a delay of almost one month before the children presented to one of the three hospitals involved in the study. The preventive strategies should be directed to this group of providers of care if we are to control acute rheumatic fever in these communities. Teaching hospitals, non-Governmental organizations, and Government programs of health education, should be actively involved in teaching primary and secondary prevention of rheumatic fever, targeting these urban communities. Our study, along with others^{24, 25, 43–45} has shown that hospital-based programmes are unlikely to control the occurrence of acute rheumatic fever.

Cure from the carditis of acute rheumatic fever is possible in those seen after the first attack, but once significant chronic valvar disease has occurred, the chances of cure are reduced. Recurrences are likely to further damage the heart, with higher morbidity and mortality.^{6, 7, 12, 46}

Thus, our prospective study has shown that acute rheumatic fever is mild when seen in those presenting with an initial attack, with severe disease limited to those suffering recurrences. The chances of cure from severe disease are small. The rate of cure from the sequels of acute rheumatic fever was significantly higher amongst those seen after their initial attack. The infections tended to occur in urban slums. Thus, it is these communities that should be targeted for preventive strategies.