Discussion

Acute rheumatic fever is seen worldwide; however, it is commonly detected in under-developed and developing countries and it still continues to be a major health problem in those countries. Rheumatic heart disease remains a significant cause of cardiovascular morbidity and mortality. Watkins et al. estimated that there were 319,400 (95% uncertainty interval: 297,300 to 337,300) deaths due to rheumatic heart disease in 2015.^{Reference Carapetis, Currie and Mathews8–Reference Watkins, Johnson and Colquhoun10} In a recent systematic review of prospective population-based studies, the reported incidence of acute rheumatic fever ranged from 5 to 51 per 100,000 population worldwide in the 5- to 15-year age group.^{Reference Tibazarwa, Volmink and Mayosi11} The lowest incidence rate of 0.5 to 3 per 100,000 population was found in Europe; however, focal outbreaks of acute rheumatic fever have been reported in industrialised countries in the last years.^{Reference Pastore, De Cunto, Benettoni, Berton, Taddio and Lepore12} The 2015 Jones criteria identified the acute rheumatic fever incidence cut-off < 2 of 100,000 school-aged children per year to distinguish between low risk and moderate- to high-risk populations.^{Reference Gewitz, Baltimore and Tani13} Still, to be able to apply the Jones criteria which was revised in 2015 to our country, the incidence and the prevalence rates should be known. There is no study on the incidence of acute rheumatic fever that covers all of our geographical regions. However, it is known that the percentage of children with acute rheumatic fever in Turkey is still higher than the global rate. The number of reports is not adequate for an accurate estimation of acute rheumatic fever incidence. Beyazova et al.^{Reference Beyazova, Benli and Beyazova14} estimated it as 56.6 per 100,000 children during 1970–1973 and 36.7 per 100,000 15 years later. Saraçlar et al.^{Reference Saraçlar, Ertuğrul and Özme ve Ajun15} found the incidence 20 per 100,000 children during 1972–1976. Örün et al.^{Reference Örün, Ceylan and Bilici16} estimated it as 37.6 per 100,000 during 1980–1989, 60 per 100,000 during 1990–1999, and 21 per 100,000 during 2000–2009 in the Central Anatolia Region of Turkey. Narin et al.^{Reference Narin, Mutlu and Argun6} estimated the lowest incidence rate of acute rheumatic fever in Kayseri which is in the Central Anatolia region (7.4 per 100,000). As being one of the developing countries, Turkey still suffers from acute rheumatic fever, and a comprehensive study about this disease still cannot be conducted due to the lack of reliable countrywide database of the patients who admitted to the hospitals in the country. Additionally, the patients who have admitted to more than one centre in our country causes problems on determining the exact incidence and prevalence rates of the disease. However, we tried to determine the frequency of acute rheumatic fever in our hospital by proportioning the number of the patients diagnosed with acute rheumatic fever to all patients who applied to the general paediatrics outpatien and PCO departments. When the distribution of patients diagnosed with acute rheumatic fever by years is examined, it is seen that there is a decrease in the number of patients with acute rheumatic fever over the years, despite the significant increase in the number of patients examined in both general paediatrics outpatien and PCO departments. In this study, the maximum numbers of the patients in a year were in 1994 and 1999 with more than 45 and 41 cases per year, respectively. In these years, when the number of patients was the highest, we found that the frequency in our hospital also increased. Also, the fact that some administrative factors could have (our hospital’s agreements with health institutions, refugees, etc.) played a role in the increase in the frequency of the disease in these 2 years could be noted. Moreover, it is clearly seen that the number of the cases per year dropped drastically, especially in the recent years of the study. This decline could be caused by certain factors including improvement in socio-economic conditions, health services, increase in antibiotic use, and hygienic conditions. Because secondary prophylaxis limits the progression of long-term cardiac sequelae, early diagnosis of acute rheumatic fever is essential for reducing morbidity and mortality. However, if we consider that the number of the paediatric centres and paediatric cardiologists increased drastically in second and third decades in our region, it can be said that this decrease on incidence rate may not be correct due to, as we mentioned, we do not have a reliable countrywide data banks of the patients. A multicentre epidemiological study is required to obtain more reliable statistical values.

Group A streptococcal pharyngitis and subsequent acute rheumatic fever are most frequent among children and adolescents between 5 and 15 years of age.^{Reference Anita, Zaidi, Goldman, Kliegman, Behrman, Jenson and Stanton17} While there were differences between the ages of the first episodes of acute rheumatic fever published in the literature, in most studies, the mean age was between 9.5–11.2 years,^{Reference Veasy and Lloyd18-Reference Currie and Carapetis21} and it was parallel with our results. In the present study, the rate of patients aged from 5 to 15 years was 97.9% and half of the patients’ age were between 9 and 12 years. There was no patient under 5 years of age.

Both genders were affected equally by acute rheumatic fever except for patients with chorea which was similar with the literature.^{Reference Saraçlar, Ertuğrul and Özme ve Ajun15,Reference Veasy and Lloyd18-Reference Bitar and Hayek20} Of the patients with chorea, 60.8% were female. Although there was a slight male predominance, no significant difference was observed between sexes in all periods. In two studies from Turkey, female/male ratio was found to have increased slightly.^{Reference Narin, Mutlu and Argun6,Reference Bostan and Cil22}

The increased rate of admission of patients with acute rheumatic fever to hospitals is in correlation with the seasonal variability of pharyngitis, which is caused by group A beta-haemolytic streptococci and is most often diagnosed in the winter and spring.^{Reference Narin, Mutlu and Argun6,Reference Örün, Ceylan and Bilici16,Reference Cleonice, Mota, Aiello, Anderson, Anderson, Baker, Penny, Redington, Rigby and Wernovsky23} Approximately two-thirds of the patients admitted in the spring and winter seasons (62.8%). The peak season for the initial presentation was winter in all periods. The least common season in the study was summer with 15 % in first period; however, the occurrence rate of acute rheumatic fever increased to 25.3% in the second period and 30% in the third period. The reason for this is that most of the people of the region spend the summer months in high altitude highlands, which are colder places. In recent years, due to climate change in our region, the temperature levels in early summer were lower. This might also explain the increase in acute rheumatic fever rates throughout summer.

Universally, the most occurring major manifestations during the first episode of acute rheumatic fever remain carditis (50%–70%) and arthritis (35%–66%).^{Reference Seckeler and Hoke4,Reference Veasy and Lloyd18,Reference Boyarchuk, Boytsanyuk and Hariyan24-Reference Grassi, Fesslovà and Carnelli29} Our data showed the frequency of rheumatic carditis of 83.6%, which is higher than the other published data and there was a slight increase (80.6%, 85.7%, and 88.6%, respectively); however, in last periods, there was no statistically significant difference (p = 0.20). Also, the number of subclinical carditis gradually increased and this was statistically significant (p < 0.001). All patients with two or three major manifestations were accompanied by carditis and approximately two-thirds of the patients had both carditis and arthritis. Patients with carditis have been reported to be around 30–45% in previous studies.^{Reference Lee, Naguwa, Cheema and Gershwin30,Reference Cimen, Oran and Cimen31} All our patients had echocardiography performed at the time when acute rheumatic fever was suspected which may be the reason for such a high percentage of carditis in our cohort. The Jones criteria for the diagnosis of acute rheumatic fever were revised in 2015.^{Reference Gewitz, Baltimore and Tani13} Echocardiography is now recommended in all the patients with suspected or confirmed acute rheumatic fever. Subclinical carditis can be used as a major criterion for acute rheumatic fever in all populations.^{Reference Beaton and Carapetis32} Genetic predisposition may be a factor for development carditis and rheumatic heart disease.^{Reference Bitar and Hayek20,Reference Haydardedeoğlu, Tutkak, Köse and Düzgün33} In our population, the prevalence of valvar lesions were similar to the reports from other high-risk countries.^{Reference Azevedo, Pereira and Guilherme3,Reference Narin, Mutlu and Argun6,Reference Bitar and Hayek20,Reference Arora, Subramanyam, Khalilullah and Gupta34} Mitral involvement was predominant at diagnosis in just more than one-half of the cases (54.9%) and was associated with aortic regurgitation (34%), whereas isolated aortic involvement was rare (5.7%). Mitral insufficiency is caused by abnormal coaptation due to the annular dilatation and chordal elongation.^{Reference Arora, Subramanyam, Khalilullah and Gupta34,Reference Barash and Matityahu35} In the present study, about one-half of the patients had mild carditis during the initial diagnosis. Although severe carditis rate was higher in the first period, the incidence of mild carditis has increased steadily and recently reached 55%. There was a statistically significant difference between the periods in terms of carditis severity (p = 0.03). We thought that, early recognition of the disease, adequate treatment of penicillin, prophylaxis, and good follow-up of the patients caused carditis to be milder. There was no statistically significant relationship between valve involvement and the severity of carditis, also between age and sex (p > 0.05). Compared with older patients, children who presented before 5 years of age were more likely to have moderate to severe carditis and to present with arthritis or the rash of erythema marginatum and were less likely to have chorea.^{Reference Lloyd, Tani, George Veasy, Minich and Shaddy36} In our study, there was no patient under 5 years of age. It was not possible to tell the true relationship between carditis severity and age since we do not know the first episode ages of patients with recurrence.

Overall, 4% to 11% of all patients with acute rheumatic fever develop clinically detectable pericarditis, one of the known, but less common findings in rheumatic carditis. When present, pericarditis is most often observed with pancarditis.^{Reference Bitar and Hayek20,Reference Arora, Subramanyam, Khalilullah and Gupta34,Reference Rathore and Barton37} In the present study, pericardial effusion was detected in 5.7% of patients with pancarditis.

Various rhythm and conduction abnormalities also can develop during the course of the disease.^{Reference Kula, Olgunturk and Ozdemir38,Reference Balli, Oflaz, Kibar and Ece39} One of the most characteristic disturbances of conduction in acute rheumatic fever is first-degree heart block. First-degree atrioventricular block is a minor Jones criterion and is seen in approximately 10–75 % of the cases. In our study, excluding patients with first-degree atrioventricular block, rhythm, and conduction anomalies were observed in five patients (1.6%). All the electrocardiogram abnormities detected at onset returned to normal once the patients recovered from the acute phase, as previously described by Ballı et al.^{Reference Balli, Oflaz, Kibar and Ece39}

The reported frequency of arthritis in children with acute rheumatic fever ranges from 60 to 80%, depending on the region.^{Reference Seckeler and Hoke4,Reference Martins, Veasy and Hill40} In our study, it was observed in 74% of the cases. There was a slight decrease (75.2% to 67.5%) in arthritis rates between first and last periods that could be related with the common use of anti-inflammatory drugs. Some authors described arthritis as the most common criterion (up to 70% of the cases).^{Reference Joseph, Madi, Kumar, Nelliyanil, Saralaya and Rai41} In the present study, the most commonly affected joints were larger ones such as knees (72.8%), ankles (53.8%), and 43 (15.4%) patients had small joint (fingers) involvement. Monoarthritis may be important as a clinical manifestation of acute rheumatic fever in selected high-risk populations.^{Reference Cann, Sive, Norton, McBride and Ketheesan26,Reference Carapetis and Currie42-Reference Sanyal, Thapar, Ahmed, Hooja and Tewari45} In the high-risk indigenous Australian population, monoarthritis has been found to be present in 16% to 18% of confirmed cases of acute rheumatic fever.^{Reference Carapetis and Currie42} In this study, monoarthritis was found in 97 (34.8%) patients and all of them were accompanied by carditis.

Chorea was seen in 13.5% cases in this study and it varied from 2.7 to 18.8% cases in other studies.^{Reference Arora, Subramanyam, Khalilullah and Gupta34,Reference Rayamajhi, Sharma and Shakya46-Reference Vasan, Shrivastava, Vijayakumar, Narang, Lister and Narula51} Some publications indicate rates of 6–31% and even 49%.^{Reference Carceller, Taoiero, Rubin and Miro52} The difficulties in the acute rheumatic fever diagnosis were described in the cases of isolated rheumatic chorea, when other major criteria of acute rheumatic fever are absent. Fourteen patients (27.5%) had isolated chorea and others (72.5%) had chorea combined with carditis and arthritis.

The low rates of detected subcutaneous nodules and erythema marginatum in acute rheumatic fever patients found in this study were likewise reported in literature.^{Reference Seckeler and Hoke4,Reference Joseph, Madi, Kumar, Nelliyanil, Saralaya and Rai41} Subcutaneous nodules are often observed in patients who also have carditis, and erythema marginatum, subcutaneous nodules almost never occur as the sole major manifestation of acute rheumatic fever. In our patients, only two had subcutaneous nodules with carditis. Two patients had erythema marginatum, although one of them occurred as the sole major manifestation.

In the present study, the rates of minor manifestations were consistent with the literature. The rates of arthralgia were reported as 54.6–81.1%, fever as 40–62%, prolongation of PR interval as 15.9–23%, elevated sedimentation rates as 81.8–95%, and C-reactive protein as 72–81.8%.^{Reference Saraçlar, Ertuğrul and Özme ve Ajun15,Reference Arora, Subramanyam, Khalilullah and Gupta34,Reference Olguntürk, Canter, Tunaoğlu and Kula53,Reference Ozer, Hallioğlu, Ozkutlu, Celiker, Alehan and Karagöz54}

Secondary prophylaxis limits the progression of long-term sequelae, and early diagnosis of acute rheumatic fever is essential for reducing morbidity and mortality. For the valve regurgitation detected during the initial diagnosis, valvar lesions completely improved for three (1.8%) patients in the first period, 14 (17.9%) patients in the second period, and 11 (15.5%) patients in the last period at the follow-up period (minimum 9 months and maximum 3.6 years). This was found statistically significant (p = 0.0002). Early diagnosis, prevention of recurrence with good secondary prophylaxis, regular follow-up and treatment of patients were effective in the improvement of valvar lesions. A total of 15 patients underwent surgical intervention (11 patients mitral valve replacement, 4 patients mitral valve replacement + aortic valve replacement), and 2 of them operated during acute attack because of uncontrolled heart failure and others at the follow-up. None of the patients in our population presented valve stenosis at the follow-up, because valve stenosis is considered to be a chronic and advanced-stage lesion.^{Reference Remenyi, ElGuindy, Smith, Yacoub and Holmes55} The mortality rate during the acute attack in developed countries was reported to be lower than 1%; however, higher rates were reported in developing countries.^{Reference Qurashi56} The mortality rate was reported to be 0.5% by Ürün et al. and 1% by Özer et al.^{Reference Örün, Ceylan and Bilici16,Reference Ozer, Hallioğlu, Ozkutlu, Celiker, Alehan and Karagöz54} There were no deaths during the acute period in the series of Narin et al.^{Reference Narin, Mutlu and Argun6} In the present study, the mortality rate was 1.2% in the first period and no deaths in the second and third periods.