ETIOLOGY AND CLINICAL PROFILE OF FEVER OF UNKNOWN ORIGIN IN CHILDREN-A PROSPECTIVE OBSERVATIONAL STUDY IN PEDIATRICS DEPARTMENT OF DMCH, LAHERIASARAI, BIHAR

Background: Fever of unknown origin (FUO) is an important cause of morbidity and mortality in children, especially in tropical countries with varied aetiology and clinical presentation. Aim of this study is to determine the aetiology and outcome of FUO in Indian children. Study design is Prospective, observational study.
Methods: We enrolled 75 children aged 1 to 12 years who were admitted with fever >8 days to the pediatric department of DMCH, Laheriasarai, Bihar from November 2019 to June 2020. Initial evaluation included complete blood count, peripheral smear, urine analysis, chest radiography, blood culture and tuberculin test.
Results: In 72 children (96%), a definitive diagnosis could be established, whereas 3 children (4%) remained undiagnosed. Most common aetiology of FUO was infectious disease (90.6%) followed by malignancy (4%) and collagen vascular disease (1.3%). Among the infections group, scrub typhus was found to be the commonest aetiology (52%).
Conclusions: Infections were the most common etiological factor for FUO in children aged 1-12 years of our region.


INTRODUCTION
Fever is the chief complaint in up to one third of patients presenting to pediatricians. FUO represents a prolonged febrile illness without an established etiology despite thorough evaluation. A consensus denition for FUO is lacking in children as different studies have used different denitions, thus preventing meaningful comparisons and rm conclusions regarding the etiology of FUO. The etiology of FUO in a given population varies with geographical and socio-economic factors, age distribution, sex difference, time of study, local disease patterns and epidemiology. It is difcult to compare studies of FUO in developed countries with those from resource-poor settings where there is limited availability of expensive serological and other tests.

METHODS
This prospective observational study was conducted in Department of Pediatrics, Darbhanga Medical College & Hospital, Laheriasarai, Bihar from November 2019 and June 2020. Informed consent was taken from the parents prior to inclusion of individual into the study.
The primary objective was to determine the etiology and outcome of FUO. The secondary objectives were to correlate the clinical prole with the etiological factor and to ascertain whether the etiology can be identied in all cases with the available laboratory work up.

Study Definitions
FUO was dened as fever of at least 8 days duration in children with no apparent diagnosis with the initial workup in the hospital or as an outpatient. This is the most accepted clinical denition of FUO in children.

Inclusion And Exclusion Criteria
All Children of 1-12 years of age, presenting to the pediatric emergency or outpatient department with fever more than or equal to 8 days duration were included. Children with diagnosed immuno compromised states and children on treatment with immuno suppressive medications were excluded.

Procedure
A detailed history was taken from the patients, including the history of travel. After history, detailed general and systemic examinations were performed, which was repeated on daily basis to look for any evidence of underlying etiology. Clinical and biochemical data were recorded on a structured proforma. Initial investigations included complete blood count with erythrocyte sedimentation rate, peripheral blood smear, complete urine analysis, routine biochemistry, chest radiography and urine/stool/blood culture. Further diagnostic tests include, serological tests for bacteria and virus, tuberculin skin test, immunological markers like ANA, C3, C4, C-ANCA, HIV, abdominal/lymph node USG, Contrast CT or MRI, echocardiography etc. was done as and when needed. They were based on the results of initial tests or clue from daily examination. More invasive investigation including lumbar puncture, pleural, ascitic uid analysis, lymph node aspiration or excision, bone marrow aspiration, biopsy and culture or liver biopsy were done if indicated. Children were classied into four categories of FUO, (5) viz. infections, malignancy, autoimmune and undiagnosed. Statistical analysis was carried out using SPSS version 19.0 (IBM SPSS, US) software Chi-square test of signicance was used to assess the correlation between pairs of categorical variables. A p-value <0.05 was considered as signicant.

KEYWORDS
Infections caused by Rickettsial pathogens dominated the group with (52%) followed by bacterial (38%), viral (6%) and protozoan (4%) pathogens. The most common cause was scrub typhus which was seen in 35 of 68 (52%) patients in the infectious group. Enteric fever (13%), pneumonia (7%) and urinary tract infection (7%) were the predominant bacterial infections seen in this study. Malaria was reported in 3 (4%) cases based on Rapid malarial card test (LDH based card tests) and thin & thick smear analysis by a trained microbiologist (2 cases of plasmodium vivax and 1 case of plasmodium falciparum). Our region is not endemic for malaria and not surprisingly all the 3 children with malaria had a history of recent travel to endemic areas. Hepatitis A infection was the commonest viral infection presenting as FUO (6 % of cases). Only 2 children had tuberculous lymphadenitis.
One of the malignant diseases causing FUO, acute leukemia (2 children) was the most common followed by Neuroblastoma (1 child). One child was diagnosed to have systemic lupus erythematosus based on presence of both anti-dsDNA and anti-nuclear antibodies.   Table 4 illustrates the diagnostic yield of various tests performed in the study. The clinical suspicion of Tuberculosis remains high thereby resulting in chest X ray and Mantoux being done for most of the children with prolonged fever. But as is evident from the study, the diagnostic yield is low.
All the patients, were diagnosed, treated and discharged successfully with no mortality. This is the rst attempt for a reasonably large prospective study of FUO in Indian children using the recent denition of FUO i.e. fever of more than or equal to 8 days duration. We have been able to establish a specic etiological diagnosis for FUO in 96% of cases.
Scrub typhus has been reported as an emerging infectious disease for the last 5 decades in this country. In recent years, the high prevalence of Scrub typhus has also been noticed in South India. Our study also supports this fact. It is necessary to evaluate the epidemiology and clinical prole of Scrub typhus in large population based studies.
The traditional way of clinical diagnosis of scrub typhus was by detection of eschar and rashes with a history of outdoor activity. In the present study, only 16 children (45%) of the cases with scrub typhus presented with eschar. This is comparable with studies done in South Vietnam and Taiwan, where eschar was documented in 46%, and 60% of the patients respectively. In most of the Indian studies, eschar was reported only in very few patients and this may be because usually, the eschar goes unnoticed in dark skinned individuals. The Gold standard test for diagnosis of scrub typhus is immune uorescence antibody test (IFA) but IFA test is technically very demanding, subjective, needs an experienced observer and not commonly available in India. ELISA on the contrary is simple, easy interpret, can be performed in laboratories and fairly comparable with IFA in terms of sensitivity and specicity.
In this study, 29 children were diagnosed both on clinical and serological studies based on IgM ELISA; and 6 were diagnosed based on the presence of eschar and other clinical symptoms and signs like myalgia, lymphadenopathy and hepatomegaly. In all children with Scrub typhus, an excellent and immediate response (dereference of fever within 48 hours) was observed to specic antibiotic treatment with either Azithromycin or Doxycycline. Among the 35 children with scrub typhus as a cause of FUO, only four had complications-three in the form of scrub typhus meningo-encephalitis and one had migratory polyarthritis. All these children had these complications at presentation. None of them deteriorated after treatment or during hospitalization. Thus, it underlines the fact that early diagnosis and timely antimicrobial therapy prevents complications in children with scrub typhus.
In this study, the second most common cause of FUO was Enteric fever accounting for 6.7% of cases. This is similar to many other studies. The main reason for increased incidence of enteric fever in India is mainly due to inadequate sanitation and poor hygiene. Most cases responded to intravenous Ceftriaxone therapy, which is the rst line drug for Enteric fever. In spite of widespread vaccination campaigns against Typhoid, still, there is a rising trend in the incidence of Enteric fever mainly because of poor town planning, population expansion and poor vaccination rates.
In the undiagnosed group (4% of FUO), one child exhibited a clinical course similar to a viral illness and eventually resolved spontaneously. Though two cases had a course similar to enteric fever their diagnosis could not be proven. The fever resolved with an empirical therapy with third generation cephalosporin. Several series have reected this trend with undiagnosed groups contributing 10-30% of all cases in children.
A majority of the cases (85%) were diagnosed by serology, blood and urine culture. The tests that have been performed as a routine in all patients had a very poor diagnostic yield in contrast to tests that have been done based on some clinical or laboratory indication. There was a lesser yield for the blood culture noticed in this study. One probable reason for a relatively lesser yield of blood culture may be the prior initiation of antibiotics in the primary care centers before coming to tertiary care centre.
A primary concern of the pediatricians in evaluating a child with FUO should be limited use of investigations that are case-specic especially in developing countries such as India where availability and affordability are the limiting factors. Hence, formulation of regional guidelines for evaluation of children with FUO would be useful in this regard. Data from our study would be useful to formulate such regional guidelines and management strategies for children presenting with fever of >8 days. Our study has some limitations. The study was a hospital based study. The sample may not be representative of the entire community. For children in whom a denitive diagnosis could not be reached, work up for viral etiologies could not be done due to economical constraints (especially EBV).

CONCLUSION
To conclude, infections have been found to be the most common etiological factor for FUO in children. Scrub (Rickettsial) typhus was the leading cause of the infections diagnosed. Initiation of relevant investigations and targeted antimicrobial treatment in the management of fever and familiarizing primary care physicians about the common causes of FUO are the key elements in the diagnosis and management of FUO in developing countries Further studies done at multiple centers and community level would be necessary to see if the results of the present study are replicable. If similar data is found, regional guidelines can be formulated for workup and management of a child with FUO >8 days.