Identifying the Probable Etiology of Acute Undifferentiated Fever through Inflammatory Markers

INTRODUCTION

Acute undifferentiated fever (AUF) is defined as any febrile illness with a duration of ≤14 days without evidence of localized infection by history, physical examination, complete blood count, chemistry profile, urinalysis, or chest radiography at the time of initial presentation. ¹ The majority of outpatient services and a significant inpatient load in India are contributed by AUF. Unlike Western countries, tropical diseases such as dengue, malaria, and scrub typhus, pose a serious health challenge to the already fragile health system in our country. Most of these diseases are diagnosed by rapid diagnostic tests (RDTs), especially in the rural parts of our country where facilities for microbiological culture or high-end investigations such as polymerase chain reaction (PCR) are not available. It is a recognized fact that diagnosis by RDT is unreliable and has significant false negatives.

While the common etiologies of AUF include dengue, malaria, typhoid, and scrub typhus, COVID-19 has recently added to the list of existing woes. In our study, we aim to identify the probable etiology of AUF through inflammatory markers. Also, the study looks for the predictors of adverse outcomes by using both clinical and laboratory parameters. This not only results in better diagnosis but also prepares the physician for close monitoring and pooling of resources.

MATERIALS AND METHODS

RESULTS

A total of 400 consecutive patients with AUF during the period from 1 ^st July 2021 to 30 ^th June 2022 were included in the study.

DISCUSSION

In our study, 400 cases of AUF from the Armed Forces Hospital in New Delhi were included. The common etiologies in our study contributing to AUF were dengue (31.5%), COVID-19 (18.5%), enteric fever (12.7%), scrub typhus (9.0%), and malaria (6.0%). In 76 cases (19%), the fever was undiagnosed ( Fig. 1 ). A small percentage of cases (3.25%) were diagnosed with other diseases such as extrapulmonary tuberculosis (nine cases), leptospirosis (two cases), amoebic liver abscess, and chikungunya (one case each).

Fig. 1:

Serum ferritin, D-dimer, and CRP are pro-inflammatory markers. They were often studied to prognosticate and as markers of severity among acute febrile illnesses. In our study, we determined a CRP value of >30, 20–30, and 10–20 mg/L as highly elevated, moderately elevated, and mildly elevated, respectively. A D-dimer value of >750 ng/mL was considered highly elevated, 500–750 ng/mL as moderately elevated, and 200–500 ng/mL as mildly elevated. Similarly, a ferritin level of more than four times the ULN was considered highly elevated, and two to four times ULN and more than two times ULN were considered moderately and mildly elevated, respectively. Procalcitonin level of >5 ng/mL was considered highly elevated, 1–5 ng/mL as moderately elevated, and <1 ng/mL as mildly elevated.

In our study, CRP, D-dimer, and ferritin were found to be raised in all etiologies of AUF ( Table 1 ). When compared to other identified etiologies, enteric fever showed the highest quantum of rise in CRP, and scrub typhus showed the highest rise in ferritin and CRP. Both nonsevere dengue and COVID-19 showed the highest rise in D-dimer. However, CRP level was highly elevated in severe dengue. Similarly, among the identified etiologies of AUF, nonsevere dengue had the least rise in CRP, enteric fever had the least rise in ferritin, and scrub typhus had the least rise in D-dimer. A summary of the probable etiology of AUF based on the levels of inflammatory markers as per our study is given in Table 2 .

Our study supports the results of Vuong et al., where higher CRP levels above 34.0 mg/L were associated with severe dengue. ² In another study by Idhayu et al., a median CRP of 1.65 and 53 mg/dL was seen in uncomplicated dengue and typhoid illness, respectively. ³ Our results substantiate the results by Idhayu et al., where the mean CRP levels are 12.4 and 34.6 mg/L in dengue and typhoid illness, respectively. Similarly, Williams et al. found hyperferritinemia in scrub typhus, the results of which are consistent with our study. ⁴

Our study was correctly able to identify 90.8% of nonsevere dengue (109 out of 120), 87.8% of typhoid (43 out of 49), 83.6% of COVID-19 (56 out of 67), and 91.4% of scrub typhus (32 out of 35) patients based on the inflammatory markers level. The study is a novel attempt to identify the etiology of AUF based on inflammatory markers level, and there are no similar studies in the available literature to compare.

A total of 5.25% of the cases developed adverse outcomes during the course of illness ( n = 21). Of the 21 patients who had adverse outcomes, there were three fatalities (one dengue, one COVID-19, and one undiagnosed).

While studies showed varying data on the increased severity of dengue among male or female genders, our study did not show any gender predilection for the emergence of adverse outcomes. ^{5 , 6} Comparison of variables among different etiologies of AUF is summarized in Table 1 . The presence of diabetes was suggested to be a predictor of severe dengue and COVID-19 by Carrasco et al. and Kristan et al. ^{7 , 8} A similar observation is made in our study where obesity and diabetes (both types 1 and 2 diabetes mellitus) were found to have a significant association with emergence of adverse outcome in AUF. Among patients who had adverse outcomes, 50% had diabetes in dengue, and it was 7 and 75% in cases of COVID-19 and undiagnosed AUF, respectively. Similarly, among patients with adverse outcomes, obesity [body mass index (BMI) > 30 kg/m ² ] was seen in 66.7% of dengue, 85.7% of COVID-19, and 75% of undiagnosed AUF.

Hypertension (55.4%) and coronary artery disease (CAD) (12.4%) were found to be associated with severe COVID-19 by Sanyaolu et al. ⁹ Similarly, while 28.6% of COVID-19 patients with hypertension and another 28.6% with CAD had adverse outcome, only 11.9 and 13.4% had hypertension and CAD in COVID-19 patients without adverse outcome. Also, in our study, hypertension was found to be significantly associated with adverse outcomes in dengue.

However, while Zhao et al. suggested that chronic obstructive pulmonary disease (COPD) and smoking contribute to worse outcomes in COVID-19, only COPD but not smoking was significantly associated with adverse outcomes in COVID-19 patients. ¹⁰ Also, malignancy and chronic kidney disease (CKD) were seen in a higher proportion of AUF patients with adverse outcomes across various etiologies, similar to the study results of Sanyaolu et al. ⁹

Ledika et al. suggested that delay in admission (≥5 days of onset of fever) was significantly higher among the patients with severe dengue. ¹¹ Similarly, in our study, when the mean duration from fever onset to first hospital visit was 2.7 days in dengue patients without adverse outcomes, it was 6.1 days in dengue patients with adverse outcomes. Also, a significant delay in visiting the hospital after the onset of fever was found in other etiologies of AUF, which had adverse outcomes. Ledika et al. also proposed thrombocytopenia (<50,000) to be a predictor of severe dengue. However, the decreased number of platelets was due to the delayed presentation of patients as they are in a critical phase of illness when platelet levels are expected to be at their nadir.

Elevated liver enzymes were reported in dengue, scrub typhus, malaria, and enteric fever. Elevated levels of aspartate aminotransferase (AST) and alanine aminotransferase (ALT) were seen in 11 and 56%, respectively, during the 1st week of typhoid illness in a study by Morgenstern and Hayes. ¹² In our study, 10.2% of patients with enteric fever had elevated AST and ALT levels, majority of them tested in the first week of illness ( Table 1 ). However, in two patients with enteric fever who had adverse outcomes, both had elevated serum transaminases. Elevated AST levels were also a part of the scoring system suggested by Mitra et al. to distinguish dengue and scrub typhus. ¹³ In our study, elevated serum transaminases (AST and ALT) were seen in all etiologies of AUF having adverse outcomes, including undiagnosed AUF when compared to AUF without adverse outcomes. However, elevated AST levels were more conspicuous in typhoid and dengue when compared to other etiologies ( Table 1 ).

Results of the study by Mishra and Sorabjee showed a median value of serum procalcitonin in enteric fever to be 0.22 ng/mL. ¹⁴ Our study differs from the results of Mishra and Sorabjee, and the mean serum procalcitonin level in enteric fever is 3.21 ng/mL, significantly higher than that of the other etiologies of AUF. However, 50–100% of all cases of AUF with adverse outcomes had mildly elevated serum procalcitonin, though not manifold times of normal values as seen in bacterial sepsis.

CONCLUSION

Dengue, enteric fever, COVID-19, scrub typhus, and malaria are the common causes of AUF in urban India. A sizeable proportion of AUF remains undiagnosed due to various reasons, the most common being the nonavailability of a testing facility.

There are subtle differences in the levels of pro-inflammatory markers such as CRP, D-dimer, and ferritin among different AUF etiologies. Careful consideration of the same will help distinguish the possible etiology at an early stage when diagnosis is a challenge. Our study was correctly able to identify 90.8% of nonsevere dengue, 87.8% of typhoid, 83.6% of COVID-19, and 91.4% of scrub typhus patients based on the inflammatory markers level.

The presence of diabetes (both types 1 and 2 diabetes mellitus), obesity, COPD, CAD, and CKD is associated with the emergence of adverse outcomes in patients with AUF. Our study indicates that rising or elevated levels of pro-inflammatory markers in AUF indicate a likely adverse outcome. Resources and monitoring should be intensified in patients with AUF who have elevated CRP, D-dimer, and ferritin.