Challenges of acute febrile illness diagnosis in a national infectious diseases center in Rio de Janeiro: 16-year experience of syndromic surveillance

Abstract

Introduction: Acute febrile illnesses (AFI) are a frequent chief complaint in outpatients. Because the capacity to investigate the causative pathogen of AFIs is limited in low- and middle-income countries, patient management may be suboptimal. Understanding the distribution of causes of AFI can improve patient outcomes. This study aims to describe the most common etiologies diagnosed over a 16-years period in a national reference center for tropical diseases in a large urban center in Rio de Janeiro, Brazil.

Methods: From August 2004-December 2019, 3591 patients > 12 years old, with AFI and/or rash were eligible. Complementary exams for etiological investigation were requested using syndromic classification as a decision guide. Results. Among the 3591 patients included, endemic arboviruses such as chikungunya (21%), dengue (15%) and zika (6%) were the most common laboratory-confirmed diagnosis, together with travel-related malaria (11%). Clinical presumptive diagnosis lacked sensitivity for emerging diseases such as zika (31%). Rickettsia disease and leptospirosis were rarely investigated and an infrequent finding when based purely on clinical features. Respiratory symptoms increased the odds for the diagnostic remaining inconclusive.

Conclusions: Numerous patients did not have a conclusive etiologic diagnosis. Since syndromic classification used for standardization of etiological investigation and presumptive clinical diagnosis had moderate accuracy, it is necessary to incorporate new diagnostic technologies to improve diagnostic accuracy and surveillance capacity.

Fig 1. Flowchart with patient enrollment and the acute febrile illness etiologies from August 2004 to December 2019.

1. Laboratory-confirmed dengue (521 patients), zika (227), chikungunya (759), yellow fever (25); co-infections: dengue/zika (1), dengue/chikungunya (12), zika/chikungunya (6), dengue/malaria (7) and chikungunya/malaria (11). 2. Lab-confirmed P. vivax (248), P. falciparum (97), P. malariae (3), P. ovale (5), mixed (P. vivax and P. falciparum) (2); co-infection malaria/dengue (7), co-infection malaria/chikungunya (11), not specified (5). 3. Lab-confirmed acute infection caused by Epstein-Barr Virus (EBV), cytomegalovirus (CMV), HIV or Toxoplasma gondii, resulting in acute infectious mononucleosis syndrome. 4. Includes common cold and acute bacterial rhinosinusitis. 5. Clinical or lab-confirmed bacterial pneumonia (15), bacterial pharyngitis/tonsillitis (11), secondary syphilis (9), urinary tract infection (7), tuberculosis (3), typhoid fever and sepsis (2), cholecystitis (3) and brucellosis (1). 6. Acute viral hepatitis (A and B) (11), parvovirus B19 (3), Varicella-Zoster infection (including chickenpox) (4), nonspecific viral meningitis (1).

Fig 2. Etiologic diagnosis of the Acute Febrile Illnesses (AFI) over time, between August 2004 and December 2019 (N = 3591).

Fig 3. Comparison of the frequency (%) of clinical and laboratory findings reported at the first visit, according to the result of the etiological investigation (N = 3360).

Co-infections and diagnosis with n ≤ 200 not included in this clinical description. Chi-square test showed significant differences among study groups: p<0.001 for all variables except for diarrhea (p = 0.003). * See S1 Table for the definition of anemia, leukopenia, lymphopenia and thrombocytopenia definitions used.

Fig 4. Clinical Features associated with the absence of a definitive diagnosis for the Acute Febrile Illness (AFI).