Use of Scatterplot Patterns Derived from Automated Haematological Analysers in the Diagnosis of Acute Febrile Illnesses EC06-EC10
Dr. Debdatta Basu,
Professor, (Senior Scale), Department of Pathology, JIPMER, Puducherry-605006, India.
Introduction: Acute Febrile Illness (AFI) is common in the tropics and caused by diverse pathogens. Dengue, malaria, enteric fever and scrub typhus are few common causes of AFI in India. Peripheral blood smear and complete blood counts are important investigations carried out in the investigation of acute febrile illness. Cell abnormalities in AFI produce unique scatterplot diagrams which may aid in diagnosis.
Aim: To differentiate the scatter plot patterns of normal blood samples from those in AFI and to assess whether the abnormal scatter plot patterns (if they exist in AFI) can be used for predicting the common acute febrile illness.
Materials and Methods: The descriptive study was conducted in JIPMER for a period of two months on EDTA blood samples of patients reported for complete blood count and peripheral blood smear examination in the haematology laboratory. Consecutive blood samples of patients with acute febrile illness were collected for a two month period and analysed for scatterplot patterns. Controls included samples with normal counts, scatterplots and no history of fever. A total of 323 samples of AFI and 100 control samples were assessed. AFI and control samples were processed in SysmexXT2000i haematological analyser and simultaneous serological and culture investigation was carried out to detect the nature of AFI. Sensitivity and specificity were calculated based on differences between AFI and control blood samples.
Results: Out of 323 AFI samples, 158 were bacterial, 2 were of enteric fever, 18 were fungal, 17 were viral, 9 were mycobacterial, 5 were malarial, 1 was helminth, 1 was rickettsial while 139 were serology and culture negative. Negative y-axis monocyte deviation was a common sensitive finding in bacterial (sensitivity-0.87), fungal (sensitivity-0.88) and viral aetiology (sensitivity-0.82) while neutrophilic leukocytosis was sensitive in bacterial (sensitivity-0.79), fungal (sensitivity-0.78) and mycobacterial aetiology (sensitivity-0.78). Lymphopenia was also found in fungi (sensitivity-0.78). Malaria showed splitting of eosinophils (specificity-0.98), RBC granularity increase (specificity-0.99) and increased neutrophil-eosinophil merge (specificity-0.83). Other common findings included monocytosis, left shift of neutrophils and presence of blue areas in RBC region.
Conclusion: Scatterplot patterns obtained from a complete blood count can give a clue in the diagnosis of the aetiologies of acute febrile illnesses.