Haematologic profile and predictors of outcome in paediatric scrub typhus: A prospective observational study

INTRODUCTION

Scrub typhus (ST), caused by Orientia tsutsugamushi, is a re-emerging infectious disease in the Asia Pacific regions, with approximately 1 million newly diagnosed cases annually.^[1] ST is transmitted to humans by the bite of the infected chiggers which are larval forms of trombiculid mites and this inoculation site later turns into a necrotic eschar. This acute febrile illness, presenting with varying severity during outbreaks, has now turned into an important cause of hospitalisation amongst children in rural India in the post-monsoon period.^[2] This infection can be complicated with hepatitis, shock, pneumonia, acute kidney injury and meningoencephalitis requiring intensive care admissions.^[3,4] ST can turn lethal if left untreated, with a mortality rate of 6% and as high as 24% with multi-organ dysfunction.^[5]

This tropical infectious disease is truly neglected with underdiagnosis and underreporting due to inadequate awareness, dearth of clinical suspicion and non-availability of specific diagnostic tests like serology in rural areas, where the disease is endemic.^[6] The suspicion of ST in early disease (<7 days) is of paramount importance to initiate appropriate empirical antibiotics and prevent severe forms of it.^[3,5] Serology, which is the most commonly done diagnostic test, typically requires 5–7 days to yield positive results. Thus, a simple laboratory test is required at the primary care setting to suspect ST at first contact and to predict severe forms to plan early referral to a higher centre during outbreaks.^[7]

The hepatic and haematological abnormalities were the most consistent, and early findings support the diagnosis of ST.^[8] Most of the haematological studies from the Indian subcontinent are on the adult population or on a small cohort of the paediatric population.^[9,10] Hence, this study aimed to describe the frequency of the abnormal haematology parameters and examine their relation with the outcome measures in a large prospective paediatric ST cohort.

MATERIAL AND METHODS

This prospective observational cohort study was conducted in the department of paediatrics in a tertiary care teaching hospital in southern India for 4 years. The study was conducted after obtaining approval from the institutional ethical committee. The cohort included all children <12 years of age who had a fever for more than 5 days with confirmed ST infection. The diagnosis of ST was confirmed by immunoglobulin M (IgM) enzyme-linked immunosorbent assay (ELISA) (InBios International, Inc., USA) serology (in a sample taken after day 7 of fever) when the optical density was more than 0.5. Other tropical infections such as dengue fever, typhoid fever and malaria were excluded using dengue serology (non-structural Protein 1 (NS1) and IgM), blood culture and smear for malarial parasites. The demographic details of all enrolled patients were documented. Clinical features such as duration of fever, signs such as lymphadenopathy, presence/absence of eschar and system involvement were recorded using detailed questioning and clinical examination.

A detailed haematology workup was done on samples taken on the day of admission, which included haemoglobin level, total leucocyte count, differential count, absolute counts and platelet count. These parameters were analysed using 2 mL of ethylenediaminetetraacetic acid anticoagulated blood in a fully automated cell counter (Sysmex xn-500).

The patients included in the study were monitored during their entire hospital stay for outcome measures such as fever clearance time (FCT), occurrence of severity and duration of hospital stay. FCT is the time between the first dose of antibiotics and when body temperature drops to ≤37.5°C and remains the same for at least 24 h without the influence of antipyretics. The subjects included in the study were classified into ‘Severe’ and ‘Non-severe’ groups based on the previously published definitions of severe ST.^[3] The children were treated as per standard treatment guidelines.^[11]

Demographic details, clinical features, haematological investigations and outcomes were computed on Excel spreadsheets. Statistical analysis was done using the Statistical Package for the Social Sciences (SPSS) software, version 23.0 (SPSS Inc., Chicago, Ill., USA). The continuous variables are presented as means with standard deviations, and the categorical variables are presented as percentages. An appropriate statistical test was used to determine the association between haematological and clinical parameters. The ‘t-test’ and Chi-square test were used, and their results were expressed as mean difference with 95% confidence interval (CI), P-value and odds ratio (OR) with 95% CI, and P-value, respectively. The P < 0.05 was considered statistically significant.

RESULTS

A total of 634 ELISA serology-proven cases of ST were included in the study. Infants ≤1 year were 58 (9.2%), children between 2 and 5 years were 224 (35.3%), and those between 6 and 12 years were 352 (55.5%). We noted that in our study, there was a male preponderance, with 337 boys and 297 girls, with a male-to-female ratio of 1.13:1.

All the patients presented with fever as a common symptom, but the duration varied. The fever lasted for 5–7 days in 304 (47.9%) children, 8–14 days in 273 (43.1%) and more than 15 days in 57 (9.0%). Hence, the early cases were 47.9% and late cases (fever lasting for more than 7 days) accounted for 52.1%. The most frequent sign was eschar, which was seen in 316 (49.8%) and generalised lymphadenopathy, which was noted in 293 (46.2%) children with ST. Gastrointestinal system involvement was seen in the majority (335/634, 52.8%) of ST cases, followed by involvement of respiratory (241/634, 38%), central nervous (30/634, 4.7%) and cardiovascular (28/634, 4.4%) systems.

In the study, the FCT was ≤48 h in 243 (38.3%) and delayed fever clearance of FCT >48 h in 391 (61.7%) children. Features of severe ST were found in 191 (30.1%) children. The duration of hospitalisation was ≤7 days in 427 (67.4%) cases and >7 days in 207 (56.9%) children.

The haematological profile of the study population is shown in Tables 1 and 2. The clinical parameters and their association with haematologic parameters were studied [Tables 3 and 4].

The neutrophil/lymphocyte (N/L) ratio was calculated and found to be ≤2 and >2 amongst 481 and 153 children with ST, respectively. Of whom 136 (28.3% of 481) and 55 (35% of 153) children had features of severe ST, 283 (58.8% of 481) and 108 (70.5% of 153) children had delayed fever defervescence (i.e., FCT >48 h).

The Chi-square test was used to find the association of the N/L ratio with severity and FCT. This analysis of the N/L ratio association with severe ST revealed an OR of 1.423, with a 95% CI of 0.969–2.093 and a significant P = 0.036. Furthermore, the N/L ratio with FCT demonstrated a stronger association, with an OR of 1.679, a 95% CI of 1.134–2.486 and a significant P = 0.004.

DISCUSSION

The organism O. tsutsugamushi primarily affects the endothelial cells (EC) during its systemic infection, resulting in widespread endothelitis. This endothelial injury is a consequence of EC activation, leucocyte infiltration and inflammatory cytokine production triggered by this pathogen. In some cases, the excessive activation of EC leads to an uncontrolled influx of neutrophils and monocytes, causing microvascular leakage and vascular compromise. Such compromise in end organs such as the lungs, heart, brain and kidneys leads to severe ST.^[12]

The study shows that anaemia, monocytopenia, eosinopenia, thrombocytopenia and leucocytosis (neutrophilic preponderance) were the commonest haematological abnormalities found in children with ST. Anaemia, leucocytosis and thrombocytopenia are also prevalent in other paediatric ST studies and have been found to differentiate ST from other causes of tropical fever.^[2,13-15] Microvascular leaks, acute haemolysis and complications such as haemophagocytic lymphohistiocytosis (HLH) were plausible explanations for anaemia in ST. While platelet aggregation was the commonest cause for thrombocytopenia apart from HLH.^[16] Severe anaemia and severe thrombocytopenia were found in 20 (3.1) and 11 (1.7) children, respectively, but neither of them presented with bleeding tendencies nor required platelet transfusion. However, in our study, 3 (out of 20) critically ill children required packed red blood cells transfusion for severe anaemia.

Our study showed that with increased duration of untreated fever in children, the severity of anaemia and neutrophilic leucocytosis increases. The unchecked multiplication of the bacterium, dissemination within the host, brisk inflammatory response by the host and prolonged cytokine activation could be the possible explanations for this finding.^[17] With increasing duration of fever, the occurrence of severe ST increases and these haematological markers have been shown to predict severe ST, particularly in children with central nervous system involvement.^[18,19]

Our study showed that anaemia and neutrophilic leucocytosis were associated with children who lack the diagnostic cutaneous sign of ST, the eschar. The occurrence of eschar is variable (30–60%) in children, and the absence of this vital clue is associated with delayed initiation of therapy.^[20] Thus, anaemia and neutrophilic leucocytosis can be considered as a haematological marker in resource-limited settings to start empirical therapy in febrile children without eschar, particularly hailing from endemic areas with end-organ damage.

In the present study, anaemia and neutrophilic leucocytosis were associated with the delayed defervescence of fever (>48 h) and prolonged duration of hospital stay. The occurrence of severe forms of ST in these children could be the possible reason for the delay. Similar delayed fever defervescence was found in other paediatric studies.^[19] Thus, the treating physician should expect a delay in fever defervescence if these haematological features are present and should also be vigilant about the mimickers or co-infections like malaria. Few paediatric studies have documented thrombocytopenia as a haematological marker of severe ST and delayed fever defervescence; our study has no association.^[21,22]

The study analysed the N/L ratio in ST children and found that a value of >2 is associated with the occurrence of severity and delayed fever defervescence. Thus, the N/L ratio can serve as an adjuvant marker of ST severity and delayed fever defervescence. Only two studies have studied the N/L ratio in ST and shown that it is useful as a predictor of ST early and in children with encephalitis Syndrome.^[23,24] Thus, this ratio can be used to guide empirical therapy in suspected ST in resource-limited settings. There was no mortality reported in our study.

The large paediatric cohort of ST and confirming ST using serology were the strengths of the study. While this is a hospital-based study and may not represent the community, it is our limitation.

CONCLUSION

Anaemia, leucocytosis and thrombocytopenia can be used as haematology markers to predict ST in children presenting with undifferentiated febrile illness from rural endemic areas. These can also serve as early markers to predict delayed fever defervescence and severity. An N/L ratio >2 should prompt the treating physician about impending severity. Thus, the primary physician can consider the abnormal haematology parameters in suspecting ST for empirical treatment, expecting delayed fever defervescence and predicting ST in the paediatric age group.