Utility of Hematological Parameters in Early Diagnosis of Neonatal Sepsis in Comparison to C-Reactive Protein

Introduction

Neonatal septicemia refers a systemic infection in the infants in the first 28 days of life.¹ It is often challenging to diagnose neonatal sepsis due to the subtle nature of presenting symptoms at various gestational ages.² Most important factor about neonatal sepsis is that most of the cases are preventable with appropriate antibiotic therapy and aggressive supportive care.

Neonatal sepsis is the most common cause of neonatal mortality and morbidity. It accounts for 26% of all neonatal deaths worldwide.³ In India, 11 to 24 neonates out of 1,000 neonates die due to neonatal sepsis.⁴ Malnutrition, lower socioeconomic status, unhygienic delivery conditions, poor health infrastructure, and traditional and cultural practices in community are some of the important causes of high mortality in developing countries.⁵ Neonatal sepsis can be classified into two categories based on onset of signs and symptoms of sepsis. When the signs and symptoms occur within 72 hours of life, it is termed as early onset neonatal sepsis (EONS). Whereas if the signs and symptoms occur after 72 hours of life, it is termed as late onset neonatal sepsis (LONS). In the Indian subcontinent, the distinction between EONS and LONS is blurred.⁶ Signs and symptoms of neonatal sepsis are often vague and subtle. Hence, it is challenging for the treating pediatrician to confirm the diagnosis and the appropriate line of management. Also, over-treating the neonate with empirical broad-spectrum antibiotics results in emergence of resistant strains of bacteria. It also has an effect on the neonatal health, as many of the antibiotics have their own adverse effects. Moreover, a child who is supposed to live his/her first few hours of life in his/her mother's arms will be separated from the mother and shifted to neonatal intensive care unit (NICU) for further medical management. Hence, it is necessary to recognize the onset of sepsis and treat appropriately. The gold standard for diagnosis of neonatal sepsis is isolation of the microorganism in blood culture. But the time that is required for the culture to be reported as positive is long enough for the infection to spread and cause fatal consequences. Also, the yield of positive culture is low.⁷⁸ Hence, an effective screening test is necessary to predict the presence of sepsis in the neonate. Sepsis screening in cases of suspected neonatal sepsis includes total leucocyte count (TLC), absolute neutrophil count, immature to total neutrophils ratio (I:T ratio), micro-erythrocyte sedimentation rate and C-reactive protein (CRP). A hematological scoring system (HSS) is designed to score the various hematological parameters and predict the presence of sepsis.⁹¹⁰

Materials and Methodology

We conducted a cross-sectional analytical study with a total of 100 cases, which included neonates admitted to NICU, Department of Paediatrics, Justice K S Hegde Medical Hospital, Mangalore, during the study period from January 2019 to June 2020. Neonates with signs and symptoms of sepsis and risk factor for developing sepsis were included in the study. Neonates with major life-threatening congenital anomalies were excluded from the study. The hematological parameters, CRP, and blood culture were sent as a part of sepsis screening for all the neonates admitted with suspicion of neonatal sepsis. The hematological parameters included TLC, absolute neutrophil count, immature neutrophil count (INC), I:T ratio, immature: mature neutrophil ratio, platelet count, and peripheral smear examination. Serum CRP estimation was done by quantitative nephelometry method. Blood culture was processed by BACTEC method. HSS was used to calculate the score and predict the presence of sepsis. The obtained data was then tabulated and analyzed using SPSS software version 22. The individual hematological parameters, HSS, and CRP were compared to the gold standard test that is blood culture.

Results

Among the 100 neonates studied, the males were predominant compared to females, with a male to female ratio of 1.6:1. Majority of the neonates presented with early onset neonatal sepsis (91%). Most of the neonates were delivered by normal vaginal delivery, were term, and were appropriate for gestational age. A wide range of clinical presentations were noted and are shown in Fig. 1 that included risk factor for neonatal sepsis as well as signs or symptoms of neonatal sepsis. In the study population, blood culture was positive in 17 neonates. The most common organism isolated in blood culture was Staphylococcus aureus. Klebsiella pneumoniae, Pantoea agglomerans, Pseudomonas species, Serratia marcescens and Streptococcus agalactiae were the other isolated organism (Fig. 2). The sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV) of the tests are tabulated in Table 1. Out of all the hematological parameters studied, I:T ratio had the highest sensitivity and NPV, followed by HSS. In addition, we noted that INC values were significantly correlated with the presence of neonatal sepsis. The presence of schistocytes, nucleated red blood cell (nRBC), and shift to left in neutrophils (leucoerythroblastic blood picture) can be seen in Figs. 3 and 4 that is the common finding seen in neonatal sepsis.

Close

Fig. 1 Distribution of clinical presentation/risk factors in the study population.

Export to PPT

Close

Fig. 2 Distribution of microorganisms isolated on blood culture.

Export to PPT

Close

Fig. 3 Peripheral smear showing hemolysis suggested by schistocytes (arrows) and nucleated red blood cells.

Export to PPT

Close

Fig. 4 Peripheral smear showing leucoerythroblastic blood picture.

Export to PPT

Discussion

In our study, we found that among the various hematological parameters analyzed, I:T ratio was a better predictor of neonatal sepsis. Similar results were noted in other studies conducted by Nayana and Sreenivas,¹¹ Nayak et al,¹² and Makadia and Shah¹³ (Table 2).

However, no single test was sufficient to diagnose neonatal sepsis. We also analyzed diagnostic utility of other parameters CRP, TLC, and HSS in neonatal sepsis.

In this study, CRP had a sensitivity of 47.1%, specificity of 75.9%, NPV of 87.5%, and a PPV of 28.6%. These results are comparable with Nayak et al. The discordance on comparing with study performed by Makadia and Shah¹³, Jeyaganguli et al,¹⁴ and Saboohi et al¹⁵ may be due to variation in selected study population.

In this study, TLC had a sensitivity of 35.3%, specificity of 77.1%, NPV of 85.3%, and a PPV of 24%. These results are comparable with studies performed by Nayak et al¹² and Makadia and Shah.¹³

In this study, HSS had a sensitivity of 64.7%, specificity of 15.7%, NPV of 68.4%, and a PPV of 13.6%. The obtained sensitivity and NPV of HSS are comparable with similar other studies performed on neonatal sepsis by Nayak et al¹² and Makadia and Shah.¹³

In this study, I:T ratio had a sensitivity of 94.1%, specificity of 46%, NPV of 83.3%, and a PPV of 17%. These results are comparable with studies performed by Nayak et al¹² and Sundarapandian et al.¹⁶

In addition, a few other studies showed that nRBC count is also increased in neonatal sepsis.¹²¹⁷ nRBC count in a term neonate is said to increase in hypoxia, severe anemia, acute stress, and sepsis. In neonatal sepsis, nRBC count is increased due to release of cytokines and is independent of hypoxia.¹⁸¹⁹

Conclusion

HSS, I:T ratio, and CRP are a good tool to diagnose neonatal sepsis at early stages and have a faster turnaround time than the gold standard test, blood culture. However, combination of above parameters would give better predictive value for neonatal sepsis. The study also found significant correlation of INC with blood culture. Therefore, further larger multicentered prospective studies are required for validation of INC as a predictor of diagnosis of neonatal sepsis.