Home Print this page Email this page Small font sizeDefault font sizeIncrease font size
Users Online: 364
About us Editorial board Search Ahead of print Current issue Archives Submit article Instructions Subscribe Contacts Advertise Login 

 Table of Contents  
Year : 2016  |  Volume : 5  |  Issue : 3  |  Page : 193-198

Correlation of blood culture results with the sepsis score and sepsis screen in the diagnosis of early-onset neonatal septicemia

Department of Pediatrics, Government Medical College, Bhavnagar, Gujarat, India

Date of Web Publication28-Sep-2016

Correspondence Address:
Dr. Rajeshri Rajendra Mehta
Room No. 801, PG-3 Building, Dhanvantari Hostel, Sir T. Hospital, Bhavnagar, Gujarat
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/2249-4847.191263

Rights and Permissions

Background and Objectives: The objective was to study the correlation of blood culture with the sepsis screen (based on six laboratory parameters) and Takkar and Bhakoo sepsis score (based on six perinatal risk factors) in the diagnosis of early-onset neonatal septicemia. Materials and Methods: In this prospective study, blood samples from 100 clinically suspected early-onset neonatal septicemia cases were selected randomly during the primary observer's study period, i.e., from January 2013 to August 2013, and subjected to blood culture and sepsis screen tests such as C-reactive protein (CRP), micro erythrocyte sedimentation rate (m-ESR), white blood cell count, absolute neutrophil count, I/T ratio, and platelet count. The culture results were correlated with the sepsis score and the sepsis screen tests. Results: Of the 100 cases studied, 63 were blood culture-positive. Higher proportions of septic babies were male (63.4%), preterm (65.1%), low birth weight (76.2%), inborn (52.4%), and those delivered spontaneously (80.9%). Gram-negative organisms (84.13%) were predominant such as Klebsiella (54%), followed by Pseudomonas (15.9%) and Escherichia coli (11.1%). Sepsis screen laboratory parameters such as CRP and m-ESR were significantly associated with culture-proven sepsis. CRP followed by thrombocytopenia and I/T ratio had more sensitivity and high negative predictive value. Leukopenia followed by m-ESR and neutropenia had high specificity and positive predictive value. On applying sepsis risk score, 81.3% were found to be culture-positive in the high-risk category; however, 49% of those with low sepsis risk score were also culture-positive. Conclusion: High-risk newborns identified on the basis of perinatal history by Takkar and Bhakoo sepsis score should be started on antibiotic treatment early. However, the present study illustrates less reliability of this score in low- and moderate-risk categories. A negative septic laboratory screen, in addition to low or moderate sepsis risk score, would be more justifiable in withholding antibiotics in such patients. Blood culture remains the definitive deciding factor.

Keywords: Blood culture, early-onset neonatal septicemia, sepsis score, sepsis screen

How to cite this article:
Hassan HR, Gohil JR, Desai R, Mehta RR, Chaudhary VP. Correlation of blood culture results with the sepsis score and sepsis screen in the diagnosis of early-onset neonatal septicemia. J Clin Neonatol 2016;5:193-8

How to cite this URL:
Hassan HR, Gohil JR, Desai R, Mehta RR, Chaudhary VP. Correlation of blood culture results with the sepsis score and sepsis screen in the diagnosis of early-onset neonatal septicemia. J Clin Neonatol [serial online] 2016 [cited 2022 May 19];5:193-8. Available from: https://www.jcnonweb.com/text.asp?2016/5/3/193/191263

  Introduction Top

Neonatal septicemia refers to a clinical syndrome characterized by systemic signs and symptoms due to generalized bacterial infection with a positive blood culture in the first 4 weeks of life.[1] In developing countries, sepsis is the most common cause of mortality responsible for 30%-50% of the 5 million total neonatal deaths each year.[2] The National Neonatal Perinatal Database (NNPD, 2002-2003) from India has reported an incidence of 23/1000 live births for septicemia.[3] Gram-negative organisms are more commonly reported from India.[4] Fulminant and fatal course of infection mandates early diagnosis for timely treatment.[1],[5],[6]

The clinical presentation is often subtle or nonspecific and usually mimicked by several other disorders. A high index of suspicion is required for the early diagnosis of neonatal septicemia. As without treatment, the case fatality rate is high, certain perinatal risk factors have been evaluated as indicators to predict neonatal septicemia to facilitate management. The scoring system by Takkar et al.,[7] consisting of six perinatal risk factors, is one such tool commonly used by clinicians to screen and treat the neonates for septicemia. It uses perinatal history to classify newborns into three risk categories and then gives recommendations for their management. High-risk groups of newborns are identified on the basis of history alone and treatment initiated early while moderate- to low-risk group can be monitored and investigated as required.

The gold standard for the diagnosis of neonatal septicemia is a positive blood culture.[1],[8] Definitive culture results take at least 48-72 h, resulting in treatment delays. Hence, certain rapid diagnostic tests such as C-reactive protein (CRP), micro erythrocyte sedimentation rate (m-ESR), buffy coat smear examination, total white blood cell (WBC) count, absolute neutrophil count (ANC), immature/total neutrophil count ratio, and platelet (PL) count collectively termed as the "Sepsis Screen"[9],[10] can be used.

The present study was carried out to isolate the organisms responsible for neonatal septicemia from blood and correlate the blood culture results with the sepsis score and the sepsis screen tests for the early diagnosis of neonatal septicemia.

  Materials and Methods Top

This prospective study was conducted at a teaching tertiary care hospital. Informed consent was taken from the parents/guardians of all patients. Detailed history and clinical findings were recorded in the pro forma. All newborn babies aged 0-72 h presenting with one or more perinatal risk factors such as prematurity, low birth weight (LBW), birth asphyxia, foul-smelling liquor, unclean per vaginal examination before delivery, prolonged rupture of membranes, and prolonged labor were included in the present study. Neonates with clinical features suggestive of septicemia and already receiving antibiotics were excluded from the study.

A score was assigned to each of the risk factors based on the perinatal infection risk scoring of Takkar and Bhakoo [Table 1] and categorized into three risk groups accordingly [Table 2].
Table 1: Takkar and Bhakoo perinatal infection risk score

Click here to view
Table 2: Takkar and Bhakoo risk score and risk group category with suggested intervention

Click here to view

All neonates included in the study were also screened using CRP and various other hematological parameters with predetermined cutoff value according to sepsis screen, and at the same time, blood culture was sent.

Sepsis screen

  1. CRP: >1 mg/dl
  2. m-ESR: >15 mm in the 1st h
  3. Total leukocyte count (TLC) (leukopenia): <5000 cells/cumm
  4. ANC (neutropenia): <1800 cells/cumm
  5. Band cell count to total neutrophil count ratio (I/T ratio): >0.2
  6. PL count (thrombocytopenia): <1.5 lakhs/cumm.

Sepsis screen was considered positive when two or more parameters were positive. If the initial screen was negative, it was repeated after 12 h if there was a strong clinical suspicion of infection.

Sample collection

About 4 ml of blood was drawn using an intravenous cannula, of which 1 ml of the blood sample was inoculated aseptically into a culture bottle, 1 ml of the blood was allowed to clot in a sterile bottle to collect serum for estimation of CRP, 1 ml was collected in a sterile bottle containing the anticoagulant ethylenediaminetetraacetic acid (EDTA) for estimation of the total WBC count, ANC, band cell count (I/T ratio), and 1 ml for m-ESR.

A volume of 4 ml blood was required for conducting all the tests in the newborn at a time as microassay tests and vaccuates were not available at our setup.

Blood culture

About 1 ml of blood was drawn aseptically and inoculated into a blood culture bottle containing 10 ml of Hartley's Infusion Broth, thus making a dilution of 1 in 10 to nullify the natural bacteriostatic/bactericidal activity of blood. The broth was distributed into 10 ml quantity in McCartney bottles and sterilized by autoclaving at 121°C for 15 min. After inoculation, the blood culture bottles were incubated at 37°C under aerobic conditions in the incubator for 7 days. The first subculture was done after 24 h of incubation, the second on the 3rd day, and the final on the 7th day. The inoculated plates were incubated aerobically in the incubator at 37°C for 24 h, and the plates were observed for growth. The growth was identified by colony characteristics, Gram's stain, and standard biochemical tests described in Mackie and McCartney and Practical Medical Microbiology which did not yield any growth, following three subcultures were reported negative at the end of 7 days.

C-reactive protein assay

This test is done using the diagnostic kit for in vitro detection of CRP in human serum by the rapid slide latex agglutination qualitative method. The sensitivity of the antigen in the kit for visible agglutination is 0.6 mg/dl. In our study, a CRP value of >1 mg/dl was taken as a positive test.

Other hematological tests

A drop of EDTA blood was taken on a clean dry slide, and a thin tongue-shaped smear was made, air dried, and stained with Leishman's stain. The TLC, differential count, ANC, band cell count, I/T ratio, and PL count were calculated as per standard hematological method.

All the findings were recorded and comparisons were drawn between blood culture results, sepsis score, and the sepsis screen tests.

Data were analyzed using the SPSS software for Windows version 11 (Statistical Presentation System Software, SPSS Inc., 1999, New York, USA), and categorical tables, Chi-square values, probability coefficients, sensitivity, specificity, positive predictive values, and negative predictive values of the three diagnostic methods were derived and the results were correlated. Conclusions were drawn from the tabulated results.

  Results Top

Of the 100 cases studied, 63 cases yielded a positive blood culture and 37 cases yielded a negative blood culture. 62 were male and 38 were female with the ratio of 1.63:1. Among culture-positive cases, 40 (63.4%) were male and 23 (36.6%) were female. Culture-proven septicemia was more common among 41 (65.1%) of preterm neonates, 48 (76.2%) of LBW neonates, 51 (80.9%) neonates with spontaneous vaginal delivery, 38 (60.3%) in primigravida, and 33 (52.4%) of the hospital inborn neonates.

[Table 3] and [Table 4] show a higher proportion of babies had an unclean vaginal examination, preterm, birth asphyxia, and LBW as a risk factor, and they were significantly associated with culture-proven sepsis. The majority of babies with culture-proven sepsis had more than two factors.
Table 3: Perinatal risk factors among neonates with early-onset sepsis

Click here to view
Table 4: Correlation of perinatal risk group with the blood culture status

Click here to view

As per septic score, those falling in high- and moderate-risk categories had a high number of culture-positive patients, 81.3% and 74.9%, respectively. However, even 49% of the low score patients had a positive culture report [Table 4].

[Table 5] shows CRP; thrombocytopenia and I/T ratio were positive in a higher proportion of culture-positive cases. Neutropenia and leukopenia were in higher proportion in culture-proven cases compared to culture-negative cases. CRP, m-ESR, and two or more tests positive were statistically significant with respect to culture-proven sepsis.
Table 5: Correlation of sepsis screen parameters with the blood culture status

Click here to view

[Table 6] shows that CRP had the highest sensitivity and negative predictive value. Neutropenia and leukopenia had the highest specificity and positive predictive value. If two or more of the above tests were positive, specificity and positive predictive value of the screening tool increased to >90%.
Table 6: Sepsis screen parameters and blood culture

Click here to view

Gram-negative organisms (84.13%) were a predominant causative agent for sepsis. Klebsiella (54%) followed by Pseudomonas aeruginosa (15.9%) and  Escherichia More Details coli (11.1%) were the most common isolates. A higher proportion of mortality was associated with a Gram-negative organism like P. aeruginosa (50%), followed by Klebsiella (32.4%) and E. coli (28.6%).

A higher proportion of mortality was associated with male (35%), preterm (24.4%), intramural (30.3%), and LBW (31.3%) in culture-proven sepsis.

  Discussion Top

The present study shows male preponderance (62%) in the development of early-onset neonatal sepsis which was comparable to other study, making ratio 1.63:1 as compared to admission ratio of 1.54:1; however, this was not statistically significant (P = 0.42). The reported male preponderance in neonatal septicemia may be linked to the X-linked immunoregulatory gene factor contributing to the host's susceptibility to infections in males.

Forty-two percent of cases having an unclean vaginal examination before delivery developed culture-positive septicemia. This is comparable with a study by Yancey et al.[11] The proportion of culture-positive septicemia cases was higher (71%) among the preterm and birth weight <2.5 kg. This is comparable with a study by Dawodu et al.,[12] Tallur et al.,[13] and Roy et al.[2] Preterm neonates are more susceptible to infections due to the lack of inherent defensive mechanism, both humoral and cellular. The results are comparable with the other studies. Higher incidence of sepsis in LBW, both preterm and term, small for gestational age (SGA) is because they have low maternally acquired IgG and they are inherent susceptible to infection. While placental transport of IgG from maternal to fetal circulation increases with maturity, this transport is hampered in SGA neonates who are often the products of placental insufficiency.

The present study clearly shows a nearly equal proportion of cases having prolonged rupture of membranes for 24 h and prolonged labor for >24 h in developing definitive septicemia. This is comparable with a study conducted by Dawodu et al.[12] and Tallur et al.[13]

Apgar score <6 suggestive of birth asphyxia was a major risk factor associated with sepsis in the present study compared to the other studies. These variations probably reflect differences in the rates of occurrence of the predisposing risk factors in the various studies. Clinical chorioamnionitis diagnosed by the presence of foul-smelling liquor was present in 24% of the culture-proven sepsis in the present study which is comparable with studies conducted by St. Geme et al.[14] and Yancey.[11]

While individual risk factors were found to be frequently associated with culture-proven neonatal septicemia, the high score also showed a positive predictive value of 81.3%. However, 74.9% and 49% of patients in the moderate- and low-risk categories were found to be culture-positive, signifying the need to alter the treatment recommendations for these categories.

A single CRP value has a sensitivity of 95.2% and negative predictive value of 89.3% which is comparable to the observation made by other studies. The differences in the results of this parameter shown by the different studies are due to variations in the diagnostic criteria, and different methods of CRP estimation and different cutoff value used in the qualitative test (kit). We had higher sensitivity because our cutoff was 1 mg/dl. The CRP concentration increases physiologically in newborns within the first few days after birth.[15] The CRP concentration also increases in certain sterile conditions after birth such as birth asphyxia and meconium-stained liquor.

In the present study, leukopenia, i.e., total WBC counts <5000 cells/cumm, and neutropenia, i.e., ANCs <1850 cells/cumm, were taken as the diagnostic criteria for detecting neonatal septicemia. Leukopenia has high specificity and positive predictive value but low sensitivity and negative predictive value. This is comparable with the results quoted by Ghosh et al.[16] Neutropenia has high specificity but low sensitivity in the present study which is comparable with the observation made by Ghosh et al.[15] These variations shown by the different authors may be due to differences in the blood sampling time, the severity of infection, the age of the neonates, and the reduced sensitivity of this test in the 1st week of life.

I/T ratio ≥0.2 was also used as the diagnostic criteria for detecting neonatal septicemia. It has high sensitivity and high negative predictive value low specificity and positive predictive value, which is comparable with the observation made by Gerdes and Polin.[10] Thrombocytopenia, i.e., PL counts <1.5 lakhs/cumm, was found to have high specificity and positive predictive value as sepsis screening tool which is comparable with the observation made by Ghosh et al.[16]

Considering any two combination tests (hematological parameters and CRP) being positive as a screening tool for sepsis, the sensitivity and the negative predictive values decreased to 81% and 74.5%, respectively, while the specificity and the positive predictive values increased to 94.6% and 96.2%, respectively, and was found to be statistically significant in detecting septicemia compared to the individual sepsis screen tests in this study. Similar observations were made by the other authors.

In the present study, 63 of 100 cases studied were culture-positive, giving a positivity rate of 63%, which is comparable with the studies conducted by Tallur et al.[13] while a study conducted by Joshi, Madhu Sharma, and NNPD showed very low culture positivity. The culture positivity depends on the time of sampling, extent of bacteremia, and prior antibiotic treatment in the neonate.

Klebsiella pneumonia (54%) was the predominate isolate in this study, followed by P. aeruginosa (15.9%). Gram-negative organisms formed the majority of the isolates as compared to Gram-positive organisms (84.1% vs. 15.9%, respectively) in the present study. This is comparable with studies conducted by others. A Gram-negative organism like E. coli and other enteric bacilli are killed by complement-mediated bactericidal activity, and the protective antibodies are of the IgM class which is deficient in neonates who acquire only maternally derived IgG class antibodies.

  Conclusion Top

The scoring system based on perinatal risk factors suggested by Takkar et al. for treatment protocol in suspected cases of septicemia has good specificity and positive predictive value for high-risk group. However, it failed to identify 49% of patients classified as low risk, who turned out to be culture-positive and needed antibiotics. Septic laboratory screen, on the other hand, was found to have high sensitivity and negative predictive value in babies regardless of their risk factor scores.

Thus, in a neonate who is stable otherwise and suspected of sepsis because of maternal risk factors, it is desirable to await results of sepsis screen before initiation of antibiotics. The combination of Takkar and Bhakoo sepsis risk score in high-risk category and results of septic laboratory screen (esp. in low- and moderate-risk categories) would thus be helpful in ensuring timely initiation of antibiotics and decrease neonatal morbidity and mortality.


I am thankful to my teacher and guide Dr. Jayendra R. Gohil, to my family and colleagues, and last but not the least, to the patients.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.

  References Top

Gotoff SP, Behrman RE. Neonatal septicemia. J Pediatr 1970;76:142-53.  Back to cited text no. 1
Deorari AK. Neonatal sepsis: Manageable daunting issue for India. J Neonatol 2009;23:7-11.  Back to cited text no. 2
Report 2002-2003. National Neonatal Perinatal Database Network. New Delhi: National Neonatology Forum of India; 2004.  Back to cited text no. 3
Mathur NB. Neonatal sepsis. Indian Pediatr 1996;33:663-74.  Back to cited text no. 4
Roy I, Jain A, Kumar M, Agarwal SK. Bacteriology of neonatal septicaemia in a tertiary care hospital of Northern India. Indian J Med Microbiol 2002;20:156-9.  Back to cited text no. 5
[PUBMED]  Medknow Journal  
Chacko B, Sohi I. Early onset neonatal sepsis. Indian J Pediatr 2005;72:23-6.  Back to cited text no. 6
Takkar VP, Bhakoo ON, Narang A. Scoring system for the prediction of early neonatal infections. Indian Pediatr 1974;11:597-600.  Back to cited text no. 7
Sankar MJ, Agarwal R, Deorari AK, Paul VK. Sepsis in the newborn. Indian J Pediatr 2008;75:261-6.  Back to cited text no. 8
Paul VK, Singh M. Diagnosis and treatment of neonatal sepsis. Indian Pediatr 1986;23:1023-35.  Back to cited text no. 9
Gerdes JS, Polin R. Early diagnosis and treatment of neonatal sepsis. Indian J Pediatr 1998;65:63-78.  Back to cited text no. 10
Yancey MK, Duff P, Kubilis P, Clark P, Frentzen BH. Risk factors for neonatal sepsis. Obstet Gynecol 1996;87:188-94.  Back to cited text no. 11
Dawodu A, Umran K, Twum-Danso K. A case study of neonatal sepsis: Experience from Saudi Arabia. J Pediatr 1997;43:84-8.  Back to cited text no. 12
Tallur SS, Kasturi AV, Nadgir SD, Krishna BV. Clinico-bacteriological study of neonatal septicemia in Hubli. Indian J Pediatr 2000;67:169-74.  Back to cited text no. 13
St. Geme JW Jr., Murray DL, Carter JO, Hobel CJ, Leake RD, Anthony BF, et al. Perinatal bacterial infection after prolonged rupture of amniotic membranes: An analysis of risk and management. J Pediatr 1984;104:608-13.  Back to cited text no. 14
Kawamura M, Nishida H. Th e usefulness of serial C-reactive protein measurement in managing neonatal infection. Acta Paediatr 1995;84:10-3.  Back to cited text no. 15
Ghosh S, Mittal M, Jaganathan G. Early diagnosis of neonatal sepsis using a hematological scoring system. Indian J Med Sci 2001;55:495-500.  Back to cited text no. 16
  Medknow Journal  


  [Table 1], [Table 2], [Table 3], [Table 4], [Table 5], [Table 6]

This article has been cited by
Meghana SK, Praveen Kumar Reddy Mopuri, A.K. Bhardwaj
[Pubmed] | [DOI]
2 A Deep Learning Approach for Sepsis Monitoring via Severity Score Estimation
Tunç Asuroglu,Hasan Ogul
Computer Methods and Programs in Biomedicine. 2020; : 105816
[Pubmed] | [DOI]
3 Evaluation of Procalcitonin, C-Reactive Protein, and Interleukin-6 as Early Markers for Diagnosis of Neonatal Sepsis
Emad A. Morad,Rehab A. Rabie,Mohamed A. Almalky,Manar G. Gebriel
International Journal of Microbiology. 2020; 2020: 1
[Pubmed] | [DOI]
Siddu Charki,Trimal Kulkarni,Vijayakumar S Biradar,S S Kalyanshettar
[Pubmed] | [DOI]


Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
Access Statistics
Email Alert *
Add to My List *
* Registration required (free)

  In this article
Materials and Me...
Article Tables

 Article Access Statistics
    PDF Downloaded737    
    Comments [Add]    
    Cited by others 4    

Recommend this journal