• 817.pdf

Download the ebook

Original Articles
EVALUATION OF RISK FACTORS FOR FATAL NEONATAL SEPSIS
N.B. Mathur, A. Singh, V.K. Sharma and L. Satyanarayana
From the Departments of Pediatrics and Microbiology, Maulana Azad Medical
College, New Delhi 110 002.
Reprint requests: Dr. N.B. Mathur, 187, Rouse Avenue, Deen Dayal Upadhyaya Marg,
New Delhi 110 002.
Received for publication: February 6,1996; Accepted May 9,1996
Objective: To evaluate risk factors for fatal neonatal sepsis. Design: Prospective study.
Setting: Referral neonatal unit of a teaching hospital. Subjects: 171 neonates admitted
with sepsis. Methods: Clinical examination and investigations on the day of admission
were recorded and the neonates followed up to determine the final outcome. Results:
The overall fatality was 48.5%. In the univariate analysis, the factors significantly
associated with death were weight, gestational age, age at onset of sepsis, hypothermia,
requirement of lPPV, presence of refractory septic shock, neutropenia, metabolic
acidosis and raised prothrombin time. However, in the multivariate analysis, only
neutropenia, metabolic acidosis, increased prothrombin time and refractory septic
shock retained their, significance. The adjusted odd's ratio (95% confidence interval)
were 0.095 (0.04 - 0.22), 1.14 (1.04 - 1.25), 1.04 (1.002 - 1.08) and 11.82 (5.47 -
69.40), respectively. Conclusion: Even in a setting with high fatality rates, high risk of
mortality in neonatal sepsis can be identified and targeted for intensive intervention.
Key words: Neonatal sepsis, Mortality, Septic shock, Metabolic acidosis, Neutropenia.
S EPSIS is a frequent and serious event
which threatens survival during the
neonatal period. The morbidity and
realized that inability to identify patients at
greatest risk is causing problems for
researchers who design clinical trials for the
mortality rate from neonatal sepsis innovative therapies to fight sepsis,
continues to be high the world over inspite systemic inflammatory response
of the development of broad spectrum syndrome and multiple organ
antibiotics and technological advances in dysfunction syndrome(13). The present
life support therapy(l-7). The reasons that study was therefore designed to evaluate
some neonates with sepsis die while others the risk factors for fatal neonatal sepsis
survive is still not clear. A reduction in using multivariate logistic regression
sepsis related mortality may be possible by analysis.
identifying high risk neonates and
Subjects and Methods
targeting them for intensive therapy.
Earlier reports on mortality in neonatal This prospective study was done on
sepsis have been simple correlation studies neonates admitted to the Referral
done in an uncontrolled way(8-12). There Neonatal Unit of Lok Nayak Hospital,
is a paucity of reports on risk factors in New Delhi, which predominantly serves
fatal neonatal sepsis using multivariate the poorer strata of the society. The
statistical analysis to establish risk factors neonates are mostly referred from hospitals
for fatality with adjustment for potential with poor facilities for neonatal care and
confounders. It has also been recently are brought late in critically sick condition.
INDIAN PEDIATRICS 817 VOLUME 33-OCTOBER 1996
MATHUR ET AL. RISK FACTORS FOR FATAL NEONATAL SEPSIS
Inclusion criteria consisted of all of the by univariate analysis. Chi Square and
following: (i) Clincial feature of neonatal Fischer's exact test were used to analyze
sepsis; (ii) Physical examination the significance of the differences. The
demonstrating either circulatory or factors found to be significant on univariate
respiratory dysfunction. Circulatory analysis were subsequently subjected to a
dysfunction was evidenced by the presence
stepwise multiple logistic regression
of tachycardia (heart rate more than
160/minute) or bradycardia (less than analysis to evaluate the independent
100/minute) or capillary refill time more predictors of mortality due to neonatal
than 3 seconds. Respiratory dysfunction sepsis.
was evidenced by the presence of grunting, Results
flaring retractions, tachypnea (more than
60 breaths/minute) or apnea lasting more Out of 171 septic neonates, 83 (48.5%)
than 15 seconds; and (iii) One or more of died (cases) and 88 were discharged after
the following laboratory criteria: (a) they improved (controls). The stages of
Positive blood culture, (b) systemic inflammatory response syndrome
Polymorphonuclear band cells more than at admission was sepsis in 49, sepsis
20%, (c) Elevated C-reactive protein(14). syndrome in 66, early septic shock in 19
and refractory septic shock in 37 neonates.
One hundred and seventy one neonates None of the neonates had multiple organ
were enrolled for the study. Findings of dysfunction at admission.
clinical examination and investigations on The significant risk factors for mortality
the day of admission were recorded in a due to sepsis on univariate analysis were
structured proforma. Prothrombin time was weight, gestational age, age of onset of
estimated 6 hours after administration of sepsis, hypothermia, requirement of IPPV
injection vitamin K in all neonates. The and presence of refractory septic shock
neonates were divided into various stages (Table I). Significant investigative findings
of systemic inflammatory response included neutropenia, metabolic acidosis
syndrome namely sepsis, sepsis syndrome, and raised prothrombin time.
early septic shock, refractory septic shock
and multiple organ dysfunction On multivariate analysis, the significant
syndrome(15). Interpretation of neutrophil independent risk factors for mortality were
counts was done as suggested earlier(16). neutropenia, increased negative base
excess and prothrombin time and presence
Management of the septic neonates of refractory septic shock (Table II).
included administration of parenteral
Age of onset of sepsis was an
fluids, vasoactive agents and oxygen for
independent clinical predictor. However,
hemodyanamic stability and oxygenation
this parameter did not remain statistically
of vital tissues (whenever required) and the
significant when laboratory variables were
use of adequate antimicrobial agents and
also included in the model. Refractory
early drainage or removal o£ purulent foci
septic shock remained an independently
to achieve bacterial eradication(15).
significant predictor of mortality even in
The neonates were followed up to the presence of laboratory variables (Table
determine the final outcome. Those who III).
died served as cases while those discharged Discussion
after recovery were taken as controls. To
find out the risk factors associated with Sepsis is a major cause of neonatal
mortality in neonatal sepsis, all clinical and mortality. The present study highlights the
investigative variables were first evaluated identification of refractory septic shock,
INDIAN PEDIATRICS 818 VOLUME 33-OCTOBER 1996
MATHUR ET AL. RISK FACTORS FOR FATAL NEONATAL SEPSIS
neutropenia, metabolic acidosis and with neutropenia, improvement in
increased prothrombin time as significant neutrophil count and function was
independent predictors of mortality in neo- documented after exchange transfusion.
natal sepsis.
Base deficit and refractory septic shock
In the present study, neutropenia was were the other independent risk factors for
the most significant risk factor for death fatality in sepsis in our study. Refractory
from neonatal sepsis. In an earlier septic shock occurs due to vasodilatation,
report(17), 16 out of 26 neonates with capillary leak and endothelial damage.
sepsis and neutropenia had severe Current evidence indicates that these
depletion of the neutrophil storage pool. pathophysiological effects appear to be
Mortality was high when neutrophil mediated by proinflammatory cytokines
storage pool was depleted and granulocyte activated in response to microbial
transfusion led to improved survival. In a components in the vascular compartment
randomized controlled trial(18) of (19-21). High values for tumor
exchange transfusion in neonatal sepsis necrosis factor, interleukin-1 and also
INDIAN PEDIATRICS 819 VOLUME 33-OCTOBER 1996
MATHUR ET AL. RISK FACTORS FOR FATAL NEONATAL SEPSIS
interleukin-6 in plasma of Multivariate analysis limited to clinical
children with sepsis have been correlated factors yielded refractory septic shock and
with increased fatality rates(14). age of onset of sepsis as independent
Interleukin-1 beta plasma levels during predictors of death. However, age of onset
sepsis in neonates appear to be correlated of sepsis lost significance when laboratory
with the decrease in diastolic tension variables were also included in the
according to birth weight(21). However, it analysis. Early onset of sepsis has been
is not feasible to estimate the levels of associated with high mortality in earlier
these cytokines routinely. studies(8,22) while other reports did not
observe this association(10-23). In a
series(24), higher mortality was
Increased prothrombin time was documented in early onset sepsis due to
another independent risk factor for fatality group B streptococcus but not in sepsis due
in our study. In the septic neonates who to E. coli.
had received Vitamin K, increased
In conclusion, even with routinely
prothrombin time probably reflected
available clinical and laboratory
activation of the coagulation cascade. A
parameters, risk of fatality can be
complete coagulogram was not feasible. In
identified in neonatal sepsis. There is an
this context, it is pertinent to note that the
urgent need for further studies to identify
present study was intended to evaluate only
risk factors before the disease reaches a
those risk factors available routinely for
critical stage. Newer modalities of treat-
septic neonates.
INDIAN PEDIATRICS 820 VOLUME 33-OCTOBER 1996
MATHUR ET AL. RISK FACTORS FOR FATAL NEONATAL SEPSIS
ment of neonatal sepsis need evaluation 7. Ferrieri P. Neonatal susceptibility and
using high risk factors as indicators of immunity to major bacterial pathogens.
severity. Rev Infect Dis 1990,125: 394-400.
REFERENCES 8. Mathur NB, Khalil A, Sarkar R, Puri RK.
Mortality in neonatal septicemia with
1. Erickson M. Neonatal septicemia. Acta involvement of mother in management.
Pediatr Scand 1983, 72: 1-5. Indian Pediatr 1991, 28:1259-1263.
2. Singh M. Hospital based data on perinatal 9. Mondal GP, Raghavan M, Bhat BV,
and neonatal mortality in India. Indian Srinivasan S. Neonatal septicemia among
Pediatr 1986, 23: 579-584. inborn and outborn babies in a referral
3. Bhatia BD, Mathur NB, Chaturvedi P, hospital. Indian J Pediatr 1991, 58: 529-
Dubey AP. Neonatal mortality pattern in 533.
rural based medical college hospital. 10. Agrawal M, Chaturvedi P, Dey
Indian J Pediatr 1984, 51: 309-312. SK, Narang P. Coagulase negative
4. Hill HR. Phagocyte transfusion-ultimate staphylococcal septicemia in
therapy of neonatal disease. J Pediatr newborn. Indian Pediatr 1990, 27: 163-
1981,98:59-61. 169.
5. Bennet R, Erickson M, Zetterstrom R. In 11. Christensen RD, Rothstein G. Exhaustion
creasing incidence of neonatal of mature marrow neutrophils in neonates
septicemia: Causative organisms and with sepsis. J Pediatr 1980, 96: 313- 318.
predisposing risk factors. Acta Pediatr 12. Dawodu AM, Efflong CE. Neonatal
Scand 1981, 70: 207-210. mortality effects of selective pediatric
6. Yoder M, Polin R. Immunotherapy of interventions. Pediatrics 1985, 75: 51-57.
neonatal septicemia. Pediatr Clin North 13. Bone RC, Sibbald WS, Sprung CL. The
Am 1986, 33: 481-499. ACCP-SCCM Consensus Conference on
INDIAN PEDIATRICS 821 VOLUME 33-OCTOBER 1996
MATHUR ET AL. RISK FACTORS FOR FATAL NEONATAL SEPSIS
sepsis and organ failure. Chest 1992, 101: circulating levels of interleukin-6 in
1481-1483. patients with septic shock: Evaluation
14. Sullivan JS, Kilpatrick L, Costarino AT, during sepsis, prognostic value and
Lee SC, Harris MC. Correlation of interplay with other cytokines. Am J Med
plasma cytokine elevations with mortality 1991, 91: 23-29.
rate in children with sepsis. J Pediatr 20. Dinavello CA. The pro-
1992, 120: 510-515. inflammatory cytokines interleukin 1 and
15. Saez-LLorens X, McCracken Jr. GH. tumor necrosis factor and treatment of
Sepsis syndrome and septic shock in the septic shock syndrome. J Infect Dis
pediatrics: Current concepts of 1991, 163: 1177-1184.
terminology, pathophysiology and
21. DeBont ESJM, Martens A, Van Raan J,
management. J Pediatr 1993,123: 497-
Samson G, Felter WF, OkkenA, Deleij
508. LHFM. Tumor nercrosis factor
16. Manroe BL, Weinberg AG, Rosenfeld alpha, interleukin-1 beta and interleukin-6
CR, Browne E. The neonatal blood count plasma levels in neonatal sepsis. Pediatr
in health and disease. I. Reference values Res 1993, 33: 380-383.
for neutrophilic cells. J Pediatr 1979, 95:
89- 98. 22. Bhakoo ON, Agarwal KC, Narang A,
Bhattacharjee S. Prognosis and the
17. Christensen RD, Rothstein G, Anstall treatment of neonatal
HB. Granulocyte transfusions in septicemia—A clinicobacterial study of
neonates with bacterial infection, 100 cases. Indian Pediatr 1974,11:519-
neutropenia and depletion of mature 528.
marrow neutrophils. Pediatrics 1982,
70:1-6. 23. McCracken GH, Shinefield H. Changes
in the pattern of neonatal septicemia and
18. Mathur N.B, Subramanian BKM, Sharma meningitis. Am J Dis Child 1966, 112:
VK, Puri RK. Exchange transfusion in 33-39.
neutropenic septiceminc neonates: effect
on granulocyte functions. Acta Pediatr 24. Vesikari R, Janas M, Gronroos P, et al.
1993, 82: 939-943. Neonatal septicemia. Arch Dis Child
1985, 60: 542-546.
19. Calandra T, Garian J, Heumann D. High
NOTES AND NEWS
INTERNATIONAL SYMPOSIUM: ADVANCES IN PEDIATRICS
The Department of Pediatrics, All India Institute of Medical Sciences, New Delhi, is
organizing this event on November 29, 30 and December 1, 1996. Guest Lectures,
Dialogue Sessions, Symposia and Case vignettes from various pediatric sub-specialties
will be held. The registration fees for Indian Delegates is Rs. 750/- and for Foreign
Delegates US $ 75/-. The amount may be sent by a cheque or Demand Draft in the name
of "International Symposium, AIIMS, New Delhi". For further details please contact Dr.
Veena Kalra, Organizing Secretary, Department of Pediatrics, All India Institute of
Medical Sciences, New Delhi 110 029. Fax: 011-6862663.
INDIAN PEDIATRICS 822 VOLUME 33-OCTOBER 1996


Use: 0.4962