Neonatal sepsis remains a major problem in NICUs, as possibly the most commonly diagnosed/suspected and treated illness in NICUs around the world. Being a developing country and with its own limited resources, sepsis-related mortality and morbidity are perhaps the most prevalent in India. We often use a number of “sepsis variant” diagnoses in clinical practice- definite sepsis, probable sepsis, early onset neonatal sepsis, late onset neonatal sepsis, suspect sepsis and even culture negative sepsis!
One drawback of being “over familiar” with sepsis is that many of us tend to attribute each and every symptom and illness to sepsis in some way. We often resort to our own clinical acumen or gut instinct rather than using evidence-based methods to reach such a diagnosis. Many of us are hypocritical in that we take supporting lab investigations at face value only when the results favour our preferred diagnosis.
If lab evidence of sepsis/ SIRS is lacking, we often fail to revise the diagnosis or treatment rather than using novel strategies or terms to reach a unequivocal diagnosis of sepsis (diagnosing culture negative sepsis, for example).
Though we all know that blood culture remains the “gold standard” for the diagnosis of neonatal sepsis, our approach to blood culture practices is far from ideal. We often do not send a proper blood culture, or do not trust a culture report.
There are many categories of neonatologists with regard to their approach towards blood culture!
1. The pessimist: altogether doubts the practical usefulness of blood culture in treatment of neonatal sepsis, and chooses not to send this particular investigation at all. They argue that the culture yield is too low to be helpful to them. The fact that culture is not a cheap investigation may also influence their conviction.
2. The hypocrite: sends the blood culture, often just because it is in their protocol or routine work up list. They have lost their trust in blood culture results a long time ago, and choose to treat neonates purely based on their clinical experience/ trust. They may accept a positive blood culture if it matches with their clinical diagnosis, but other- wise choose to ignore the report!
3. The optimist: sends blood cultures religiously, and although the yield is low, chooses to hope for the best and gets excited every time a positive report is available.
4. The steadfast: ensures a proper blood culture is sent in every suspected case of blood culture, and trusts the report fully. They show courage to take action based on the blood culture, whether the report is positive or negative.
Though in practice there is so much variation in the attitudes of neonatologists (as mentioned above) we all like to be the steadfast example in an ideal world. The question is whether all of us can alter/ improve our clinical practice to match such standards or not. There are a lot of valid questions that we have to answer and convince ourselves of before we have any scope in changing existing practices:
1. What is the supposed sensitivity of the blood culture in neonatal sepsis?
2. How can we improve the sensitivity of blood cultures?
3. How can we avoid false positive reports?
4. How early can I take action based on the culture report?
5. Should I need to step down treatment if the baby is clinically improving?
Contrary to popular belief, blood culture in neonatology is supposedly 100% sensitive!¹ That relies on at least 1ml of blood being inoculated, and the neonate having a bacteremia level of a minimum 4 CFU (colony forming units) per milliliter. Fortunately, or unfortunately, in a typical case of neonatal sepsis, the extent of bacteremia is > 100 CFU per milliliter.² The detection rates of modern blood culture methods are becoming even more accurate, retaining high sensitivity even when the bacteremia is low level, such as at 1-4 CFU/ ml. Though it can be argued that, in the unusual case of a neonate with ultra-low bacteremia (that is < 1 CFU/ml), the sensitivity decreases significantly, such a situation is often clinically irrelevant. In all practical aspects, for a baby with ultralow bacteremia, the natural body defense mechanisms would be sufficient to clear the bacterial infection, and an inevitable dose of antibiotics (for 48 hours or so) should do the job.³
This naturally raises the question of “why are culture yields traditionally so low?”. The simple answer is that we are miserly with respect to the amount of blood we are inoculating! Believe it or not, we often end up submitting < 0.5 ml of blood for blood culture. Decreasing the amount of blood from 1 ml to 0.5 ml would decrease the sensitivity by as much as 40%. A 0.1 or 0.2 ml sample of blood sent for the purpose of blood culture almost makes the process a waste of time and resources. One of the perceived reasons behind not sending a 1 ml blood sample for culture is the practical difficulty in bloodletting. Though 1 ml is a significant amount in very preterm babies, the clinical implications make it an absolute necessity. The problem can be partially solved by taking the first 1 ml of blood after accessing the IV cannula/ or during sampling for the culture, with all other investigations being given second priority. This means that, for an occasional baby, the RFT or PT/ INR etc. may have to be postponed. A clear written order in the case sheet mentioning this and the amount of blood needed is also found to be helpful.
Another condition that makes blood culture less sensitive is prior antibiotic use. Ensuring that blood culture samples are taken at first contact or before administering the first dose of antibiotics may be the only way around this problem. This problem is greatest in outpatient births and referred cases of neonates. However, even in those cases, ensuring a proper amount of blood for culture often greatly improves the situation.
False positives or contamination in blood culture can be easily avoided by ensuring proper aseptic precautions are taken during sampling. Proper use of skin disinfectants such as 2% chlorhexidine gluconate in 70% isopropyl alcohol (2% alcoholic chlorhexidine), or even 10% povidone iodine is mandatory. Creating a specific “cannulation set” is a rewarding and simple practice (a sterile tray with cotton swabs and other necessities for cannulation), as this ensures a sterile surface/field which significantly decreases the contamination risks.
Trusting blood culture reports and taking the necessary actions are steps which are as important as ensuring proper blood culture collection technique. This means that, once a positive culture is received and sensitivity is available, we should downgrade antibiotics to the narrowest spectrum antibiotic clinically appropriate. This is necessary even if the baby’s condition was very poor to begin with and this is currently improving on an empirically sensitive broad-spectrum antibiotic.
Trusting the blood culture if the report is sterile is often a more demanding process. There is plenty of evidence to show that it is safe to take action over a 48-hour blood culture report.4,5 A practical approach is to stop antibiotics if the blood culture is sterile and the baby is doing well after 48 hours. If the baby is still symptomatic at 48 hours (and you are still not able to rule out infection as a cause for the symptoms), it is wise to consider inflammatory markers (CRP, Procalcitonin etc) and to stop antibiotics if these fall within an acceptable range.
Preventing the unnecessary use of antibiotics is the need of the hour, and an essential step in providing quality care for sick neonates. Antibiotic usage in a culture negative vulnerable neonate has been shown to increase the incidence of necrotizing enterocolitis, bronchopulmonary dysplasia and death.6
Though a 100 percent theoretical sensitivity of blood culture in neonatal sepsis is difficult to achieve in routine clinical practice, we can definitely improve neonatal outcomes with the above-mentioned quality improvement initiatives. This is like targeting a zero incidence of health care associated infection (HAI) rate in the NICU: we won’t succeed unless we try, we won’t try if we don’t believe, and if we don’t believe then nothing will ever happen.
1. Schelonka RL, Chai MK, Yoder BA, Hensley D, Brockett RM, Ascher DP. Volume of blood required to detect common neonatal pathogens. J Pediatr. 1996;129(2):275–278.
2. Sabui T, Tudehope DI, Tilse M. Clinical significance of quantita- tive blood cultures in newborn infants. J Paediatr Child Health. 1999;35(6):578–581.
3. Cantey JB and Baird SD. Ending the Culture of Culture-Negative Sepsis in the Neonatal ICU. Pediatrics. 2017;140(4):e20170044
4. Kaiser J, Cassat J, Lewno M. Should Antibiotics be Discontinued at 48 Hours for Negative Late-Onset Sepsis Evaluations in the Neonatal Intensive Care Unit? J Perinatol. 2002;22:445–447.
5. Dretvik T, Solevåg AL, Finvåg A, Størdal EH, Størdal K, Klingen- berg C. Active antibiotic discontinuation in suspected but not confirmed early-onset neonatal sepsis – a quality-improvement initiative. Acta Paediatr. 2020; Accepted Author Manuscript: doi:10.1111/apa.15202.
6. Ting JY, Synnes A, Roberts A, Deshpandey A, Dow K, Yoon EW et al. Association between antibiotic use and neonatal mortal- ity and morbidities in very low-birthweight infants without culture-proven sepsis or necrotizing enterocolitis. JAMA Pediatr. 2016;170(12):1181–1187.