Transcription:
Late Onset Sepsis: Evaluation and Management

Julie Weiner, DO (Host): Hello everyone. And welcome to this edition of Neonatal Review Isolette to Crib. The purpose again of this podcast is to review high yield common topics in neonatology. While our focus is geared towards the perinatal neonatal boards, anyone learning or study neonatology will find this podcast helpful.

I will be one of your hosts for this podcast. I am Dr. Julie Weiner, one of the Neonatologists and Medical Director of the NICU at Children's Mercy Hospital in Kansas City, Missouri. We're located here in the Heartland where we are known for our barbecue. And another fun fact is Kansas City has more barbecue restaurants per capita than anywhere else in the world.

So, to help us prepare for boards, we have a collection of neonatologists to help us along the way. All have either passed boards. Been there, done that, or are currently studying for boards, same boat as you. We are joined today again by Dr. Joti Sharma, the other host for this podcast, who is the Fellowship Program Director here at Children's Mercy Hospital. Dr. Sharma tell everybody hello, again.

Joti Sharma, MD (Guest): Hello, everyone from rainy Kansas City today. It's raining here today.

Host: She is joining us to finish up our discussion on neonatal infections, late onset sepsis. We will discuss the evaluation and treatment for late onset sepsis. As you remember, in our last podcast, we discussed the epidemiology and pathogenesis of late onset neonatal sepsis. Okay, Dr. Sharma let's get started. What are the things to consider for evaluation of late onset sepsis in newborns?

Dr. Sharma: Hi, Dr. Weiner. Before we move to discussing about the evaluation of late onset sepsis, I just want to let our audience know that a lot of aspects of neonatal sepsis evaluation and treatment is also covered in the podcast we did previously on early onset sepsis. So, I would refer you to that also.

So, moving back to late onset sepsis. So, it' similar to evaluation of early onset sepsis. When we are evaluating an infant for late onset sepsis, we need to consider three aspects. These include risk factors for sepsis, which include perinatal risk factors. We all are aware of choriamnionitis, maternal fever, maternal GBS colonization, rupture of membranes more than 18 hours, and of course, prematurity as being the important perinatal risk factors. The second risk factor is infant risk factors. And this is more applicable to preterm infants who are in the NICU, who are undergoing medical interventions, or may have medical devices such as IV catheters or be on mechanical ventilation. And they are predisposed to healthcare related infections or nosocomial infections. The other two aspects to consider as part of evaluation are the clinical status of the infant, and lastly, the lab investigations.

Host: So, part of the evaluation, is considering the signs and symptoms of sepsis. Dr. Sharma, can you discuss some of the signs and symptoms of neonatal sepsis and more specifically as it relates to late onset sepsis?

Dr. Sharma: So, while clinical signs and symptoms is an important aspect of the evaluation of sepsis, I think what is very important to note is that with neonatal sepsis and more so with late onset sepsis, the clinical manifestations of sepsis, are usually nonspecific. And also the clinical signs and symptoms vary with gestational age and severity of the infection. So, as an example, we may have a term infant who presents from home at 10 days of life with fever, not feeding well or in respiratory distress and who may have late onset sepsis. And this is a scenario we may encounter with late onset GBS infection. As opposed to we may have a 10 day old former 25 weeks gestation pre-term infant in the NICU who may present with apneic spells, lethargy, or even hypothermia as a sign of sepsis. In general, the clinical signs and symptoms of late onset sepsis may range from temperature instability. You may have fever or hypothermia, oxygen desaturation, bradycardia, poor perfusion, reduced capillary refill, hypotension, and even septic shock.

Host: So, Dr. Sharma, as we know, preterm infants are a group that's at high risk for late onset sepsis. Can you discuss more specifically about the signs for this group?

Dr. Sharma: Hey, Dr. Weiner, thank you for that question. And you are true. The clinical manifestations of late onset sepsis in preterm infants generally depends on the gestational age. So, one of the NICHD studies on very low birth weight infants, found that about 50% of very low birth weight infants with late onset sepsis presented with apnea, about 40% had feeding intolerance, abdominal distension, about a third had increased need for respiratory support and about 20% had lethargy and hypotonia.

So, as you see the signs and symptoms is wide ranging. Also interesting is that patients with late onset sepsis due to gram negative organisms or fungal infections, have more severe signs and symptoms compared to gram-positive infections such as coagulase negative staph. But again, this is very difficult to distinguish clinically. So, we should assume the worst if the baby is really sick.

Host: So, recently there's been a lot of literature supporting the heart rate characteristics, or HRC. Can you spend a little bit more time just discussing this specific clinical sign and if there's others used more recently for the evaluation of sepsis.

Dr. Sharma: Oh, yes, Dr. Weiner, there's a lot more coming out about HRC. So, HRC monitoring is actually a more recent technology that has been utilized in some NICUs for early diagnosis of late onset sepsis. And it's based on the findings that infants develop decreased heart rate variability and transient decelerations prior to clinical manifestations of sepsis. We are aware of this through our fetal heart rate monitoring. So, it's a similar concept. This abnormal HRC pattern may occur more than 24 hours before the onset of other clinical signs of sepsis. And the exact mechanism as to why this pattern occurs, is not known, but it is thought to be mediated by cytokines. They're in a study of nine NICUs, where they looked at the very low birth weight infants. They found that babies that had HRC monitoring had a lower sepsis associated mortality compared to the controls. So, in short, HRC monitoring can be utilized as a clinical tool for the evaluation of late onset sepsis.

Host: Thank you for that clarification, Dr. Sharma. Okay. So, as we know, lab investigations are a very important part of neonatal sepsis. Dr. Sharma, can we talk a little bit more about what labs are needed in late onset sepsis?

Dr. Sharma: Dr. Weiner, yes. Lab investigations are always an important part of a sepsis workup, but again, it is very important to perform evidence-based investigations. So, the typical sepsis evaluation for a neonate in whom we suspect late onset sepsis, is very similar to the evaluation for patient with early onset sepsis, except in late onset sepsis, we do a urine culture. So, in short, for the evaluation of late onset sepsis, we should get a complete blood count with white cell and differential, peripheral blood culture, spinal tap for cell count and culture and urine culture.

Host: Are there other investigations to consider as part of the workup for late onset sepsis?

Dr. Sharma: Dr. Weiner, very important question, because remember most of our babies in the NICU undergoing different medical interventions and may have different devices in place. Therefore, other site cultures as relevant should be considered such as if you have a patient who is on the ventilator, you may consider doing a tracheal aspirate for gram stain, cell count and culture.

And if a baby has any skin lesions, especially if it's a pustual you may consider deroofing it and sending it for culture. The other investigations that might be considered include imaging studies based on your clinical assessment. You may consider a chest x-ray for pneumonia or a abdominal x-ray or abdominal ultrasound if necrotizing enterocolitis is suspected. Of course, how can we forget the acute phase reactant right, CRP is also a part of the evaluation of late onset sepsis. I do want to mention, actually more recently, molecular based techniques, such as PCR with sequencing of species specific gene regions is being used as a tool for diagnosis of neonatal sepsis. Most of it is still in the research area, but it will help clinically once it's fully available. It actually helps differentiate contaminant cultures versus true infection. And this is, will be very helpful for us when it comes to coagulase negative infection.

Host: So, I know we discussed the usefulness and limitations of white blood cell count and it's indices as part of the evaluation of neonatal sepsis in the previous discussion on early onset sepsis evaluation. Can you for our listeners to summarize some of these points again?

Dr. Sharma: Sure. I think we discussed this the last time. I think a CBC is probably the most common lab test that we do in a neonate besides the newborn screen of course. So, as part of that, we do a white cell count, differential, absolute neutrophil count, or ANC. And the IT ratio or the immature to a total neutrophil ratio. All these are widely used as screening tests for neonatal sepsis, but unfortunately none of these tests have been particularly useful in identifying the majority of septic infants. So, the total white cell count has a poor predictive value for sepsis. Neutropenia ANC less than a thousand actually has a greater specificity for neonatal sepsis. So, it's more useful. The IT ratio, which we use often is actually more useful in ruling out infection then in ruling in infection. In clinical practice, we also look at platelet counts, but platelet counts are not very sensitive or specific for the diagnosis of neonatal sepsis.

Host: Thank you, Dr. Sharma for summarizing that. Again, if you guys want to learn more about the use of white blood cell counts and its components for neonatal sepsis evaluation, check out the podcast on early onset sepsis evaluation and management. Okay, we're going to move on. So, we know that blood culture is the gold standard lab investigation for sepsis. Is there anything different with regards to late onset compared to early onset sepsis?

Dr. Sharma: So, with regards to late onset sepsis in infants with suspected late onset sepsis, we should ideally do two blood culture. And this is more important for preterm infants in the NICU, where coagulase negative staph is not an uncommon cause of late onset sepsis. And if you have two positive blood cultures with CONS, it is more likely to suggest a true infection as compared to a contaminant.

So, I just helpful in that. So, we prefer two peripheral blood cultures, but if that's not possible, then you may be able to do one peripheral and one central line blood culture, knowing that if it is colonized, it may be an issue, but in that respect, hopefully the second blood culture will help. Now, what is important is that the volume of blood is important. And actually one of the studies has shown that if you do one blood culture with one ml of blood, you are likely to get a true organism grown as opposed to if you only use 0.5 mls. But again, all of this has to be taken into account when we are dealing with a very low birth weight infant, because the volume of blood and the ability to access peripheral sites will always be a challenge. Therefore, we have to use clinical judgment as regards to how many blood cultures we can do.

Host: Okay, Dr. Sharma let's move on to acute phase reactants. What are they?

Dr. Sharma: So the most commonly used acute phase reactant is the CRP and the other one is procalcitonin. And these are the two that we most commonly talk about in neonatal sepsis.

Host: Okay. So first let's focus and talk more about the CRP and how it's used in late onset sepsis.

Dr. Sharma: So, again, we covered CRP in more detail in the podcast on evaluation of early onset sepsis, but in short, CRP rises within six to eight hours of infection and it peaks around 24 hours. A value of less than one milligram per DL or 10 milligram per liter is considered normal. And CRP has its best predictive value if it's measured within 24 to 48 hours after the onset of infection. And we all know that an increasing CRP is a better predictor of infection than a single value. And for that reason, if you decide to use CRP as a marker of infection, you do need to do a follow up, rather than just base your clinical decisions on just one level.

Host: So, you just mentioned procalcitonin. Can you briefly talk about this and how it has more recently been considered useful for the evaluation of suspected sepsis?

Dr. Sharma: So, Dr. Weiner, procalcitonin is a propeptide of calcitonin, and it's actually produced mainly by the monocytes and hepatocytes. And it's elevated during infections. It's half-life is about 24 hours and it is more sensitive for early detection of sepsis than CRP is. And interestingly, procalcitonin level is more likely to be elevated during bacterial infections than during viral infection and it declines rapidly with therapy. And currently it's mainly used in research. And more recently it has begun to be utilized in clinical practice. So, we will be probably seeing more of the use of procalcitonin in our practice.

Host: Thank you, Dr. Sharma. So, what about spinal fluid for evaluation of late onset sepsis. Do we always need to do an LP as part of late onset sepsis evaluation?

Dr. Sharma: Yeah. So, with regards to the spinal tap, remember when we talked about early onset sepsis, we talked about how some consider it controversial for evaluation of early onset sepsis. But I think it's important to remember that a spinal tap is a mandatory work up for late onset sepsis. And I mentioned that, because approximately 20% of neonates with bacteremia will also have meningitis and approximately 35 to 40% of those with meningitis will have a negative blood culture. Therefore, a spinal tap should be a component of every neonatal sepsis evaluation.

So, in theory. Yes. It's easy for us to kind of discuss that an LP is mandatory, but remember sometimes it's not always feasible in clinical practice. Ideally, LP should be done prior to the new session of antibiotics, but if infant is hemodynamically unstable or has a coagulopathy, or has thrombocytopenia, or for some reason that there will be a delay in antibiotic administration, then you should go ahead and start antibiotics and do the LP later. But keeping in mind that it can give a false negative CSF culture, therefore you will have to rely on other inflammatory markers in the CSF, such as cell count to guide you.

Host: Okay. Before we move to discuss treatment of late onset sepsis, can you discuss why the urine culture should be part of the evaluation for late onset sepsis?

Dr. Sharma: Yes. Dr. Weiner. Isn't it always, when we talk about in our practice, let's do a sepsis work up, the question a;ways comes up, should we do a urine culture or not. So, as opposed to the early onset sepsis evaluation, we talked about urine culture is not only needed, but urine cultures should be part of an evaluation for late onset neonatal sepsis. And this is why. The risk of UTI in infants is about 10 to 15%, with the incidence being higher in the preterm and the risk increases with postnatal age. So, at about a week of life, the risk of UTI in a neonate is about 1%, but it increases to about 25% by about two to three months of life. And when we talk about a urine culture, there is always another question. How do we get the urine? So, suprapubic bladder aspiration or suprapubic tap is considered the best method for obtaining urine culture in those less than six months. But as you know, more and more bladder catherization specimens are being done.

And in as much aseptic conditions as we can. And I would say that in the last decade or so, suprapubic aspirations have markedly decreased. And if appropriate, to do a urine bladder catheterization and send off a urine culture, but you need to be aware that it does have a high false positive. So the colony count, a single organism and clinical signs symptoms need to be considered for a diagnosis of UTI.

Host: Thank you, Dr. Sharma. Okay. So, clinically you suspect late onset sepsis. The sepsis workup has been completed. The next decision is the choice of empiric antibiotic therapy in late onset sepsis. Dr. Sharma, can you discuss the antibiotic choices and treatment for late onset sepsis for us?

Dr. Sharma: Okay. So yes, as you said, you know, we do the sepsis work up, we send the cultures and then we start the antibiotic. So, empiric antibiotic choice for late onset sepsis actually depends on the likely organisms and the sensitivity patterns at your local NICUs. And this is more so specifically with regards to the preterm or the very low birth weight infants. Generally most do the use of a combination of vancomycin and gentamycin as the empirical antibiotic for late onset sepsis. Because the most common organism, especially in the preterm NICU patients are coagulase negative staph and gram negative organisms. And these antibiotics will cover them.

But with the advent of antibiotic stewardship and the CDC recommendations to use, to limit use of vancomycin as empiric antibiotic choice, due to concerns of development of vancomycin resistant organisms, we do need to consider some alternatives.

Host: So, with the concerns for development of vancomyocin resistance, what are the alternative empiric antibiotic therapies to consider?

Dr. Sharma: So, the alternate empiric antibiotic choices that you can consider for late onset sepsis are ampicillin. Especially if you consider GBS. If you are considering clinically, the patient has GBS infection or oxacillin, where if the patient does not have any central lines and gentamycin. And this more so in term or late preterm infants, or as I mentioned earlier, those without central lines or other devices. A little bit about cephalosporins. So, cefotaxime, cefepime or ceftazidime should only be considered as empiric antibiotics, if meningitis is clinically considered or suspected. So, that's kind of the only role due to development of resistance and increased risk of fungal infection, cephalosporins should be reserved.

Host: Okay. So, we have a positive blood culture. How is antibiotic therapy now modified?

Dr. Sharma: So, once you get your blood culture results back, based on the blood culture and other site cultures that you may have performed, the antibiotic therapy is modified to pathogen directed therapy, and also you need to know the sensitivities. So, this would be for coagulase negative staph. And if the patient has methicillin resistant staph aureus infection, you would continue your vancomycin. For MSSA or methicillin sensitive staph aureus, you can give oxacillin. For GBS, we have our good old ampicillin. For gram negative infections based on which gram negative organism it is, and also have you need to consider the ESBL status, antibiotic choices could be ampicillin, cefdesitime or cefapime. And for ESBL, meropenem is the choice.

Host: Now that we've talked a little bit about which antibiotics, how about the duration of antibiotic therapy for proven infection?

Dr. Sharma: Dr. Weiner, so in general, for uncomplicated bloodstream infections, the duration of antibiotic therapy is 10 to 14 days. Usually for uncomplicated gram-positive infections, such as GBS or MSSA, we usually do at 10 days. And with gram negative infections is we usually do 14 days.

Host: How does the duration of antibiotic therapy for late onset sepsis with meningitis change or differ?

Dr. Sharma: So, with late onset sepsis with meningitis, the duration of antibiotic therapy is around 14 to 21 days for gram-positive infections. And for the gram negatives, you have to give at least a minimum of 21 days, but sometimes with follow-up LPs, you may have to extend it to 28 days.

Host: Okay. So, we have just discussed treatment when there is proven infection, but more often, the clinical questions that usually come up is how long to continue antibiotics when sepsis markers are benign and patient clinically is doing well, how long do we wait for cultures? It appears most of us use 48 hours, but sometimes it could be longer.

Dr. Sharma: Great question. And this is a concept that we sometimes struggle with in our practice. And again, it's important to consider this, given all the other issues with unnecessary use of antibiotics. So, in clinical practice, we normally wait for 48 hours for bacterial cultures and there is actually evidence behind it. The negative predictive value that is ruling out infection for the isolation of a definite bacterial pathogen is around 99.8% at 48 hours. And the growth in the culture medium after 72 hours of incubation is more likely to be a contaminant. And this is from the study that was published in 2001 in the Archives of Diseases of Childhood, by Kumar and colleagues. So, as you can see, waiting 48 to 72 hours to get your cultures, and if they are negative and clinically, the patient is doing well; it would be reasonable to stop antibiotics. I do want to talk a little bit about antibiotic stewardship in the NICU because it involves our very premature infants. So, the 2002 NICHD study of very low birth weight infants with late onset sepsis, showed that 44% of infants, more than three days of life received a treatment course of vancomycin. What is more interesting is that 30% of infants, more than three days of life without a proven infection received a treatment course of vancomycin. So, you can see that even without a positive blood culture, 30% of infants, of very low birth weight infants actually did receive a course of vancomycin. And that is concerning, especially when we are always worried about a development of resistance. The rate of antibiotic use was inversely related to gesstational age and birth weight.

So, the smaller you are and the lighter you are, the more likely you are to get antibiotics. And interestingly, vancomycin was also inversely related to gestational age. Again, the more premature, the more likely use of vancomycin. So, while the decision to continue antibiotic therapy in an infant with negative culture is based on clinical judgment of the clinical team taking care of the patient, at the same time, it is important to consider appropriate use of antibiotics to reduce not only antibiotic resistant bacteria, but also decrease the risk of fungal infections.

Host: Thank you, Dr. Sharma. Before we move to summary, I just want to tell the audience that prevention of late onset sepsis and the use of probiotics in the NICU population will be covered in a later podcast as well as neonatal fungal sepsis. Okay. Dr. Sharma, can you summarize what we have been discussing in this podcast regarding late onset sepsis evaluation and treatment?

Dr. Sharma: Thank you, Dr. Weiner. So, to summarize, in the evaluation of late onset sepsis, you should consider the perinatal and infant risk factors, the clinical status of the infant and the laboratory investigations. The typical complete sepsis evaluation in a neonate suspected of late onset sepsis consists of a complete blood count with white cell differentials, preferably two peripheral blood cultures, a spinal tap, other site cultures, urine culture, acute phase reactants, such as CRP and imaging studies as indicated clinically. The empiric antibiotic of choice most often is a combination of vancomycin and gentamycin, because the most common organisms in the NICU population are coagulase negative staph and gram negative organisms. But remember that the development of vancomycin resistance is a concern. Alternative empiric antibiotic choice include ampicillin and oxacillin with gentamycin, moreso in the term or late preterm infants or those infants without a central line. Based on your blood cultures and other site cultures, the antibiotic therapy should be modified to pathogen directed therapy. If your patient is clinically doing well and 48 to 72 hour blood cultures are negative, then you should consider stopping antibiotics. For uncomplicated bloodstream infections, the duration of antibiotics is 10 to 14 days. And late onset sepsis with meningitis, the duration is usually 14 to 21 days.

Host: Okay, Dr. Sharma, are we ready to wrap up? Do you have a question for our listeners?

Dr. Sharma: Yes, Dr. Weiner. I do have one question. So, you have a three-week old former 27 weeks gestation infant on noninvasive ventilation who develops worsening apnea with need for increased FIO2 from 28 to 45% and also needing to increase respiratory support. She had been tolerating enteral feeds at 90 per kg per day, and is also on partial parental nutrition via a PICC line. And in the last six hours has had two episodes of emesis. On a blood gas this morning, there is a base deficit of 10, which is actually new. Late onset sepsis is on your differential diagnosis. You go ahead and do a sepsis work up and start antibiotic. What is your empiric antibiotic of choice? A. Ampicillin and gentamycin. B. Choice is marapenem and gentamycin. C. Is oxacillin and gentamycin. D. Is vancomycin and gentamycin and D. Is ceftaz and gentamycin. Dr. Weiner, you want to answer this question?

Host: Yes, Dr. Sharma. So, since I was paying so closely attention to the podcast, I think the answer is D I would've chose vancomycin and gentamycin for the treatment in this three week old, former 27 weeker, who I worried about late onset sepsis, who also has a PICC line in place. Dr. Sharma. What's the answer? And can you explain?

Dr. Sharma: Yes. I think you are true that this, because this is a baby in the NICU, a 27 weeker, has a central line in place and we talked about this, the most common bacteria is coagulase negative staph. So, while I'm concerned about the overuse of vancomycin, I would go ahead and put the patient on vancomycin and gentamycin pending my blood cultures and the patient's clinical status.

Now I do have another question. Now if this same three week old patient was a 40 weeker, that actually came from home with let's say hypothermia and respiratory distress. And we did not know his perinatal history, except mom had said that he went home at two days of life. Then I would consider ampicillin and gentamycin because GBS would be one of the bacteria I would be concerned about and it should also cover my gram negatives. Don't you agree?

Host: Yes. Dr. Sharma. I was just thinking that ampicillin would have been my choice in that baby.

Dr. Sharma: Great.

Host: All right. Thank you, Dr. Sharma, hopefully this information has helped regarding late onset sepsis for the evaluation and treatment. Our next podcast will focus on intra-amniotic infections. This is Neonatology Review Isolette to Crib. I'm Dr. Julie Weiner and thank you for listening to us.