The Febrile Baby: Introduction

00:01 In this lecture, we're going to discuss the febrile baby.

00:05 If you've done a pediatric rotation and have spent some time in the inpatient setting, you've probably encountered a baby who was febrile who was admitted for what we call a rule out sepsis making sure that there's not a more significant cause of illness.

00:20 I'm going to try and break that down for you here and make it clear so we can understand how and why we manage the babies that way we do.

00:28 A febrile infant is defined as any infant with a fever over 38 °C or 100.4 °F.

00:36 We use to define serious bacterial infection a little bit differently than we do now.

00:41 And that's because we've discovered that urinary tract infections really aren't as bad as we thought they were.

00:46 So, serious bacterial infection is typically defined as meningitis or bacteremia.

00:52 But urinary tract infection is no longer considered an SBI because the outcome of these infants is actually fairly mild.

01:01 The AAP febrile infant guideline breaks down infants into three distinct groups.

01:06 The first group is 8 to 21 days.

01:09 Notice we're skipping 0 to 7, realistically those infants just go home from the newborn nursery or they may come back to the newborn nursery if they're febrile that quickly after birth.

01:19 Then there's the 22 to 28 day group.

01:22 And then there's the 29 to 60 day group.

01:24 And these are treated a little bit differently.

01:27 Let me go through them one at a time.

01:29 For infants in the first three weeks of life from 8 to 21 days, all these infants should get a urinalysis, a urine culture.

01:37 Keep in mind that 10% of infants in this group will have a urinary tract infection.

01:43 We will obtain that by catheter or suprapubic tap rather than the clean catch.

01:47 And we'll use a very strict cut off 10,000 CFU as defined as a urinary tract infection.

01:55 Note that's a little bit different than two months to two years of age where we typically use 50,000 CFU as a cutoff.

02:04 We should get a blood culture.

02:06 Keep in mind that between about 4 and 5% of infants in this group have bacteremia.

02:11 Now 20% of the infants with UTI have bacteremia.

02:14 So a bunch of them are from UTI.

02:18 But also keep in mind that between 60 and 90% of these kids have fake positive cultures.

02:25 These are contaminants.

02:26 So, when you get a blood culture, chances are it's a contaminant.

02:31 If it's not a contaminant, it's probably from a UTI.

02:34 But a reasonable percentage around 1 or 2%, it will be true bacteremia absent of UTI so we have to look for this.

02:42 Also in the first three weeks of life, we really should get a spinal tap.

02:45 And when we send the spinal tap fluid for testing, we're going to get a white count, a protein, a glucose, a gram stain and a culture.

02:54 We're also going to do an enterovirus PCR.

02:57 Because during enterovirus season, if that's positive, it can really simplify care.

03:02 If there's a pleocytosis or if this infant is at high risk for HSV, which we will talk about in another lecture.

03:08 We will also send the HSV PCR.

03:11 Some centres are sending these children's CSF for something called a multiplex PCR.

03:16 This is a panel that looks for lots of bacteria that might be in the child's spinal fluid.

03:22 The problem is, we are really instructed by the AAP, we should not use this test.

03:27 And the reason is, is because there are false negatives and false positives.

03:31 And also listeria isn't often on the panel and listeria can cause meningitis in infants.

03:37 Additionally, one thing that some people are taught is that, hey, if you get a bloody tap, you can correct for the amount of blood to estimate the number of white cells.

03:45 And there's various formulas 400 to 1, you do the math.

03:49 In fact, we should not correct for blood.

03:51 And the reason is, is because if you do that you will actually miss some infants with pleocytosis and bacterial meningitis.

03:59 So for these kids, up to the first three weeks of life, we're going to admit them to the hospital.

04:04 We're going to start ampicillin gentamicin.

04:07 Unless the CSF is consistent with bacterial infection, then we would start Ceftazidime and Gentamicin.

04:12 We do not need to be doing antibiotics if the enterovirus PCR is positive.

04:17 Enterovirus PCR is really interesting test and it's really beyond the scope of this lecture to get why.

04:22 But the false positive rate of this test is essentially zero.

04:26 So if you see enterovirus, that's what it is, you can stop the antibiotics.

04:30 You might not discharge the kid if they're very sick, but you can certainly stop the antibiotics.

04:35 Then you will discharge the patient home after the cultures are negative for 24 to 36 hours and the baby's looking better.

04:43 Now, it's a little bit different for this fourth week of life and I want to sort of get into why because I think it's pretty interesting.

04:50 So you will still get a urinalysis, urine culture both again by catheter or suprapubic tap And if you want, you can get a bag urine and if that's negative, skip the cap or the suprapubic tap.

05:04 However, it's positive, then you do have to repeat the culture.

05:07 And the reason is, is that bag urines have a very high false positive rate for bacterial culture.

05:13 Then you'll get a blood culture.

05:14 Keeping in mind that still a fairly substantial number of these infants have bacteremia.

05:19 And also remember that the huge number who have UTI, about 7.5 to 10% of them have bacteremia.

05:26 So that's fairly common.

05:28 Here's the kicker.

05:30 You also should assess an inflammatory marker in this child.

05:33 And what many have been taught is that the white count is an accurate inflammatory marker.

05:39 And that is in fact, untrue.

05:42 The WBC count for infants does not reflect whether they're likely to have bacterial versus viral infection.

05:49 The abnormal inflammatory markers you can include are the ANC, which if it's over 4000 is abnormal.

05:57 Keep in mind, that's the white count times the percentage of bands plus segs.

06:01 So if you have 50% segs and a white count of 10, that's 5000.

06:06 And that would be abnormal.

06:08 That's not as good as a marker as the CRP or the procalcitonin.

06:13 For CRP, we'll use an abnormal number of greater than 20.

06:17 And for procalcitonin, we'll use an abnormal number of 0.5 ng/mL.

06:23 So let me explain why white count is not accurate.

06:26 Because there's a lot of students out there who are still being told white count is the way to go.

06:30 And it really isn't.

06:32 And to get it that you have to use something called an ROC curve.

06:35 This is a little bit off topic.

06:36 But I think, it's incredibly important for us to understand what an ROC curve is.

06:40 We use an ROC curve to figure out what that number should be.

06:44 If you recall, we said what the CRP should be more than 20 to be abnormal.

06:50 Well, where do we get that number 20? Well, what we do is we look at the test and we say, what is the true positive rate and what is the false positive rate? Remember, true positive rate is the number of positive tests in sick people.

07:02 And the false positive rate is number of positive tests in the healthy people.

07:06 It makes sense.

07:07 And we're just going to graph it here.

07:09 So let's say these are a bunch of children who are getting a white count, and the blue ones do not have disease and the red ones are sick.

07:17 And these are what their white counts are.

07:19 Well, what should we use as a cut off? Let's pretend I said, any white count over one is abnormal.

07:27 Well, every single patient has an abnormal white count.

07:29 So the true positive rate is 100%.

07:33 But also the false positive rate is 100%.

07:36 So we'll plot that point right there in the top right corner.

07:39 What if I said 20 was abnormal, then nobody on this panel of patients has an abnormal and therefore the true positive rate is zero and the false positive rate is zero plotted right where that 20 is.

07:52 But what if I used a cut off of 10? What you can see is that about 3 out of 7 of the babies would be abnormal, who are healthy.

08:02 And unfortunately, about 4 of the 7 babies who have the disease would be abnormal.

08:10 And so this is not a particularly accurate test.

08:14 And you see sort of like a straight line on that graph.

08:18 But let's look see what a good test look like, let's pretend we're looking at procalcitonin.

08:24 And we say, okay, well, of our healthy babies, these were the numbers they had ranging from between 0.2 and 0.6.

08:31 And of our sick babies, it's between somewhere between 0.4 and 1.2.

08:35 Well, if we set an abnormal test was 0.1, everyone would be positive.

08:41 if we set an abnormal test was 2, everyone would be negative.

08:45 But what if we used an abnormal cutoff of 0.5? Well, that's kind of cool.

08:51 You could see that only 1 child of the 7 would have a false positive, and only 1 child in the sick group would have a false negative.

09:02 So we plot that you can see that 0.5 number is a little bit closer the top left corner of the graph.

09:09 So what you want to see in a test is a curve where it bends up the top left, and then we look at the area under the curve.

09:16 And remember, the white count was a straight line.

09:18 So that was half of the graph and half is useless.

09:22 One would be a perfect test.

09:24 So the closer the area under the curve is to one, the better the test is.

09:28 Well, here's the data for children.

09:30 And what you can see is the procalcitonin gets very close to one, the CRP and the ANC are a little bit less, and then the white count is absolutely useless.

09:40 So this is why we do not recommend a white count in children who have fever who are infants when we're testing them for bacterial disease.

09:48 Let's get back to the guideline.

09:50 In the fourth week of life, the CSF may be obtained if the inflammatory markers are negative.

09:56 You've got the urine in the blood.

09:58 The child is hospitalized and the UA result doesn't affect this recommendation.

10:02 In other words, they have a UTI or they don't, kids in the hospital inflammatory markers are normal, and you don't have to get an LP.

10:11 You should get an LP however, if there's an elevated inflammatory marker, so that's why we're getting that marker.

10:17 And specifically, the white count isn't one of them, you really should go a CRP or procalcitonin if you can get it at your institution.

10:25 Let's go through some examples.

10:26 So if we had a 26-day old well appearing infant in the ER with fever and upper respiratory infection, not really bronchiolitis.

10:34 His temp is 100.7. His CRP is 2 mg/L.

10:37 That's really low. Remember, 20 is our cutoff.

10:40 You have a urine and a blood culture pending, therefore, you can skip the LP, but you should do it if the parent or you are still otherwise concerned.

10:48 You were like, "Maybe that was a seizure when he was shaking, I should just double check. I think I'll do the LP." That's fine.

10:55 But example B is a 27-day old well appearing infant in the ER with fever and nonbloody diarrhea.

11:01 His temp is 100.6. But as CRP is high at 24 mg/L, you should do the LP.

11:06 So I hope you get the difference between how we think about those two circumstances.

11:11 So in that four with fourth week of life, we are going to hospitalize them if the CSS CSF testing is abnormal, and then we're going to treat either with ceftriaxone, not ceftazidime, which is what we're using at the little infants because that is subtraction is contraindicated in jaundice babies, or you can do ceftazidime and vancomycin to cover resistant strep pneumo.

11:34 You're also going to hospitalized that the CSF is normal, the UA is normal, but the CRP is abnormal, and you put them on ceftriaxone.

11:41 If all the labs are normal, and somebody is worried parent, doctor or nurse, whoever, then sure, go ahead and observe them.

11:48 You can either observe them on or off ceftriaxone.

11:51 And if all labs are normal, and the parents want to go home, you can send them home.

11:56 So this is an interesting guideline because it's allowing parents to help make a decision which I kind of like.

12:01 Now what about 4 to 8 weeks, this is a little bit simpler.

12:04 You're gonna get a UA and a urine culture.

12:06 You're going to send a blood culture.

12:08 You're going to check inflammatory markers.

12:10 If it's high, consider an LP but you don't have to do it.

12:13 If it's normal, consider discharging home but you don't have to do it.

12:17 And you may treat a well appearing infant with a Urinary Tract Infection with oral antibiotics and send them home.

12:23 You do not have to hospitalized an infant with a urinary tract infection at this age.