Febrile Infant

Management of the Febrile Infant Janet Wong, M.D.

1

Most Common Non-serious Febrile Infections Fever is defined as a significant elevation in your body temperature while in a neutral thermic environment, which is mediated by

€

Respiratory viral infections (Fall, Winter, Spring)

€

Enteroviral infections (Summer)

differentiate an elevated body temperature in an unusually hot

€

Roseola infantum

environment from a true fever. We see them in the hot weather

change in your hypothalamic temperature, it is important to

from the sun, and we see it in a normal newborn nursery, in children that are in radiant warmth.

We see children, occasionally, with severe cerebral palsy, who will have temperatures of 106, 107, 108 degrees, after their surgery or other procedures for maintenance of a child with severe CNS problems. The temperature is elevated because their hypothalamic mechanisms are not perfectly normal, so they tend to have body temperatures which are higher than normal.

Fevers in general are mediated by the hypothalamic set point, within a range which usually does not exceed about 106 degrees, and area manifestation of disease.

Babies that are less than two years of age but are out of the newborn period, are relatively often seen with positive cultures. Those babies that have temperatures above 39.4 or 39.5 have a much higher risk of having a positive blood culture. Those babies that had a white count with significant elevation, also had a positive blood culture, and the combination of age and elevated temperature, and elevated white count, is actually a very small sub-group of febrile children. There was only 29 of them that have the combination of age, the white count, and temperature, and those babies have a 28% prevalence of positive blood cultures.

We all worry about the patients that have meningitis. In this case of meningococcus, the problem that we have is that when we see children in the office or the emergency room setting, we do not want to fail and to miss the diagnosis. And the real issue about all of this is how far you go with the child that comes in to see you in order to try to be sure that they do not have one of these more terrible situations, because one viewpoint about the whole thing is that if you have a child that you miss this problem in, maybe you will not have an opportunity to get a second opinion for their rapidly developing disease.

Many of the children that we see that have these problems are not septic, and they do not look terrible, and, in this particular case.

First of all, most children with fever have non-serious illnesses, and most of those fevers, as we are told, are viral. In the fall and spring they are mostly respiratory viral. In the summer they are mostly enteroviral, although we see periods of type III virus infections in the summer as a respiratory viral infection. And then we have, throughout the year, roseola noted virals with typical occult febrile presentation. If you did viral cultures of the respiratory tract of every kid that ever came into your office with febrile illness, I guarantee you would not find more than 10% or 20% that had a positive viral

2

Organisms Isolated from Blood in Outpatients with Bacteremia

culture.

How many times do you find a child with high fever that really does have viremia? The answer to that is, “That’s very rare.” Here is a study which guided viral cultures of blood in every febrile child which came through a setting with very high fevers and actually hardly any of them had positive viral cultures from the blood. They

1994 (%)

had two enteroviral infections that were in the blood. Most respiratory infections are mucosal so you do not see viral as often.

S. pneumoniae

85%

H. Influenzae b

4%

The other common cause of febrile illness in the office practice is

N. meningitis

1%

otitis media.

Salmonella

4%

Other

5%

3

Identification of an "Unstable" Child What is really important in managing febrile babies is identifying a

• Airway patency • Breathing

• Circulation

child that is seriously ill. Rapid cardiopulmonary assessment involves assessment of the patency, vital signs, including chest

Heart rate

movement, respiratory rate, and presence or absence of cyanosis.

Rate

Blood pressure

And a circulatory evaluation which includes pulse, blood pressure,

Air Entry

Peripheral pulses

Mechanics

Skin perfusion

pulses; or perfusion is, how quickly turns to normal after you press,

Color

CNS perfusion

that will tell me that then I ask what is the blood capillary refill time,

peripheral pulses, that is you palpate them, say “Yes, I can feel the dorsalis pedis, radial pulse; Yes, I can feel the posterior tibial

what are the peripheral pulses, what is the ultrastat.

4

Identification of a “Toxic” Child Yale Acute Illness Observation Scales (AIOS)

The other way of differentiating the sick child from the not-so-sick child is covered by the term “toxic”. What is a toxic child? If they are toxic, that is bad. They go to the hospital, the blood culture,

€ Quality of cry

temperature, urine, IV antibiotics, the whole works. For anybody that is toxic, that is what you are supposed to do. If the kid is non-

€ Reaction to Parent Stimulation

toxic, then you go through this little algorithm of white count is this,

€ State Variation

temperature is that, urine culture is this, stool is that, and then the

€ Color

following things.

€ Hydration € Response to Social Overtures € Neurologic State of Consciousness

5

Yale AIOS: Percentage of Serious Illness This system of determining whether a child is toxic or not. That was the one by Yale. Which is basically a 34 point scoring system for how sick a child is. It is based on 1, 3, or 5 points for each of the

Score

% "Serious Illness"

quality of child’s cry, the reaction to stimulation from the parent, the

< 10

3%

seek variation, which is basically the child’s neurologic state of

11-15

26%

>16

92%

consciousness, their color, their hydration, and the response to social overtures, and the quality of the cry, ranging from either there is no cry or a lusty, hearty cry which is a good score, that is a 1, to a whimper or no vocalization at all which would be a 5. Reaction

McCarthy et al: Pediatrics 70:802, 1982

to parental stimulation means that patient can be calmed down or comforted by a parent. The degree of state variation: is the child awake, is the child immediately awake, is the child out of it, is the child aware of the surroundings. That is also scored from 1 to 5. Color: Pink to blue or pink to ashen would be a score of 5, and pink would be a score of 1. Hydration: The response to social overtures: Social overtures turned out to be the most important quality predictors of serious problems, you have to learn how to elicit a smile from a little baby. This is an important element in determining whether the child has a serious illness or not. Any response is very important.

6

Association Between Hyperpyrexia and Bacteremia in Febrile Children <24 Months of

If you are a private pediatrician seeing patients, the number of the

Age

disease based on these kinds of things, is relatively small. And out

patients that come into your office with the appearance of serious

of these 369, maybe 10 children had such an episode in period of two years. The private pediatrician does not really see a very high

Temperature

% with Bacteremia

104-104.90

7%

105-105.90

13%

0

>106

26%

prevalence of these types of situations in his practice, at least according to the data from some of you who are in practice.

When you get down to URIs and otitis media, they are relatively few in our population with an overall number of about 3%. So people try to focus on what the bugs are and what their importance is to the child and how you can identify a child that is at higher risk than the

McCarthy el: al: Am J Dis Child 130:849,1976

overall average to have that disease. The majority of bacteremic episodes are caused by S. pneumoniae. The prevalence of hemophilus influenzae type B as a cause factor in bacteremic disease is going down. In 94 and 97, it was down pretty close to zero. I think the last kid I saw with bacterial meningitis was about four years ago and that was a 3 month old, not quite old enough to get the h. flu vaccine, but h. flu is disappearing.

So we are really increasingly talking now about pneumococcal bacteremia. We know that the higher the fever, the higher the probability of bacteremic disease and below 104 temperature Fahrenheit, the figure is probably 3% overall, between 102 and 104 it is probably 3 or 4%, but it is not in this particular study. This study was for extremely high temperatures, and what they found was that the higher the temperature, the higher the probability of bacteremic disease. It was thought for a while that if you had bacteremic disease or severe bacterial disease, that you would not respond as well to Tylenol therapy, but this has been studied, and it turns out that children that are bacteremic and children that are highly febrile but not bacteremic care essentially identical.

7

Association Between Leukocytosis and Bacteremia

The higher the white count, the higher the probability that the patient will have bacteremic diseases. Above 20,000, you are really getting into a very high risk category. But between 15000 and 20000 it is lower. This negative predicted value aspect is probably

WBC (per mm3)

% Bacteremia

<10,000

0%

your white count, what your criteria are, you kind of never get the

10,000-14,999

6%

odds of about 25%, and that is the maximum percent that will be

15,000'19,999

9%

>20,000

26%

as important as the positive predicted value, but no matter what

bacteremic children, otherwise they are not toxic and will treat well in the clinic. 25% of bacteremic children will have disease. But those are the patients that have the high probability of complicating bacterial disease.

Bass et al: Ped Infect Dis J 12:466, 1993

8

Bacteremia and Subsequent Complications by Organism

When we talk about complications, we are interested in the child that goes along to develop meningitis or goes along to develop septic shock. The child then goes on to develop major focal

% with Complications

bacterial infections like pneumonia from hemophilus influenzae, septic arthritis, cellulitis. So what we conclude from this is,

S. pneumoniae

6%

remember these are untreated children. What we conclude from

Hib

31%

this is that 94% of children have Streptococcus pneumoniae in their

N. meningitidis

60%

blood recover from it without being treated. I find that many people feel that if they do a white count as part of a work-up of a child, they are either going to have to send blood to the microbiology lab for blood culture or they are going to have to do the CBC and come back later to do the blood culture. My favorite method for both the CBC and a blood culture is to draw blood for a CBC and then draw blood for a blood culture into a citrate sterile tube, and if your CBC gives an indications that you may have a problem with bacteremia, then submit your yellow top tube to blood culture, ascertaining that the child has a probability of a significant bacteremia.

9

Rochester Low-risk Criteria for Serious Bacterial Illness in Febrile Infants <3 Months

All of the things that I have been talking to you and the data that I have presented to you are focused on babies between the ages of 90 days and year 2 or 3 of age. What elevates, and little babies are an area that has recently kind of come to be incorporated into the

€

Previously healthy, "non-toxic", AND

€

No clinical evidence of focal infection (ears, skin, bones, joints, lungs)

to do with children, either managing them as out patients or

€

WBC 5,000-15,000/mm3

inpatients and treating them or not treating them.

thinking about looking for serious disease and trying to decide what

€

Absolute band count <1500/mm

€

Normal UA

for management of febrile babies as published in Pediatrics 1993.

€

Fecal leukocytes <5/hpf (if diarrhea)

Basically any child that is “toxic”, everybody agrees that they are

3

What I want to do to conclude is to just go over the two algorithms

admitted to the hospital, complete sepsis work-up, parenteral antibiotics. If the child does not appear toxic, you can enter a very complex management scheme. If the temperature is less than 39 degrees, that is considered to be a good indicator that the child has a “viral illness”, and can be managed with Tylenol and encouragement. With temperatures higher, then you get into how you should work up such a child, and urine cultures are recommended. Urinary tract infections are very common causes of occult bacteria infection in this category of children. Urine cultures are probably going to count for 10% of these children if you go looking for urinary tract infections. If the child has diarrhea, you watch it, proceed with stool cultures. If there is respiratory distress, get a chest x-ray. Recommendation that you could do blood cultures in children with temperatures above 39 regardless of white count or if you were a cost conscious person and kind, you could consider blood cultures only on children with elevated white counts. Tylenol is encouraged, note the big dose, 10 per kilo is the average dose, 15 kilo is the max empirical dose. Then follow up, and follow-up as needed.

10

Rochester Low-risk Criteria Percentage of Serious Illness

Then the last one is the algorithm aforementioned previously healthy infant of less than 90 days of age, which is bigger too. There again, the issue that is paramount in deciding what to do with children is whether they are toxic or non-toxic. If they are less

% Serious

than a month of age, it is hard to determine whether they are toxic,

Category

Bacterial Illness

% Bacteremia

and so, I think what you do is, I think there is pretty good reading on

Low-Risk

1%

0%

this, although what I hear now is that some people feel that the line

High-Risk

25%

10%

of demarcation should not be 28 days, but should be 7 days. But, at any rate, the leading recommendation is that 28 days. If the child is younger than that with a fever of unknown source, regardless of

Dagan et al: J Pediatr 107:US, l985

any lab work that you do, regardless of family history of viral illness or febrile illness, probably get the kid to the hospital and manage both diagnostically and therapeutically aggressively.

11

Metananalysis of Rochester Low-risk Criteria for Serious Bacterial Illness in Febrile Infants

The baby that is older than 28 days but younger than 90 days, there

<3 Months

have to decide whether the child is a high risk infant or a low risk

is room for some difference of opinion on that, but basically you

infant. Any child that is a low risk infant running towards a high risk infant, goes into the category that is the same as the other baby

Category

% Bacteremia

Low-Risk

1%

1%

0.5%

previously healthy, non-toxic by clinical appearance, normal

High-Risk (Nontoxic)

9%

2%

1%

physical exam, has white counts between 5000 and 15000, the

17%

11%

4%

High-Risk (Toxic)

% Meningitis

and is admitted to the hospital for aggressive treatments. So if we

% SBI

are only talking about the non-toxic 28-90 day old baby who is

band count is less than 1500, has a normal UA and has a fewer than 5 white cells in their stool without diarrhea, there is general agreement they can be managed as outpatients.

12