Pediatric fever guidelines are woefully out of date, leaving emergency physicians with too much uncertainty when facing a febrile infant.
It’s night shift and the patient flow has been steady. You sneak off to the lounge to grab a quick bite to eat and when you come back, you have not one, but two 6-week-olds with a chief complaint of fever. No problem, right? Aren’t there protocols for that? Well, yes and no.
Up until the 1980s, you didn’t really have to think, just do. The accepted standard of the day was a full sepsis work-up and admission for febrile infants three months of age or younger. However, the practice was increasingly called into question. Research came out describing the iatrogenic risks and costs of this practice. Studies documented a lack of adherence to this plan in pediatric offices and academic centers. So in the 1990s, researchers in the emergency departments of the University of Rochester, Boston Children’s Hospital and the Children’s Hospital of Philadelphia came out with decision strategies to help determine the appropriate management of febrile infants. The Rochester criteria, Boston criteria, and Philadelphia criteria all provided rubrics by which a clinician could determine which infants were at high risk of serious occult bacterial infection and which were not. Baraff and associates performed a meta-analysis and developed a practice guideline for the management of febrile infants without a source of fever by history and physical exam. All of these protocols were meant to help a physician treating children in the emergency department decide who needed to be admitted and who could go home, with or without antibiotics.
Most infants who have a fever, are not septic-appearing, and have no source for the fever identified by history or exam have viral illnesses. But which of them do not?
The guidelines are supposed to help us answer this question. But there is a problem. A recent survey of pediatric emergency departments that have PEM fellowship programs found that only about half have any policy about the management of febrile infants. Of those that do have policies, 80% of those policies differ from the guidelines mentioned above. In addition, 85% of PEM program directors feel that new guidelines are needed. When a guideline exists but the majority of clinicians don’t stick to it, is it still a guideline?
Why don’t clinicians want to adhere to the guidelines? Because of a marked change in the prevalence of serious bacterial infections in the well-appearing baby. The hidden infections that everyone was looking for were UTIs, bacteremia and meningitis. While the phenomenon of hidden pneumonia has been described, studies have shown that chest X-rays are not likely to be helpful in infants that have no respiratory signs or symptoms. Other infections, such as those involving the skin and soft tissues, are not hidden, they are out there to be found by a careful physical exam. Leaving aside UTIs for the moment, what has happened to occult bacteremia and meningitis since the guidelines came out in the 1990s?
The arrival of the conjugate vaccines for Haemophilus influenza (1988) and Streptococcus pneumoniae (2000) markedly decreased the incidence of bacteremia and meningitis in infants. The expanded vaccine for Streptococcus pneumoniae introduced in 2010 is expected to continue this trend. In the mean time, studies have also documented a low risk of bacteremia in infants with RSV and influenza. So it doesn’t make sense to stratify risk of bacteremia in infants based on outdated prevalence data, which is what the 1990s protocols do. Currently the likelihood of a falsely positive blood culture substantially outweighs the likelihood of a truly positive culture.
So back to our babies. What would you do?
Baby number one is a term infant with a temperature this evening at home of 38.3 rectally. She has been a healthy baby thus far, with no significant problems around her birth or nursery course, and is feeding well and gaining weight. Her two year-old sister has had several days of cough and runny nose and low grade fevers. This morning the patient developed a runny nose and slight cough and she felt warm this evening, so Mom took her temperature. She doesn’t have a fever at the present time. Her exam is normal and she looks great. So what do you want to do?
A reasonably conservative approach is to get a CBC and blood culture in this infant. After all, she has not received any of the conjugate vaccines yet although she should benefit from herd immunity. Since she has a runny nose, you can send it for RSV. You could also send it for flu but you haven’t seen a case of the flu in a good while and you know that flu tests have a high enough rate of false positives that they have poor predictive value when the prevalence of disease is low. Studies have shown that infants with viral illnesses have a very low rate of bacteremia and meningitis but can still have UTIs, so you send urine for UA and gram stain. And because this infant has respiratory symptoms (cough and coryza) you get a chest X-ray as well.
Are you going to do a lumbar puncture? No one would fault you if you did, but this infant most likely has a viral infection. You decide to hold off. If the baby has a high white count, pneumonia or a UTI, you plan to start antibiotics. At that point, an LP would be a good idea so that you don’t mask evolving meningitis if the child comes back sicker and you need to re-evaluate her.
The results come back and there is nothing alarming. Her urine is clean, her chest film is negative, she doesn’t have RSV, her white count is about 11,000 with no left shift. She continues to look great. Mom is reliable, has transportation to return if needed. You send the baby home from the hospital with no LP and no antibiotics. The plan is for her to follow-up the next day with her pediatrician.
Baby number two is a little boy. He is also a term infant, also low risk. He felt warm this evening so mom brought him in. At triage, his temperature was 39.2. He received Tylenol from the triage nurse and by the time you see him, he looks good, is nursing avidly. He was a little fussy while he was hot, but has had no other symptoms. His exam is normal, no source for his fever.
Again, there are options as to how to work him up. He is different from the other infant in that he has no URI symptoms and his fever is higher. You decide to get a CBC and blood culture and urine studies. You forego the swab for RSV and the chest X-ray because of the complete lack of respiratory symptoms. Then you observe the baby and wait for the results.
The baby continues to look good in the emergency department but the urine comes back looking infected. You are going to have to treat this baby with antibiotics. This is a potential game changer. Meningitis can be very tricky to detect in infants at six weeks of age. Giving this patient antibiotics could mask the subtle signs of early meningitis or risk creating a situation where the baby has a tap later on but is “partially treated” and has a CSF pleocytosis that is difficult to interpret. Preferring not to take these risks, you perform an LP. The spinal fluid looks clean. You admit this infant for IV antibiotics for presumed pyelonephritis.
Two 6 week olds, two fevers, two different management strategies. Would your colleague manage them the same way? Maybe, maybe not. Your institution does not have a fever protocol and the published guidelines seem out of date.
Is there someone out there working on a new guideline?
- Dagan R, Powell KR, Hall CB, Menegus MA. Identification of infants unlikely to have serious bacterial infection although hospitalized for suspected sepsis. J. Pediatr 1985;107:855-860.
- DeAngelis C, Joffe A, Wilson M. Iatrogenic risks and financial costs of hospitalizing febrile infants. Am J Dis Child 1983;137:1146-1149.
- Baraff LJ, Bass JW, Fleisher, GR, Klein JO, McCracken GH, Powell KR, Schriger DL. Practice guideline for the management of infants and children 0 to 36 months of age with fever without source. Ann Emerg Med 1993;22:1198-1210.
- Ferguson CC, Roosevelt G, Bajaj L. Practice patterns of pediatric emergency physicians caring for young febrile infants. Clin Pediatr 2010;49(4);350-354.
- Schneider C, Blumberg S, Crain EF. A survey of the management of febrile infants in pediatric emergency departments. Pediatr Emerg Care 2012;28(10);1022-1026.
- Jhaveri R, Byington CL, Klein JO, Shapiro ED. Management of the non-toxic-appearing acutely febrile child: a 21st century approach. J. Pediatr 2011;159:181-185.
- Krief WI, Levine DA, Platt SL, Macias CG, Dayan PS, Zorc JJ, Feffermann N, Kupperman N. Influenza virus infection and the risk of serious bacterial infections in young febrile infants. Pediatr 2009;124(1):30-39.
- Byington CL, Enriquez R, Hoff C, Tuohy R, Taggart EW, Hillyard DR, CarrollKC, Christenson JC. Serious bacterial infections in febrile infants 1 to 90 days old with and without viral infections. Pediatr 2004;113:1662-1666.
Neither of these infants management varies from current accepted practice. The most common pathogens still are GBS and E. coli sepsis, Late onset GBS is still a significant illness and its incidence has not changed over the past several decades. The management here is really not controversial. Again the most important consideration is reliability of the parents and follow up.
And how many babies with identical stories and findings stayed at home and did just fine?
I would assume the “UTI baby” would not do fine if they stayed at home.
I am a junior pediatric resident and I also struggle with this dilemma in the 4-8wk age group. I completely agree with your treatment strategies and think it makes complete sense. Being a junior however poses difficulties in implementing such an approach when common practice in a department is to “be safe” and local area guidelines don’t recommend such an approach. As low as the risk might be, if the kid comes back with meningitis, can I justify not performing the LP. Probably not given they are probably as harmful as the 3 attempts at an IVC i just had.