Potential Downstream Harms of Routine Temperature Measurements in Well-Appearing Infants

A 14-day-old patient, who was born at 39 weeks’ gestation by spontaneous vaginal delivery, is brought to the clinic for a routine weight evaluation. Maternal medical history is benign with no pregnancy complications. Prenatal laboratory test results were within normal limits; however, the mother tested positive for group B streptococcus that was adequately treated with penicillin. The mother has no known history of herpes simplex virus (HSV), and there were no vaginal lesions noted at delivery. At this appointment, the patient has a rectal temperature of 96.5°F (35.8°C), appears well, is nursing on demand, and has gained weight appropriately. Due to the recorded hypothermia creating concern for infection, you refer the infant to the emergency department.

In the emergency department, an intravenous line is placed and a complete blood cell count; procalcitonin and C-reactive protein levels; and liver function tests are obtained. Empiric ampicillin, ceftazidime, and acyclovir are started. After multiple attempts at lumbar puncture, cerebrospinal fluid (CSF) is obtained. The fluid is bloody with 20,000 red blood cells/μL. Cultures are obtained from the blood, urine, and CSF, and HSV polymerase chain reaction testing of the skin, eye, mouth, serum, and CSF is performed.

It is unclear how the practice of routine temperature measurements at well-child visits began. Some hypothesize that this practice arose because of the need to detect serious bacterial illnesses before the modern vaccination era.¹ Others suspect that it originated as a way to improve workflow efficiency.¹ However, there is no evidence-based reason for this practice. The American Academy of Pediatrics (AAP) does not make specific recommendations regarding routine temperature measurements at well-child visits.² As a result, there is significant variability regarding this practice, with temperature measured at only one-half of well-child visits.³

Although the AAP has published clinical practice guidelines for the care of febrile patients ages 8 to 60 days, there are no formal guidelines for the care of hypothermic infants, which has led to uncertainty regarding evaluation and treatment.^4–6 The definition of hypothermia also lacks consensus, with some using the World Health Organization threshold of 97.7°F (36.5°C), and others using the International Pediatric Sepsis Consensus Conference threshold of 96.8°F (36.0°C).^7,8

Rationale for Routine Temperature Measurement

Clinicians have been trained that abnormal temperatures could suggest serious infections with high morbidity and mortality rates in infants (eg, HSV, bacteremia and meningitis from group B streptococcus, Escherichia coli, Streptococcus pneumoniae, and Listeria monocytogenes).^4,9–13 But prevalence of most of these diseases is declining because of vaccines as well as screening and treatment of mothers with group B streptococcus or HSV.^8,13,14

Hypothermic infants represent a unique population. Without clear clinical guidelines for the management of hypothermia in neonates and young infants, many clinicians use decision tools for febrile infants that were not designed for hypothermic infants.^15,16 A secondary analysis of 314 infants 90 days or younger with temperatures of 96.8°F (36.0°C) or lower determined that the most common febrile infant decision tools had high sensitivity but low specificity for serious bacterial infections.¹⁶

Although hypothermia can be caused by bacterial or viral infections, there are many noninfectious causes, such as exposure to low environmental temperatures, immature thermoregulation, and decreased fat stores, especially for premature and low-birth-weight infants.^15,17,18 Studies show that serious infection rates in hypothermic infants 90 days or younger range from 2.0% to 8.3%, which is similar to febrile infants.^17–19 However, these studies are limited by single-site data, small sample sizes, and variable temperature cutoffs. These studies also do not reflect the true risk for healthy term infants with incidentally detected hypothermia because they evaluated all infants, including symptomatic and ill-appearing infants.