The serious implications of high fever in infants during their first three months. Six years' experience at Yale-New Haven Hospital Emergency Room

Hyperpyrexia and high fever as a predictor for serious bacterial infection (SBI) in children-a systematic review

It is not clear if children with high fever are at increased risk for serious bacterial infection (SBI). Our aim was to systematically review if children suffering from high fever are at high risk for SBI. Our data sources were Embase, Medline, and Pubmed; from their inception until the last week of March 2017. The study selection were of cohort and case control studies comparing the incidence of SBI in children with hyperpyrexia with children with fever of 41 °C or less, and children with a temperature higher than 40 °C, with children with fever of 40 °C or less. Two reviewers independently pooled studies for detailed review using a structured data-collection form. We calculated the odds ratio and 95% confidence intervals (CI) for SBI, assuming a random-effects model. A sub-group analysis was conducted. In our results, 11 studies met the inclusion criteria. Two studies showed that children with hyperpyrexia are at higher risk for SBI (OR 1.96 95% CI 1.3-1.97). An increased risk for SBI in children with high fever (OR 3.21 95% CI 1.67; 6.22). SBI in infants with temperature over 40 °C was higher compared to infants with lower degree of fever (OR 6.3 95% CI 4.44; 8.95). On older children, the risk for SBI was only slightly higher in children with fever above 40 °C. The limitation of the study is the small amount of studies and that the heterogeneity of the studies was very high.

CONCLUSION

Young infants with temperature higher than 400 °C are at increased risk for SBI. Risk of SBI in older children with temperature > 400C is minimal. What is known: • An association between high fever and increased risk for SBI was reported in young infants. • Based on only two studies from the 1970s and 1980s, hyperpyrexia is associated with increased risk for SBI. What is new: • Infants under the age of 3 months with fever > 40 °C were found to have increased risk for SBI. • Risk of SBI in older children with temperature > 40 °C is minimal.

[Children less than 3 months hospitalised due to acute febrile syndrome. 5 years clinical experience]

INTRODUCTION

Acute fever of unknown origin (AFUO) is established when the anamnesis and physical examination cannot identify the cause. In infants less than 3 months-old this is situation for concern, due to the risk of a serious bacterial infection.

OBJECTIVE

To describe the clinical and laboratory variable of patients with AFUO, in order to look for clues in order to base studies on the decisions arising drom this problem.

PATIENTS AND METHODS

A report is presented on a retrospective study conducted on a cohort of children less than three months-old admitted to the Hospital Roberto del Río (2007-2011) due to an AFUO. Clinical histories were reviewed and the patients were grouped, according to the severity of the admission diagnosis, into severe and non-severe. They were compared in strata determined by the variables of clinical interest.

RESULTADOS

A total of 550 children were admitted with AFUO during the study period. There was low agreement between the severity on admission and at discharge (kappa=0.079; P=.26). There were 23.8% of children in the severe group and 76.2% in the non-severe group. Urinary tract infection predominated in the severe group (68.7%) and 40.7% with acute febrile syndrome in the non-severe group. The cut-off levels for C-reactive protein, white cells, and neutrophils per mm(3), to calculate the fixed and variable indices, only showed negative predictive values of some use for ruling out serious bacterial infection. The ROC curves with white cell and neutrophil counts and C-reactive protein, did not provide andy fixed indices of clinical use. More than one-third (34.6%) of lumbar punctures were traumatic or failures.

CONCLUSIONS

According to the results of this study, there is an obvious excess of hospital admissions, little usefulness in the examinations to identify serious bacterial infection, a high percentage lumbar punctures traumatic and lumbar punctures failures, and an excess of antibiotic treatments. A review of clinical criteria and procedures is needed.

Signs and symptoms in children with a serious infection: a qualitative study

Background

Early diagnosis of serious infections in children is difficult in general practice, as incidence is low, patients present themselves at an early stage of the disease and diagnostic tools are limited to signs and symptoms from observation, clinical history and physical examination. Little is known which signs and symptoms are important in general practice. With this qualitative study, we aimed to identify possible new important diagnostic variables.

Methods

Semi-structured interviews with parents and physicians of children with a serious infection. We investigated all signs and symptoms that were related to or preceded the diagnosis. The analysis was done according to the grounded theory approach. Participants were recruited in general practice and at the hospital.

Results

18 children who were hospitalised because of a serious infection were included. On average, parents and paediatricians were interviewed 3 days after admittance of the child to hospital, general practitioners between 5 and 8 days after the initial contact.

The most prominent diagnostic signs in seriously ill children were changed behaviour, crying characteristics and the parents' opinion. Children either behaved drowsy or irritable and cried differently, either moaning or an inconsolable, loud crying. The parents found this illness different from previous illnesses, because of the seriousness or duration of the symptoms, or the occurrence of a critical incident. Classical signs, like high fever, petechiae or abnormalities at auscultation were helpful for the diagnosis when they were present, but not helpful when they were absent.

Conclusion

behavioural signs and symptoms were very prominent in children with a serious infection. They will be further assessed for diagnostic accuracy in a subsequent, quantitative diagnostic study.

Hyperpyrexia among infants younger than 3 months

OBJECTIVE

To determine the prevalence of serious bacterial infection in infants younger than 3 months with fever > or =40 degrees C.

METHODS

We retrospectively identified all infants younger than 3 months with fever who presented to a pediatric emergency department. The medical records were reviewed. The prevalence of serious bacterial infection (SBI) among those patients with hyperpyrexia was compared with febrile infants with lower fever.

RESULTS

5279 infants younger than 3 months with fever were reviewed. Ninety-eight patients (1.7%) had triage temperature > or =40 degrees C rectally. Median age, temperature, and white blood count for those with hyperpyrexia were 58 days (interquartile range [IQR] 36-78 days), 40.2 degrees C (IQR, 40.0-40.4 degrees C), and 10,800/mm3 (IQR, 7900-14,600/mm3), respectively. Diagnostic studies included blood culture (100%), urine culture (100%), lumbar puncture (100%), chest radiographs (34%), and stool cultures (11%). SBI was found in 38% infants with hyperpyrexia: urinary tract infection was the most common SBI (71%). Among patients with hyperpyrexia, patients with SBI had similar mean white blood cell counts (14,000 vs. 10,200 cells/mm3) and age (54 vs. 53 days) as those with hyperpyrexia but no SBI. The prevalence of SBI among febrile infants with temperatures > or =40.0 degrees C was 38% (95% CI 27-48%) compared with those with fever < or =40 degrees C: 8.8% (95% CI 8.1-9.5%).

CONCLUSION

Hyperpyrexia is rare among febrile infants younger than 3 months. One-third of infants with temperature > or =40.0 degrees C had SBI. Future management algorithms might include hyperpyrexia as a risk factor for serious infection.

Predictors of physician compliance with a published guideline on management of febrile infants

BACKGROUND

Previous studies have demonstrated clinicians' poor compliance with published management strategies and protocols, but the reasons why physicians often choose to vary their management of the febrile infant from published guidelines are poorly understood.

OBJECTIVE

We conducted a study of physicians to learn more about the issues that influence their decisions in the management of febrile infants.

METHODS

A survey study of pediatricians, emergency physicians and family physicians randomly selected from a list of licensed physicians in the United States. Chi square and Kruskal-Wallis tests were used to measure differences in responses by specialty. Odds ratios from logistic regression were used to measure differences in compliance with a recently published guideline.

RESULTS

We received 193 completed surveys from pediatricians, 177 from emergency physicians and 104 from family physicians. After controlling for other variables, odds for compliance with a recently published guideline were higher for pediatricians [odds ratio (OR) = 9.13] and emergency physicians (OR = 2.5) than for family physicians (P < 0.001). Factors associated with decreased odds of compliance included more years since graduation from medical school (OR = 0.93), a higher proportion of office visits by children < 1 year of age (OR = 0.97) and increased comfort diagnosing serious bacterial illness (OR = 0.35). Factors associated with increased odds of compliance included a higher perceived likelihood of serious bacterial illness in febrile infants (OR = 1.01) and better reported knowledge of the recently published guideline (OR = 2.01).

CONCLUSIONS

We found that specialty as well as other factors were associated with physician compliance with a recently published guideline. This information may facilitate guideline development and implementation by providing a better understanding of what motivates physicians in their clinical decision making.

Evaluation of the infant with fever without source: an evidence based approach

The infant with fever without an obvious source upon physical examination offers a challenging clinical problem. A combination of detailed history, physical examination, and selected laboratory tests allows the clinician to discern which infants are at lower risk for bacterial illness. Implications for management and future research are discussed herein.

The clinical characteristics and infectious outcomes of febrile infants aged 8 to 12 weeks

We reviewed 356 consecutive cases of febrile infants aged 8 to 12 weeks who received outpatient evaluation for sepsis over 4 years. Thirty-three infants (9.3%) had a serious bacterial infection (SBI), including bacterial meningitis, bacteremia, urinary tract infection (UTI), and Salmonella enteritis. The SBI rate, which was directly proportional to fever height, was significantly greater for infants with hyperpyrexia (35%) than those with lesser degrees of fever (7.7%) and for infants with peripheral blood leukocytosis (total WBC count > or = 15,000/mm3; 25%) than those with lesser total WBC counts (5.8%). An attending-level physician judged that 67% of infants with SBI appeared to be "well," including five or eight cases (63%) of bacteremia, 14 of 17 cases (82%) of UTI, and all three cases of Salmonella enteritis, whereas all five patients with bacterial meningitis appeared to be "ill." Urinalysis abnormalities indicative of UTI were present in 15 of 17 infants (88%) who had this infection. SBIs are not uncommon in febrile infants aged 8 to 12 weeks and occur significantly more often in those with either hyperpyrexia or peripheral blood leukocytosis.

Association of hyperpyrexia with serious disease in children

In a prospective study over 7 years, 105 consecutive pediatric patients with hyperpyrexia (temperature > or = 41.1 degrees C [106 degrees F]) were evaluated to determine the incidence, sensitive indicators, and types of illnesses encountered. The incidence of hyperpyrexia in a large urban pediatric emergency department was 0.36 per 1,000 visits or approximately one in 2,759 visits. In patients with temperature > or = 41.1 degrees C, 65 (61.9%) had a serious illness. Pneumonia (33 lobar, three interstitial, two clinical) was the most common diagnosis (36.2%), followed by probable viral illness in 20 (19.0%) of the patients. Bacteremia (6.7%) and bacterial meningitis (5.7%) were less commonly found. Four (3.8%) patients died. The admission rate was 62.9%. Eighteen patients (17.1%) also had seizures. Sensitive indicators to help distinguish those with serious illness, with the exception of clinical appearance, were not found. Pneumonia is commonly found in children with hyperpyrexia. Temperature > or = 41.1 degrees C was associated with a high rate of serious disease.

Fever: causes and consequences

The present review distinguishes pathogenic, neurogenic, and psychogenic fever, but focuses largely on pathogenic fever, the hallmark of infectious disease. The data presented show that a complex cascade of events underlies pathogenic fever, which in broad outline - and with frank disregard of contradictory data - can be described as follows. An invading microorganism releases endotoxin that stimulates macrophages to synthesize a variety of pyrogenic compounds called cytokines. Carried in blood, these cytokines reach the perivascular spaces of the organum vasculosum laminae terminalis (OVLT) and other regions near the brain where they promote the synthesis and release of prostaglandin (PGE2). This prostaglandin then penetrates the blood-brain barrier to evoke the autonomic and behavioral responses characteristic of fever. But then once expressed, fever does not continue unchecked; endogenous antipyretics likely act on the septum to limit the rise in body temperature. The present review also examines fever-resistance in neonates, the blunting of fever in the aged, and the behaviorally induced rise in body temperature following infection in ectotherms. And finally it takes up the question of whether fever enhances immune responsiveness, and through such enhancement contributes to host survival.

What is fever? Normal temperature in infants less than 3 months old

To determine the normal temperature of healthy infants, we studied 691 infants less than 3 months of age being seen for regularly scheduled well-baby visits. All temperatures were taken rectally with an electronic thermometer. The mean temperature was 37.5 degrees C +/- 0.3 degrees C. Using a cutoff of 2 standard deviations (SD) above the mean, fever would be defined as a temperature > or = 38.1 degrees C. The 95th percentile was 38.0 degrees C, and > or = 38.1 degrees C would correspond to values above the 95th percentile. The most widely used definition of fever at present is a temperature > or = 38.0 degrees C; by this definition, 6.5% of these well infants would be considered to have a fever. A significant rise in temperature with age was noted. For infants from birth to 30 days old, 2 SD above the mean was 38.0 degrees C; for those 31 to 60 days old, it was 38.1 degrees C; and for those 61 to 91 days old, it was 38.2 degrees C. Similarly, the 95th percentile was 37.9 degrees C, 38.0 degrees C, and 38.1 degrees C, respectively. Temperature also varied significantly with the season of the year, being higher in the summer (2 SD above a mean = 38.3 degrees C) than in the winter (2 SD above a mean = 38.0 degrees C). Fever should be defined as a temperature > or = 38.0 degrees C in infants less than 30 days of age, > or = 38.1 degrees C in 1-month-olds, and > or = 38.2 degrees C in 2-month-olds.

Relationship of fever magnitude to rate of serious bacterial infections in infants aged 4-8 weeks

We correlated the height of fever with underlying infectious etiology in 683 consecutive febrile infants aged four to eight weeks who received outpatient evaluation for sepsis during a five-year period. The relative number of infants with fever was inversely proportional to fever height, as 51% had a temperature 38.1-38.9 degrees C, 45% had a temperature 39-39.9 degrees C, and 4% had a temperature greater than or equal to 40 degrees C [hyperpyrexia]. There were 34 cases of serious bacterial infections [SBI], including 16 cases of urinary tract infection, 8 cases of bacteremia, 6 cases of bacterial meningitis, and 4 cases of Salmonella enteritis. The rate of SBI increased in direct proportion to fever height, being 3.2% in those with a temperature 38.1-38.9 degrees C, 5.2% in those with a temperature 39-39.9 degrees C, and 26% in those with a temperature greater than or equal to 40 degrees C. The 6.8% rate of SBI in those with fever greater than or equal to 39 degrees C was significantly greater than the 3.2% rate in those with fever less than 39 degrees C [p less than 0.035]; and the 26% rate of SBI in those with hyperpyrexia was significantly greater than the 4.1% rate in those with fever less than 40 degrees C [p less than 0.000004]. In identifying those with SBI, the presence of hyperpyrexia had a sensitivity of 21%, specificity of 97%, positive-predictive value of 25%, and negative-predictive value of 96%.(ABSTRACT TRUNCATED AT 250 WORDS)

Serious bacterial infections in children. When can outpatient treatment be used?

Several studies now support outpatient treatment of many serious bacterial infections in children, such as periorbital or buccal cellulitis, urinary tract infection, pneumonia, and abscess. However, an appropriate agent, that is, a third-generation cephalosporin with a long half-life, must be available and its effectiveness properly researched. In addition, children must be free of other illnesses and able to ingest fluids and maintain hydration, and their parents must be willing and able to cooperate with an outpatient treatment regimen. Family physicians can maintain the close patient and family contact needed to facilitate this form of therapy.

Comparison of rectal, axillary, and tympanic membrane temperatures in infants and young children

STUDY OBJECTIVE

To evaluate the reliability of a tympanic membrane thermometer in detecting fever in young children presenting to the emergency department.

SETTING

Pediatric emergency department in an urban teaching hospital, DESIGN/MEASUREMENT/PARTICIPANTS: Temperature measurements were obtained sequentially at three body sites in children less than 3 years old presenting to the pediatric ED. Axillary and rectal temperatures were obtained with an electronic thermistor probe (Diatek 500), and tympanic membrane temperatures were obtained with a noncontact, infrared sensing device (First TEMP). Patients were stratified by age, ear canal patency, presence of otitis media, and rectal temperature.

RESULTS

Of 224 patients enrolled, 87 (39%) were febrile. Overall correlation of axillary and tympanic membrane measurements to rectal for all strata was .75 (P = .001) and .81 (P = .001), respectively. Sensitivity in detecting fever for axillary and tympanic membrane sites was .48 and .55, respectively. Otitis media and ear patency did not influence correlation of tympanic membrane measurements. Low tympanic membrane temperature sensitivity may be a result of probe configuration.

CONCLUSION

Tympanic membrane and axillary temperatures should be viewed with caution in children less than 3 years old as neither can detect fever reliably.

Febrile infants less than eight weeks old. Predictors of infection

Febrile infants less than eight weeks old frequently are admitted and receive parenteral antibiotics for treatment of possible sepsis. The authors assess 52 infants less than eight weeks old with a rectal temperature of 38.1 degrees C or higher as having either a readily identifiable focus of infection by physical examination, appearing "toxic" without a focus, or appearing well. The authors screened patients by using white blood cell (WBC) counts, erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), and chest radiographs in addition to blood, cerebrospinal fluid and urine cultures. The authors found a 9.6% incidence of bacteria in the 52 infants evaluated, with a 4.3% incidence in those febrile infants who appeared well. Five patients had positive blood cultures with Group B B Hemolytic streptococcus (four patients), and Viridans streptococcus (one patient). A clinical assessment of toxicity and a total band count greater than or equal to 0.5 x 10(3) cells/uL together were sensitive indicators of bacteremia, as were toxicity and a positive CRP. A "toxic" appearance, a WBC count greater than or equal to 15 x 10(3) cells/uL and an ESR greater than 30 were specific indicators of bacteria. Based on these data, identification of bacteremia in febrile infants may be possible with clinical assessment and screening laboratory tests. Because of the relatively small sampling size of this study, the authors feel that evaluation of a larger number of patients is warranted to evaluate these sensitivities in a more diffuse patient population.

Extreme hyperpyrexia in childhood. Presentation similar to hemorrhagic shock and encephalopathy

The authors report the cases of five previously well children, aged 8 to 33 months, who were seen over a 14-year period, with admission temperatures in excess of 42.0 degrees C (107.6 degrees F). Four of the patients died. Each child had a similar clinical illness in which the hyperpyrexia played a critical role. Negative blood, cerebrospinal fluid, and stool cultures excluded bacterial sepsis as a possible etiology. This illness is similar, if not identical, to the newly described syndrome of hemorrhagic shock and encephalopathy (HSES) reported in European and American infants.

Ambulatory care of febrile infants younger than 2 months of age classified as being at low risk for having serious bacterial infections

We prospectively examined whether febrile infants younger than 2 months of age who were defined as being at low risk for having bacterial infection could be observed as outpatients without the usual complete evaluation for sepsis and without antibiotic treatment. A total of 237 previously healthy febrile infants were seen at the Pediatric Emergency Room over 17 1/2 months. One hundred forty-eight infants (63%) fulfilled the criteria for being at low risk: no physical findings consisting of soft tissue or skeletal infections, no purulent otitis media, normal urinalysis, less than 25 white blood cells per high-power field on microsopic stool examination, peripheral leukocyte count 5000 to 15,000/mm3 with less than 1500 band cells/mm3. One infant appeared too ill to be included, and had sepsis and meningitis. None of the 148 infants at low risk had bacterial infections, versus 21 of 88 (24%) of those at high risk (P less than 0.0001); eight of 88 (9%) had bacteremia. Of the 148 infants classified as being at low risk for having bacterial infection, 62 (42%) were discharged to home, and 72 (49%) were initially observed for less than or equal to 24 hours and then discharged. Seventeen infants (11%) were hospitalized: in six, low risk became high risk; six had indications other than fever; and five because the study physicians could not be found. The 137 nontreated infants were closely observed as outpatients. The duration of fever was less than 48 hours in 42%, and less than 96 hours in 91%. All infants were observed for at least 10 days after the last examination. The fever resolved spontaneously in all infants but two, with otitis media, who were treated as outpatients. Our data suggest that management of fever in selected young infants as outpatients is feasible if meticulous follow-up is provided.

Maturation of pyrogen-elicited fever in the kitten

The febrile response to the endotoxin Salmonella typhosa was studied in developing kittens. We found that kittens younger than 30 days of age generated only a small rise in temperature in response to a standardized endotoxin challenge that consistently causes fever in adult cats. Some degree of febrile response was present at birth, but the dose of pyrogen necessary to elicit a fever was 10-15 times greater than that required in the adult. There was a gradual increase in both the magnitude and duration of fever as a function of age with the largest change occurring after 30 days of age. There was a direct relationship between the ability of the kitten to maintain its body temperature (Tbo) at the room ambient (Ta) and the magnitude of the elicited fever. However, increasing the Ta to thermoneutral (Ta = 30-32 degrees C) did not enhance the thermal response indicating that the failure to elicit the fever is not due to passive effects of Ta. These data suggest that the febrile response to an endotoxin develops over the first 6-7 weeks of the kitten's life and are discussed in relation to other physical variables of development.

Weight loss and body temperature changes in breast-fed and bottle-fed neonates

Among 1138 newborns in a Level II nursery, breast-fed and formula-fed infants were comparable in terms of sex, mode of delivery, gestational age, birth weight, and birth temperature. Breast-fed neonates subsequently lost more weight and a greater percentage of their birth weight (mean, 7.4% vs. 4.9%) than did formula-fed infants. Loss of more than 10 percent of birth weight was associated with short gestation and low birth weight and with breast feeding. Birth weight loss of greater than or equal to 3 percent was associated with a risk of fever (greater than or equal to 37.5 degrees C) among breast-fed and formula-fed infants, but there was no gradient of increasing risk of fever with increasing percentage weight loss beyond 3 percent. After weight loss and other significant variables were adjusted for in a multivariate analysis, breast feeding was not independently predictive of fever. Although breast feeding may be associated with weight loss, it is not prudent to assume that this is the cause of fever in a breast-fed neonate.

Identification of infants unlikely to have serious bacterial infection although hospitalized for suspected sepsis

During a 2-year period, 233 infants younger than 3 months were prospectively studied to determine whether physical examination, white blood cell and band count, and urinalysis could identify infants unlikely to have serious bacterial infections. Only previously healthy infants (born at term, no perinatal complications, no previous or underlying diseases, no previous antibiotic therapy) were studied. One hundred forty-four (62%) of the 233 infants were considered unlikely to have serious bacterial infections, because they did not have physical findings consistent with ear, soft tissue, or skeletal infection, had between 5000 and 15,000 white blood cells/mm3, had less than 1500 bands/mm3, and urinalysis yielded normal findings. Eighty-nine (38%) infants did not meet one or more of these criteria and were classified as being at high risk for serious bacterial infection. Only one (0.7%) of the 144 infants in the low-risk group had a serious infection, compared with 22 (25%) of the 89 infants in the high risk group (P less than 0.0001). None of the infants in the low-risk group had bacteremia, compared with nine (10%) of the 89 infants in the high-risk group (P less than 0.0005). Neither traditional risk factors, such as age, sex, and temperature, nor other signs, symptoms, or laboratory findings were adequate predictors of serious bacterial infection. We conclude that previously healthy infants younger than 3 months with an acute illness are unlikely to have serious bacterial infection if they have no findings consistent with ear, soft tissue, or skeletal infections and have normal white blood cell and band form counts and normal urine findings.

Association of temperature greater than 41.1 degrees C (106 degrees F) with serious illness

In a 9-month prospective study conducted in an urban emergency room, 15 children with rectal temperature greater than 41.1 degrees C (106 degrees F) were evaluated. Seven of the 15 patients were admitted to the hospital. Two children who were discharged home required subsequent admission, and six were managed on an ambulatory basis. Eight (53.3%) children had serious disease: two bacterial meningitis, two bacteremia without meningitis, two pneumonia, one pericarditis with effusion, and one Kawasaki disease. In four, the final diagnosis indicated a much more serious illness than was considered initially. The laboratory studies did not correlate reliably with the final diagnosis or need for admission. Children with a rectal temperature greater than 41.1 degrees C are at high risk for a life-threatening illness and should be evaluated for sepsis and meningitis.

Variability in the management of the febrile infant under 2 months of age

The management in the emergency department of febrile infants less than 2 months of age is influenced by the standard of practice in the community. We sought to determine if uniform practices existed across the United States. Individual academically based faculty from 154 (61%) United States pediatric residency programs responding to a questionnaire on the emergency department management of febrile infants less than 2 months of age showed great variability. Twenty-nine respondents reported written policies and 103 reported informal but defined guidelines for the evaluation and management of infants seen at their institutions. There was little consensus among the respondents as to the definition of fever in this age group. Those at institutions with formal policies reported using more laboratory tests in the evaluation. Respondents differed on the number and types of tests used and on antibiotic administration. University affiliation, type of population served, or presence of advanced training programs in ambulatory pediatrics were not related to the type of policy. The care of the young febrile infant varies greatly.

Febrile infants: predictors of bacteremia

One hundred and seventy-five infants less than 8 weeks of age, presenting to the pediatric emergency room of the Bronx Municipal Hospital Center with rectal temperature greater than or equal to 38 degrees C (100.4 degrees F), were studied. House officers recorded their impressions of the infants on a number of variables prior to performing a lumbar puncture and obtaining laboratory data. All infants were admitted for parenteral antibiotic therapy pending culture results. Culture-positive bacterial infections occurred in 6.3% (n = 11); the incidence of bacteremia was 3.4% (n = 6). Of special concern were the 134 infants who had no visible source for their fever during the first examination. A major goal was to determine whether there were any early predictors of bacteremia in this group. The individual variables of white blood cell count greater than or equal to 15,000/mm3, band count greater than or equal to 500/mm3, temperature, impression of irritability, tone, cry, and activity level were not related to the presence of bacteremia. An erythrocyte sedimentation rate greater than or equal to 30 and the examiner's impression of sepsis were significantly associated with bacteremia but did not correctly identify all cases. However, the combination of impression of sepsis, white blood count greater than or equal to 15,000/mm3, and erythrocyte sedimentation rate greater than or equal to 30 identified all infants with bacteremia and excluded 82% of the infants who were eventually shown not to have bacteremia.

Management of febrile outpatient neonates

To determine whether all febrile outpatients below the age of two months should be routinely hospitalized and given parenteral antibiotics because of the possibility of occult sepsis or meningitis, 147 such admissions were reviewed over four years. None of the neonates had occult bacterial meningitis, while one may have had occult sepsis. Because serious infections may have been missed in outpatients not hospitalized, the outcome of all such illnesses seen in the clinic during a subsequent six months was monitored; none of 17 hospitalized or 20 followed as outpatients experienced sepsis or meningitis. If our findings are confirmed in other studies, febrile outpatient neonates with a well appearance, normal cerebrospinal fluid and physical examination, and dependable follow-up may not require universal hospitalization and antibiotics.

Is overheating a factor in some unexpected infant deaths?

Clinical and pathological evidence of overheating was sought in a consecutive series of infants dying in Newcastle and Gateshead. In 8 of 33 cases of cot death investigated pathologically histological changes in the small intestine of the kind described in association with heatstroke were seen. Such changes were not seen in any of 12 deaths from chronic or congenital conditions, and in only 1 of 8 acute explained deaths--in a baby who died with necrotising enterocolitis. 15 of 34 cot-death babies investigated clinically were judged to have been excessively clothed or covered at the time of death. 7 babies were unusually hot when found dead, and 4 others had been noticed to be hot shortly before death. 10 more babies had evidence of a terminal infection without observed fever. The possibility that overheating contributes to some cot deaths has important implications for health education.

Fever in the first six months of life: risks of underlying serious infection

The age-specific rate of elevated temperature over 37.8 C was evaluated in all infants less than 6 months of age (n = 1341) seen from July 1, 1974 to June 30, 1978 in a family practice clinic. Mild elevations (37.8 C-38.3 C) were common even in the first few months of life, and accounted for 20.7 per cent of infant visits. Temperatures greater than 38.3 C are uncommon in the first months of life but are seen more frequently with each succeeding month. Temperature elevation over 38.3 C was associated with a significantly higher rate of meningitis (p less than .01), otitis media (p less than .001) and lower respiratory infection (p less than .05). Significantly higher laboratory usage was documented in infants less than 3 months and for infants with temperature more than 38.3 C. The high rate of mild temperature elevations in young infants suggests that a selective diagnostic strategy directed at high-risk infants is important. Infants less than three months of age with a fever exceeding 38.3 C are calculated to have 21.5 times the risk of a serious underlying infection as infants older than three months with a similar temperature elevation. Clinical evaluation must remain an important tool in determining which febrile infants should be evaluated by further laboratory and diagnostic tests.