Initial Management of the Febrile Child in Family Practice Residency Clinics
A study of fever management in very young patients being treated at a family practice residency clinic.
By Damon A. Schranz, DO
Initial management of fever is important in very young children due to the possibility of significant systemic disease. For example, a recent report from the Pediatric Research in Office Settings (PROs) Network demonstrated that in febrile infants less than three months of age, meningitis occurred at a rate of 2.5 per 500 patients, and bacteremia at a rate of 8.5 per 500 patients.(1)
Although several practice guidelines exist on how to evaluate fever,(2-9) controversy continues regarding appropriate evaluation and management of febrile children.(10)
A lack of awareness of management protocols,(11) and a lack of compliance with protocols – even when there is awareness(12, 13) – has been documented in pediatricians. Little is known, however, about the management practices of other primary care specialties, including family practice.(14)
This study analyzed whether the degree of increased temperature affects antibiotic prescription practices and laboratory test utilization in a family practice setting. Use of residency training sites (i.e., residency programs) was indicated, as these places are where customary protocols are disseminated to residents, and practical experience in clinical medicine occurs.
Preschool children often present to primary care physicians for evaluation of fever. This fever can be reported by a patient or guardian, or documented during medical examination of a sick child. Furthermore, while recent literature indicates that most studies of initial evaluation of febrile children come from pediatric or emergency room settings, little information exists on the management of febrile children in family practice settings.
Acute febrile illness in preschool children is a common clinical scenario that can be a diagnostic challenge. The physician’s primary task is to identify the infant or child who is at risk for serious bacterial or viral infection. Meningitis, though not common, should be included in the differential diagnosis of the febrile child. While viral infection is the most common cause of fever in most infants and children, there remains a sense of worry in the physician’s mind on how to treat the fever and alleviate parental anxiety.
Parental misconceptions about fever continue to exist without regard to race, education level or socioeconomic status,(15) resulting in a common problem all family physicians will face. Unfortunately in today’s health care market, there is a trend to see more patients as reimbursements decline. This does not allow adequate education of parents on fever, and in some instances empiric antibiotic therapy for fever may be initiated. A meta-analysis(16) of studies involving the use of empiric antibiotics in older infants and children with fever concluded that treatment for possible occult bacteremia did not demonstrate a significant outcome advantage for antibiotic therapy.
Methods
Design
The basic design of this study was a prospective multi-site, cross-sectional descriptive study of usual practices undertaken by family practice physicians (residents, faculty) and non-physician providers (physician assistant, nurse practitioner) in the initial management of febrile children. No interventions were proposed.
Subjects
The primary subjects were health care providers in family practice training sites, including resident, fellow and faculty physicians, physician assistants and nurse practitioners. The secondary subjects were all infants and children under six years of age who presented to one of the 10 residency training sites with a complaint of temperature or a temperature of 37.5° C (99.5° F) documented in the clinic. The final sample consisted of 578 febrile children. The childrens’ ages range from one month to 71 months, with a median age of 20 months and an interquartile range of 11 to 36 months.
Inclusion Criteria of the Febrile Child
Data Collection
Data collection occurred over a 13-week period, starting in November 2000 and finishing in February 2001. Ten residency programs from the southern to the northern United States participated in the study. See Table 1 for data card and all collected variables. Data were entered on the data recording pocket cards, using a new card each week of the study. A number identified each participant in the data collection process.
On completion of data collection, the cards were sent to a central location (University of Wisconsin) and entered into a database. The database was checked twice for accuracy. The complete database was subsequently returned to the site principal investigators. Data were collected over a 13-week period by participating faculty, residents, and other health care providers, Data points collected included degree of temperature elevation, whether antibiotics were prescribed or not, and sum total number laboratory tests ordered at different degrees of temperature (see Table 2).
Statistical Analysis
ANOVA (Analysis of Variance test) was used to test for statistical significance between the dependent continuous variable of presenting temperature and the independent variable number of lab tests ordered. The relationship between initial degree of temperature and antibiotic use was investigated via chi-square analysis. Initial temperature was transformed into a categorical variable with six categories representing less than 99° F, 99-100° F, 100-101° F, 101-102° F, 102-103° F, and greater than 103° F. Statistical analyses were performed using the JMP software package. All statistical tests were two-tailed and a prior level of statistical significance was p<0.05. Not all data cards were complete. All available data variables were used. This will explain the differences in the sum data points.
Results
Two hundred seventy-five (49 percent) of the 560 febrile children were female. Two hundred fifty-one (45 percent) of the children were Caucasian; 208 (37 percent), Hispanic; 72 (13 percent), African-American; and 25 (five percent) Asian-American. Temperature followed a normal curve with a minimum temperature of 35.7° C (96.2° F) and a maximum temperature of 40.6° C (105.1° F). Mean temperature was 38.3° C (100.9° F).
Of the 560 febrile children, 460 were diagnosed without the utilization of laboratory tests. Sixty-one children received one lab test, 23 received two lab tests, and 16 required three or more lab tests. The most frequent laboratory test ordered was a White Blood Cell (WBC) count. Upper Respiratory Infection was the most common diagnosis (see Table 3). Treatment with antipyretics was the most common prescription of the treating physician (see Table 4).
Analysis of variance statistically described the mean difference of temperature for different levels of laboratory use. The model value of 12.57 exceeded the F value of 2.6, with a p<0.0001 (see Table 5). Chi-square was used to determine an association between degree of temperature and whether an antibiotic was prescribed. Temperature was divided into six categories and for each category a distribution of antibiotics prescribed and not prescribed was noted. There was not a clinically or statistically significant trend noted p=0.3770.
Discussion
A first aim of this study was to determine whether the degree of temperature influences antibiotic prescription in febrile children. It has been concluded from the literature search that parents do become anxious when their child runs a fever. These concerns can be translated to the physician through direct verbal conversation with the parent.
Most physicians, at one point or another, have been faced with the situation of prescribing antibiotics against the norm of traditional medical training to satisfy the anxious parent or patient. While parental anxiety is certainly a stressor to prescribe antibiotics, what is the general feeling of the practitioner initially evaluating a febrile child?
As physicians we are told never to prescribe unless a reasonable diagnosis can be elicited. Diagnosis requires history and physical examination. Unfortunately, young patients often cannot verbally describe the symptoms they have and in the event of a relatively normal physical exam the physician is left perplexed. Fever in children is usually suspicious for infection. The analysis of the febrile database has brought about some interesting observations.
First, there is no statistically significant evidence to support the thought that physicians may react to fever by prescribing antibiotics (see Table 6). This would not support the hypothesis that temperature stressors are affecting the physician’s prescribing habits. However it should be noted that in a febrile child population of 560, there were 310 antibiotic prescriptions written. The two major diagnoses were upper-respiratory infection (256) and otitis media (192), of which the majority are secondary to viral infections. This contradicts current medical prescribing guidelines.
A second aim of this study was to determine whether the degree of temperature influences laboratory test utilization in febrile children. Physician anxiety in the approach to the febrile child is normal and expected. We, as physicians, are expected to properly diagnose and treat medical problems. A majority of the time we succeed. However, there is always the possibility of a physician missing a tiny clue or detail leading to a catastrophic consequence.
For instance, occult bacteremia if not caught early can lead to death in a very short period of time. Physicians have thus become reliant on laboratory testing to help in their decision-making. Analysis of the database revealed that as fever increases there is a statistically significant trend to order more laboratory tests (see Table 5). The temperature differences are of such a degree that clinical significance can be implied.
Conclusion
Over-utilization of antibiotics and laboratory tests has a wide array of consequence, both economically and medically. The overuse of antibiotics has been implicated in the creation of a new class of bacteria, the so-called “super bugs.” Now that methicillin resistant staph aureus (MRSA) and various other “super bugs” have come to the forefront of modern medicine, there is a shift in ideology on how to control and combat this ever-increasing list.
Foremost is the education of physicians on the harms of antibiotic misuse. From the analysis of the febrile child dataset we can infer that much more education is needed especially in our training programs. Likewise, the use of laboratory tests in direct relation to temperature is troubling. A higher degree of temperature does not predict the likelihood of bacterial over viral infection. Prudent use of laboratory tests in select situations will ensure the economic viability of medicine in the future.
As physicians, we should rely more on our own education and instinct in the formulation of a diagnosis, rather than diagnostic tests.
Damon A. Schranz, DO, is in practice in Ft. Worth, Texas. Dr. Schranz also acknowledges the assistance of Elizabeth Palmorozzi, DO, in the preparation of this article.
References: