Fever is the thermoregulated increase in body temperature above normal as the result of a coordinated response to a pathologic insult. The two essential features of fever are that it represents an abnormal elevation in body temperature and that it results from a coordinated physiologic response. The first feature differentiates fever from normal regulated elevations in body temperature (e.g., elevations associated with the circadian rhythm), whereas the latter distinguishes it from conditions in which the regulatory mechanisms are overwhelmed or dysfunctional (e.g., heat stroke). In children, the pathologic insult most likely to result in fever is infection. A variety of other conditions, including malignancies and autoimmune diseases, also may result in this phenomenon, however.
Normal Body Temperature
Although the general public and physicians alike often refer to “the” body temperature, the implication that a single number can represent the thermal state of the entire body is inaccurate. Depending on the site of measurement, body temperature may vary by 1°C or more. These regional variations in temperature do not have a fixed relationship to each other. Although axillary temperatures are consistently lower than rectal temperatures, the absolute difference between the two varies greatly. In addition, even in the healthy state, body temperature is not constant; it varies depending on numerous factors, such as time of day, level of activity, and phase of the menstrual cycle. Generally, clinicians have been most interested in the core body temperature, defined as the temperature of the internal organs of the trunk and head. Under normal circumstances, core temperature is higher than the temperature of more superficial tissues such as skin. Even within these two anatomic regions, temperature gradients exist, however.
The most widely accepted definition of normal body temperature is 37°C (98.6°F). This number is derived from studies performed in the 19th century by Wunderlich. He reportedly arrived at this figure based on the result of several million measurements conducted in approximately 25,000 individuals. Other more recent studies have found slightly lower mean temperatures in healthy individuals, despite the fact that these more recent studies are based on oral or rectal temperatures whereas Wunderlich’s studies relied on axillary temperatures. Mackowiak and colleagues determined the mean oral temperature in adults to be 36.8°C (98.2°F), with the upper limit of normal ranging from 37.2°C (98.9°F) at 6:00 am to 37.7°C (99.9°F) at 4:00 pm . Given the limitations imposed by the technology available at the time, however, perhaps the most surprising aspect of the value reported by Wunderlich is how closely it approximates these more recent determinations.
As noted previously, body temperature fluctuates depending on numerous normal physiologic factors. Core temperature shows a diurnal variation of 1°C, the nadir occurring in the early morning hours and the peak in the late afternoon. After exercise and in the postprandial state, body temperature increases. In addition, variations in the normal body temperature of women associated with the menstrual cycle are well described, with increases in baseline temperature occurring after ovulation.
Thermoregulation
Humans, similar to other mammals, are homeothermic, indicating that they regulate body temperature within a narrow range despite wide variations in the ambient temperature. Regulation of temperature is mediated by a variety of physiologic (e.g., vasoconstriction, sweating) and behavioral (e.g., moving to a warmer environment, putting on additional clothing) responses.
The principal thermoregulatory area is located within the brain in the preoptic area and anterior hypothalamus. Although it frequently is conceptualized as a single center, no single neuronal structure seems to control all aspects of temperature regulation. Rather, a complex interplay occurs among a variety of neural pathways, with the final result being the maintenance of the body’s temperature within a narrow range. Regardless of the precise nature of central regulation, body temperature ultimately is a function of the balance between heat gain and heat loss.
Heat energy is a by-product of the inefficiency of the body’s normal metabolic processes. It is this “waste” heat that renders the homeothermic state possible. During exercise, the increased metabolic activity in muscle tissue results in increased production of heat, leading to an increase in the body’s temperature. Shivering, an involuntary form of muscle activity, is the primary means by which the body generates additional heat under conditions of cold stress. Heat also may be generated by a process known as nonshivering thermogenesis. Originally described in rats, nonshivering thermogenesis has been found to occur in a variety of mammals, including humans. It seems to be of greater importance in neonates than in adults. Although this process has been shown to occur in a variety of tissues, brown adipose tissue seems to be the most important site for this phenomenon. Under the control of the adrenergic system, production of free fatty acid is increased, resulting in an uncoupling of oxidative phosphorylation and the production of large amounts of heat.
Four mechanisms are responsible for heat transfer: radiation, conduction, convection, and evaporation. Heat loss owing to radiation occurs when heat is transferred directly between two objects not in direct contact. Conduction involves the transfer of heat energy between two objects in contact with each other. Convection is the result of the movement of a fluid or gas across the surface of the body (e.g., as the result of fanning). Evaporative heat loss occurs in association with the energy required to convert liquid to gas form.
Under normal conditions, radiation accounts for most of the body’s heat loss. In contrast, conductive losses are smaller under normal circumstances. Conductive losses may become substantial, however, under conditions in which a large portion of the individual’s body surface is in direct contact with a cooler object (e.g., an unclothed infant in an unheated bassinet). Convective losses are proportional to the amount of air moving over the body surface; these losses are greatest in windy conditions. Conductive and convective heat losses are particularly important in infants and children because of their relatively greater body surface area compared with that of adults. Evaporative losses occur when fluids such as sweat evaporate from the skin’s surface. In addition, substantial evaporative losses are associated with respiration.
Pathogenesis of Fever
In the classic model of fever pathogenesis, exogenous pyrogens stimulate the release of circulating endogenous pyrogens, which act via prostaglandins to increase the set point of the hypothalamic thermoregulatory center. In this model, exogenous pyrogens are substances extrinsic to the body, primarily various bacterial microorganisms or the products of those microorganisms. Conversely, endogenous pyrogens are a varied group of proteins produced within the human body that share the intrinsic ability to induce fever. The first endogenous pyrogen was described originally more than 60 years ago and was derived from leukocytes, primarily granulocytes, and now is known as interleukin-1 (IL-1). Numerous other cytokines that qualify as endogenous pyrogens, including tumor necrosis factor, interferon-α, interferon-γ, and IL-6, have been identified.
A variety of alternative and complementary theories of fever pathogenesis have been proposed. The actual mechanisms involved likely are more varied and complex than just described. A variety of intrinsically produced substances (e.g., antigen-antibody complexes) may act as “exogenous” pyrogens. Murine models suggest that neuromodulatory endocannabinoid lipids play a role in lipopolysaccharide-induced fever. The classic model provides a reasonable framework, however, for understanding most of the observed phenomena associated with the febrile response.
Regardless of the precise pathogenesis, the height of fever seems to be limited. Retrospective studies of hyperpyrexia in children have found that it is unusual for the body temperature to rise above 41.1°C (106°F), and it rarely rises above 41.7°C (107°F). Children with temperatures exceeding this range almost always have an element of heat illness.
Effects of Fever
Attempts to treat fever often are predicated on the assumption that fever has harmful effects and that reduction in temperature would abrogate such harm. The evidence with regard to these premises is mixed, however.
Adverse Effects
Some animal studies have found that high fever may impair certain immunologic responses, including phagocytosis of staphylococci by polymorphonuclear leukocytes and lymphocyte transformation in response to mitogens. Whether these isolated in vitro phenomena observed in animal models are relevant to human infection is unknown.
Fever may cause seizures, a phenomenon observed most frequently in young children. The onset generally occurs in infants 6 to 30 months of age. Recurrence is common, occurring in approximately one third of children experiencing an initial febrile seizure. The primary adverse consequences of febrile seizures are the emotional distress experienced by patients and their families and the need for medical evaluation, which may involve invasive testing and substantial expense. Febrile seizures do not cause brain injury and are not associated with subsequent intellectual or neurologic deficits. Furthermore numerous studies have failed to demonstrate that use of antipyretics is effective in preventing febrile seizures.
An important nonphysiologic adverse effect of fever is caregiver anxiety, reflecting an underlying concern for harm due to untreated fevers despite a lack of supporting data. The unfounded anxiety is not limited to the lay public. In addition, fear phobia may be inadvertently reinforced by the actions of medical personnel; for example, it has been observed fever may routinely be treated more promptly than pain in the pediatric emergency room. Caregiver distress also crosses cultural boundaries, as suggested by a recent Turkish study that revealed that 3 out of 4 parents would wake their child from sleep to administer antipyretics.
Beneficial Effects
Fever may be beneficial by enhancing the host response to infection and by directly inhibiting the infecting agent. Several studies have demonstrated that the immune system responds to mildly elevated body temperature by increasing migration of leukocytes, production of interferon, and lymphocyte transformation and phagocytosis. Studies of bacterial infection in reptiles and fish have shown an increased survival rate in groups maintained at approximately 4°C (reptiles) and 2.5°C (fish) above baseline. Kluger and Vaughn showed that rabbits infected with Pasteurella multocida had improved survival rates at body temperatures of approximately 4.5°C above normal. Although these data are impressive, their clinical significance with regard to humans has yet to be determined.
Fever also may inhibit the growth and survival of some infectious agents. One potential mechanism for this inhibition is the decrease in serum iron and increase in ferritin that are associated with fever, coupled with the increased iron requirement of many bacteria at higher temperatures. Fever therapy was used historically to treat neurosyphilis and gonococcal urethritis, correlating with more recent studies that show that certain gonococci and Treponema are eradicated at temperatures of 40°C (104°F) and greater. Finally, growth of some pneumococci and viruses seems to be impaired at higher temperatures.
Several studies have found that the treatment of fever with antipyretics is associated with adverse consequences, providing indirect evidence of a beneficial effect of fever. Ahmady and Samadi reported that the use of aspirin in children with measles prolonged the duration of the illness and was associated with an increase in prevalence of respiratory complications and diarrhea. Other investigators have reported that the length of time to total scabbing in varicella was significantly longer in children treated with acetaminophen compared with children treated with placebo. Several studies in animal models and in humans have found prolongation of viral shedding, depressed neutralizing antibody response, and increased nasal symptoms in association with the use of antipyretics. Although these studies establish that an association exists between reduction of fever and adverse outcomes, they do not prove a causal relationship. The adverse effects possibly are mediated by some direct physiologic effects of antipyretics rather than indirectly by their impact on fever.
Clinical Thermometry
Types of Thermometers
For many years, glass thermometers containing mercury were the most common type of thermometer used to measure body temperature. This type of thermometer is reasonably accurate for most clinical purposes. Although they are still available, use of mercury-containing thermometers has diminished greatly because of environmental concerns about mercury exposure from broken or discarded thermometers. These thermometers have been replaced largely by digital thermometers or glass thermometers containing liquids other than mercury.
Electronic thermometers (often referred to as digital thermometers) previously were used primarily in the hospital and office setting. As their cost has decreased, they are used more frequently in the home as well. Electronic thermometers have the advantage over mercury thermometers of requiring a significantly shorter dwell time—that is, the time they must remain in situ to obtain an accurate reading. Hospital-grade electronic thermometers typically have two modes: monitor and predictive. In the monitor mode, these thermometers function similarly to mercury thermometers in that they must remain in place until equilibration occurs, a process that may require several minutes. In the predictive mode, a complex algorithm is used to estimate the final temperature based on measurements made during the first few seconds. Because the predictive mode produces a temperature reading within seconds, it is the mode used most often in clinical settings. Determinations of temperatures using these two modes have been found to correlate well.
Infrared thermometers are a more recent addition to the clinician’s armamentarium. Devices that determine the temperature by detecting infrared radiation emitted from the eardrum are used most frequently. Tympanic temperature should provide an accurate estimation of the core temperature because its blood supply is derived from the carotid artery. Additional advantages of this type of thermometer are its speed, acceptance by patients, and decreased risk of cross-contamination compared with oral or rectal thermometers. Studies of the accuracy of tympanic thermometers have yielded mixed results, with numerous studies finding tympanic thermometers inaccurate compared with mercury in glass or electronic thermometers. A recent meta-analysis showed pooled sensitivity of 70% and specificity of 86% as compared to rectal thermometry in diagnosing pediatric fever. Discrepancies seem to be particularly common in infants who are in the first few months of life. Tympanic thermometers should not be used in young infants because of the importance of fever in making management decisions in these patients.
Even more recently, the temporal artery thermometer has been introduced. These thermometers use an infrared sensor to determine skin temperature as the device is passed across the forehead and temporal area. The site of highest measured temperature is assumed to represent that of the temporal artery. An algorithm is applied to the measured temperature to estimate the core temperature. Studies to date suggest that temporal artery thermometer temperatures correlate significantly better with rectal and core temperatures than with temperatures determined by tympanic thermometers. The data also suggest that these thermometers are more sensitive at detecting fever in children than in adults. Temporal artery thermometers do not seem to correlate well enough with rectal or core temperature measurements, however, to replace rectal thermometry in clinical situations in which accurate measurement of fever is crucial for making decisions about management. Correlation with rectal thermometry may be particularly limited in children younger than 36 months of age, with a reported sensitivity as low as 27% to 53%. The accuracy of temporal artery thermometer readings is adversely affected by sweating and may be affected by vascular constriction or dilation. In addition, data comparing temporal artery and axillary thermometry are lacking.
Several other types of thermometers have been developed. Among them are electronic pacifier thermometers, which are used for obtaining an oral temperature in infants. Although they are appealing in theory, this type of thermometer has the disadvantage of requiring a prolonged dwell time and has not been found to be sufficiently accurate to recommend its use. Another approach to measuring temperature is the use of liquid crystal thermometers that are applied to the skin of the forehead. Results generally have been disappointing when these thermometers have been compared with more standard techniques.
Despite ready availability of thermometers, many parents continue to utilize palpation, generally of a child’s forehead, to detect the presence of fever. A study of infants under the age of 3 months found the positive predictive value of this practice to be only 33%, although the negative predictive value was as high as 95%. These results support the conclusion that palpation should not be used as the sole means of fever assessment.
Measurement Site
The most common locations for measuring body temperature are the mouth, rectum, axilla, and tympanic membrane. Because of the previously noted regional variations in body temperature, each of these sites has its own range of normal temperatures. The oral cavity historically has been the preferred site for measuring temperature in older children and adults. When taking a temperature orally, one should place the thermometer in the sublingual space because the blood supply for structures in this region is derived from branches of the carotid arteries and should reflect the core temperature accurately. Younger children usually are unable to cooperate adequately to permit the use of oral thermometers. In addition, the oral temperature may be affected by recent ingestion of hot or cold liquids, and it may be altered by tachypnea.
Rectal temperatures are used frequently in younger children. The rectal temperature correlates well with the core body temperature. The rectal temperature may exceed the core temperature (using the pulmonary artery temperature as the reference standard), however, possibly because of the effects of bacterial activity in the rectum. The use of rectal thermometry has the disadvantage of causing the patient discomfort and is contraindicated in patients with neutropenia because of the risks of causing invasive infection via trauma to the rectal mucosa.
Axillary temperatures are appealing because of the ready accessibility of the axillae. Considerable variability occurs in the readings obtained, however, particularly in younger children. One should not rely on axillary temperatures, particularly in neonates and young children.
Treatment
As noted previously, fever may have numerous beneficial effects, convincing evidence of harm owing to fever is lacking, and treatment of fever may be associated with undesirable effects. Therefore routine intervention to reduce fever is not warranted. Rather physicians should individualize the decision to treat fever and the specific method chosen to do so. This approach is in accordance with the recommendations of the American Academy of Pediatrics.
Indications
Antipyretic therapy often is considered for children who have an increased risk of having febrile seizures, either because of age or because of a history of febrile seizures. Although treatment seems rational in this situation, studies of antipyretic therapy to prevent febrile seizures have repeatedly failed to demonstrate its efficacy. Even this indication should be considered relative rather than absolute.
Prophylactic administration of acetaminophen prior to routine immunizations to prevent vaccine-related fever has been debated. The practice is generally not recommended due to lack of supporting evidence. While the American Academy of Pediatrics Red Book recommends considering acetaminophen administration before and after DTaP immunization for children with a personal or family history of seizures, studies have generally failed to produce results to support this practice.
Antipyretic therapy also should be considered for children with poorly compensated underlying cardiac or pulmonary disease, significant neurologic impairment, or sepsis and for children with significant alterations of fluid and electrolyte balance. Definitive controlled trials to support these indications are lacking. Rather the recommendations are based on the metabolic consequences of fever and their potential adverse impact on the underlying disease.
Perhaps the most frequent indication for the use of antipyretics is to improve the patient’s comfort. Despite the absence of definitive studies to support this practice and the potential adverse consequences of using antipyresis, such an approach is reasonable in the absence of definitive evidence to the contrary. In some circumstances, improving the patient’s comfort may enhance the ability to assess the seriousness of the patient’s illness accurately.
Antipyretics
A wide variety of antipyretic agents is available. In the United States, the drugs used most frequently for treatment of fever in children are acetaminophen and ibuprofen. Previously, aspirin was the antipyretic used most frequently. Aspirin has fallen into disuse for the management of fever in children, however, primarily because of its association with Reye syndrome, particularly when used in managing children with varicella or influenza. In addition, aspirin has a variety of other adverse effects, including inhibition of platelet function, gastritis and gastrointestinal bleeding, and provocation of asthma exacerbations, although this third complication occurs more frequently in adults. Aspirin has greater toxicity in situations of overdose than do acetaminophen and ibuprofen.
Each of these agents acts to restore normal body temperature by reducing the set point of the temperature regulatory center in the hypothalamus. The specific mechanism of action seems to be interference with prostaglandin synthesis in the preoptic anterior hypothalamus. When selecting among the available antipyretic agents, one should consider efficacy and potential toxicity.
Similar to aspirin, ibuprofen inhibits prostaglandin synthesis in a variety of tissues outside the central nervous system. It shares many of the toxicities associated with aspirin. One exception is that ibuprofen lacks the association with Reye syndrome. Ibuprofen inhibits platelet function because of its effect on prostaglandin synthesis, but this effect is reversible with discontinuation of the drug, and platelet dysfunction is short lived compared with the effect of aspirin. Because prostaglandins are important to the integrity of the gastrointestinal mucosa, inhibition by ibuprofen may result in gastrointestinal upset and bleeding. Although ibuprofen has been associated with exacerbations of asthma in some children, the risk seems to be small and may not be greater than that associated with the use of acetaminophen.
Acetaminophen has a lengthy track record of safety. When used in the usual therapeutic doses, it has few adverse effects. Acetaminophen inhibits prostaglandin synthase activity, but this action is inhibited by peroxide. Because peroxide is generated at sites of inflammation, acetaminophen has little anti-inflammatory activity. It also lacks the adverse gastrointestinal and platelet effects of aspirin and ibuprofen. A growing body of evidence suggests that acetaminophen use is associated with an increased risk of subsequent asthma. Despite the fact that studies have consistently found this association, evidence of a causal relationship has yet to be conclusively established. Recommended dosing of ibuprofen is 5 to 10 mg/kg every 6 hours as needed. Acetaminophen is administered at a dose of 10 to 15 mg/kg every 4 hours but no more frequently than five times per day. An important note for individuals administering these agents is that a variety of over-the-counter combination medications contain one or the other of these agents. Co-administration may result in inadvertent overdosing.
In addition, recognition that ibuprofen and acetaminophen come in a variety of formulations is important. Acetaminophen, once available as infant drops (concentration 10 mg/mL) and children’s liquid (concentration 32 mg/mL) is now only available in a single concentration (32 mg/mL). The voluntary removal from the market of the more concentrated formulation was prompted by reports of inadvertent overdose from substitution of infant drops for the children’s liquid without adjusting the volume administered to reflect the difference in concentration. Acetaminophen also is available in suppository form. Absorption varies, however, and is delayed compared with oral administration. In addition, the medication is not distributed uniformly throughout the suppository, resulting in potential dosing errors if the suppositories are divided before use. The use of acetaminophen in suppository form should be discouraged. Ibuprofen continues to be available in two concentrations: a children’s liquid (20 mg/mL) and infant drops (40 mg/mL). The dual concentrations create the potential for incorrect dosing seen previously with acetaminophen.
The antipyretic efficacy of acetaminophen and ibuprofen has been the subject of numerous clinical trials. These trials have shown uniformly that both are effective antipyretic agents. Ibuprofen seems to result in a greater decrease in temperature than does acetaminophen, however. In addition, the antipyretic effect of ibuprofen is more prolonged, not surprising in light of ibuprofen’s longer half-life compared with acetaminophen. These observations were confirmed in a meta-analysis of trials comparing ibuprofen and acetaminophen.
Acetaminophen and ibuprofen frequently are used in combination. Use of alternating doses often is advocated by practicing physicians. The pathways for metabolism of these drugs are distinct, and, theoretically, metabolism of one should not affect the metabolism of the other. Controlled trials documenting safety and efficacy of combination or alternating use are sparse, however. In one study, using alternate doses of acetaminophen (12.5 mg/kg per dose) and ibuprofen (5 mg/kg per dose) every 4 hours was found to be associated with a more rapid reduction in fever, lower mean temperature, and fewer caregiver days absent from work and infant days absent from daycare compared with use of either agent alone. However, the groups assigned to a single agent received dosing that was either on the low end or infrequent compared with usual practice in the United States. In a more recent study utilizing full therapeutic doses and comparing ibuprofen alone with alternating and combined schedules of ibuprofen and acetaminophen, the alternating and combined schedules were found to provide greater antipyresis. Recent systematic reviews of combination versus single-agent treatment with acetaminophen and ibuprofen concluded that the antipyretic effect of combined therapy was statistically superior but the clinical relevance was marginal at best. Furthermore the studies found no evidence of increased toxicity with combined treatment but were not designed to definitively address the issue. Evidence is also lacking regarding any beneficial effects of combined therapy on child discomfort.
Another approach to antipyretic treatment frequently employed is the use of a second antipyretic when the initial agent is judged to have resulted in an inadequate response. Theoretically, some individuals may have a better antipyretic response to one agent than another. Although the premise is reasonable, sequential use of acetaminophen and ibuprofen remains unproved in terms of efficacy and safety.
Acetaminophen and ibuprofen have been proved to be remarkably safe when used in the recommended doses. Although ibuprofen may be slightly more efficacious in producing and sustaining fever reduction, acetaminophen remains the antipyretic of choice because of its longer track record and more favorable side-effect profile. Because acetaminophen lacks significant antiinflammatory activity, ibuprofen or another nonsteroidal antiinflammatory drug may be preferred in febrile conditions for which antiinflammatory activity is desired (e.g., juvenile arthritis). Limited data suggest that use of acetaminophen and ibuprofen in combination may be safe and more efficacious than is either agent alone. Prudence suggests, however, that combined therapy seldom is warranted for the treatment of fever, a generally benign condition. Perhaps most important, patients and their parents should be educated about the benign nature of fever and the lack of evidence to indicate that routine treatment, particularly complete suppression, is either necessary or beneficial.
External Cooling
The use of external cooling in the management of fever has a long history. Compared with standard antipyretics, sponging with tepid water is inferior in fever reduction at 2 to 3 hours, although sponging was found to reduce the temperature more quickly than did antipyretics in one trial. External cooling without concomitant administration of antipyretics makes little sense from a physiologic standpoint, however. In a febrile patient whose temperature regulatory center set point has not been reset by administration of an antipyretic agent, external cooling inevitably results in an increase in the body’s heat-production mechanisms.
When used in conjunction with an antipyretic agent, the usual goal of external cooling is to reduce the body’s temperature more rapidly or to a greater degree. Several studies have compared the use of external cooling combined with antipyretics with antipyretics alone. Results have been mixed; some studies showed no difference in efficacy of the two approaches, whereas others found combination therapy to be superior. Even in the studies in which a difference was found, the superiority of combination therapy was shown primarily in the very early phase of treatment. A recent randomized controlled study performed in an adult critical care unit population concluded that external cooling decreased vasopressor requirements and early mortality in septic shock. However, it is unclear if these results can be translated to the pediatric population.
In addition, the use of external cooling usually is uncomfortable for the patient. When external cooling is to be used, sponging with tepid water is preferred. Alcohol and solutions containing alcohol should not be used for this purpose. Absorption of alcohol vapors via the lungs may occur in sufficient quantities to produce toxicity and even death.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
