Program Number: 2021-09-04
Approved Dates: October 1, 2021-April 1, 2022
Approved Contact Hours: One Hour(s) (1) CE(s) per session
Kids Survive: Evidence Based Management of Pediatric Sepsis
Author: Victoria H. Anderson, PharmD; PGY-1 Pharmacy Resident (2020-2021)
Mentor: Jordan E. Anderson, PharmD, BCPS, BCCPS; Pharmacy Clinical Coordinator/PGY-1 Residency Program Director
University of Missouri Women’s and Children’s Hospital – Columbia, MO
Take Assessment Quiz
The Surviving Sepsis Campaign published updated pediatric sepsis guidelines in February 2020. Prior to these guidelines, pediatric recommendations shared a guideline document with adult recommendations in 2004, 2008, and 2012, and the Surviving Sepsis Campaign only published adult guidelines in 2016.
Sepsis Considerations: Definitions and Pharmacokinetics
Early recognition of sepsis and evidence-based care of pediatric sepsis patients are of utmost importance to improve patient outcomes as a study in the UK found that over 50% of pediatric sepsis fatalities occurred in the first 24 hours and approximately half of those fatalities were prior to admission to pediatric intensive care units.1 Adherence to guidelines can improve outcomes and development of protocols including order sets and standardized education can reduce mortality.1 A study of emergency department teams found that only 45% correctly adhered to treatment metrics, so there is certainly room for improved implementation of guideline recommended care.1
Sepsis is defined as systemic inflammatory response syndrome (SIRS) plus suspected or known infection. SIRS criteria include hypo- or hyperthermia, leukopenia or leukocytosis, brady or tachycardia, and tachypnea (see Figure 1).2 At least two criteria must be met to diagnose SIRS and at least one of those criteria must be the patient’s temperature or leukocyte count.2 Of note, the SIRS criteria and sepsis definitions do not apply to premature infants.2 To qualify for severe sepsis, a patient must have cardiovascular or respiratory dysfunction or dysfunction of at least two other organ systems such as neurologic, hematologic, renal, and hepatic.2 Septic shock is defined as sepsis with cardiovascular organ dysfunction.2 At least 40mL/kg of fluids must be administered over an hour prior to assessing for cardiovascular dysfunction to meet the organ dysfunction criteria.2 It is important to recognize that hypotension may not be present with cardiovascular dysfunction until a patient is near collapse, as pediatric patients often have strong compensatory mechanisms.1 As a result, fluids are very important to management in even normotensive patients as delays can increase ICU and hospital length of stay and increase risk of acute kidney injury(AKI).1 Rather than titrating to blood pressure goals which is often practiced in adults, fluids should be titrated to increased urine output, improved mental status, and decreased capillary refill time as long as patients remain without hepatomegaly or rales.
Organ function changes are important to note not just for diagnosis of SIRS and subsequently sepsis, but because they also can change the pharmacokinetics of medications administered to these children. Critically ill children often have changes in both synthesis and binding affinity of albumin and alpha-1 acid glycoprotein which can impact unbound fraction of medications.3 Changes in pH whether acidosis or alkalosis can also impact drug levels by impacting ionization which can impact both effective concentration and elimination rates of medications.3 Fluid shifts impact volume of distribution which can lead to changes in concentration, particularly of hydrophilic medications such as vancomycin, so it is important when monitoring vancomycin to consider the impact of fluid resuscitation and maintenance fluids on serum concentrations.3 Renal dysfunction can impact enzyme activity, pH, total body water, and drug clearance, but conversely, sepsis can also lead to augmented renal clearance.3 Mild hepatic dysfunction can modify hepatic blood flow, enzyme activity, drug transport, protein binding, and total body water, impacting the disposition of medication.3 More severe hepatic dysfunction is marked by capillary leak, coagulopathy, renal dysfunction, and hypoglycemia.3 Additionally, inflammatory mediators can decrease P450 enzyme metabolism.3 Changes in blood flow to drug clearing organs may decrease elimination of many drugs.3 Cardiovascular changes can also lead to fluid overload and edema, changing volume of distribution, and increases in alpha-1 acid glycoprotein.3 Abdominal venous congestion can also lead to a decrease in enteral absorption of medication.3 Finally, therapies to compensate for dysfunctional organs such as extracorporeal membrane oxygenation (ECMO) and continuous renal replacement therapy (CRRT) also can impact pharmacokinetics.3 The ECMO circuit tubing binds some medications requiring increased dosing initially until binding sites are saturated.3 ECMO also may increase volume of distribution of some medications secondary to added volume required for the circuit.3 The impact of CRRT varies primarily based on the level of protein binding of the medication with lower protein binding leading to increased removal.3
It is also important to consider comorbidities that may impact medication therapy. Forty-nine percent of children with sepsis have a comorbidity making them more vulnerable to infection.1 A common comorbidity is cancer which both the condition itself and the treatment can impact pharmacokinetics.3 For example, fluconazole trough concentrations have been found to be significantly lower in pediatric cancer patients.3 All of these changes make careful use and monitoring of medications in these patients of utmost importance.
Initial Care: Antibiotic and Hemodynamic Management
Once we have made the presumptive diagnosis of sepsis, rapid protocol-driven treatment is the next important step. Per the Surviving Sepsis Campaign International Guidelines for the Management of Septic Shock and Sepsis-Associated Organ Dysfunction in Children, in patients presenting without shock, antimicrobial therapy should be started within three hours of recognition.4 While blood cultures should be obtained prior to initiation of antibiotics if this will not cause delay, it is important to start treatment with antimicrobials as soon as possible, so timing of antibiotics is not dependent upon availability of blood cultures.4 The guidelines do not address obtaining multiple blood cultures, however, it is common practice to obtain two blood cultures to enhance our ability to detect causative organisms, and to help determine if bacteria found are pathogenic versus a contaminant. If patients have pre-existing intravenous access devices it is prudent to obtain a set of cultures both from the device and from a peripheral site. In addition to blood cultures, cultures should also be obtained from any suspected non-blood site of infection such as urine, cerebrospinal fluid, or wound drainage.4 If a patient presents with shock, guidelines recommend treatment with antimicrobials within one hour of recognition, but for patients without shock, treatment within the first hour has not been found to have a significant impact on mortality as compared to treatment by the end of the third hour.4 This is in contrast to the 2016 adult and 2012 pediatric guidelines which recommended antimicrobials within the first hour for all patients with sepsis regardless of the presence of shock, and the 2008 pediatric guidelines which recommended antibiotics within one hour of recognizing severe sepsis, but after inidicated cultures have been taken without the caveat in more recent guidelines not to delay evidence-based treatment while waiting for cultures.5,6,7 Giving antibiotics sooner than required will clearly not cause harm as long as this is not done at the expense of fully evaluating the patient, so following the older practice of administration within the first hour is not an inappropriate practice when practical.
These initial antimicrobial agents should be broad spectrum to cover all likely pathogens without excess coverage.4 A previously health child presenting from home should receive a third-generation cephalosporin such as ceftriaxone.4 A child at risk for methicillin-resistant Staphylococcus aureus (MRSA) or presenting from a community with high prevalence of ceftriaxone-resistant pneumococci should additionally receive vancomycin.4 In communities where ceftriaxone-resistant gram negative rods are prevalent, addition of aminoglycosides or substitution of a carbapenem in place of the third generation cephalosporin is recommended.4 If patients have influenza-like illness and influenza is unable to be ruled out, an anti-viral should additionally be given.4 Immunocompromised patients or those with hospital acquired sepsis should receive a beta-lactam covering Pseudomonas aeruginosa.4 Patients with suspected intra-abdominal source of sepsis should receive an antibiotic covering anaerobes, and those with toxic shock syndrome or necrotizing fasciitis should receive clindamycin or lincomycin.4 Neonates should receive ampicillin and if HSV is suspected should additionally receive acyclovir.4 Synergy may be required for device-associated infections, Enterococcal or Staphylococcal endocarditis, group B Streptococcus infections, and carbapenem-resistant Enterobacteriaceaea, but the only patients who should receive double coverage are cancer and transplant patients who are unstable and come from a community with a gram-negative resistance rate of greater than ten percent.4 These patents should receive a second agent covering gram-negatives.4 This is in contrast to the 2016 adult guidelines which recommended empiric combination therapy for patients in shock with de-escalation within the first few days if the patient is clinically improving regardless of results of cultures, but recommended against routinely double covering any particular pathogen in neutropenic patients unless they were presenting with shock.5
The antimicrobials selected should be dosed based on published pharmacokinetic and pharmacodynamic information to ensure safe and effective use.4 It is important to note that because vancomycin and beta-lactams have time dependent antimicrobial activity, extended infusion times may enhance therapeutic effect in patients with augmented renal clearance in early sepsis.4 Once this initial treatment is selected, it should be re-assessed daily for opportunities to optimize coverage.4 Narrowing of therapy should occur based on culture and sensitivity results, or if unavailable, on site of infection and other patient specific factors.4 Source control, which is physical removal of infection such as draining an abscess, and site and etiology of infection, as well as the patient’s response to treatment should guide duration of treatment for pediatric sepsis, although the 2016 adult guidelines suggest a duration of 7-10 days unless clinical evidence suggests that a longer or shorter course could be indicated.4,5
Sepsis can cause an array of physiologic changes which may lead to hemodynamic instability and poor organ perfusion. Fluids are therefore another important component of the initial management of a septic patient. Although caution is required in patients presenting with or developing signs of fluid overload, generally in a health care setting with access to intensive care resources, patients should receive up to 40-60mL/kg in 10-20mL/kg doses over the first hour of treatment.4 In settings where intensive care is unavailable, the maximum volume of bolus fluid a child should receive is 40mL/kg in the first hour if hypotension is present, and if hypotension is not present, fluids should be given at a maintenance rate rather than administered as boluses.4 Signs to assess for hydration status include blood pressure, capillary refill and heart rate.4 Caution is required to avoid fluid overload which may present as pulmonary edema or worsening hepatomegaly.4
While most recommendations on selection of fluids for resuscitation are weak, generally balanced buffered crystalloids such as Lactated Ringer’s or PlasmaLyte solutions are recommended over albumin, normal saline, starches, or gelatin-derived fluids.4 These recommendations balance both cost and safety with cost being the driving factor in recommending against albumin, and safety being the driver of the only strong recommendation to avoid starches as adult data has shown significant risk of both AKI and mortality with the use of hydroxyethyl starch. The adult guidelines, however, give the option of normal saline in place of balanced crystalloids, and recommend use of albumin if patients are requiring significant crystalloids.5 The recommendation to use balanced crystalloids such as Lactated Ringer’s over normal saline in the pediatric guidelines is based on risk of hyperchloremic acidosis, AKI, and coagulopathy with the increased chloride content of normal saline.4,8 Interestingly, the American Academy of Pediatrics clinical practice guideline for maintenance intravenous fluid in children recommends isotonic fluids for maintenance and specifically mentions normal saline and PlasmaLyte, while excluding Lactated Ringer’s solution from their discussion.9 Further studies have evaluated the risks of the supraphysiologic chloride concentrations in normal saline and found that the risk is more significant in bolus fluid administration than slower administration rates, so outside of the bolus phase of treatment switching to normal saline is a reasonable choice although the selection of this fluid for boluses is less clear.8 PlasmaLyte is similar though not identical to Lactated Ringer’s, but is used less often secondary to cost. Per the guidelines, albumin is not a preferred fluid in children secondary to evidence that there is not a mortality benefit and there is a significant cost difference between albumin and crystalloids.4
For septic shock resistant to fluids or in children who are already at risk of or showing signs of fluid overload, epinephrine and norepinephrine are good first line options recommended over dopamine for blood pressure support.4 This is a change from 2008 and 2012 pediatric guidelines when dopamine and inodilators were recommended prior to initiation of norepinephrine or epinephrine.6,7 While central access is preferred for vasoactive medications, peripheral administration is reasonable during stabilization if central access is not available, and the 2008 and 2012 pediatric guidelines further recommend intra-osseous access if central access is not readily available.4,6,7 If children are on high doses of catecholamines and still in shock, vasopressin may be added, and if hypoperfusion is still present, addition of an inodilator such as milrinone or dobutamine may be a reasonable option.4 This is similar though not identical to the 2016 adult guidelines which recommended norepinephrine first line followed by epinephrine or vasopressin then inodilators.5
Some patients may also require hydrocortisone. Hydrocortisone is a corticosteroid that decreases inflammation and reverses capillary permeability. In both the 2016 adult guidelines and the 2020 pediatric guidelines, hydrocortisone is recommended in patients who are still hemodynamically unstable after administration of fluids and vasoactive agents, but not those who are stable on those medications.4,5 While hydrocortisone can enhance the activity of norepinephrine by decreasing reuptake, steroids have numerous adverse effects including immunosuppression, hyperglycemia, and neuromuscular weakness, so routine use is clearly not benign.4 Additionally, while hydrocortisone is an evidence-based treatment for primary adrenal insufficiency, stimulation testing or random cortisol levels in the setting of sepsis are not recommended unless there is evidence to enhance concern for adrenal insufficiency such as significant unexplained hyponatremia and hypoglycemia.4
Continued Care: Drugs with and without Routine Indications
Patients with sepsis may also require intubation. In these patients, etomidate is not a recommended induction agent secondary to risk of adrenal insufficiency following etomidate exposure.4 The 2008 pediatric guidelines were a bit more specific discouraging etomidate only in meningococcal sepsis.7 This recommendation was broadened to address all pediatric sepsis patients in the 2012 guidelines in which a mortality benefit to avoiding etomidate in meningococcal sepsis was mentioned, and avoiding both etomidate and dexmedetomidine in all sepsis patients was recommended secondary to risk of adrenal suppression.6 The recommendation to avoid dexmedetomidine was not carried forward into the 2020 pediatric sepsis guidelines.4 Future evidence seeks to strengthen recommendations regarding length of neuromuscular blockade for intubated patients which is currently a grey area. Life-threatening Acute Respiratory Failure in Children: to Breathe or Not to Breathe Spontaneously, That’s the Question, is a current study in the Netherlands which started in December 2019 to evaluate rocuronium versus placebo in mechanically ventilated children.10 This study is expected to be completed in mid-2024.10
While nutrition is important in recovery from any illness, enteral nutrition, particularly gastric nutrition, is preferred over parenteral nutrition for both adult and pediatric sepsis patients.4 The 2008 pediatric guidelines did not address nutrition, but the 2012 pediatric guidelines agree that enteral is preferred if tolerated and additionally address that if not tolerated parenteral feeding with dextrose 10% in a sodium containing solution is recommended.6 If patients are not tolerating full enteral feeds, prokinetic agents such as erythromycin or metoclopramide may be tempting options, however this is a practice that is not evidence-based, not risk-free, and not guideline recommended for pediatric patients despite being recommended in the 2016 adult guidelines.4,5 Even if patients are not tolerating full enteral feeds, total parenteral nutrition (TPN) is not recommended during the first seven days in the PICU.4 Delaying initiation of TPN may improve neurocognitive development without negative impacts on survival or health status.4 The guidelines are not designed to fully address neonatal sepsis, so neonates expected to be unable to tolerate enteral feeding should be an exception to this seven-day delay in TPN. While the benefit of focusing on enteral nutrition seems clear, blood glucose and calcium goals are less well-defined. A blood glucose goal of less than 140mg/dL is not recommended in order to avoid hypoglycemic events, however the pediatric sepsis guidelines do not define a specific blood glucose target.4 American Diabetes Association guidelines for care of hospitalized diabetic patients recommend a blood glucose target of 140-180mg/dL in most critically ill and non-critically ill patients, so this range is a reasonable target.11 This is corroborated with the 2016 adult guidelines which suggest initiating insulin if the blood glucose is greater than 180mg/dL as opposed to more stringent blood glucose targets and the 2012 pediatric guidelines recommending a blood glucose goal less than 180mg/dL.5,6 Calcium goals are even more poorly defined. The prevalence of hypocalcemia in pediatric sepsis may be as high as 75%, yet calcium supplementation has not been evaluated nor is there consensus in the pediatric healthcare community on the management of hypocalcemia in these patients.4 This is further muddied as in the adult critical care population, calcium supplementation has been associated with worsening organ dysfunction, yet in the pediatric population it is known that calcium is associated with improved hemodynamics, particularly in infants with immature cardiomyocytes.4,12
The pediatric sepsis guidelines specifically recommend against certain nutritional supplements. These include enteral lipid emulsions such as fish oil supplements, selenium, glutamine, arginine, zinc, vitamin C, and thiamine.4 Studies have not found benefit to using these supplements, and some studies have even found harm with arginine.4 Additionally, acute vitamin D replacement is not recommended despite vitamin D deficiency being associated with organ dysfunction.4 While this seems counter-intuitive, vitamin D levels appear falsely low post-resuscitation, and hypervitaminosis D can cause complications leading to death.4 If patients are known to be vitamin D deficient prior to onset of sepsis it is recommended to provide supplementation as recommended outside of sepsis, but measuring of levels during sepsis to guide treatment would not be an acceptable standard of care.4 The 2016 adult guidelines do not comment on as many supplements, but do agree in recommending against selenium, arginine, and glutamine.5 The adult guidelines were unable to make a recommendation for or against carnitine, which is in agreement with pediatric guidelines as carnitine is not mentioned.5
There is also limited evidence on temperature management in pediatric sepsis patients. The guidelines discuss use of antipyretics or a permissive strategy for fevers.4 Elevated temperature dose have some positive effects on the immune system which are lost if temperature is reduced with antipyretics, however it is not benign, which is most obvious in terms of patient comfort.4 Additionally, comparing antipyretics to physical cooling strategies favors antipyretics for early mortality, so it is clear that cooling is generally not a recommended approach in a pediatric sepsis patient.4 Acetaminophen and ibuprofen are commonly used antipyretics, but acetaminophen is not recommended in patients less than 3 months of age and ibuprofen in patients less than 6 months of age as the difference between effective and toxic doses is much smaller in these patients leading to heightened risk of adverse events. These are not absolute contraindications, and in fact a study in 2018 found that ibuprofen use in patients less than 6 months of age was not associated with greater adverse events than use in patients over 6 months.13 Risk of gastrointestinal and renal adverse events was heightened modestly in patients receiving ibuprofen versus those receiving only acetaminophen, but this risk difference was the same in both age groups.13 However, it is still prudent to use these medications cautiously in our youngest patients.
In sepsis patients, T3 and T4 level turnover increases and de-iodination of T4 to T3 decreases.4 Additionally, TSH may not be elevated, however, levothyroxine is not recommended for routine use in children with sick euthyroid.4 Although theoretically there may be a role for levothyroxine, studies in critically ill children have found no difference in patients treated with levothyroxine as compared to those who did not receive levothyroxine.4
Pediatric stress ulcer prophylaxis is a grey area as little research has been published in this area. Hospitals have published protocols, but there are no recent guidelines. A recent observational study found that age, mechanical ventilation, and not receiving enteral nutrition were all independently associated with increased likelihood of the patient receiving stress ulcer prophylaxis.14 These authors also acknowledged that use of vasopressors, corticosteroids, and NSAIDs, presence of coagulopathy, and higher PRISM III score on admission may also be factors in determining need for stress ulcer prophylaxis, but there was not a uniform prescribing practice.14 As stress ulcer prophylaxis is not benign, potentially leading to increased risk of pneumonia and Clostridium difficile infections as well as decreased bone mineral density with longer durations of treatment, it is important to consider risks and benefits of stress ulcer prophylaxis in determining which patients require these medications. It is clear that simply being diagnosed with sepsis or even septic shock is not enough to warrant use of stress ulcer prophylaxis.4 While there is better data in the adult population, the adult guidelines agree that stress ulcer prophylaxis is only indicated if patients have risk factors that would indicate requiring stress ulcer prophylaxis outside of sepsis.5 The 2008 and 2012 pediatric guidelines had no graded recommendations for or against stress ulcer prophylaxis.6,7 As histamine 2 receptor antagonists are more studied in the pediatric population than proton pump inhibitors and have lower rates of adverse events, choosing histamine 2 receptor antagonists over proton pump inhibitors is a good choice though drug selection is not addressed in the sepsis guidelines.
Venous thromboembolism (VTE) prophylaxis is not routinely indicated in pediatric sepsis patients as VTE is much less common in pediatric patients than it is in adults, but some patients may benefit from VTE prophylaxis.4 This is in contrast to the adult guidelines which recommend both mechanical and pharmacologic VTE prophylaxis in sepsis patients.5 Pediatric sepsis patients most at risk for VTE and therefore most likely to benefit from VTE prophylaxis are those who are post-pubertal and have a central venous catheter.4 As much data on pediatric VTE has been extrapolated from adult data, other risk factors are similar in the two populations including obesity, cancer, use of exogenous estrogens, and renal and cardiac comorbidities.15 Previous pediatric guidelines had similar grey areas in terms of VTE prophylaxis with the 2008 guidelines stating to use VTE prophylaxis in post-pubertal children and primarily addressing heparin bonded central venous catheters rather than administration of LMWH or UFH, but there was no mention of management of children who have not yet reached puberty.7 The 2012 update simply stated no graded recommendations for VTE prophylaxis in pre-pubertal children, again discussed heparin bonded catheters, and did not mention post-pubertal children.6 The medication most often used for VTE prophylaxis in pediatric patients is enoxaparin.
Intravenous immune globulin (IVIG) may also be indicated in select patients because it enhances passive immunity but evidence has not supported routine use of IVIG in patients with sepsis.4 While IVIG is not routinely recommended for sepsis in pediatric patients, there are a subset of patients for whom IVIG may be a reasonable treatment option.4 Patients with streptococcal toxic shock syndrome are likely to benefit from IVIG.4 Additionally, it may be beneficial to consider IVIG in immunodeficient or immunocompromised patients with low immunoglobulin levels and patients with necrotizing fasciitis.4 While the adult guidelines recommend against IVIG without this caveat of potentially appropriate use, they also suggest that further research may be needed to evaluate the efficacy of IVIG, so with four additional years of research these two guideline statements are essentially in agreement.5 The 2012 pediatric guidelines stated that use of IVIG in toxic shock was unclear, which is a bit incongruent with 2008 pediatric guidelines which suggested IVIG in patients with severe sepsis, not just those with toxic shock.6,7
Medications that no longer appear in the pediatric sepsis guidelines but have previously been recommended against include recombinant human activated protein C in the 2008 guidelines secondary to findings in clinical trials of no difference between patients receiving activated protein C versus placebo.7 Based on these findings it would be reasonable to assume that despite not being mentioned, continuing to not utilize activated protein C would be appropriate. The 2012 guidelines recommended against long-term propofol for sedation in patients under three years of age secondary to concerns for metabolic acidosis.6 This is another recommendation that while not being mentioned in other iterations of the guidelines continues to seem prudent to follow as it is known that propofol related infusion syndrome is a risk for patients of all ages with extended use of propofol, making propofol a less than ideal agent for patients requiring long-term sedation.
Organ dysfunction is useful not only in differentiating sepsis from severe sepsis, but also in considering changes to medication pharmacokinetics. When shock is present antibiotics should be given within one hour. Fluids are an important component of care even in normotensive patients and balanced buffered crystalloids are recommended. Unless patients require hydrocortisone at baseline, it should only be given if fluids and vasoactive medications are insufficient to maintain hemodynamic stability. Nutrition should be provided enterally if possible, and many supplements including zinc, thiamine, and vitamin C are not routinely indicated. Finally, it is important to carefully consider indications for IVIG, stress ulcer prophylaxis, and VTE prophylaxis.