Reducing Invasive Care for Low-riskFebrile Infants ThroughImplementation of a Clinical PathwayKathryn E. Kasmire, MD, MS,a,b,c Eric C. Hoppa, MD,a,d Pooja P. Patel, BS,b Kelsey N. Boch, MD,b Tina Sacco, RN, BS, CPHQ,a

Ilana Y. Waynik, MDa,d

abstractBACKGROUND AND OBJECTIVES: Significant variation in management of febrile infantsexists both nationally and within our institution. Risk stratification can beused to identify low-risk infants who can be managed as outpatients withoutlumbar puncture (LP) or antibiotics. Our objective was to reduce invasiveinterventions for febrile infants aged 29 to 60 days at low risk for seriousbacterial infection (SBI) through implementation of a clinical pathwaysupported by quality improvement (QI).

METHODS: The evidence-based clinical pathway was developed andimplemented by a multidisciplinary team with continuous-process QI tosustain use. Low-risk infants who underwent LP, received antibiotics, andwere admitted to the hospital were compared pre- and postpathwayimplementation with SBI in low-risk infants and appropriate care for high-risk infants as balancing measures.

RESULTS: Of 350 included patients, 220 were pre- and 130 were postpathwayimplementation. With pathway implementation in July 2016, invasiveinterventions decreased significantly in low-risk infants, with LPs decreasingfrom 32% to 0%, antibiotic administration from 30% to 1%, and hospitaladmission from 17% to 2%. Postimplementation, there were 0 SBIs in low-risk infants versus 29.2% in high-risk infants. The percentage of high-riskpatients receiving care per pathway remained unchanged. Improvement wassustained for 12 months through QI interventions, including order-setdevelopment and e-mail reminders.

CONCLUSIONS: Implementation of a clinical pathway by using QI methodsresulted in sustained reduction in invasive interventions for low-risk febrileinfants without missed SBIs. Clinical pathways and QI can be key strategies inthe delivery of evidence-based care for febrile infants.

Fever is a common reason foremergency department (ED) visits ininfants.1,2 Fever in infants can bea presenting sign of a serious bacterialinfection (SBI), which must beidentified and treated promptly toprevent morbidity and mortality.Numerous strategies for the riskstratification of febrile infants havebeen proposed and studied, with

common criteria being referred to asthe Rochester, Boston, and Philadelphiacriteria.3–7 These criteria vary in regardto use of lumbar puncture (LP),empirical antibiotic administration, andhospital admission, with LP andantibiotics not being recommended forlow-risk infants by Rochestercriteria.3,4,8 All of these algorithms havegood sensitivity for SBI but lack

aConnecticut Children’s Medical Center, Hartford,Connecticut; dDepartment of Pediatrics, bSchool of Medicine,University of Connecticut, Farmington, Connecticut; andcPenn State Health Milton S. Hershey Medical Center,Hershey, Pennsylvania

Dr Kasmire conceptualized and designed the study,conducted the initial analyses, drafted the initialmanuscript, and reviewed and revised themanuscript; Drs Hoppa and Waynik conceptualizedand designed the study, conducted the initialanalyses, coordinated and supervised datacollection, and reviewed and revised the manuscript;Dr Boch, Ms Patel, and Ms Sacco designed the datacollection instruments, collected data, assisted withthe initial analyses, and critically reviewed themanuscript; and all authors approved the finalmanuscript as submitted and agree to beaccountable for all aspects of the work.

DOI: https://doi.org/10.1542/peds.2018-1610

Accepted for publication Nov 29, 2018

Address correspondence to Kathryn E. Kasmire, MD,MS, Penn State Health Milton S. Hershey MedicalCenter, 500 University Dr, H043, Hershey, PA 17033.E-mail: kkasmire@pennstatehealth.psu.edu

PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online,1098-4275).

Copyright © 2019 by the American Academy ofPediatrics

FINANCIAL DISCLOSURE: The authors have indicatedthey have no financial relationships relevant to thisarticle to disclose.

FUNDING: No external funding.

POTENTIAL CONFLICT OF INTEREST: The authors haveindicated they have no potential conflicts of interestto disclose.

To cite: Kasmire KE, Hoppa EC, Patel PP, et al.Reducing Invasive Care for Low-risk FebrileInfants Through Implementation of a ClinicalPathway. Pediatrics. 2019;143(3):e20181610

PEDIATRICS Volume 143, number 3, March 2019:e20181610 QUALITY REPORT by guest on October 2, 2020www.aappublications.org/newsDownloaded from

specificity, resulting in a number ofinfants without SBI undergoingextensive testing and treatment.Although the risks of missing SBI aresignificant compared with the risks ofevaluation and treatment, aggressiveevaluation and intervention can causeunnecessary pain to infants, stress tocaregivers, adverse events, andsignificant cost to the health caresystem.9 The risks associated withevaluation and empirical treatmentlikely contribute to the significantvariability in care provided to febrileinfants, resulting in many febrileinfants not receiving appropriateevaluation for SBI.10–13

The epidemiology of SBIs in younginfants has changed since thedevelopment of the most commonlyused criteria with a predominanceof urinary tract infection(UTI)–associated pathogens,a decrease in bacteremia, and anoverall low incidence of bacterialmeningitis.14,15 In keeping with thischanging epidemiology, evaluation forUTI and bacteremia remain essential;however, criteria exist to identify low-risk infants in whom LP may beavoided. Growing evidence revealsthat the use of stepwise evaluation byusing risk stratification without initialLP in infants ages 29 to 60 days issafe, with LP and empirical antibioticadministration being reserved forhigh-risk infants.8,16,17 In particular,the Rochester criteria, with the use ofwhite blood cell count, band count,and urinalysis to risk stratify infants,has been shown to have $97%negative predictive value forSBI.3,4,18–21

Clinical pathways are a usefulstrategy to improve the deliveryof evidence-based care, withnumerous examples of successfulimplementation seen in the pediatricED setting.22–25 Given the widespreadvariation in the management offebrile infants, the use of a clinicalpathway could be a beneficialstrategy to reduce variability.Byington et al17 showed that

implementing a pathway based onmodified Rochester criteria decreasedoverall cost while improvingappropriate testing and treatment ofhigh-risk infants without increasingmissed SBIs. Similar pathways are inuse at other institutions.17,26–28

At our institution, a clinical pathwayfor the evaluation and management ofneonates #28 days old with feverwas implemented in 2014, but themanagement of febrile infants.28 days of age was notstandardized; providers at ourinstitution varied greatly in their careof this group. In particular, care forlow-risk infants had high variability,with low-risk infants undergoing LP100% of the time in some monthsversus 0% in other months, andmonthly antibiotic use ranged from0% to 100% (Fig 1). To improve carefor febrile infants aged 29 to 60 days,an evidence-based clinical pathwaywas developed and implemented atour institution, and continuousquality improvement (QI) was used tosupport the implementation of thepathway, including promoting useand sustainability. Our aim inimplementation was to reduce LP,antibiotic administration, andhospital admission for infants at lowrisk for SBI to ,10% in 12 monthswithout missing SBIs.

METHODS

Context

This QI project was conducted ata single urban, tertiary children’shospital with an annual ED volume of61 000 patients. Our ED evaluatesa number of infants with fever,including patients referred fromprimary care providers and transfersfrom numerous community EDs.Many infants are admitted for furthercare, whereas some are dischargedif pediatrician follow-up can bearranged. The institution usesnumerous other clinical pathways,which are reviewed and approved bya clinical effectiveness committee.

This QI project was deemed exemptfrom full review by our institutionalreview board.

Study Population

The study population consisted offebrile infants aged 29 to 60 daysmeeting inclusion criteria for thefebrile-infant clinical pathway: fever$38.0°C (rectal; before arrival or inthe ED) and gestational age $37weeks. Exclusions includedevaluation initiated at an outsideED or recent previous ED visit,history of immunodeficiency,identified focal infection, underlyingchronic medical disease, currentantibiotic therapy, gestational age,37 weeks, or a clinical diagnosis ofbronchiolitis.

Planning the Intervention

Before our QI initiative, significantvariability existed in care for febrileinfants aged 29 to 60 days at ourinstitution. To standardize andimprove care for febrile infants, anevidence-based clinical pathway(based on modified Rochestercriteria)3,17,18 was developed overa period of 17 months (starting inFebruary 2015) with input from keystakeholders from the divisions ofemergency medicine, hospitalmedicine, and infectious disease. Theinitial development phase engagedstakeholders in the ED, includingpediatric emergency medicinephysicians and fellows, and pediatrichospitalists. Journal club sessionswere held for hospitalists and EDphysicians (in February 2015 andApril 2015, respectively) to reviewevidence in febrile-infantmanagement, and after discussion,consensus was reached on a pathwayinvolving risk stratification forSBI based on modified Rochestercriteria (shown in Fig 2).Collaboration between ED andhospitalist providers during thedevelopment period allowed forthe creation of a multidisciplinaryteam and coordination of careacross the hospital continuum. Final

2 KASMIRE et al by guest on October 2, 2020www.aappublications.org/newsDownloaded from

FIGURE 1Statistical process control chart with the percentage of low-risk infants who received an invasive intervention. A, Percentage of low-risk infants whounderwent LP. B, Percentage of low-risk infants who received an antibiotic. C, Percentage of infants who were admitted to the hospital. Dotted linesrepresent control limits.

PEDIATRICS Volume 143, number 3, March 2019 3 by guest on October 2, 2020www.aappublications.org/newsDownloaded from

FIGURE 2Clinical pathway for the evaluation and treatment of infants aged 29 to 60 days with fever. ALT, alanine aminotransferase; AST, aspartate aminotrans-ferase; CBC, complete blood cell count; CSF, cerebrospinal fluid; hx, history; HSV, herpes simplex virus; IM, intramuscularly; IV, intravenously; PCP, primarycare provider; PCR, polymerase chain reaction; PE, physical exam; q6h, every 6 hours; q8h, every 8 hours; q12h, every 12 hours; RSV, respiratory syncytialvirus; sxs, symptoms; WBC, white blood cell count.

4 KASMIRE et al by guest on October 2, 2020www.aappublications.org/newsDownloaded from

pathway review and approvalwas obtained following standardhospital process through theinstitution’s clinical effectivenesscommittee. A hospital-wide grandrounds presentation in January2016 on febrile-infant managementwas given, which was focused onpresentation of the proposedpathway and the supportingevidence.

Interventions

The clinical pathway (Fig 2) wasimplemented in July 2016. As withall of our clinical pathways, aneducational PowerPoint presentationwas provided to pathway end usersvia e-mail and posted on the hospitalintranet at time of pathwayimplementation. QI interventionsoccurred in plan-do-study-act cyclesto improve and sustain pathwayadherence, including implementationof a pathway order set in theelectronic medical record inNovember 2016, monthly e-mailreminders about the pathway forresidents rotating through the ED(starting in January 2017), anda ceftriaxone order panel (which wasintroduced in April 2017).

Study of the Interventions

Evaluation and management of febrileinfants aged 29 to 60 days wasassessed retrospectively for 2 yearsbefore implementation (from May2014 to June 2016) and comparedwith the postimplementation period,which was assessed prospectively for1 year. Data collected includedcompletion of LP, administration ofantibiotics, hospital admission, andadherence to modified Rochestercriteria as presented in ourinstitution’s pathway (Fig 2), withadherence being defined as thecollection of all necessarylaboratories and/or cultures, correctuse of LP, the correct use ofantibiotics, and hospital admissionwhen indicated. Additional data werecollected as necessary for riskstratification (white blood cell andband counts, urinalysis results, andgeneral appearance) as well theultimate presence of SBI anddemographic information. Afterpathway implementation, data werecollected weekly by our qualitydepartment and reviewed bya collaborative ED and hospitalistteam. A clinical data coordinatordesignated patients as low or highrisk on the basis of the treating

provider’s documentation andlaboratory results. If a discrepancywas found, physician consensus wasused (according to the pathwaydefinitions of low and high risk). Datawere analyzed periodically to driveadditional plan-do-study-act cyclesand interventions.

Eligible patients were identified fromall ED patients 29 to 60 days of age byusing chief complaint of fever ora relevant International Classificationof Diseases, 10th Revision code(including fever, UTI, bacteremia,meningitis, and pneumonia). Chartswere reviewed for inclusion andexclusion criteria.

Measures

The primary outcome measure wasthe percentage of low-risk infantswho received LP or antibiotics orwere admitted to the hospital. Beforepathway implementation, there wasfrequent use of LP and antibioticadministration for infants who couldbe classified as low risk (Fig 3A). Thespecific goal was to decrease LP,antibiotic use, and hospital admissionfor low-risk infants aged 29 to60 days to ,10% by increasing thepercentage of febrile infants managedaccording to the clinical pathway

FIGURE 3A, Pareto chart of reasons for deviation from modified Rochester criteria preimplementation. B, Pareto chart of reasons for deviation from modifiedRochester criteria postimplementation. Cumulative percentages are plotted as solid lines, and bars reveal the total number for each reason. The left-handaxes have different scales.

PEDIATRICS Volume 143, number 3, March 2019 5 by guest on October 2, 2020www.aappublications.org/newsDownloaded from

from 60% to 90% in 12 months. EDlength of stay (LOS) was assessed asa secondary outcome. Processmeasures included pathwayadherence and order-set use.Pathway adherence was defined ashaving all of the following: collectionof complete blood count, urinalysis,and blood and urine cultures as wellas appropriate use of LP, antibiotics,and hospital admission based on riskclassification.

SBI in low-risk infants wasassessed pre- and postpathwayimplementation as a balancingmeasure, with SBI being defined asa true-positive blood or cerebrospinalfluid culture result or a positive urineculture result ($100000 colony-forming units [CFUs] per mL ofa pathogenic bacteria or $10 000CFUs/mL with positive urinalysisresults). The lower cutoff for urinecultures was used because ourlaboratory does not specify countsbetween 10000 and 100000 CFUs/mL,and recent studies have revealedthat true UTIs in infants commonlyhave counts ,100000 CFUs/mL.29,30

To detect a doubling of SBIs in low-risk infants with 80% power and95% confidence, a sample size of∼3000 would be required, which wasnot feasible in our study time frame.In light of this limitation, we alsotracked care for high-risk infants (LP,antibiotics, and hospital admission)as balancing measures.

Analysis

Demographic and clinicalcharacteristics, LOS, processmeasures, and balancing measureswere compared between the pre- andpostimplementation groups by usingt tests and x2 tests or Fisher’s exacttests for comparisons with n ,5 andwith statistical significance set ata = .05. Statistical process controlcharts were used to assess the impactof the intervention on outcome andprocess measures. Control limits wereset at 3 SDs from the mean. Standardrules were used to determine special

cause variation, including $8 valuesabove the baseline centerline.31

RESULTS

Outcome Measures

During the 3-year study period, 350febrile infants aged 29 to 60 dayswere included, 220 before (62.9%)and 130 after (37.1%) pathwayimplementation (Table 1). Weexcluded 120 infants (total pre- andpostpathway), most commonlybecause of prematurity (n = 52) andbronchiolitis (n = 43). There wereno significant differences indemographic and clinicalcharacteristics between the pre- andpostimplementation groups,including in age, sex, maximumtemperature, proportion of low-riskinfants, or prevalence of SBI(Table 1).

Before pathway implementation, low-risk infants frequently underwenttesting and treatment not indicatedby modified Rochester criteria (Fig 3).The most common reasons fordeviation from modified Rochestercriteria were LP performed andantibiotics administered in low-riskinfants and a lack of hospitaladmission for high-risk infants(Fig 3A). Of 136 low-risk infantsbefore pathway implementation, 44(32%) underwent LP, 41 (30%)received antibiotics, and 23 (17%)were admitted to the hospital.Improvements were seen in allmeasures after pathwayimplementation, with 0 (0%) low-riskinfants undergoing LP, 1 (1%)receiving antibiotics, and 2 (2%)being admitted to the hospital out of

82 low-risk infants, all of which werebelow the aim of 10% (Fig 1). Thecontrol charts show special causevariation at the time of pathwayimplementation with decreasedinterventions after pathwayimplementation for LP, antibiotics,and admission for low-risk infants(Fig 1). LPs, antibiotic use, andhospital admission remained at lowrates after pathway implementation,during which time additional QIcycles were conducted, includingorder-set implementation, e-mailreminders, and a ceftriaxone orderpanel to further improve pathwayadherence (Figs 1 and 4).

Average ED LOS decreased from295 minutes before pathwayimplementation to 272 minutes afterimplementation for all patients(P = .03). For patients dischargedfrom the ED, LOS decreasedsignificantly from 279 minutesbefore implementation to 237minutes after (P , .001). ED LOSincreased for admitted patients froma mean of 320 minutes beforeimplementation to 359 minutes after(P = .01).

Process Measures

Modified Rochester criteria werefollowed more often after pathwayimplementation, increasing from55.4% at baseline to 76.2% afterpathway implementation (P , .001;Fig 4). The most common deviationsfrom the pathway afterimplementation were in the care ofhigh-risk infants, including incorrectantibiotic dosing and a lack ofhospital admission when indicated(Fig 3B). The statistical process

TABLE 1 Demographic and Clinical Characteristics

Characteristics Prepathway (N = 220) Postpathway (N = 130) P

Age, d, mean (IQR) 45.3 (38–53) 45.4 (38–52) .92Male sex, n (%) 122 (55.5) 74 (56.9) .88Tmax, °C, mean 38.5 38.5 .82Low risk, n (%) 136 (61.8) 82 (63.1) .90SBI, n (%) 19 (8.6) 14 (10.8) .64SBI in high-risk infants, n (%) 18 (21.4) 14 (29.2) .62

Tmax is measured either in the ED or before arrival. IQR, interquartile range; Tmax, maximum temperature.

6 KASMIRE et al by guest on October 2, 2020www.aappublications.org/newsDownloaded from

control chart for pathway adherencereveals an upward shift in meanadherence in November 2015 beforepathway implementation (whichcould be attributable to educationalinterventions occurring beforepathway implementation) and againafter implementation in November2016, indicating special causevariation (Fig 4). An order setbecame available in the electronicmedical record in November2016 and was used for 32% ofpatients postpathwayimplementation.

Balancing Measures

After pathway implementation, therewere no missed SBIs. There was 1missed SBI before pathwayimplementation: an infant classifiedas low risk on the basis of wellappearance and low-risk laboratoryresults then diagnosed withCitrobacter bacteremia, which was

subsequently treated successfully. Inthe postimplementation period, 1questionable case was ultimatelyconsidered a contaminant (urineculture with 10 000 CFUs/mLEnterococcus faecalis with traceleukocyte esterase on urinalysis)because the infant recovered quicklywithout antibiotics.

Appropriate care for high-riskinfants remained consistent despiteour effort to reduce invasiveinterventions for low-risk infants.For infants classified as high riskby using our pathway, rates of allrecommended interventions were notsignificantly different after pathwayimplementation: LP performed in73% vs 83%, appropriate antibioticadministered in 77% vs 85%, andhospital admission for 70% vs 77%pre- and postimplementation,respectively (P . .05 for allcomparisons).

DISCUSSION

By using continuous-process QI tosupport the implementation of anevidence-based clinical pathway, wesuccessfully achieved our goal ofreducing LPs, antibiotic use, andhospital admission for febrile infantswho are at low risk for SBI on thebasis of modified Rochester criteria.This was achieved without detrimentto the identification of SBIs or carefor high-risk infants. Multiple QIinterventions contributed tosustained improvement during the 1-year postimplementation period.

This QI initiative had a number ofstrengths that contributed toachieving our goal. Most importantly,engagement of ED stakeholders(including pediatric emergencymedicine physicians and fellows) andpediatric hospitalists occurredthroughout the pathway developmentperiod. The plan for pathway

FIGURE 4Statistical process control chart with adherence to modified Rochester criteria pre- and postpathway implementation for all infants, including low- andhigh-risk infants. Dotted lines represent control limits.

PEDIATRICS Volume 143, number 3, March 2019 7 by guest on October 2, 2020www.aappublications.org/newsDownloaded from

development was initiated bya multidisciplinary team, educationon the evidence and rationale waspresented at several stages ofdevelopment, and input from bothpediatric emergency medicinephysicians and pediatric hospitalistswas sought and incorporated. Whenthe pathway was implemented, thestandardized hospital process forpathway use was followed, allowingfor dissemination throughout thehospital. In addition, prospectiveweekly tracking of outcomes afterimplementation allowed forcontinuous process improvement andfurther interventions aimed atsustaining improvement.

Our results emphasize the utility ofthe implementation of a clinicalpathway supported by QI efforts toachieve improvement in providingevidence-based care. Although we didnote improvements in care for low-risk febrile infants before the studyinitiation and during the pathwaydevelopment period, improvement inreductions in invasive care for low-risk infants reached a consistent levelonly with pathway implementation,which was sustained throughadditional QI cycles (Figs 1 and 4).Some of the discussion and educationfor providers (including journal clubsand grand rounds) may have led tothe adoption of modified Rochestercriteria before pathwayimplementation.

Previous studies have also revealedbenefit in the use of clinical pathwaysfor the reduction of unnecessarytesting or interventions in pediatrics,including reducing broad-spectrumantibiotic use for pneumonia,computed tomography scans forappendicitis and head injury,catheterization to test for UTIs, anddecreased use of antibiotics for low-risk febrile infants.22,25,32–34

The implementation of the clinicalpathway reduced ED LOS for patientswho were discharged from the ED.Factors that may have contributed to

this include earlier initiation ofworkup by residents due to comfortwith pathways and order sets,reduced time spent on LPs andantibiotic administration, and earlierdisposition due to the standardizationof risk-stratification and admissioncriteria. Although we did not trackthis during the study period, standingnursing triage orders aligned with thepathway allowed for nurses to obtainblood and urine testing beforeprovider evaluation. Evaluation ofnursing triage order use couldprovide further opportunity forimprovement in LOS. However, EDLOS remained high for admittedpatients, and although outside factorsunrelated to the pathway (such asinpatient bed availability) may havecontributed, the improvement fordischarged patients may have come atthe expense of longer stays foradmitted patients, possibly due towaiting for laboratory results to berisk stratified before performing LPrather than completing all tests oninitial evaluation.

Care for low-risk infants waspositively impacted by our QIinitiative. However, room forimprovement remains for high-riskinfants, with persistent variability inantibiotic dosing and administration,use of LP, and hospital admissionbeing observed. In addressing thesedifferences in management, furtherdiscussion with providers regardingreasons for not following the pathwaywill help to plan furtherinterventions. The low specificity ofRochester and other risk-stratification criteria may encourageproviders to make exceptions forhigh-risk infants given the stillrelatively low likelihood of SBI.Newer risk-stratification tests, suchas procalcitonin, hold promise,20 andonce available, modification of ourpathway to incorporate advancescould help further standardize care.For antibiotic dosing variability,additional plan-do-study-act cycleswith interventions aimed at antibiotic

dosing standardization were initiated,including the development of anorder panel for ceftriaxone to aidin appropriate dose selection.Pathway order-set use remainedlow at the end of our tracking period,presenting another area forimprovement.

This study does have severalimportant limitations. Although wedid show significant improvementswith our QI interventions, includingpathway implementation, outsidefactors (such as national trendstoward fewer interventions forinfants .28 days old with fever) mayhave contributed to this improvementgiven the trend in reducedinterventions for low-risk infantsbefore the study’s conception. Thepathway was successfullyimplemented with no missed SBIsafter implementation; however, thestudy is not powered to detecta difference in missed SBIs. Intracking the results after pathwayimplementation, bias could have beenpresent in classifying patients ashigh-risk on the basis of appearance,although data collection by ourquality department may have helpedto minimize this bias, and theproportion of low- and high-riskpatients was similar pre- andpostpathway implementation.Because this project was conductedat an academic children’s hospital,the results may not be generalizableto other institutions orcommunity EDs.

CONCLUSIONS

The implementation of an evidence-based clinical pathway supported bycontinuous-process QI resulted ina sustained reduction in invasiveinterventions (LP, antibioticadministration, and hospitaladmission) for febrile infants aged 29to 60 days who were at low risk forSBI while reducing ED LOS andwithout missed SBIs. Our studyreveals that using QI processes to

8 KASMIRE et al by guest on October 2, 2020www.aappublications.org/newsDownloaded from

support clinical pathwayimplementation can be a key strategyin the delivery of evidence-based carefor febrile infants.

ABBREVIATIONS

CFU: colony-forming unitED: emergency departmentLOS: length of stayLP: lumbar punctureQI: quality improvementSBI: serious bacterial infectionUTI: urinary tract infection

REFERENCES

1. Krauss BS, Harakal T, Fleisher GR. Thespectrum and frequency of illnesspresenting to a pediatric emergencydepartment. Pediatr Emerg Care. 1991;7(2):67–71

2. Nelson DS, Walsh K, Fleisher GR.Spectrum and frequency of pediatricillness presenting to a generalcommunity hospital emergencydepartment. Pediatrics. 1992;90(1, pt 1):5–10

3. Dagan R, Powell KR, Hall CB, MenegusMA. Identification of infants unlikely tohave serious bacterial infectionalthough hospitalized for suspectedsepsis. J Pediatr. 1985;107(6):855–860

4. Dagan R, Sofer S, Phillip M, Shachak E.Ambulatory care of febrile infantsyounger than 2 months of ageclassified as being at low risk forhaving serious bacterial infections.J Pediatr. 1988;112(3):355–360

5. Baskin MN, O’Rourke EJ, Fleisher GR.Outpatient treatment of febrile infants28 to 89 days of age with intramuscularadministration of ceftriaxone. J Pediatr.1992;120(1):22–27

6. Baker MD, Bell LM, Avner JR. Outpatientmanagement without antibiotics offever in selected infants. N Engl J Med.1993;329(20):1437–1441

7. Baker MD, Bell LM, Avner JR. Theefficacy of routine outpatientmanagement without antibiotics offever in selected infants. Pediatrics.1999;103(3):627–631

8. Huppler AR, Eickhoff JC, Wald ER.Performance of low-risk criteria in theevaluation of young infants with fever:review of the literature. Pediatrics.2010;125(2):228–233

9. Paxton RD, Byington CL. An examinationof the unintended consequences of therule-out sepsis evaluation: a parentalperspective. Clin Pediatr (Phila). 2001;40(2):71–77

10. Pantell RH, Newman TB, Bernzweig J,et al. Management and outcomes ofcare of fever in early infancy. JAMA.2004;291(10):1203–1212

11. Aronson PL, Thurm C, Alpern ER, et al;Febrile Young Infant ResearchCollaborative. Variation in care of thefebrile young infant ,90 days in USpediatric emergency departments.Pediatrics. 2014;134(4):667–677

12. Greenhow TL, Hung YY, Pantell RH.Management and outcomes ofpreviously healthy, full-term, febrileinfants ages 7 to 90 days. Pediatrics.2016;138(6):e20160270

13. Jain S, Cheng J, Alpern ER, et al.Management of febrile neonates in USpediatric emergency departments.Pediatrics. 2014;133(2):187–195

14. Greenhow TL, Hung YY, Herz AM.Changing epidemiology of bacteremiain infants aged 1 week to 3 months.Pediatrics. 2012;129(3). Available at:www.pediatrics.org/cgi/content/full/129/3/e590

15. Greenhow TL, Hung YY, Herz AM, LosadaE, Pantell RH. The changingepidemiology of serious bacterialinfections in young infants. PediatrInfect Dis J. 2014;33(6):595–599

16. Baraff LJ. Management of fever withoutsource in infants and children. AnnEmerg Med. 2000;36(6):602–614

17. Byington CL, Reynolds CC, Korgenski K,et al. Costs and infant outcomes afterimplementation of a care processmodel for febrile infants. Pediatrics.2012;130(1). Available at: www.pediatrics.org/cgi/content/full/130/1/e16

18. Jaskiewicz JA, McCarthy CA, RichardsonAC, et al; Febrile Infant CollaborativeStudy Group. Febrile infants at low riskfor serious bacterial infection–anappraisal of the Rochester criteria and

implications for management.Pediatrics. 1994;94(3):390–396

19. Byington CL, Enriquez FR, Hoff C, et al.Serious bacterial infections in febrileinfants 1 to 90 days old with andwithout viral infections. Pediatrics.2004;113(6):1662–1666

20. Gomez B, Mintegi S, Bressan S, Da DaltL, Gervaix A, Lacroix L; European Groupfor Validation of the Step-by-StepApproach. Validation of the “step-by-step” approach in the management ofyoung febrile infants. Pediatrics. 2016;138(2):e20154381

21. Garra G, Cunningham SJ, Crain EF.Reappraisal of criteria used to predictserious bacterial illness in febrileinfants less than 8 weeks of age.Acad Emerg Med. 2005;12(10):921–925

22. Murray AL, Alpern E, Lavelle J, Mollen C.Clinical pathway effectiveness: febrileyoung infant clinical pathway ina pediatric emergency department.Pediatr Emerg Care. 2017;33(9):e33–e37

23. DeLaroche AM, Sivaswamy L, Farooqi A,Kannikeswaran N. Pediatric strokeclinical pathway improves the time todiagnosis in an emergency department.Pediatr Neurol. 2016;65:39–44

24. Mohan S, Nandi D, Stephens P, MʼFarrejM, Vogel RL, Bonafide CP.Implementation of a clinical pathwayfor chest pain in a pediatric emergencydepartment. Pediatr Emerg Care. 2018;34(11):778–782

25. Ramarajan N, Krishnamoorthi R, BarthR, et al. An interdisciplinary initiative toreduce radiation exposure: evaluationof appendicitis in a pediatricemergency department with clinicalassessment supported by a stagedultrasound and computed tomographypathway. Acad Emerg Med. 2009;16(11):1258–1265

26. Scarfone R, Gala R, Murray A, Funari M,Lavelle J, Bell L; Children’s Hospital ofPhiladelphia. ED pathway forevaluation/treatment of febrile younginfants (0-56 days old). Available at:www.chop.edu/clinical-pathway/febrile-infant-emergent-evaluation-clinical-pathway. Accessed June 8, 2017

27. Bishop J, Ackley H, Beardsley E, et al;Seattle Children’s Hospital. Neonatal

PEDIATRICS Volume 143, number 3, March 2019 9 by guest on October 2, 2020www.aappublications.org/newsDownloaded from

fever pathway. 2017. Available at: www.seattlechildrens.org/pdf/neonatal-fever-pathway.pdf. Accessed September 17,2017

28. Aronson PL, Thurm C, Williams DJ, et al;Febrile Young Infant ResearchCollaborative. Association of clinicalpractice guidelines with emergencydepartment management of febrileinfants #56 days of age. J Hosp Med.2015;10(6):358–365

29. Swerkersson S, Jodal U, Åhrén C, Sixt R,Stokland E, Hansson S. Urinary tractinfection in infants: the significance of

low bacterial count. Pediatr Nephrol.2016;31(2):239–245

30. Tullus K. Low urinary bacterial counts:do they count? Pediatr Nephrol. 2016;31(2):171–174

31. Benneyan JC, Lloyd RC, Plsek PE.Statistical process control as a tool forresearch and healthcare improvement.Qual Saf Health Care. 2003;12(6):458–464

32. Ambroggio L, Thomson J, MurtaghKurowski E, et al. Quality improvementmethods increase appropriate

antibiotic prescribing for childhoodpneumonia. Pediatrics. 2013;131(5).Available at: www.pediatrics.org/cgi/content/full/131/5/e1623

33. Jennings RM, Burtner JJ, Pellicer JF,et al. Reducing head CT use for childrenwith head injuries in a communityemergency department. Pediatrics.2017;139(4):e20161349

34. Lavelle JM, Blackstone MM, Funari MK,et al. Two-step process for ED UTIscreening in febrile young children:reducing catheterization rates.Pediatrics. 2016;138(1):e20153023

10 KASMIRE et al by guest on October 2, 2020www.aappublications.org/newsDownloaded from

DOI: 10.1542/peds.2018-1610 originally published online February 6, 2019; 2019;143;Pediatrics 

Ilana Y. WaynikKathryn E. Kasmire, Eric C. Hoppa, Pooja P. Patel, Kelsey N. Boch, Tina Sacco and

of a Clinical PathwayReducing Invasive Care for Low-risk Febrile Infants Through Implementation

ServicesUpdated Information &

http://pediatrics.aappublications.org/content/143/3/e20181610including high resolution figures, can be found at:

Referenceshttp://pediatrics.aappublications.org/content/143/3/e20181610#BIBLThis article cites 32 articles, 15 of which you can access for free at:

Subspecialty Collections

subhttp://www.aappublications.org/cgi/collection/quality_improvement_Quality Improvemente_management_subhttp://www.aappublications.org/cgi/collection/administration:practicAdministration/Practice Managementsubhttp://www.aappublications.org/cgi/collection/emergency_medicine_Emergency Medicinefollowing collection(s): This article, along with others on similar topics, appears in the

Permissions & Licensing

http://www.aappublications.org/site/misc/Permissions.xhtmlin its entirety can be found online at: Information about reproducing this article in parts (figures, tables) or

Reprintshttp://www.aappublications.org/site/misc/reprints.xhtmlInformation about ordering reprints can be found online:

by guest on October 2, 2020www.aappublications.org/newsDownloaded from

DOI: 10.1542/peds.2018-1610 originally published online February 6, 2019; 2019;143;Pediatrics 

Ilana Y. WaynikKathryn E. Kasmire, Eric C. Hoppa, Pooja P. Patel, Kelsey N. Boch, Tina Sacco and

of a Clinical PathwayReducing Invasive Care for Low-risk Febrile Infants Through Implementation

http://pediatrics.aappublications.org/content/143/3/e20181610located on the World Wide Web at:

The online version of this article, along with updated information and services, is

by the American Academy of Pediatrics. All rights reserved. Print ISSN: 1073-0397. the American Academy of Pediatrics, 345 Park Avenue, Itasca, Illinois, 60143. Copyright © 2019has been published continuously since 1948. Pediatrics is owned, published, and trademarked by Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it

by guest on October 2, 2020www.aappublications.org/newsDownloaded from