BACKGROUND AND OBJECTIVES

Bronchiolitis is a leading cause of pediatric hospitalization in the United States, resulting in significant morbidity and health care resource use. Despite American Academy of Pediatrics recommendations against obtaining chest radiographs (CXRs) for bronchiolitis, variation in care continues. Historically, clinical practice guidelines and educational campaigns have had mixed success in reducing unnecessary CXR use. Our aim was to reduce CXR use for children <2 years with a primary diagnosis of bronchiolitis, regardless of emergency department (ED) disposition or preexisting conditions, from 42.1% to <15% of encounters by March 2020.

METHODS

A multidisciplinary team was created at our institution in 2012 to standardize bronchiolitis care. Given success with higher reliability interventions in asthma, similar interventions affecting workflow were subsequently pursued with bronchiolitis, starting in 2017, by using quality improvement science methods. The primary outcome was the percent of bronchiolitis encounters with a CXR. The balancing measure was return visits within 72 hours to the ED. Statistical process control charts were used to monitor and analyze data obtained from an internally created dashboard.

RESULTS

From 2012 to 2020, our hospital had 12 120 bronchiolitis encounters. Preimplementation baseline revealed a mean of 42.1% for CXR use. Low reliability interventions, like educational campaigns, resulted in unsustained effects on CXR use. Higher reliability interventions were associated with sustained reductions to 23.3% and 18.9% over the last 4 years. There was no change in ED return visits.

CONCLUSIONS

High-reliability workflow redesign was more effective in translating American Academy of Pediatrics recommendations into sustained practice than educational campaigns.

Bronchiolitis is a major contributor to cost and morbidity in pediatrics, with an estimated cost of $1.73 billion annually in the United States.1  The American Academy of Pediatrics (AAP) introduced evidence-based guidelines for treatment in 2006,2  which were revised in 2014,3  to help support clinicians. Unfortunately, these have made little change in clinical practice with persistent variation,4,5  contributing to ∼$545 million in direct expenditures without improved outcomes.1 

One consistent recommendation within these guidelines has been the avoidance of routine radiography, given its limited diagnostic value in bronchiolitis.6  Radiography has also been associated with increased antibiotic use without a change in clinical outcome.7  Efforts to reduce radiography in bronchiolitis have gained momentum by inclusion in Choosing Wisely, an initiative of the American Board of Internal Medicine Foundation to encourage evidence-based practice and reduce waste.8  A cost-effectiveness analysis of the use of radiography in bronchiolitis revealed savings, without compromising diagnostic accuracy for alternate diagnoses, such as bacterial pneumonia.9  Although pediatric emergency departments (EDs) have fared better than general EDs in chest radiograph (CXR) use, use within pediatric hospitals continues to remain high, with reported averages ranging from 42% to 55%.5,10 

Multiple quality improvement (QI) projects have been performed to improve compliance with evidence-based bronchiolitis practices by using interventions, including social psychology, educational campaigns, and standardization, like clinical practice guidelines (CPGs).11,12  Reliability science measures these types of modifications as essential for initial improvement work, but, often, they are unlikely to have sustained impacts. System redesign with an emphasis on processes and the structure in which processes operate may be more effective because it creates an environment in which it is easier for providers to perform intended actions.13,14  At our institution, similar strategies of education and CPG development were initiated in 2013, without sustained CXR reduction. In 2015, a separate QI project began with the aim to reduce CXR use in the management of asthma by removing CXR as a default option (a higher reliability intervention) in a commonly used dyspnea order set.15  These findings informed our QI efforts to focus on one area of overused therapies with high-reliability interventions.

This study was a QI project designed to reduce CXR use in the management of bronchiolitis, in accordance with reported achievable benchmarks of 32% to 42%.16,17  Unlike previous studies, in our study, we did not limit analysis to the bronchiolitis season alone. We also captured institutional performance for all cases of bronchiolitis over the course of 8 years. Critical care, acute care, and ED patients were all included in an effort to examine the hospital system as a whole, instead of singular points of care. Specifically, our objective was to decrease the percentage of CXRs obtained for patient encounters at our institution with a primary diagnosis of bronchiolitis from a historical baseline of 42.1% to <15% by March 2020.

Monroe Carell Jr Children’s Hospital at Vanderbilt (MCJCHV) is a 305-bed quaternary care pediatric referral center in Nashville, TN. MCJCHV currently has >16 000 inpatient admissions and ∼53 000 ED visits annually. The ED team includes pediatric emergency medicine (PEM) trained physicians, general emergency medicine physicians, PEM fellows, emergency medicine and pediatric residents, general pediatricians, pediatric nurse practitioners, pediatric nurses, and respiratory therapists. Children are admitted to the acute care floors where care is directed by either a pediatric hospitalist or general pediatrician. All inpatient teams have pediatric residents and, occasionally, nonpediatric residents. The PICU is continuously staffed by a pediatric critical care physician, nurse practitioners, PICU fellows, pediatric and emergency medicine residents, PICU nurses and respiratory therapists. Providers in both inpatient and emergency settings use an electronic medical record (EMR) system and computerized physician order entry.

We included all children <2 years old presenting to MCJCHV’s ED with a primary billing code (International Classification of Diseases, Ninth Revision [ICD-9], International Classification of Diseases, 10th Revision [ICD-10]) for bronchiolitis (Supplemental Table 1). This included patients who required admission for critical care and those with preexisting conditions, including but not limited to congenital heart disease, prematurity, congenital airway anomalies, and chronic lung disease.

A multidisciplinary team was formed to standardize care in accordance with AAP bronchiolitis guidelines in Fall 2012. The team involved nurse educators, PEM and hospitalist physicians, a PEM fellow, respiratory therapists, and pediatric residents. Initial interventions were composed of level 1 reliability interventions, including educational campaigns and the creation of a CPG.

In July 2017, a subset of this group began focusing on CXR reduction and developed 4 key drivers to meet our stated aim (Fig 1). After appreciating the impact of higher level of reliability interventions in asthma, further emphasis was placed on second and third level reliability interventions. The key drivers were focused on a standardized approach to bronchiolitis care, electronic ordering infrastructure to encourage evidence-based practices, continued engagement of the interprofessional care team regarding bronchiolitis care, and transparency of performance by providing feedback to clinicians. Interventions included electronic order infrastructure modification, with order set creation and best practice advisory (BPA) alerts, in addition to provider practice feedback and updating the formatting of the CPG.

FIGURE 1

Reduction of CXR use in bronchiolitis: key driver diagram. CPOE, computerized physician order entry; LOR1, level 1 reliability; LOR2, level 2 reliability; LOR3, level 3 reliability; RN, registered nurse; RT, respiratory therapist; SMART, specific, measurable, achievable, realistic, and timely.

FIGURE 1

Reduction of CXR use in bronchiolitis: key driver diagram. CPOE, computerized physician order entry; LOR1, level 1 reliability; LOR2, level 2 reliability; LOR3, level 3 reliability; RN, registered nurse; RT, respiratory therapist; SMART, specific, measurable, achievable, realistic, and timely.

Close modal

Data were collected from January 2012 to March 2020 via an electronic dashboard for all discharged bronchiolitis encounters from ED or inpatient settings. Baseline data were also obtained during the 2012 calendar year before CPG implementation in January 2013. This dashboard populated new data weekly. Statistical process control charts were used to study CXR use and monitor any temporal relationships between interventions and use.18 

Primary Measure

The primary outcome was the percentage of patients with bronchiolitis with a CXR at any time during their encounter, which included ED treat-and-release visits and inpatient hospitalizations. After study completion, we completed subset analysis, analyzing ED treat-and-release and hospitalizations separately.

Balancing Measure

Return visits to the ED within 72 hours were monitored starting in 2017, after transition to a new EMR.

CXR use and return visits to the ED with 72 hours were monitored by using annotated statistical process control charts by use of QI Charts software (Process Improvement Products, Austin, TX). Given the highly seasonal nature of bronchiolitis, patients were organized in groups of 40 consecutive discharges, with the percentage of CXRs performed plotted on the p-chart. This provided rapid data analysis opportunities as well as a lower control limit that was above the x-axis when improvement work began. Additional retrospective analysis to better characterize CXR use was performed at the end of the study period by separately analyzing patients who were admitted and those who were treated and discharged from the ED. Annotations of interventions were added to the p-chart to highlight the relationship of changes in outcomes. Rules of interpretation for special cause variation for Shewhart p-charts were used.18 

Our study was approved by our institutional review board as nonresearch QI.

During the study period, there were 12 120 encounters for bronchiolitis meeting ICD-9 and/or ICD-10 criteria. A total of 3635 CXRs were ordered. The 2012 calendar year (1800 encounters) was used to determine baseline ordering rates.

Before our focus on CXRs, a CPG was created and introduced in 2013, which included recommendations for ED and inpatient settings regarding many aspects of care. The recommendation against routine CXR use was 1 of many recommendations provided in the CPG, which was posted in workspaces and published online. To improve CPG use, educational campaigns were conducted in October 2013 and during the winter months of 2016–2017. There was no meaningful sustained reduction in CXR use. These first steps of building reliable systems through education and guideline creation are considered as level 1 reliability interventions.

A CXR-specific team was formed in July 2017 and introduced changes to the CPG in 2018. Revisions included more user-friendly formatting and respiratory distress scoring to stratify bronchiolitis severity. From previous CXR work done at our institution for asthma, we also learned that providing specific reasons to order a CXR in place of general recommendations had a stronger impact on provider behavior.15  Rather than simply discouraging CXRs for bronchiolitis in general, the team created a list of indications for CXRs, including persistent focal lung findings, airway compromise, worsening condition, and new murmur.

The improvement team also created a control chart analyzing CXR use in patients with bronchiolitis over time. The team noted that there was a shift indicating improvement in July 2015 that was temporally unrelated to interventions of the bronchiolitis workgroup. Further exploration uncovered that this was associated with the change in the dyspnea order set meant to reduce CXR use in asthma,15  but this workflow change also led to an unintended yet sustained consequence of reducing CXRs in bronchiolitis from the baseline of 42.1% to 23.3%. Our improvement team appreciated the effectiveness of this high-reliability intervention, particularly given the absence of CXR reduction in bronchiolitis with lower-level reliability interventions.

In April 2018, a multidisciplinary team involving ED and hospital medicine physicians began designing a modified electronic ordering infrastructure with a bronchiolitis order set for both ED and inpatient settings. These order sets are considered level 2 reliability interventions. The ED order set was placed in a commonly used complaint-based panel within the EMR to encourage use (Supplemental Fig 5). In addition, this workgroup created BPA alerts (Supplemental Fig 6) for unwarranted tests or therapies in patients with a charted diagnosis of bronchiolitis within the EMR or in whom the inpatient bronchiolitis order set was used. These 2 interventions went into production in Fall 2018. This is considered a level 3 reliability intervention. BPA alerts served as real-time reminders and the default choice was purposefully set to remove the order. The work involved in designing and implementing these high-reliability changes was associated with a reduction in CXR use to 18.9% (Fig 2).

FIGURE 2

Bronchiolitis encounters undergoing CXR: p-chart with annotations. LCL, lower control limit; LOR1, level 1 reliability; LOR2, level 2 reliability; LOR3, level 3 reliability; UCL, upper control limit.

FIGURE 2

Bronchiolitis encounters undergoing CXR: p-chart with annotations. LCL, lower control limit; LOR1, level 1 reliability; LOR2, level 2 reliability; LOR3, level 3 reliability; UCL, upper control limit.

Close modal

Educational meetings were continued with nurses, respiratory therapists, and housestaff to review our CPG (level 1 reliability). Efforts were focused on the challenges of an academic institution, given the frequent turnover of staff and rotating trainees. Specific education or discussion also responded to practice variation trends or concerns when they were noted.

Transparency of performance was introduced in November 2019 by posting overall performance in work areas as well as offering feedback to individual providers. Feedback was given monthly in anonymous fashion through February 2020 by using funnel plots to show individual performance compared with that of colleagues (Supplemental Fig 7). We also continued regular review of the literature to ensure feedback was accurate with current recommendations and guidelines during the duration of the study period. Provider feedback, a level 1 reliability intervention, to emergency medicine physicians and nurse practitioners during the 2019–2020 winter months was not associated with further changes in CXR use.

In the baseline data, it was indicated that 42.1% of bronchiolitis encounters within our institution underwent a CXR at some point during their encounter. After a change in a dyspnea order set removing CXR as a choice in July 2015, special cause variation indicated a decrease to 23.3% (Fig 2). After efforts were focused on high-reliability interventions, special cause variation was again achieved in June 2018 to a new mean of 18.9% and was sustained for 20 months (Fig 2). A retrospective review revealed that the care of discharged patients from the ED was significantly impacted by the dyspnea order set modification, with a reduction in CXR use from 32.9% to 15.2%, and the ED order set and BPA alert interventions again reduced ED treat-and-release use to 9% (Fig 3). Inpatient use decreased from 62.5% to 35.1%, associated with the removal of the dyspnea order set (Fig 4).

FIGURE 3

Subset p-chart analysis: bronchiolitis CXR use in ED treat-and-release patients only. LCL, lower control limit; LOR1, level 1 reliability; LOR2, level 2 reliability; LOR3, level 3 reliability 3; UCL, upper control limit.

FIGURE 3

Subset p-chart analysis: bronchiolitis CXR use in ED treat-and-release patients only. LCL, lower control limit; LOR1, level 1 reliability; LOR2, level 2 reliability; LOR3, level 3 reliability 3; UCL, upper control limit.

Close modal
FIGURE 4

Subset p-chart analysis: bronchiolitis CXR use for hospitalized inpatients only. LCL, lower control limit; LOR2, level 2 reliability; UCL, upper control limit.

FIGURE 4

Subset p-chart analysis: bronchiolitis CXR use for hospitalized inpatients only. LCL, lower control limit; LOR2, level 2 reliability; UCL, upper control limit.

Close modal

We observed no change in the percent of children initially discharged from the ED with a bronchiolitis diagnosis who returned within 72 hours (data are not shown).

Our use of QI methodology led to sustained reduction in CXR use in pediatric patients with bronchiolitis, without any increase in 72-hour ED reuse. Although provider education was likely necessary, it was not sufficient to impact CXR use. Interventions that led to a significant reduction in CXRs were those that incorporated high-reliability theory: the removal of a CXR option in a dyspnea order set and creation of physician support tools (BPAs and order sets) that were within providers’ EMR workflows. Provider feedback and continued education did not lead to further reduction.

Our findings are consistent with systematic reviews, including one specific to QI projects in bronchiolitis, in which it was concluded that educational interventions, such as CPGs, require voluntary involvement and often produce smaller changes in outcomes, compared with forced functions or workflow changes.19,20  Similarly, AAP bronchiolitis guidelines, which were updated in 2014 during our study period and accompanied by local educational campaigns, had no effect on CXR use. Additionally, PEM providers report familiarity with AAP guidelines.21  Thus, continued application of education to a problem that does not involve knowledge deficits appears to have limited benefit.22 

Notably, all of these interventions would be considered level 1 reliability by the Institute for Healthcare Improvement.13  Thus, understanding improvement science would indicate that these steps were necessary but were never going to be sufficient for our team to reach our goals, no matter how much time and effort were put into these education campaigns. Level 2 and 3 reliability interventions are focused on identifying and/or mitigating failure and redesigning the system to mitigate these failures. Level 2 and 3 reliability interventions include making the desired action the default option and building decision aids into the system.13  On the basis of this methodology, it is unsurprising that a separate asthma QI project at our institution that removed a default option for a CXR in a dyspnea order set (a second level reliability intervention) also made an unintentional but significant impact on CXR use for bronchiolitis because these patients also present with dyspnea. This unexpected finding speaks to the effectiveness of workflow interventions.15  Given this unintended but serendipitous observation, our focus was changed toward developing workflow tools that could match this level of reliability.14  Our order set was updated with recommended orders that complied with AAP guidelines, and it was placed in a more commonly used complaint-based panel.

The third level of reliability interventions further investigate the persistence of failure and emphasize redesigning the system to eliminate these failures.13  Some providers chose not to use the order set, thereby working around the intervention. Additionally, order sets were only impactful during the ED stay, prompting an additional need to further redesign the system to reduce CXRs in bronchiolitis. Creating a BPA addressed this need by impacting time during hospitalization as well, also helping to guide subsequent decision-making in the ED if the order set was used.

Our last intervention involved giving specific feedback to providers on practice patterns. After manual chart review in 2018, we observed that 80% of all CXRs were ordered in the ED, theorizing that focused effort in the ED had significant potential. Provider-specific feedback did not change overall rate of CXRs. This finding contradicts previous studies in which researchers reported a positive impact.12,23,24  Although feedback mechanisms are considered first level reliability interventions, we believe that there is a component of second level reliability with a shared agreement or buy-in of doctors and nurse practitioners to follow and learn from standard processes through this regular feedback. We hypothesize that this intervention may have had more impact if it had been done earlier in the QI project, before already reducing our mean to 18.9%.

Our achieved percent of 18.9% was considerably lower than the achievable benchmark of care of 32% to 42% for pediatric hospitals,16,17  especially considering we did not exclude higher risk patients. Tyler et al12  reported a CXR use of 13.6% outside of an ICU setting. Given these studies, it is possible that achievable benchmarks for typical previously healthy patients with acute bronchiolitis might be lower than previously described. Pediatric hospitals have different practice patterns than general hospitals and EDs. Achievable benchmarks of care should reflect reasonable, achievable goals for each setting.10,25 

Our improvement project was done at a single institution that specializes in pediatric care. Our findings may not be generalizable to all institutions. The patient population was determined by billing codes solely, which is subject to inaccuracies. In addition, certain patients with preexisting conditions, such as extreme prematurity, congenital heart disease, tracheostomies, or other conditions, were included in our study population. These patients are often excluded from AAP bronchiolitis guidelines and may have clinically warranted CXRs. Given our inclusion of these patients, the percentage of previously healthy, noncritically ill children receiving CXRs may have been lower than our mean of 18.9%. This decision was made because we were unable to build such logic into the real-time dashboard, and manual review of such large patient numbers would have inhibited our ability to learn as quickly from our data.

Our goal to reduce CXRs to <15% may have been unrealistic, given the inclusions mentioned previously. In reviewing samples of charts to determine failure points, we found patients with preexisting conditions and ICU admission likely led to higher rates of radiography. Additionally, we found evidence of other reasons for ordering CXRs in this patient population, such as confirming placement of tubes or catheters, which is standard at our institution and falls outside of “routine” care for bronchiolitis. Appropriately excluding these would have required manual review of all 12 120 encounters.

The impact of high-reliability interventions, applied through QI methodology, were associated with significant, sustained reductions in CXR use and were more effective than educational campaigns. Successful implementation of CPGs may rely more heavily on workflow redesign and less on educational campaigns in the hospital setting. Reliance on higher reliability interventions instead of education alone may provide other institutions an improved opportunity to reduce overuse and improve care for children at their hospitals.

We thank Adam Vukovic, Pradeep Mummidi, Nicholas Jones, and Lesley Worsley-Hynd for their leadership involvement in the QI initiative.

Drs Frazier and Johnson conceptualized and designed the study, collected data, conducted analyses, and drafted the initial manuscript; Drs Plemmons, Walls, and Jain conceptualized and designed the study and collected and analyzed the data; and all authors reviewed and revised the manuscript, approved the final manuscript as submitted, and agree to be accountable for all aspects of the work.

FUNDING: No external funding.

AAP

American Academy of Pediatrics

BPA

best practice advisory

CPG

clinical practice guideline

CXR

chest radiograph

ED

emergency department

EMR

electronic medical record

ICD-9

International Classification of Diseases, Ninth Revision

ICD-10

International Classification of Diseases, 10th Revision

MCJCHV

Monroe Carell Jr Children’s Hospital at Vanderbilt

PEM

pediatric emergency medicine

QI

quality improvement

1
Hasegawa
K
,
Tsugawa
Y
,
Brown
DFM
,
Mansbach
JM
,
Camargo
CA
 Jr
.
Trends in bronchiolitis hospitalizations in the United States, 2000-2009
.
Pediatrics
.
2013
;
132
(
1
):
28
36
2
American Academy of Pediatrics Subcommittee on Diagnosis and Management of Bronchiolitis
.
Diagnosis and management of bronchiolitis
.
Pediatrics
.
2006
;
118
(
4
):
1774
1793
3
Ralston
SL
,
Lieberthal
AS
,
Meissner
HC
, et al;
American Academy of Pediatrics
.
Clinical practice guideline: the diagnosis, management, and prevention of bronchiolitis. [published correction appears Pediatrics. 2015;136(4):782]
.
Pediatrics
.
2014
;
134
(
5
).
4
Parikh
K
,
Hall
M
,
Teach
SJ
.
Bronchiolitis management before and after the AAP guidelines
.
Pediatrics
.
2014
;
133
(
1
).
5
Florin
TA
,
Byczkowski
T
,
Ruddy
RM
,
Zorc
JJ
,
Test
M
,
Shah
SS
.
Variation in the management of infants hospitalized for bronchiolitis persists after the 2006 American Academy of Pediatrics bronchiolitis guidelines
.
J Pediatr
.
2014
;
165
(
4
):
786
792.e1
6
Schuh
S
,
Lalani
A
,
Allen
U
, et al
.
Evaluation of the utility of radiography in acute bronchiolitis
.
J Pediatr
.
2007
;
150
(
4
):
429
433
7
Swingler
GH
,
Hussey
GD
,
Zwarenstein
M
.
Randomised controlled trial of clinical outcome after chest radiograph in ambulatory acute lower-respiratory infection in children
.
Lancet
.
1998
;
351
(
9100
):
404
408
8
Quinonez
RA
,
Garber
MD
,
Schroeder
AR
, et al
.
Choosing wisely in pediatric hospital medicine: five opportunities for improved healthcare value
.
J Hosp Med
.
2013
;
8
(
9
):
479
485
9
Yong
JHE
,
Schuh
S
,
Rashidi
R
, et al
.
A cost effectiveness analysis of omitting radiography in diagnosis of acute bronchiolitis
.
Pediatr Pulmonol
.
2009
;
44
(
2
):
122
127
10
Gong
C
,
Byczkowski
T
,
McAneney
C
,
Goyal
MK
,
Florin
TA
.
Emergency department management of bronchiolitis in the United States
.
Pediatr Emerg Care
.
2019
;
35
(
5
):
323
329
11
Sprecher
E
,
Chi
G
,
Ozonoff
A
,
Cox
J
,
Patel
N
,
Conroy
K
.
Use of social psychology to improve adherence to national bronchiolitis guidelines
.
Pediatrics
.
2019
;
143
(
1
):
e20174156
12
Tyler
A
,
Krack
P
,
Bakel
LA
, et al
.
Interventions to reduce over-utilized tests and treatments in bronchiolitis
.
Pediatrics
.
2018
;
141
(
6
):
e20170485
13
Nolan
T
,
Resar
R
,
Haraden
C
,
Griffin
FA
.
Improving the Reliability of Health Care
.
Boston, MA
:
Institute for Healthcare Improvement
;
2004
14
Luria
JW
,
Muething
SE
,
Schoettker
PJ
,
Kotagal
UR
.
Reliability science and patient safety
.
Pediatr Clin North Am
.
2006
;
53
(
6
):
1121
1133
15
Watnick
CS
,
Arnold
DH
,
Latuska
R
,
O’Connor
M
,
Johnson
DP
.
Successful chest radiograph reduction by using quality improvement methodology for children with asthma
.
Pediatrics
.
2018
;
142
(
2
):
e20174003
16
Parikh
K
,
Hall
M
,
Mittal
V
, et al
.
Establishing benchmarks for the hospitalized care of children with asthma, bronchiolitis, and pneumonia
.
Pediatrics
.
2014
;
134
(
3
):
555
562
17
Ralston
S
,
Parikh
K
,
Goodman
D
.
Benchmarking overuse of medical interventions for bronchiolitis
.
JAMA Pediatr
.
2015
;
169
(
9
):
805
806
18
Provost
L
,
Murray
S
.
The Health Care Data Guide: Learning From Data for Improvement
. 1st ed.
Hoboken, NJ
:
John Wiley & Sons, Inc
;
2011
19
Ralston
S
,
Comick
A
,
Nichols
E
,
Parker
D
,
Lanter
P
.
Effectiveness of quality improvement in hospitalization for bronchiolitis: a systematic review
.
Pediatrics
.
2014
;
134
(
3
):
571
581
20
Forsetlund
L
,
Bjørndal
A
,
Rashidian
A
, et al
.
Continuing education meetings and workshops: effects on professional practice and health care outcomes
.
Cochrane Database Syst Rev
.
2009
;(
2
):
CD003030
21
Rivera-Sepulveda
AV
,
Rebmann
T
,
Gerard
J
,
Charney
RL
.
Physician compliance with bronchiolitis guidelines in pediatric emergency departments
.
Clin Pediatr (Phila)
.
2019
;
58
(
9
):
1008
1018
22
Soong
C
,
Shojania
KG
.
Education as a low-value improvement intervention: often necessary but rarely sufficient
.
BMJ Qual Saf
.
2020
;
29
(
5
):
353
357
23
Meeker
D
,
Linder
JA
,
Fox
CR
, et al
.
Effect of behavioral interventions on inappropriate antibiotic prescribing among primary care practices: a randomized clinical trial
.
JAMA
.
2016
;
315
(
6
):
562
570
24
Gerber
JS
,
Prasad
PA
,
Fiks
AG
, et al
.
Effect of an outpatient antimicrobial stewardship intervention on broad-spectrum antibiotic prescribing by primary care pediatricians: a randomized trial
.
JAMA
.
2013
;
309
(
22
):
2345
2352
25
Ho
SW
,
Huang
KY
,
Teng
YH
,
Ku
MS
,
Chiou
JY
.
Practice variations between emergency physicians and pediatricians in treating acute bronchiolitis in the emergency department: a nationwide study
.
J Emerg Med
.
2015
;
48
(
5
):
536
541

Competing Interests

POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.

FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.

Supplementary data