Clinical Research

Nutrition in Clinical Practice

Volume 34 Number 6
Clinical Decision Support Tools and a Standardized Order Set December 2019 916–921

C 2019 American Society for

Enhances Early Enteral Nutrition in Critically Ill Children Parenteral and Enteral Nutrition
DOI: 10.1002/ncp.10272
wileyonlinelibrary.com

Keegan J. Ziemba, MD1,2 ; Rajeswari Kumar, PhD1 ; Kathryn Nuss, MD1 ;

Maria Estrada, DO1 ; Ada Lin, MD1 ; and Onsy Ayad, MD1

Abstract
Background: Critically ill children in a pediatric intensive care unit (PICU) have unique nutrition needs that are challenging to
achieve and thus are at high risk of malnutrition. There is increasing evidence that children who reach caloric goals early have
improved outcomes. The purpose of this initiative was to implement an enteral nutrition (EN) algorithm in a tertiary care PICU
utilizing clinical decision support tools (CDSTs) and a standardized order set within an electronic health record. Methods: A quality
improvement initiative was undertaken to implement an EN feeding protocol using electronic CDSTs, including a new standardized
order set. Results: In a historical cohort of 376 patients, only 18% met goal EN in the first 48 hours of admission. The EN protocol
was implemented in 272 patients who met 88% goal feed volume within 48 hours of intensive care unit admission. Median time to
start EN (1.7 vs 1.3 days, P < 0.0001) and time to goal nutrition (2.8 vs 2.2 days, P < 0.001) improved after project implementation.
Length of stay in the PICU was significantly reduced following protocol implementation (202 hours pre-implementation vs 156
hours post implementation, P < 0.0001). Conclusions: We used CDSTs and standardized order sets to implement a nutrition
algorithm to facilitate and likely improve the nutrition care of critically ill children. (Nutr Clin Pract. 2019;34:916–921)

Keywords
child; critical illness; electronic health records; enteral nutrition; nutrition support; pediatric intensive care unit; pediatrics; quality
improvement

Background admission. The goal is to provide >50% of calories in the

Computerized provider order entry (CPOE) has been inte-
grated into routine patient care over the last 20 years. Its
use has been associated with reduced rates of medication
errors and adverse drug events.1-4 There is also evidence
revealing its effectiveness within order sets to influence
and benefit patient outcomes in the adult hospital setting.
Order sets within the intensive care setting have been
utilized to benefit patients for congestive heart failure and
pneumonia.5,6
An area of increasing importance for patient outcomes is
nutrition, especially in the critically ill. Evidence is mounting
that early enteral nutrition (EN) is beneficial. Critically
ill patients, both adults and children, are at high risk of
malnutrition during their intensive care unit (ICU) stay.
Addition of enteral feeds as early as 48 hours has been
associated with improved outcomes.7-10
In respiratory illnesses, when positive pressure ventila-
tion for support is indicated, patients are unable to tolerate
oral nutrition and must be supplemented either enterally
or parenterally. EN is the preferred route according to
the American Society for Enteral and Parenteral Nutrition
(ASPEN) guidelines.11,12 In fact, ASPEN guidelines recom-
mend that enteral feeds be initiated within 24–48 hours of
first week of admission. Additionally, evidence exists that
the use of bedside algorithms and protocols may assist with
earlier initiation of EN.13-15
Given the wide use of the electronic health record (EHR)
in hospital systems, it follows that clinical decision support
tools (CDSTs) and CPOE can be utilized to improve patient
outcomes. In a pediatric critical care setting, there is a
paucity of research that describes the utilization of CDSTs.
Twenty percent of pediatric ICU (PICU) patients are mal-
nourished at the time of admission.9 Therefore, it makes
sense to target this high-risk population to enhance the
bedside deliver of EN in a timely fashion in our facility. In
From the 1 Nationwide Children’s Hospital, Columbus, Ohio, USA;
and the 2 Brody School of Medicine, East Carolina University,
Greenville, North Carolina, USA.
Financial disclosure: None declared.
Conflicts of interest: None declared.
This article originally appeared online on April 1, 2019.
Corresponding Author:
Keegan J. Ziemba, MD, ECU Brody School of Medicine, Department
of Pediatrics, 600 Moye Blvd, Room 2CH252, Greenville, NC 27834,
USA.
Email: ziemba.keegan@gmail.com
Ziemba et al
this initiative, we aimed to increase the number of children
receiving EN within 48 hours of admission to the PICU
by leveraging functionality in the EHR as part of a quality
improvement (QI) initiative.
Methods
This study was performed in a large, urban, quaternary
referral hospital. The PICU is a 42-bed medical/surgical unit
with 2700 admissions annually. In 2006, our hospital system
began implementation of an EHR that was intended for all
patient documentation, order management, messaging, and
patient tracking (Epic Systems, Corp.).
A QI initiative was undertaken to design and imple-
ment an EN feeding protocol. Children <1 year of age
with diagnoses of respiratory distress and/or insufficiency
admitted to the PICU were identified as a target population,
as these high-risk patients would likely benefit most from
early initiation of a nutrition protocol. A historical cohort
included patients that met study criteria from January
2013 to November 2015. During this phase, CPOE was
already in place. The intervention cohort included patients
from December 2015 to February 2017 when the CDSTs
and feeding order set went live. All patients admitted
to the PICU for a respiratory illness requiring positive
pressure ventilation (invasive or noninvasive) and who were
<1 year of age were included. Patients admitted to a
surgical service and children with a gastrostomy tube were
excluded.
A retrospective review of medical records established
clinical practice baselines. We sought to create an informat-
ics build that leveraged several electronic elements in the
EHR to design a CDST to achieve this goal. A SNOMED
concept diagnosis grouper was created to aid in the identi-
fication of the target patient population in the EHR. Rules
evaluated the patients for age, location, and hospital service,
as well as the time from admission to the PICU location
and service. A comprehensive standardized EN order set
(Figure 1A and B) was developed in conjunction with the
PICU clinicians. It also included a selection of trophic feeds
if pertinent for the patient. The enteral feed section included
an age-appropriate nutrition supplement with a caloric titra-
tion order and assessments for feeding intolerance. Bedside
nurses were reeducated to utilize existing flow sheets to allow
for better documentation, data capture, and success of this
intervention. We selected 18 hours from admission to the
PICU as the most optimal time to introduce the interruptive
system alert to encourage the initiation of the newly created
EN order set. This alert was a best-practice advisory (BPA)
and as such could be delayed or overridden (Figure 1C).
The BPA was presented to physicians and advanced practice
providers during the ordering activity directed at patients
that met the inclusion criteria (ࣘ1 year of age, diagnosis
of bronchiolitis, in the ICU ࣙ 18 hours, and no feeding
917
initiated). The BPA hosted the order set link and could be
invoked directly from the advisory.
For the analysis, only patients selected for full enteral
feedings were included, as our primary interest was in
assessing the capabilities of the advancing protocol. Patients
who received trophic feeds were not included. Selection of
feeding type (enteral vs trophic) was at the discretion of the
primary team and therefore varied between practitioners.
Pediatric Risk of Mortality (PRISM) scores were com-
piled from a prospective, observational database, collected
at the time of admission to the PICU. PRISM is an indicator
for severity of illness in critically ill patients, with scores <10
associated with mortality < 10%.16,17 All statistics were per-
formed in GraphPad 7.0. All continuous data were analyzed
with Mann-Whitney test. All data were reported as median
values due to the nonparametric dataset, confirmed with
Shapiro-Wilk normality test (P < 0.01). A P-value < 0.05
was considered significant.
Results
The EN feeding protocol, which included the CDSTs and
standardized order set, was instituted on December 1,
2015, with data collected through February 2017. A total
of 272 children met inclusion criteria. All children were
<1 year of age. The most common reason for admission
was acute bronchiolitis in both historical and intervention
cohorts. Eighteen hours into the PICU admission, the BPA
encouraged the use of an enteral feeding order set. Table 1
describes the patient demographics at the time of PICU
admission. Among those meeting the study criteria, EN
was started within 48 hours of PICU admission in 81%
(221/272) after implementation of the standardized EN
order set, compared with 63% (236/375) in the historical co-
hort (P < 0.01). EN was started during the implementation
period at a median of 1.3 vs 1.7 days in the historical cohort
(P < 0.01, Figure 2). Goal EN was also achieved sooner
with protocol implementation (2.2 vs 2.8 days, P < 0.01,
Figure 3). The percentage of children who reached goal EN
in the first 48 hours of admission increased from 18% to 38%
(P < 0.01).
Acuity was not significantly different between the histor-
ical and intervention cohorts, according to PRISM scores
(median 2 and 2, respectively, P = 0.3). Initial weight was
higher in the implementation cohort. However, patients
admitted after the introduction of the CDSTs and the
standardized order set (intervention cohort) had a median
weight gain of 140 g during admission vs 80 g in the
historical cohort (P = 0.001). Length of stay (LOS) from
admission to hospital discharge was not significantly differ-
ent in the 2 study periods (8.7 vs 8.4 days, P = 0.9). Yet,
LOS in the PICU was significantly shorter following the
intervention (202 vs 156 hours, P < 0.01) (Figure 4). There
were no mortalities in either cohort.
918 Nutrition in Clinical Practice 34(6)

Figure 1. (A and B) Enteral Nutrition Protocol implemented during the quality improvement initiative. (C) Best-practice advisory
created to alert 18 hours from admission. ABD, abdomen; AG, abdominal girth; EN, enteral nutrition; GRV, gastric residual
volume; HO, house officer; NEO, neonate; NG, nasogastric; NJ, nasojejunal; PICU, pediatric intensive care unit; Q4H, every
4 hours.
Ziemba et al 919

Table 1. Patient Demographics at PICU Admission. Discussion

Historical Implementation The widespread adoption of EHR is now standard in the
Cohort Cohort P-Value medical setting. Utilizing its full capabilities provides the
opportunity to impact patient care. In this QI initiative, we
Female, n (%) 141 (39) 106 (39) 0.65
Ethnicity, n (%) focused on the use of electronic CDST and a standardized
African 7 (2) 10 (4) order set to support and achieve the aim of improving the
Asian 2 (1) 3 (1) time to initiation of EN in our PICU. The informatics
Latino 21 (6) 14 (5) design and build allowed for streamlined integration of
Multiracial 22 (6) 16 (6) the EN feeding protocol as demonstrated by the increased
Black 84 (22) 54 (20) number of children who received earlier EN.
White 227 (60) 166 (61)
This QI initiative provided the opportunity to evaluate
Unknown 11 (3) 8 (3)
PRISM score 2 2 0.31 the use of CDSTs with a standardized order set as a
(median) method of implementation of an EN protocol in a high-
Age, months 2 3 0.11 risk population. We found earlier implementation of EN
Weight, kg 5.1 6.1 0.01 and achievement of goal EN in our study population at
significantly higher rates than previously. This achievement
PICU, pediatric intensive care unit; PRISM, Pediatric Risk of was our primary goal. In secondary analysis, we also noted
Mortality.
greater weight gain and shorter PICU stays after imple-
mentation of the protocol. Although that cohort did have
a higher initial weight, the measured change in weight was
also greater than the historical cohort.
In the critical care setting, use of CPOE includes in-
tegration of order sets into bedside delivery of care.1,2
Notably, order sets have been used in congestive heart
failure and pneumonia in adult critical care units, with
reduction in LOS and mortality.5,6 The results of this study
demonstrated reduction in PICU LOS, though mortality
and hospital LOS were not affected. However, we cannot
with certainty attribute this solely to our project. Regardless,
this study adds pediatric evidence for using targeted CDSTs
and standardized order sets to enhance patient care.
QI methodologies are utilized in many facets of
Figure 2. Time to initiation of enteral nutrition from time of medicine. This model has been used to implement CPOE for
admission to the PICU. * Indicates significance, P < 0.001. the administration of chemotherapy agents, care of patients
PICU, pediatric intensive care unit.
with chronic obstructive pulmonary disease, and reduction
of excessive laboratory test orders.3,4,18 To our knowledge,
this study is the first to utilize QI methodology to integrate
CDSTs to prompt users to initiate early EN in pediatrics.
By achieving the aim of increasing the number of children
who attain goal nutrition sooner in the PICU setting, this
provides evidence that CDSTs can be used to integrate new
technology to influence everyday patient care.
The use of algorithms to enhance nutrition within the
critical illness setting has been investigated in adults and
children. Several studies have shown that these algorithms
may minimize interruptions to achieve nutrition goals and
thus optimize delivery in this setting.13-15 Our results reflect
prior studies, as we demonstrated a higher percentage of
children achieving prescribed nutrition goals in a timely
manner.
Figure 3. Time to goal enteral nutrition from time of The ultimate impact of early EN remains a topic of
admission to the PICU. * Indicates significance, P < 0.001. research in pediatrics. Prior studies have shown that mal-
PICU, pediatric intensive care unit.
nutrition and suboptimal energy intake are associated with
920
Figure 4. Length of stay for (A) PICU and (B) hospital. * Indicates significance, P < 0.001. PICU, pediatric intensive care unit.
increased odds of death, whereas early implementation of
EN may reduce mortality.7,10,19 This QI initiative was not
designed to evaluate long-term outcomes, and thus the lack
of impact on time to hospital discharge was unsurprising.
However, we did see a statistically significant decrease in
PICU LOS by nearly 2 days. Although overall LOS are
low, we theorize that nutrition support during a critically ill
period may have enhanced healing and affected the growth
hormone axis, thus contributing to decreased length of
time of respiratory support. During the implementation
period, patients had greater weight gain. This may indicate
better nutrition during their PICU stay, which could have
an impact on further growth and development, especially in
the first year of life. This finding, however, must be taken
lightly, as weight is highly fluctuant in the PICU, influenced
by a multitude of factors other than nutrition, including
fluid balance. Further in-depth analysis is needed to identify
the true clinical impact.
There are a few limitations of this study. This is a
single-center experience designed to focus on a specific
patient population within the PICU. We did not include
cardiac or surgical patients, and the study’s impact cannot
be generalized to these patient populations. This study
evaluated both the use of a CDST and a standardized order
set for its implementation; thus, we cannot discern their
separate influences on our findings. Our results may have
been influenced by the tendency for better adherence to a
protocol when first introduced; however, our results were
carried through a full year beyond implementation. Future
directions of our efforts will focus on expansion of the order
set and protocol to other patient populations. This project
was also not designed to evaluate long-term developmental
or nutrition outcomes, and thus we can only postulate
about any effects this may have. We also hesitate using
weight as a marker for good nutrition for this study, given
the previously mentioned fluctuating tendencies within the
PICU. We acknowledge that these other factors may have
affected the changes in weight that patients experienced.
Nutrition in Clinical Practice 34(6)
Future studies are needed to fully address long-term
outcomes. Regardless, our QI initiative serves to demon-
strate that use of CDSTs and a standardized order set
is both feasible and potentially impactful when used in
conjunction with clinical protocols. In conclusion, this study
demonstrates that the use of specific CDST can facilitate
the implementation of a physician-ordered and nursing-
driven protocol to enhance the care of critically ill pediatric
patients.
Statement of Authorship
K. J. Ziemba, R. Kumar, K. Nuss, M. Estrada, and O. Ayad
contributed to the conception and design of the research. K.
J. Ziemba contributed to the acquisition of the data. K. J.
Ziemba, R. Kumar, and K. Nuss contributed to the analysis
and interpretation of the data. K. J. Ziemba, R. Kumar,
K. Nuss, and A. Lin drafted the manuscript. All authors
critically revised the manuscript, agree to be fully accountable
for ensuring the integrity and accuracy of the work, and read
and approved the final manuscript.
References
1. Ali NA, Mekhjian, HS, Kuehn, PL, et al. Specificity of computerized
physician order entry has a significant effect on the efficiency of
workflow for critically ill patients. Crit Care Med. 2005;33(1):110-114.
2. Maslove DM, Rizk N, Lowe HJ. Computerized physician order entry
in the critical care environment: a review of current literature. J
Intensive Care Med. 2011;26(3):165-171.
3. Saad A, Der-Nigoghossian CA, Njeim R, Sakr R, Salameh P, Mas-
soud M. Prescription errors with chemotherapy: quality improvement
through standardized order templates. Asian Pac J Cancer Prev.
2016;17(4):2329-2336.
4. White AA, McKinney CM, Hofman NG, Sutton PR. Optimizing
vitamin D naming conventions in computerized order entry to support
high-value care. J Am Med Inform Assoc. 2017:24(1):172-175.
5. Krive J, Shoolin JS, Zink SD. Effectiveness of evidence-based conges-
tive heart failure (CHF) CPOE order sets measured by health outcomes.
AMIA Annu Symp Proc. 2014;2014:815-824.
6. Krive J, Shoolin JS, Zink SD. Effectiveness of evidence-based pneu-
monia CPOE order sets measured by health outcomes. Online J Public
Health Inform. 2015;7(2):e211.
Ziemba et al
7. Mikhailov TA, Kuhn EM, Manzi J, et al. Early enteral nutrition
is associated with lower mortality in critically ill children. JPEN J
Parenter Enteral Nutr. 2014;38(4):459-466.
8. Mehta NM, Bechard LJ, Zurakowski D, Duggan CP, Heyland DK.
Adequate enteral protein intake is inversely associated with 60-d
mortality in critically ill children: a multicenter, prospective, cohort
study. Am J Clin Nutr. 2015;102(1):199-206.
9. Hulst J, Joosten K, Zimmermann L, et al. Malnutrition in critically
ill children: from admission to 6 months after discharge. Clin Nutr.
2004;23(2):223-232.
10. Compher C, Chittams J, Sammarco T, Nicolo M, Heyland DK.
Greater protein and energy intake may be associated with improved
mortality in higher risk critically ill patients: a multicenter, multina-
tional observational study. Crit Care Med. 2017;45(2):156-163.
11. Mehta NM, Compher C. A.S.P.E.N. Clinical Guidelines: nutrition
support of the critically ill child. JPEN J Parenter Enteral Nutr.
2009;33(3):260-276.
12. McClave SA, Taylor BE, Martindale RG, et al. Guidelines for the
provision and assessment of nutrition support therapy in the adult
critically ill patient: Society of Critical Care Medicine and Ameri-
can Society for Parenteral and Enteral Nutrition. JPEN. 2016;40(2):
159-211.
921
13. Martinez EE, Bechard LJ, Mehta NM. Nutrition algorithms and bed-
side nutrient delivery practices in pediatric intensive care units: an inter-
national multicenter cohort study. Nutr Clin Pract. 2014;29(3):360-367.
14. Wong JJ, Ong C, Han WM, Lee JH. Protocol-driven enteral nutrition
in critically ill children: a systematic review. JPEN J Parenter Enteral
Nutr. 2014;38(1):29-39.
15. Hamilton S, McAleer DM, Ariagno K, et al. A stepwise enteral
nutrition algorithm for critically ill children helps achieve nutrient
delivery goals. Pediatr Crit Care Med. 2014;15(7):583-589.
16. Pollack MM, Holubkov R, Funai T, et al. The pediatric risk of
mortality score: update 2015. PCCM. 2015;17(1):2-9.
17. Costa GA, Delgado AF, Ferraro A, Okay TS. Application of the
Pediatric Risk of Mortality Score (PRISM) and determination of
mortality risk factors in a tertiary pediatric intensive care unit. Clinics
(Sao Paulo). 2010; 65(11):1087-1092.
18. Brown KE, Johnson KJ, DeRonne BM, Parenti CM, Rice KL. Order
set to improve the care of patients hospitalized for an exacerba-
tion of chronic obstructive pulmonary disease. Ann Am Thorac Soc.
2016;13(6):811-815.
19. Mehta NM, Bechard LJ, Chill N, et al. Nutritional practices and their
relationship to clinical outcomes in critically ill children—an interna-
tional multicenter cohort study. Crit Care Med. 2012;40(7):2204-2211.