The Joint Commission Journal on Quality and Patient Safety 2019; 45:543–551

Prompting Rounding Teams to Address a Daily Best

Practice Checklist in a Pediatric Intensive Care Unit
Christina L. Cifra, MD, MS; Mandi Houston, MSN, RN, CCRN; Angela Otto, MSN, RN, CNML;
Sameer S. Kamath, MBBS

Background: Implementation of best practices for pediatric ICU (PICU) patients is challenging. The objective of this project
was to improve process of care outcomes and clinical outcomes by having a dedicated person (quality champion [QC]) prompt
PICU rounding teams to address a daily best practice rounding checklist.
Methods: A prospective cohort study was performed in an academic tertiary referral PICU, which implemented a daily rounding
checklist, including reminders to assess central line/urinary catheter need, sedation goals, sedative/paralytic need, enteral nutrition
readiness, and extubation readiness. Data were collected on patient characteristics, process of care outcomes, and clinical outcomes
over three periods: before and after the checklist was implemented and after the practice of prompting for checklist use was instituted.
Results: Over nine months, 444 patients were included. The QC was present on rounds 94 of 139 (67.6%) days. Checklist
adherence (all checklist items discussed daily) improved from 75.7% to 86.6% of patients. There was a reduction in urinary
catheter days across all time periods (p = 0.001), and post hoc analysis showed fewer blood draws (p = 0.049) among patients
for whom the QC was present consistently during rounds. There was also a decrease in PICU length of stay after the checklist
was implemented (p = 0.008), although this may be due to less severity of illness in the prompted cohort.
Conclusion: Prompting PICU rounding teams to address a daily best practice rounding checklist may improve some process
of care outcomes. Further study is needed to delineate long-term effects of this initiative

C ritically ill children are at high risk for adverse events ow-
ing to the complex nature of their illness and the dynamic
medical care they require. 1 Because pediatric critical care entails
At the University of Iowa PICU, a daily best practice
rounding checklist was instituted in March 2013 with the
goal of optimizing patient outcomes and avoiding harm.
multiple, often invasive interventions, constant vigilance is However, rounding teams used the checklist inconsistently.
needed to deliver high-quality and safe medical care. 2 Because Administrative mandates were insufficient for faithful adop-
there is a well-known link between best practice adherence tion of the checklist, given the many competing patient care
and outcomes, 3–5 best practice interventions such as central demands. In this project, multidisciplinary leaders and staff
line bundles 6 and extubation readiness protocols 7 have been sought to improve adherence to the rounding checklist by
adopted over the years by pediatric ICUs (PICUs). implementing real-time audit and feedback by a dedicated
However, consistent implementation of best practices is chal- QC. We hypothesized that consistent performance of key
lenging in a busy PICU environment. One way to address this best practices as prompted by the checklist would lead to im-
challenge is by using specific rounding checklists to ensure that proved process of care outcomes and clinical outcomes for
best practices are performed every day for every patient. Check- patients, and that the presence and prompting of the QC
lists have been shown to be efficient and effective for improving would be well accepted by the PICU staff.
patient safety and outcomes. The pioneering work of Pronovost
and Gawande demonstrate that checklists can help reduce cen-
tral line infections and surgical complications. 8,9 Checklists METHODS
have since been applied to many other areas of health care,
This work was reviewed and determined to be a quality im-
including ICU settings, where they have been shown to pre-
provement project, exempt from oversight of human sub-
vent errors of omission, 10,11 prevent adverse events, 12 and
jects research by the University of Iowa Institutional
increase compliance with clinical practice guidelines. 13,14
Review Board (IRB) (IRB #201603782, Iowa City, Iowa,
Recent work shows that having a dedicated person (quality
March 16, 2016).
champion [QC]) to ensure checklist adherence further de-
creased ventilator days, length of stay, and mortality among Environment and Context
ICU patients. 15
The project was conducted at the University of Iowa PICU,
1553-7250/$-see front matter
a 20-bed unit accommodating ~ 1,000 admissions per year of
© 2019 The Joint Commission. Published by Elsevier Inc. All rights re- patients aged 0 to 21 years. It serves as a tertiary referral cen-
served. ter for the state of Iowa and other nearby states. The PICU
https://doi.org/10.1016/j.jcjq.2019.05.012
Volume 45, No. 8, August 2019 544

admits a wide variety of surgical and medical cases, including (reviewing charts) and a PICU nurse (accompanying the
patients undergoing cardiac surgery or solid organ trans- QC during rounds) to ensure accuracy. Data collected in-
plants, requiring extracorporeal membrane oxygenation, or cluded patient characteristics (age, sex, case category [medical
needing intensive trauma care. vs. surgical], diagnostic category [main organ system in-
Patient safety and quality improvement initiatives are con- volved in illness—for example, respiratory, cardiovascular],
ducted and monitored by a dedicated PICU Interdisciplinary and risk of mortality scores), patient process of care out-
Quality and Safety Team (IQST). This group conducts comes (sedative/paralytic infusion days, central line/urinary
twice-monthly meetings to discuss and address tracked qual- catheter days, NPO (nil per os) days, mechanical ventilation
ity metrics and reported patient adverse events. The group days, number and volume of blood draws, and number of
has also spearheaded various PICU quality improvement laboratory tests performed), and clinical outcomes (central
projects, including implementation of the daily best practice line–associated bloodstream infection [CLABSI] rates,
rounding checklist. catheter-associated urinary tract infection [CAUTI] rates,
PICU length of stay, hospital length of stay, and mortality).
Project Design We also collected data on checklist adherence (median pro-
portion of patients per rounding day wherein all checklist
In March 2013 the PICU IQST instituted a rounding items were discussed) and rounding time per patient (time
checklist intended to be a quality improvement tool to im- from start to end of rounds per patient) for the latter two
prove adherence to best practices for every patient. The time periods. In addition to these, data were collected on
checklist provides reminders to the PICU rounding team the PICU staff’s perceived acceptability and usefulness of
to ensure that eligible patients receive interventions to reduce the checklist and the QC’s role to improve adherence. Two
harm (for example, daily assessment of central line need to staff surveys with similar questions were administered before
decrease central line days and prevent infection) and improve and after prompting was instituted.
quality of care (for example, daily assessment of enteral nutri-
tion readiness to more quickly fulfill nutrition needs and pre- The PICU Quality Champion and Conceptual
vent malnutrition). The checklist is prominently displayed at Framework
the door of every room and was intended to be used daily
during rounds for each patient in the PICU. A person with sufficient medical background to have a gen-
eral understanding of the care provided in the ICU (not nec-
Project Periods and Patients. For this project, a pro- essarily having specific ICU experience) was hired to fulfill
spective cohort design was employed to compare outcomes the PICU QC role. This role was funded internally by the
during three periods: (1) no checklist: before implementation University of Iowa through a quality and safety grant provid-
of the checklist (3 months—December 1, 2012, to February ing a full-time position ($40,000) for a year for a general
28, 2013); (2) no prompting: after implementation of the PICU quality improvement manager.
checklist but before real-time audit and feedback by a dedi- The QC underwent two weeks of orientation and training
cated PICU QC (4.5 months—December 1, 2015, to April prior to starting in the new role, during which he was intro-
14, 2016); and (3) prompting—after real-time audit and feed- duced to the PICU multidisciplinary staff and attended reg-
back by a dedicated PICU QC attempting to increase use of ular PICU IQST meetings. He was familiarized with the
the checklist (4.5 months—April 15, 2016, to August 31, PICU work flow, including the various quality improvement
2016). Although the intervention was implemented for all pa- initiatives that are in place. Specifically, he was oriented to
tients, only those admitted and discharged during each spe- the team rounding process and observed how daily rounds
cific time period were included in the analysis to enable us were conducted. He also received education and training
to isolate effects of the intervention on outcomes (that is, we on each item in the rounding checklist.
excluded patients whose PICU stay overlapped onto more After orientation, the PICU QC participated in daily
than one project time period). The following patients were rounds for every patient (weekdays only). During the no
also excluded from analysis: (1) patients readmitted to the prompting period, he audited the rounding team’s adherence
PICU during the same hospital admission and (2) patients to the checklist. He specifically noted whether the following
transferred from another ICU. These patients were excluded items were discussed, if applicable to the patient:
because outcomes may have been affected by medical man-
agement outside of the University of Iowa PICU. 1. Review of mobility goal
2. Assessment of central line need and removal if not necessary
Data Collection and Outcomes. Data collection was 3. Review of frequency of blood draws for laboratory tests
performed by the same PICU QC after a period of orienta- 4. Review of State Behavioral Scale (SBS) 16 goal for sedation
tion and training using a standard process of chart review 5. Assessment of need for sedatives/paralytics and weaning or
for all time periods and by bedside observation during interruption if decreased need
rounds for the no prompting and prompting time periods. 6. Assessment of ability to start enteral nutrition
Collected data were audited by a pediatric intensivist 7. Assessment of urinary catheter need and removal if not necessary
545 Christina L. Cifra, MD, MS, et al. Prompting Rounding Teams to Address a Daily Best Practice Checklist

8. Assessment of extubation readiness by planning to perform a also performed comparing the same variables across all time
spontaneous breathing trial periods but including only patients in the prompting phase
who were rounded on with the QC present ≫ 50% of the
During the prompting period, the QC also audited adher-
time (high QC presence). Comparisons were performed
ence to the checklist as described; however, if an item was
using the Kruskal-Wallis test for continuous variables and
not discussed, he prompted discussion using scripted ques-
chi-square test or Fisher’s exact test for categorical variables.
tions. A verbal script was used to address each item if it
Survey responses were compared using the Student’s t-test.
was missed. For example, if continued need for central lines
A p value ≪ 0.05 was considered statistically significant. Sta-
was not discussed, the QC will prompt, “Does the patient
tistical analysis was performed using Stata 12.1 (StataCorp
still require use of a central line?” Questions were directed
LLC, College Station, Texas).
to the PICU attending physician, fellow, and/or assigned res-
ident/nurse practitioner. The PICU QC did not have any
other role during rounds except to perform this audit and
RESULTS
feedback function and to collect relevant data. A conceptual
framework for how real-time audit and feedback can poten- Patient Characteristics
tially improve patient outcomes is illustrated in Figure 1. Over a nine-month period, a total of 444 patients were in-
Statistical Analysis cluded, 69 from the no checklist period, 133 from the no
prompting period, and 242 from the prompting period.
Descriptive data analysis was performed, reporting means, Among all time periods, there was no difference in age,
medians, and proportions as appropriate. PICU staff survey sex, and case category (medical vs. surgical). The top three
responses (on a Likert scale) were scored per item (0 – diagnostic categories over all time periods were respiratory,
strongly disagree; 1 – disagree; 2 – agree; 3 – strongly agree), cardiovascular, and neurologic. There were more neurologic
and the mean score per item was calculated and compared cases after the checklist was instituted than before (14.5% vs.
for the two surveys administered before and after the inter- 30.1%–33.5%; p = 0.009). Patients’ risk of mortality was
vention. We compared patient characteristics, process of care less in the prompting period compared to other periods, as
outcomes, and clinical outcomes across all time periods. To measured by standard PICU severity of illness scoring sys-
better assess outcomes in patients who received the prompt- tems (Pediatric Index of Mortality 2 [PIM 2] and Pediatric
ing intervention more consistently, a post hoc analysis was Risk of Mortality III [PRISM III]) (1.35% vs. 1.17% vs.

Figure 1: This conceptual framework relates how real-time audit and feedback during daily pediatric ICU (PICU) rounds can po-
tentially improve process of care outcomes and clinical outcomes of patients. SBS, State Behavioral Scale.
Volume 45, No. 8, August 2019 546

0.87%; p ≪ 0.001; and 0.51% vs. 0.51% vs. 0.38%; p = (29.2%) and assessing the continued need for sedatives/para-
0.047, respectively). Post hoc analysis including only the lytics (19.7%) (Table 2).
high QC presence cohort in the prompting period showed
similar findings (Table 1).
Process of Care Outcomes and Clinical Outcomes
Prompting Requirements and Checklist Adherence Across Time Periods

The QC was present on rounds 94 days out of the 139-day There were no differences in sedative/paralytic infusion
prompting period (67.6%) (all weekdays), for a total of 581 days, central line days, NPO days, ventilator days, total
rounding patient-days (sum of the number of patients rounded volume of blood drawn for point-of-care labs, and num-
on for each rounding day). There was no difference in mean ber of complete blood counts and basic metabolic panels
rounding time per patient before and after prompting was insti- drawn per patient. There was, however, a reduction in
tuted (8.4 vs. 8.2 minutes; p = 0.403), indicating that prompt- mean urinary catheter days per patient across time pe-
ing did not prolong rounds. Checklist adherence (median riods (3.5 vs. 2.0 vs. 2.1 days; p = 0.001). That remained
proportion of patients per rounding day wherein all checklist true in the post hoc analysis including only the high QC
items were discussed) improved from 75.7% to 86.6% after presence cohort in the prompting period (3.5 vs. 2.0 vs.
prompting was instituted (Figure 2). There was a decrease in 1.7 days; p ≪ 0.001). Post hoc analysis also showed a reduc-
adherence immediately after prompting was started, likely as tion in the mean number of blood draws per patient between
both the QC and the rounding teams were adjusting to the in- the no checklist/no prompting and prompting (high QC
tervention (for example, QC learning when and how to presence) time periods (15.2 vs. 14.0 blood draws; p =
prompt, rounding teams waiting for prompting initially). 0.049). Similarly, post hoc analysis showed a trend to-
Checklist adherence did not vary by day of the week. Depend- ward a reduction in mean sedative infusion days per pa-
ing on the checklist item, prompting was required on 2.8% to tient across all time periods (4.5 vs. 2.4 vs. 3.2 days; p =
29.2% of rounding patient-days. The two items that required 0.057), though this was not statistically significant
the most prompting were indicating an SBS goal for sedation (Table 3).

Table 1. Patient Characteristics

With Checklist
Prompting†
No Checklist No Prompting Prompting p (High QC Presence) p
Patient Characteristics n = 69 n = 133 n = 242 Value* n = 185 Value‡
Age, months, median (IQR) 49.5 (145.4) 31.1 (128.6) 64.9 0.365 56 (140.7) 0.552
(142.4)
Sex, male, n (%) 39 (56.5) 77 (57.9) 135 (55.8) 0.925 108 (58.4) 0.965
Case category, n (%)§
Medical 37 (53.6) 74 (55.6) 124 (51.2) 0.874 93 (50.3) 0.806
Surgical 32 (46.4) 58 (43.6) 116 (47.9) 90 (48.7)
Cardiac surgical 10/32 (31.3) 19/58 (32.8) 32/116 0.761 24/90 (26.7) 0.708
(27.6)
Top 3 diagnostic categories, n (%)
Respiratory 20 (29.0) 31 (23.3) 53 (21.9) 0.471 41 (22.2) 0.518
Cardiovascular 17 (24.6) 28 (21.1) 49 (20.2) 0.733 35 (18.9) 0.600
Neurologic 10 (14.5) 40 (30.1) 81 (33.5) 0.009 66 (35.7) 0.005
Risk of mortality (%)
PIM 2–based risk of mortality, me- 1.35 (3.23) 1.17 (2.75) 0.87 (1.50) ≪ 0.85 (1.00) ≪
dian (IQR) 0.001 0.001
PRISM III–based risk of mortality, 0.51 (1.58) 0.51 (0.78) 0.38 (0.60) 0.047 0.38 (0.44) 0.020
median (IQR)

QC, quality champion; IQR, interquartile range; PIM, Pediatric Index of Mortality; PRISM, Pediatric Risk of Mortality.
*
No checklist, no prompting, and prompting groups were compared. Continuous variables were compared using the Kruskal-Wallis test,
while categorical variables were compared using the chi-square test.
†
Post hoc analysis including only patients in the prompting group who were rounded on with the quality champion present ≫ 50% of the
time (high QC presence).
‡
Statistical tests used are the same as noted, this time comparing no checklist, no prompting, and prompting (high QC presence) groups.
§
A third case category, Trauma, was not included in the table.
547 Christina L. Cifra, MD, MS, et al. Prompting Rounding Teams to Address a Daily Best Practice Checklist

Figure 2: The proportion of patients wherein all relevant checklist items were discussed during daily rounds increased by a me-
dian of approximately 11% after prompting by the quality champion was initiated. PICU, pediatric ICU.

There were no documented CLABSIs or CAUTIs analysis (Table 3). There was no difference in mortality
among the patients included in the analysis during the across all time periods.
project time period, thus we are unable to determine if
the intervention would have affected infection rates. PICU PICU Staff Perceptions
and hospital length of stay decreased across all time periods Table 4 shows survey responses by PICU staff members be-
(median of 3 vs. 2 vs. 2 days; p = 0.08; and 8 vs. 7 vs. 5 days; fore and after institution of prompting. For both surveys,
p = 0.02, respectively), with similar findings in the post hoc most respondents were composed of nurses (65.7% and
85.3%), followed by physicians (34.3% and 14.7%), while
the rest were composed of nurse practitioners, respiratory
Table 2. Prompting Required therapists, perfusionists, nutritionists, and social workers. Fa-
miliarity with the rounding checklist increased from 91.3%
Prompting Patient-Days/
Checklist Item Rounding Patient-Days (%)* to 100% from the first to the second survey. Survey re-
sponses increased in the desired direction for all items—a
Central line need 9/319 (2.8)
greater proportion of respondents indicated that the checklist
Urinary catheter need 12/133 (9.0)
helps them improve care for patients, that the checklist is
State Behavioral Scale goal for 50/171 (29.2) used consistently during rounds for every patient, and that
sedation
prompting by the QC has helped increase consistent use of
Need for sedatives/paralytics 42/213 (19.7)
the checklist. Respondents noted that the top barriers to
Ability to start enteral nutrition 8/165 (4.8) consistent use of the checklist were distractions (36.9%
Extubation readiness 5/193 (2.6) and 39.3%), forgetting by staff (24.8% and 25.7%), lack
Need for blood draws 33/580 (5.7) of time (13.9% and 6.1%), and having no assigned person
*
This is the total number of patient-days on which a prompting in- to discuss items on the checklist (10.3% and 13.6%).
tervention was required for the particular checklist item divided by
the total number of patient-days on which rounding occurred. For
each checklist item, only rounding patient-days during which the DISCUSSION
variable was applicable were included (for example, only days on
which patients were mechanically ventilated were included as This single-site project showed that a dedicated PICU QC
rounding patient-days for the checklist item extubation readiness).
prompting rounding teams to address a daily best practice
Volume 45, No. 8, August 2019 548

Table 3. Process of Care Outcomes and Clinical Outcomes

With Checklist
Prompting†
No Checklist No Prompting Prompting p (High QC Presence) p
Outcomes n = 69 n = 133 n = 242 Value* n = 185 Value‡
Process of Care Outcomes
Sedation and Paralysis (mechanically
ventilated patients only)
Patients on at least one sedative 46 (66.7) 50 (37.6) 70 (28.9) 45 (24.3)
infusion, n (%)
Mean sedative infusion days per pa- 4.5 (6.4) 2.4 (2.7) 3.9 (6.1) 0.080 3.2 (4.3) 0.057
tient with infusion/s (SD)
Patients on at least one paralytic 18 (26.1) 20 (15.0) 37 (15.3) 22 (11.9)
infusion, n (%)
Mean paralytic infusion days per pa- 1.5 (0.6) 1.4 (0.7) 1.7 (1.1) 0.451 1.5 (0.7) 0.649
tient with infusion/s (SD)
Line and Catheter Days
Patients with at least one central 46 (66.7) 64 (48.1) 86 (35.5) 53 (28.7)
line, n (%)
Mean central line days per patient 6.9 (11.0) 6.0 (6.8) 7.8 (11.7) 0.947 5.6 (7.0) 0.307
with at least one central line (SD)
Patients with a urinary catheter, n 49 (71.0) 73 (54.9) 106 (43.8) 72 (38.9)
(%)
Mean urinary catheter days per pa- 3.5 (5.0) 2.0 (1.7) 2.1 (2.1) 0.001 1.7 (1.3) b
tient with a urinary catheter (SD) 0.001
Nutrition
Mean NPO days per patient (SD) 0.6 (0.7) 0.6 (0.7) 0.5 (0.7) 0.449 0.5 (0.7) 0.240
Mechanical Ventilation
Patients on mechanical ventila- 51 (73.9) 69 (51.9) 74 (30.6) 44 (23.8)
tion, n (%)
Mean ventilator days per 5 (6.8) 2.7 (2.3) 4.9 (7.6) 0.153 3.4 (3.8) 0.130
patient (SD)
Laboratory Testing
Mean total volume of blood no data 5.6 (7.7) 6.3 (12.8) 0.365 5.3 (11.4) 0.094
drawn per patient in mL (SD)§
Mean number of blood draws per no data 15.2 (19.6) 16.8 (32.7) 0.221 14.0 (28.6) 0.049
patient (SD)§
Mean number of complete blood no data 4.8 (10.2) 4.4 (8.9) 0.596 2.8 (4.1) 0.490
counts resulted per patient (SD)
Mean number of basic metabolic no data 4.6 (10.9) 5 (11.4) 0.214 3.2 (4.3) 0.820
panels resulted per patient (SD)

Clinical Outcomes
PICU length of stay, days, 3 (5) 2 (5) 2 (3) 0.008 2 (2) b
median (IQR) 0.001
Hospital length of stay, days, 8 (5) 7 (8) 5 (8) 0.020 5 (6) b
median (IQR) 0.001
Death, n (%) 2 (2.9) 3 (2.3) 8 (3.3) 0.926 2 (1.1) 0.545

QC, quality champion; SD, standard deviation; NPO, nil per os; PICU, pediatric intensive care unit; IQR, interquartile range.
*
No checklist, no prompting, and prompting groups were compared. Continuous variables were compared using the Kruskal-Wallis test,
while categorical variables were compared using the chi-square test.
†
Post hoc analysis including only patients in the prompting group who were rounded on with the quality champion present ≫ 50% of the
time (high QC presence).
‡
Statistical tests used are the same as noted, this time comparing no checklist, no prompting, and prompting (high QC presence) groups.
§
For point-of-care labs.
549 Christina L. Cifra, MD, MS, et al. Prompting Rounding Teams to Address a Daily Best Practice Checklist

Table 4. Survey Responses Before and After Prompting by the PICU Quality Champion
Item Means*
No Prompting Prompting Mean p
Survey Items n = 92 n = 38 Change Value†
The daily best practice checklist helps improve care for patients. 2.2 2.5 0.3 0.030
The daily best practice checklist is used consistently during rounds for every 1.8 2.2 0.4 0.003
patient.
Having a dedicated PICU quality champion prompting use of the checklist will 2.2 2.4 0.2 0.211
help/has helped to increase consistent use.
Simple management changes as discussed using the checklist are carried out 1.4 1.6 0.2 0.220
in a timely manner.
Having a dedicated PICU quality champion will help/has helped to make sure 2.0 2.2 0.2 0.093
that management changes as discussed using the checklist are carried out in
a timely manner.

PICU, pediatric ICU.

*
Each survey item was scored based on the respondent’s Likert scale response (0–strongly disagree; 1–disagree; 2–agree; 3–strongly agree),
and the mean score per item was calculated. The minimum possible score is 0, while the maximum possible score is 3.
†
Mean scores were compared before and after prompting was implemented using the Student’s t-test.

checklist improved checklist adherence and decreased uri- prompted during rounds only 67.6% of the total number
nary catheter days and blood draws, although the interven- of prompting days. Even so, checklist adherence improved
tion was implemented for only a short period. There was after the prompting intervention, with an approximately
also a trend toward decreasing sedative infusion days, though 11% increase in the median proportion of patients for whom
this did not reach significance. all relevant checklist items were discussed. Third, only pa-
Prior work has shown that one way to sustain change in tients admitted and discharged during each project period
quality improvement efforts is to employ a consistent QC were included in the analysis (to isolate effects of the inter-
in concert with an established quality and safety team. 17–19 vention), thus excluding patients with longer lengths of stay
Specifically regarding rounding checklists, one study found for which best practice implementation may be more crucial
that having a dedicated champion to ensure adherence im- in providing consistently safe and high-quality care. A longer
proved both process of care outcomes such as decreased venti- period to study the intervention will capture this population
lator days and clinical outcomes such as decreased length of and perhaps show improvements in outcome that were not
stay and mortality. 15 This finding is in line with the current demonstrated in this time-limited project.
trends in implementation science focusing on more direct, Of note, the improvement in urinary catheter days coin-
point-of-care interaction with the front lines to initiate and cided with a separate effort to decrease CAUTI rates through
sustain change. 20 Furthermore, research on effective health focused staff education, consistent querying of providers re-
care teams has shown the importance of closed-loop commu- garding the need for a urinary catheter, and training in asep-
nication and situational awareness/monitoring. 21–23 Going tic techniques of catheter insertion and maintenance. Thus,
through a best practice checklist during rounds as a team im- the improvement noted may be due mostly to these efforts
proves communication and shared awareness of the patient’s with a small contribution of the checklist prompting inter-
situation, orienting team members to changes that need to vention. At the time of the study, there were no other ongo-
be made to ensure that best practices are implemented. ing significant quality improvement activities.
In this project, lack of improvement in most of the mea- Our results are in sharp contrast to a prior study that dem-
sured outcomes may be due to several factors. First, there onstrated that prompting for adherence to a similar checklist
was reasonably good adherence to the checklist even before improved not only multiple process of care outcomes but
the prompting intervention, with each checklist item requir- mortality and length of stay. 15 This may in part be due to
ing prompting on only 2.8% to 29.2% of rounding patient- a different study design using a concurrent controlled cohort
days. It is interesting to note that the two most prompted to compare two groups (checklist only vs. checklist with
items were specifying an SBS goal for sedation (29.2%) prompting). For this project, because there was no concur-
and assessing the continued need for sedatives/paralytics rent control group, the presence of the QC during the no
(19.7%). A trend (not statistically significant) toward reduc- prompting period may have artificially increased baseline
tion in sedative infusion days was noted, which may be a rates of checklist adherence. Furthermore, the baseline pop-
small effect of the prompting intervention and may reach sig- ulation characteristics of patients included in each time pe-
nificance given longer study. Second, implementation of the riod are different, the most important being that patients
intervention was not 100%—the QC was present and included after the checklist/prompting intervention have less
Volume 45, No. 8, August 2019 550

severity of illness. This is likely due to the long time period be- in this project was supported by a grant as a pilot; it is more
tween launching the checklist and instituting the prompting practical to have an existing staff member be the prompter,
intervention. Post hoc analysis, however, showed improve- though implementation may be inconsistent due to compet-
ment in total blood draws per patient for those who received ing demands. Currently, resident physicians at the Univer-
the prompting intervention more consistently (high QC pres- sity of Iowa are oriented to its function and use prior to
ence) despite having the same severity of illness as the original their rotation and are expected to automatically include discus-
prompting cohort. Despite demonstrating positive change in sion of checklist items at the end of rounds for each patient.
only a few of the outcomes measured, the PICU staff perceived Formal sustainability studies are required to determine lasting
significant value in implementing the checklist, noting that the effects of such an intervention. Also, although we did not see
checklist helps improve care for patients and that prompting much additional benefit from prompting, a dedicated QC
increases consistent use of the checklist. These findings are can possibly add value by auditing performance on multiple
consistent with another study that showed ≫ 80% staff satis- quality improvement initiatives in a busy ICU.
faction with a similar PICU checklist. 13 Prior work shows
that staff engagement and acceptance of an intervention
greatly contributes to the intervention’s effectiveness and CONCLUSION
sustainability. 24,25 In an informal audit performed three
months after the project was completed, we found that In this project, prompting PICU rounding teams to address a
91.5% of patients had all checklist items discussed during daily best practice rounding checklist improved some process
rounds, which may point to sustainability of high checklist of care outcomes, although we did not see an effect on clinical
use even without prompting, perhaps partly due to positive outcomes. It is unclear if these were related to less severity of ill-
perceptions by the PICU staff. ness in the prompting period. The checklist and prompting in-
This project has several limitations. First, this is a single-site tervention, however, were well-received by the PICU staff and
study with no control group, thus initial data collection per- were perceived to be valuable in improving patient care and out-
formed by the same QC without prompting may have influ- comes. Further study over a longer period with more patients is
enced results after the prompting was implemented. The needed to delineate long-term effects and sustainability of this
same QC also performed chart reviews and abstracted data initiative. Further studies can also focus on the impact of such
for all patients, which could have led to bias. This was mitigated a champion on implementation of new initiatives and sustain-
by rigorous training in standard data collection, the use of stan- ability of multiple quality improvement initiatives in the ICU.
dard data collection forms, and periodic data audits. Second,
the study periods were short and included for analysis only Funding. This study was supported by a grant from the University of Iowa
those patients admitted and discharged within each time pe- Stead Family Children’s Hospital Stead Leadership Award. Dr. Christina L.
Cifra is currently supported by an NIH-NICHD (National Institutes of Health–
riod, excluding longer stay patients. This can be mitigated by Eunice Kennedy Shriver National Institute of Child Health and Human
extending the study period to include more chronic patients; Development) Institutional K12 Award #HD027748. REDCap™, a secure
unfortunately, we were unable to do so due to limited re- Web application for building and managing online databases, was used for
sources. Third, patients’ severity of illness was less after prompt- data entry and organization. Use of this software was supported by the
National Center for Advancing Translational Sciences of the National
ing was instituted, which could explain improvements in length Institutes of Health under Award Number U54TR001356 through the
of stay. There were also more patients with neurologic diagno- University of Iowa Institute for Clinical and Translational Science.
ses admitted over time, mostly composed of elective neurosur-
gical patients who are usually less severely ill and do not have Acknowledgments. The authors thank Geoffrey Obel, the pediatric ICU
quality champion who carried out the project intervention by prompting
long PICU stays. Fourth, a coordinated effort to reduce rounding teams, performed bedside and chart review data collection, and
CAUTI rates was being implemented during the study period, assisted in data entry and organization for this study. The authors are also
which may have influenced the noted reduction in urinary grateful for the assistance of the following individuals in obtaining data for the
catheter days. Last, markedly fewer PICU staff responded to study: Tonya Naughton, Deb Jansen, Oluchi Abosi, Eric Endahl, and R. Cam
Davis.
the second survey administered, which may have biased results
toward positive perceptions. However, our informal audit three Conflicts of Interest. All authors report no conflicts of interest.
months after project completion still showed good checklist ad-
herence by staff, leading us to believe that positive staff percep-
tions of the checklist are likely accurate. Christina L. Cifra, MD, MS, is Clinical Assistant Professor, Division of Critical
Implementation of prompting for a longer period may Care Medicine, Department of Pediatrics, University of Iowa Carver College
of Medicine. Mandi Houston, MSN, RN, CCRN, is Clinical Practice Leader,
yield more significant improvements, thus University of Pediatric ICU, Stead Family Children’s Hospital, University of Iowa Hospitals
Iowa PICU quality and safety leaders have continued to pro- and Clinics. Angela Otto, MSN, RN, CNML, is Nurse Manager, Pediatric
mote consistent use of the checklist even after the project Cardiac ICU and Pediatric Transitional Cardiac Unit, St. Joseph’s Children’s
Hospital, Tampa, Florida. Sameer S. Kamath, MBBS, is Associate Professor,
ended. It is likely cost-prohibitive to have a dedicated QC Division of Critical Care Medicine, Department of Pediatrics, Duke University
performing only this prompting function, although we did School of Medicine. Please address correspondence to Christina L. Cifra,
[email protected].
not perform a formal cost-benefit analysis. The QC position
551 Christina L. Cifra, MD, MS, et al. Prompting Rounding Teams to Address a Daily Best Practice Checklist

REFERENCES 14. Lane D. A systematic review of evidence-informed practices

1. Agarwal S, et al. Prevalence of adverse events in pediatric in- for patient care rounds in the ICU. Crit Care Med. 2013;
tensive care units in the United States. Pediatr Crit Care 41:2015–2029.
Med. 2010; 11:568–578. 15. Weiss CH. Prompting physicians to address a daily checklist
2. Niedner MF, Muething SE, Sutcliffe KM. The high-reliability and process of care and clinical outcomes: a single-site study.
pediatric intensive care unit. Am. 2013; 60:563–580. Respir Crit Care Med. 2011 Sep 15; 184:680–686.
3. Woolf SH. The meaning of translational research and why it 16. Curley MA. State Behavioral Scale: a sedation assessment in-
matters. JAMA. 2008 Jan 9; 299:211–213. strument for infants and young children supported on me-
4. Peterson ED, et al. Association between hospital process per- chanical ventilation. Pediatr Crit Care Med. 2006; 7:107–114.
formance and outcomes among patients with acute coronary 17. Shaw EK. The role of the champion in primary care change ef-
syndromes. JAMA. 2006 Apr 26; 295:1912–1920. forts: from the State Networks of Colorado Ambulatory Prac-
5. Pronovost PJ, et al. Developing and implementing measures tices and Partners (SNOCAP). J Am Board Fam Med. 2012;
of quality of care in the intensive care unit. Curr Opin Crit 25:676–685.
Care. 2001; 7:297–303. 18. Stanford JR, Swaney-Berghoff L, Recht K. Cardiac surgical out-
6. Miller MR, et al. Reducing PICU central line–associated blood- comes improvement led by a physician champion working
stream infections: 3-year results. Pediatrics. 2011; 128: with a nurse clinical coordinator. Med Qual. 2012; 27:5–10.
e1077–e1083. 19. Zavalkoff S, Korah N, Quach C. Presence of a physician safety
7. Foronda FK, et al. The impact of daily evaluation and sponta- champion is associated with a reduction in urinary catheter uti-
neous breathing test on the duration of pediatric mechanical lization in the pediatric intensive care unit. PLoS One. 2015
ventilation: a randomized controlled trial. Crit Care Med. Dec 14; 10e0144222.
2011; 39:2526–2533. 20. Weiss CH, Wunderink RG. Prompting physicians to address a
8. Pronovost P, et al. An intervention to decrease catheter- daily checklist for antibiotics: do we need a co-pilot in the
related bloodstream infections in the ICU. Med. 2006 Dec ICU? Curr Opin Crit Care. 2013; 19:448–452.
28; 355:2725–2732. 21. Dietz AS. A systematic review of teamwork in the intensive
9. Gawande A. The Checklist Manifesto: How to Get Things care unit: what do we know about teamwork, team tasks,
Right. New York: Metropolitan Books/Henry Holt, 2010. and improvement strategies? J Crit Care. 2014; 29:908–914.
10. Byrnes MC. Implementation of a mandatory checklist of pro- 22. Pennsylvania Patient Safety Authority. Patient Safety Is En-
tocols and objectives improves compliance with a wide range hanced by Teamwork. Pennsylvania Patient Safety Advisory,
of evidence-based intensive care unit practices. Crit Care Jun 16, 2010 http://patientsafety.pa.gov:80/ADVISORIES/
Med. 2009; 37:2775–2781. Pages/2010sup2_14.aspx.
11. Ullman A. The KIDS SAFE checklist for pediatric intensive care 23. Agency for Healthcare Research and Quality. TeamSTEPPS®.
units. Crit Care. 2013; 22:61–69. https://www.ahrq.gov/teamstepps/index.html.
12. Stroud MH. Introduction of a rounding sticker improves care 24. Taitz JM, Lee TH, Sequist TD. A framework for engaging
and reduces infection rates in the pediatric intensive care unit physicians in quality and safety. BMJ Qual Saf. 2012;
(PICU). J Ark Med Soc. 2012; 109:114–117. 21:722–728.
13. Tarrago R. Reductions in invasive device use and care costs af- 25. Kalisch BJ, Curley M, Stefanov S. An intervention to enhance
ter institution of a daily safety checklist in a pediatric critical nursing staff teamwork and engagement. J Nurs Adm. 2007;
care unit. Qual Patient Saf. 2014; 40:270–278. 37:77–84.