If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Address for reprints: Tara C. Cosgrove, MD, Division of Pediatric Cardiology, The Heart Center and Center for Clinical Excellence, Nationwide Children's Hospital, 700 Children's Dr, Columbus, OH 43205.
The prevalence of postoperative cardiac arrest (CA) increases with cardiothoracic surgical case complexity and is associated with a 40% to 50% mortality. Despite having a low overall surgical mortality rate at our center, our postoperative CA rates were higher than expected, with an observed-to-expected ratio of 2.6. Utilizing quality improvement methodology, we evaluated the influence of proactive risk mitigation on postprocedure CA in a high-risk cohort of pediatric cardiac patients.
Methods
This single-center study utilized the Institute for Healthcare Improvement model. We created and implemented our Proactive Mitigation to Decrease Serious Adverse Events program in July 2020, prospectively enrolling preidentified high-risk patients. Enrolled patients underwent scheduled multidisciplinary reviews via virtual platform at 2 periprocedural time points with discussion of patient-specific risks and the subsequent development of proactive risk mitigation plans. Primary outcome measures were derived from the Pediatric Cardiac Critical Care Consortium national registry and included rate of postprocedure CA within 7 days and an institution-specific observed-to-expected ratio for postoperative CA.
Results
Our baseline median number of high-risk cases between postprocedure CAs was 3. Following project initiation, median high-risk cases between events increased to 7. Our observed-to-expected ratio for postoperative CA decreased from 2.56 during the 12 months before Proactive Mitigation to Decrease Serious Adverse Events program implementation to 1.01 during the 12 months after Proactive Mitigation to Decrease Serious Adverse Events program implementation, and hospital length of stay decreased by ∼10 days.
Conclusions
Implementation of periprocedure-related proactive risk mitigation strategies in high-risk pediatric cardiac patients led to improvement in postprocedure CA with a 133% increase in high-risk cases between events.
Use of periprocedural proactive risk mitigation strategies in high-risk pediatric heart patients reduced postprocedure cardiac arrest and was associated with a decrease in overall length of stay.
Many consider cardiac arrest a modifiable outcome in congenital heart patients with prior local and national success in cardiac arrest prevention via interventions targeting situational awareness. We expanded this work using proactive safety principles in high-risk cardiac patients with periprocedural huddles to facilitate risk mitigation, shared decision making, and psychological safety.
Each year in the United States, more than 35,000 babies are born with congenital heart disease, which represents the most common type of birth defect.
Over the past few decades, perioperative mortality rates have improved significantly, especially for complex surgeries, shifting the target for improvement to reducing morbidity.
In the current era, approximately 2% to 3% of pediatric patients undergoing cardiothoracic surgery experience a postoperative cardiac arrest (CA) with the prevalence ranging from 0.7% to 12.7% depending on case complexity.
Given that CA rates vary substantially between pediatric cardiac intensive care units (CICUs), CA prevention has been targeted by individual institutions and national quality collaborates as a modifiable outcome.
In health care and other high-risk industries, efforts to improve safety have typically followed a Safety I approach (ie, retrospective review of adverse events and design of interventions to prevent recurrence). As safety efforts evolve, there is recognition that complex adaptive systems require an additional approach, termed Safety II, which attempts to proactively mitigate risk before an event with efforts to enhance learning from what goes right. Proactive safety huddles and communication plans are 2 ways in which Safety II theory can be operationalized in health care, aiming to enhance interdisciplinary communication by preparing mitigation plans for unique or complex patient situations thus increasing situational awareness.
Whereas Safety II theory is gaining momentum in health care, there has been limited application with measurable influence on clinical outcomes.
The Heart Center and Center for Clinical Excellence at Nationwide Children's Hospital (NCH) are committed to continuous quality improvement. In 2019, despite having a low operative mortality (2.01%), we observed a relatively high incidence of postoperative CA with an observed-to-expected (O/E) ratio of 2.6. Stemming from multidisciplinary code reviews aimed at discovering root causes, a failure to recognize and act on clinical deterioration, delayed escalation, and delay in diagnostic testing were identified as opportunities for improvement. Concurrently, there was ongoing institutional work to enhance patient safety through the application of Safety II principles.
We sought to apply novel Safety II strategies to a high-risk cohort of cardiac patients during the periprocedure period with the global aim of reducing morbidity and mortality. Our primary aim was to increase the number of high-risk cardiothoracic surgery or interventional catheterization cases between postprocedure CAs (pCAs) from a baseline median of 3 cases to ≥6 cases (a 100% increase) within 6 months and sustain indefinitely.
Materials and Methods
Context
This project was a single-center quality improvement (QI) initiative conducted at NCH, the second largest freestanding pediatric hospital in the United States. The Heart Center performs more than 450 cardiothoracic surgeries per year and 550 cardiac catheterization procedures per year.
As standard of care, all patients being considered for surgical or catheter-based intervention are formally discussed at our twice weekly multidisciplinary conference to gain consensus on global management decisions such as the need for and timing of a cardiac intervention. All postoperative care is delivered in the CICU by a specialized medical team following a bedside handoff with input on management decisions from the proceduralist available upon request. This handoff involves key periprocedure stakeholders and includes a summary of the procedure and intraoperative course with postoperative management directed by the CICU attending physician. Routine CICU management decisions occur during twice-daily bedside family-centered rounds.
Before project implementation, defined mechanisms for acute patient management discussions outside the biweekly case conference were lacking. Informal discussions did occur, but the timing, content, and attendees were variable and would often occur after a patient deviated from the expected course and/or was acutely decompensating. Additionally, when consensus could not be reached, there was a lack of shared decision making that negatively influenced culture and psychological safety. CICU provider comfort in escalating acute clinical concerns to necessary parties was variable with no formal process to facilitate communication.
Medical Ethics Approval
The Institutional Review Board at NCH determined that this project was quality improvement and not human subjects research; therefore, review and approval was not required per institutional policy.
Preintervention
During spring 2020, we assembled a multidisciplinary QI team consisting of Heart Center faculty and our service line QI coordinator. The team utilized strategies derived from the Institute for Healthcare Improvement model for improvement, including Aim Statement and plan-do-study-act cycles. Using internal historical data, we focused initial efforts on developing criteria (Table 1) to identify high-risk patients for whom interventions could be targeted. Baseline data for patients fulfilling high-risk criteria were collected retrospectively with prospective collection of measures following project initiation.
Table 1Preidentified high-risk criteria
All neonatal surgeries requiring cardiopulmonary bypass (aged ≤30 d)
All Stage 1 hybrid procedures
All new surgical aortopulmonary connections
Any patient on a vasoactive infusion at the time of the procedure
Any patient intubated at the time of the procedure
Any diagnostic or interventional cardiac catheterization procedure on patients with the following diagnoses: neonatal pulmonary atresia with intact ventricular septum and critical aortic stenosis
Cardiac catheterization with placement of a patent ductus arteriosus stent
High-risk heart transplant or mechanical circulatory support patients
Any other patient deemed higher risk for morbidity/mortality based on status
The Proactive Mitigation to Decrease Serious Adverse Events (PROMISE) program was implemented in July 2020 with prospective enrollment of preidentified high-risk patients. Eligible patients underwent scheduled multidisciplinary reviews (ie, PROMISE calls) via virtual platform with 2 required time points: preprocedure and immediately postprocedure. Additional calls could be convened at the discretion of the CICU attending physician within the context of multiple factors, including patient clinical status, projected trajectory, and outcome of previous discussions. Required participants in PROMISE calls were key stakeholders, including the primary proceduralist(s), cardiac anesthesiologist, and cardiologists as well as intensivists, imagers, interventionalists, and divisional medical and surgical leadership. Additional members were invited ad hoc depending on patient-specific risks.
To ensure consistency with high-quality calls, a checklist was created containing key elements of PROMISE calls (Figure 1). The checklist facilitates review of the procedure plan by the proceduralist and anesthesia plan by the anesthesiologist. Then, current, and anticipated patient-specific risks are discussed with subsequent development of mitigation strategies to address identified risks, including thresholds for activating invasive interventions (eg, sternal opening). Finally, follow-up plans for additional reviews are discussed. This physical checklist was completed in real time by the call facilitator (typically the CICU attending physician) with completed forms entered into Research Electronic Data Capture and tracked by the QI team as a process measure.
After observing initial encouraging improvements in our outcomes of interest, we enacted changes in the electronic medical record (EMR) to aid in the sustainability of our results. During November 2021, we implemented a proactive safety plan EMR order that flags a patient as high-risk to improve situational awareness amongst all caregivers entering the patient chart. Additionally, PROMISE call note templates were developed within the EMR to allow for documentation of call timing, call participants, and patient-specific clinical risks and mitigation strategies. The electronic notes replaced the paper key element checklist and aided in tracking compliance.
Primary outcome measures
•
High-risk cases between postprocedural cardiac arrests: High-risk cases are defined in Table 1. pCA was defined as cardiopulmonary arrest requiring (any) chest compressions and/or defibrillation, or acute respiratory compromise requiring emergency-assisted ventilation, and/or leading to CA requiring chest compressions and/or defibrillation.
Given that our interventions were limited to the acute postprocedure period, we included pCAs if they occurred within 7 days following the surgery or catheter intervention. Multiple pCAs in the same patient were not counted separately.
•
Postoperative CA O/E ratio: O/E ratio was calculated by dividing the observed rate of postoperative CA from all surgical encounters by the expected rate of postoperative CAs. The expected rate of postoperative CAs was obtained from the Pediatric Cardiac Critical Care Consortium (PC4).
A ratio >1 indicates that there were more CAs observed than expected based on case-mix. Notably this outcome metric was limited to surgical patients only.
Secondary outcome measures
•
Postprocedure 30-day mortality: This was calculated as death occurring within 30 days of high-risk procedure and included both in-hospital and postdischarge mortality. We chose 30-day mortality instead of in-hospital mortality given the acute periprocedure focus of our intervention.
•
Postprocedural extracorporeal membrane oxygenation (ECMO): Denoted if a patient required ECMO following a high-risk procedure.
•
Postprocedure length of stay (LOS): Number of days from high-risk procedure to hospital discharge.
•
Hospital LOS: Number of days from hospital admission to hospital discharge.
•
Program impact: An internal survey performed 8 months following project implementation was administered to assess perceived influence of the PROMISE program. The survey contained the following question: “The Heart Center's PROMISE program is improving communication and teamwork amongst staff” with possible answers ranging from strongly agree to strongly disagree.
Process Measures
Process measures included compliance with specific interventions:
•
Percent of high-risk patients enrolled in PROMISE: Assessed monthly to evaluate effectiveness of processes developed to identify and enroll eligible high-risk patients. A percentage was calculated by comparing the number of eligible patients per month to the number of completed PROMISE calls during the same month.
•
PROMISE call key element checklist compliance: Assessed to evaluate compliance with PROMISE call key elements. Each checklist element was assessed individually by calculating a percent compliance for documented PROMISE calls; specifically, the number of calls with a completed element divided by total calls. This measure was limited to PROMISE calls entered into Research Electronic Data Capture (REDCap Consortium).
Balancing Measures
Balancing measures were tracked to monitor for unintended consequences:
•
Unplanned interventions: Defined by PC4 as 1 or more unplanned procedures, unplanned interventions in our project included cardiac reoperation(s) and/or cardiac catheterization(s).
•
Open sternum: Defined by PC4 as any planned or unplanned event where the patient's sternum was left open, in our project this included patients who return from the operating room with an open chest and those who were reopened following primary closure.
Given the rarity of CA events, our primary outcome measure was plotted as high-risk cases between pCA on a g-chart. The g-chart is a type of statistical process control (SPC) chart utilized in QI methodology to measure improvement for rare outcomes with higher values (more cases between incidents) indicating improvement. Hospital and postprocedure LOS were plotted on an X-bar chart, which is a type of SPC chart utilized to measure improvement for continuous outcome variables. Our SPC charts were generated using QI Macros SPC Software version 2020.10 (KnowWare International), an add-in to Microsoft Excel. We applied established rules from the Healthcare Data Guide to identify signals of special cause variation.
We compared patient demographic characteristics, clinical factors, secondary outcome measures, and balancing measures for the pre-PROMISE cohort (January through June 2020) and post-PROMISE cohort (July 2020 through December 2021) with inclusion of all high-risk patients. We performed a sensitivity analysis excluding high-risk catheterization cases to understand the influence of PROMISE on patients undergoing surgery. Data were summarized using frequency (percentage) for categorical variables and median (interquartile range) for continuous variables. Differences in characteristics of the pre-PROMISE and intervention cohorts were evaluated using Pearson χ2, Fisher exact, or Wilcoxon rank sum tests. Statistical analyses were performed in R version 4.0 (R Foundation for Statistical Computing).
Results
There were 106 total patients meeting high-risk criteria; 33 in the pre-PROMISE cohort and 73 post-PROMISE. We identified no significant differences in demographic and clinical variables between cohorts (Table 2) with similar frequencies of neonates, single ventricle diagnoses, and The Society of Thoracic Surgeons–European Association for Cardio-Thoracic Surgery 4/5 surgical procedures. We observed no intergroup differences in high-risk criteria.
Table 2Demographic characteristics, clinical characteristics, and risk factors by cohort
Pearson χ2 test, Fisher exact test, or Wilcoxon rank sum test.
Characteristic
Gender
.7
Male
61 (58)
20 (61)
41 (56)
Female
45 (42)
13 (39)
32 (44)
Diagnosis physiology
.6
Single ventricle
39 (37)
11 (33)
28 (38)
Biventricular
67 (63)
22 (67)
45 (62)
STAT category for surgical patients
.7
2
5 (5.6)
1 (3.4)
4 (6.6)
3
15 (17)
6 (21)
9 (15)
4
47 (52)
13 (45)
34 (56)
5
23 (26)
9 (31)
14 (23)
Neonatal CPB (≤30 d), n = 53
Prematurity
8 (15)
1 (6.7)
7 (18)
.4
Chromosomal abnormality
16 (30)
4 (27)
12 (32)
.9
Extracardiac abnormality
12 (23)
5 (33)
7 (18)
.3
Weight at surgery (kg)
3.4 (3.1-3.8)
3.5 (3.3-3.6)
3.4 (3.0-3.8)
.4
PROMISE high risk inclusion criteria
Neonatal CPB (≤30 d)
53 (50)
15 (45)
38 (52)
.5
Intubated at time of intervention
18 (17)
6 (18)
12 (16)
.8
Inotropic/vasoactive infusion at time of intervention
18 (17)
9 (27)
9 (12)
.058
Stage 1 hybrid procedure
16 (15)
5 (15)
11 (15)
.9
Aortopulmonary shunt
4 (3.8)
1 (3.0)
3 (4.1)
.9
Cath: PDA stent placement
14 (13)
1 (3.0)
13 (18)
.059
Cath: Diagnostic or interventional with diagnosis of PA/IVS or critical AS
6 (5.7)
3 (9.1)
3 (4.1)
.4
Values are presented as n (%) or median (interquartile range). PROMISE, Proactive Mitigation to Decrease Serious Adverse Events program; STAT, The Society of Thoracic Surgeons-European Association for Cardio-Thoracic Surgery; CPB, cardiopulmonary bypass; Cath, Cardiac catheterization; PDA, patient ductus arteriosus; PA/IVS, pulmonary atresia intact ventricular septum; AS, aortic stenosis.
∗ Pearson χ2 test, Fisher exact test, or Wilcoxon rank sum test.
Our baseline median number of high-risk surgical cases between pCA events was 3 (Figure 2). Since PROMISE launch, we have observed multiple signals of special cause variation. The first occurred in October 2020 with a single point outside the upper control limit (ie, an outlier). We noted additional special cause variation after the February and August 2021 pCA events because these fulfilled the outer third rule (ie, 2 of 3 successive points on the same side of and >2 σ from the centerline). Given multiple signals of special cause variation occurring temporally with our intervention, the centerline shifted with an increase in median cases between events to 7, representing a 133% increase from baseline. Utilizing traditional statistics, we saw a clinical reduction in pCA following PROMISE implementation (21% vs 8%, respectively) (Table 3), although this did not reach statistical significance (P = .11). Sensitivity analysis excluding high-risk patients undergoing catheterization revealed similar outcomes to our full cohort (Table E1). Chart review on all pCA events during the study period revealed 100% of arrests were cardiac in etiology.
Figure 2High-risk cases between postprocedure cardiac arrests (CAs): Each individual data point on this statistical process control g-chart represents a CA event with the x-axis containing the dates of events. The y-axis contains the number of high-risk cases between CAs with improvement represented by higher values (more cases between events). The solid center line (CL) represents the theoretical median of the distribution. Baseline data revealed a median of 3 cases between CAs, which increased to a median of 7 cases following project initiation (July 2020) given multiple signals of special cause variation (circled). The red hashed line represent the upper control limit (±3 SD above the CL) with no lower control limit. PROMISE, Proactive Mitigation to Decrease Serious Adverse Events program.
Values are presented as n (%) or median (interquartile range). PROMISE, Proactive Mitigation to Decrease Serious Adverse Events program; ECMO, extracorporeal membrane oxygenation; LOS, length of stay.
Our baseline unit-level postoperative CA rate for surgical patients in the 12 months before PROMISE implementation was 3.79% with an O/E ratio of 2.56 based on a PC4-derived expected CA rate of 1.48%. In the 12 months following PROMISE implementation, our observed postoperative CA rate was 1.74% giving an O/E ratio of 1.01 based on a PC4-derived expected CA rate of 1.73%.
Of the patients who experienced CA during the study period (n = 13), 54% were neonates, 38% were intubated at the time of the procedure, 30% were undergoing stage 1 hybrid procedure, 23% were on vasoactive support, and 8% had critical aortic stenosis undergoing cardiac catheterization. Fifty-four percent of patients who experienced an arrest fulfilled 1 high-risk criterion, whereas a minority (38% and 8%) fulfilled 2 or ≥3 high-risk criteria, respectively. Regarding diagnosis/procedure type, 46% of arrests occurred in patients undergoing the first stage of single-ventricle palliation (Norwood or stage 1 hybrid procedure), whereas 31% occurred in patients with dextro-transposition of the great arteries status after arterial switch. The remaining 23% (n = 3) had the diagnosis of truncus arteriosus, double outlet right ventricle, and total anomalous pulmonary venous return with a single patient in each category.
We observed no statistically significant difference between the 2 cohorts for postoperative ECMO or 30-day mortality (Table 3). Our baseline mean overall hospital LOS for high-risk patients was 77 days, with baseline mean postprocedure LOS of 45 days. We observed a shift on SPC charts; secondary outcome measures occurred temporally with PROMISE initiation with a new mean overall hospital LOS of 33 days and mean postprocedure LOS of 19 days (Figures 3 and 4). Using traditional statistics, reductions in overall median hospital LOS reached statistical significance (29 days vs 19 days; P = .015) but differences in postprocedure median LOS did not.
Figure 3Overall hospital length of stay (LOS): Each individual data point on this statistical process control X-bar chart represents the average hospital LOS for high-risk patients in a given month by procedure date (month/y on x-axis). The y-axis represents the average LOS in days with improvement represented by lower values (decreased LOS). The solid centerline (CL) represents the average of all data points within a specific process stage with the baseline period defined as January through June 2020 (average LOS, 77 days). Following project initiation (July 2020), the CL shifted (special cause variation circled with 8 consecutive data points below the CL) with a new average LOS of 33 days. Hashed lines represent the upper and lower control limits (±3 SD above and below the CL) with the lower control limit not visible because it overlaps the x-axis. PROMISE, Proactive Mitigation to Decrease Serious Adverse Events program.
Figure 4Postprocedure length of stay (LOS): Each individual data point on this statistical process control X-bar chart represents the average postprocedure LOS for high-risk patients in a given month by procedure date (month/y on x-axis). The y-axis represents the average LOS in days with improvement represented by lower values (decreased LOS). The solid centerline (CL) represents the average of all data points within a specific process stage with the baseline period defined as January through June 2020 (average postprocedure LOS, 45 days). Following project initiation (July 2020), the CL shifted (special cause variation circled with 8 consecutive data points below the CL) with a new average postprocedure LOS of 19 days. Hashed lines represent the upper and lower control limits (±3 SD above and below the CL). PROMISE, Proactive Mitigation to Decrease Serious Adverse Events program.
Following project launch, 80% of eligible high-risk patients had at least 1 PROMISE call. Specific reasons for nonadherence for the remaining high-risk patients was not formally captured. For enrolled patients with a single PROMISE call, insufficient data were collected to determine which of the calls was missed. Compliance with individual discussion elements on the PROMISE call key element checklist was high, ranging from 80% to 100% depending on the specific element. We observed 100% compliance regarding discussion of anticipated concerns and mitigation strategies for documented calls.
PROMISE calls averaged 10 minutes with complete attendance of required participants. The program was well received, with 70% of surveyed participants in this project agreeing that PROMISE improved communication and teamwork amongst staff. Table E2 contains themes of patient management strategies that were altered because of PROMISE call discussions. There were no unintended consequences of the program with no significant differences in rates of open sternum or unplanned interventions between the pre- and postintervention cohort (21% vs 22% [P = .9] and 27% vs 12% [P = .058], respectively).
Discussion
Given the unexpectedly high rate of postoperative CAs in our CICU, our team implemented periprocedural PROMISE calls for high-risk cardiac patients. These calls have facilitated multidisciplinary discussions of current and anticipated patient-specific risks along with the development of risk-mitigation strategies. With project initiation, we observed a marked increase in the number of high-risk cases between pCAs, a reduction in the incidence of postoperative CAs, a lower postoperative CA O/E ratio, and a reduction in overall and postprocedure LOS without unintended consequences (see Video 1 for a summary of the project).
Given the small sample size coupled with the low frequency of clinical events, it is not surprising that traditional statistics failed to detect a significant difference in pCA following PROMISE initiation despite a clinically important reduction from 21% to 8%. Similarly, we observed no significant influence on 30-day postprocedure mortality, likely given our low observed mortality rate. QI methodology circumvents some of these challenges by utilizing the g-chart as a statistical tool to detect special cause when tracking rare events with a set of well-established criteria to signal a statistical change in the system (ie, a shift).
The centerline of our g-chart (median) plotting cases between pCA events shifted after observing 2 signals of special cause variation occurring temporally with PROMISE initiation that would be extremely unlikely to occur by chance. During the study period, there were no known changes to routine procedure-related or postprocedure-related care and there were no concurrent initiatives related to improving pCA rates in high-risk patients. We therefore suggest that our interventions had an influence on outcomes of interest.
We additionally observed a reduction in both mean hospital and postprocedure LOS utilizing SPC chart methodology and median hospital LOS utilizing traditional statistics. We observed multiple statistical outliers with very high LOS following CA in the pre-PROMISE cohort with fewer outliers in the post-PROMISE cohort. We postulate that the reduced incidence of CA more dramatically influenced mean postprocedural LOS as median is robust to outliers, especially considering procedure complexity, and clinical risks were similar between the cohorts.
Single-center and multi-institution quality collaboratives are working to prevent CAs because the paradigm has shifted to consider this outcome, at least in part, modifiable. Riley and colleagues
spearheaded proactive risk mitigation for CA prevention in children with cardiac disease. They implemented “high-risk precautions,” a bedside tool to promote formal recognition of high-risk patients (including both medical and postsurgical patients) with subsequent discussion points aimed at creating a shared mental model of patient specific risk(s) and escalation parameters. They did not achieve sustained improvement until multidisciplinary engagement was achieved through multiple interventions focusing on team dynamics and communication. This work informed national quality collaborative (ie, PC4) efforts with preliminary data presented by Alten and colleagues
showing a 45% reduction in CICU CA rate across 19 centers following CA prevention bundle implementation with similarly focused interventions.
Whereas our study corroborates the influence of multidisciplinary situational awareness on clinical outcomes noted in previous work, differences between the studies merit additional discussion. The PROMISE program was focused on preventing postprocedure CA and was limited to interventions during the immediate periprocedure time frame. This intentional focus stems from the fact that more than half of postoperative CA occurs during the first 24 to 48 hours following intervention, making this a particularly vulnerable time.
Moreover, when exploring local factors contributing to our unacceptable CA rates, we identified deficiencies in the immediate postoperative period resulting in a suboptimal psychological safety, communication, and shared decision making that negatively influenced situational awareness not highlighted in the previous studies. Additional disparities between studies are methodological with notable differences in outcome measures. Both Alten and colleagues
assessed the influence of their interventions by tracking overall CA rates (CA per 1000 CICU-days) despite limiting their interventions to a smaller, high-risk cohort (ie, neonatal bypass, infants after single-ventricle palliation, and medical patients intubated within 4 hours of admission). In contrast, except for our expected CA measure (provided by PC4), the influence of the PROMISE program was based on data that included only patients exposed to the intervention rather than all CICU patients.
Unique to the PROMISE program, multidisciplinary calls were structured so all team members' concerns could be heard. A virtual platform was initially chosen given the COVID-19 pandemic and has remained the preferred venue for calls given excellent attendance, engagement, and convenience. The key element checklist used to facilitate calls not only contained critical elements to review but also identified specific stakeholders as the “responsible party” for that portion of the call to ensure contribution from all members of the team and to maintain participant engagement. Survey feedback revealed overwhelming agreement that the PROMISE program positively influenced communication and teamwork within the division. Creating a psychologically safe environment allows individuals to offer ideas, ask questions, and share concerns freely and without fear.
Psychological safety recognizes that high-performing teams require flexibility and openness, especially in complex or evolving situations. The PROMISE program has positively influenced psychological safety and shared decision making during the perioperative period allowing for more robust and fruitful discussions that influence situational awareness of patient-specific risks and acute management strategies with criteria for escalating clinical concerns.
To aid in sustainability, program high-risk criteria will be revisited periodically following multidisciplinary code reviews because additional risk factors may be identified. Future opportunities exist both locally and nationally because proactive risk-mitigation strategies and Safety II principles can be applied broadly to any high-risk cohort. There are likely opportunities to expand and collaborate with other PC4 centers on perioperative-specific CA prevention. Considering the association of CA to short- and long-term complications, there are opportunities to assess the influence of CA prevention on neurodevelopmental outcomes, patient/family satisfaction, and cost savings analyses.
Study Limitations
The findings from this QI project should be interpreted mindful of several limitations. This study was limited to a single center that has a strong culture rooted in continuous improvement coupled with a dedicated Center for Clinical Excellence that allocates resources for QI. Institutions without these resources may have difficulty duplicating our work. Additionally, the small absolute number of patients experiencing our outcome of interest makes detection of statistical differences challenging. Implementation of multiple concurrent interventions combined with insufficient compliance data make determining causality a limitation of this study. Finally, other factors may have influenced our outcomes that were beyond our control, including case mix, acuity, and volume that were unaccounted for and beyond the scope of this analysis.
Conclusions
Implementation of proactive risk-mitigation strategies periprocedurally in a high-risk cohort of pediatric and congenital cardiac patients was associated with improvement in pCA with a 133% increase in the number of high-risk cases between events. We additionally observed a decrease in postoperative CA incidence and O/E ratio. Efforts to prevent CA were associated with a reduction in the mean overall hospital and postprocedure LOS utilizing SPC methodology. Future studies are needed to assess the program's influence on other variables such as neurodevelopmental outcomes.
The Journal policy requires editors and reviewers to disclose conflicts of interest and to decline handling or reviewing manuscripts for which they have a conflict of interest. The editors and reviewers of this article have no conflicts of interest.
The authors thank the following individuals for their substantial contributions to this project: Catherine Dimes, RN; Kevin Dolan, BTS, MSHA; Mariah Eisner, MS; Elizabeth Grogan, MD; and Jenna Merandi PharmD, MS, CPPS. The authors also thank the clinical faculty within the Heart Center for their outstanding participation in the Proactive Mitigation to Decrease Serious Adverse Events program and dedication to continuous quality improvement.
Dr Tara Cosgrove provides a brief summary of the background, methods, and results of the Proactive Mitigation to Decrease Serious Adverse Events program, a quality improvement project aimed at postprocedure cardiac arrest prevention. Video available at: https://www.jtcvs.org/article/S2666-2736(22)00374-6/fulltext.
Dr Tara Cosgrove provides a brief summary of the background, methods, and results of the Proactive Mitigation to Decrease Serious Adverse Events program, a quality improvement project aimed at postprocedure cardiac arrest prevention. Video available at: https://www.jtcvs.org/article/S2666-2736(22)00374-6/fulltext.
Appendix E1
Table E1Sensitivity analysis of clinical outcomes by cohort, excluding cardiac catheterization cases (surgical only)
Values are presented as (%) or median (interquartile range). PROMISE, Proactive Mitigation to Decrease Serious Adverse Events program; ECMO, extracorporeal membrane oxygenation; LOS, length of stay.
00374-6&id=fx1.jpg){kind=link}
00374-6&id=gr1.jpg){kind=link}
00374-6&id=gr2.jpg){kind=link}
00374-6&id=gr3.jpg){kind=link}
00374-6&id=gr4.jpg){kind=link}
00374-6&id=fx2.jpg){kind=link}