Introduction
CHD affects approximately 1% or 40,000 live births each year in the United States.1 Approximately 25% of those are infants with critical CHD, who require a surgical or interventional cardiac procedure in the first year of life.1 Children with CHD are at risk for neurodevelopmental impairments, and though they are often mild, some children face severe developmental challenges. Risk factors for neurodevelopmental disorders in this population begin prenatally and continue through the perioperative period and after discharge from the hospital.Reference Marino, Lipkin and Newburger2–Reference Ryan, Jones and Allen4 Along this trajectory, there are opportunities to reduce the risk factors to optimise long-term outcomes. In the cardiac ICU, an expanded approach to pain management in the perioperative period may be an opportunity to mitigate risks associated with pain and pain medications.
Both unalleviated pain and exposure to opioids have detrimental effects on the developing brain and long-term neurodevelopment outcomes.Reference Iqbal O’Meara, Miller Ferguson and Zven5–Reference Cook, De Asis-Cruz and Kim12 Most studies evaluating the effects of pain and opioid exposure in neonates involve preterm infants, not specifically infants with CHD. However, given the delayed fetal brain maturation noted in infants with CHD, data from preterm infants may be relevant to our understanding of how pain and opioid exposure affect brain development in infants with CHD.Reference Limperopoulos, Tworetzky and McElhinney13–Reference Peyvandi and Rollins15 Recently, Cook and colleagues found that untreated pain in premature neonates was associated with hyperconnectivity between the cerebellum and limbic regions and later neurodevelopmental impairments, including lower motor and language scores and higher risk of autism.Reference Cook, De Asis-Cruz and Kim12 Unfortunately, pain medications also present significant risks to the developing brain.Reference Attarian, Tran, Moore, Stanton, Meyer and Moore10,Reference O’Byrne, Baxelbaum and Tam16 Morphine exposure has been associated with apoptosis of microglial and neuronal cells.Reference Attarian, Tran, Moore, Stanton, Meyer and Moore10,Reference Hu, Sheng, Lokensgard and Peterson17 O’Byrne et al described the long-term effects of neonatal morphine exposure in infants with CHD. At 2 years of age, higher total exposure to opioids in the first year of life was associated with lower cognitive, language, and motor scores.Reference O’Byrne, Baxelbaum and Tam16
Given the risk of both pain and pain medication on the developing brain, our team sought an innovative approach to optimise comfort in the cardiac ICU. We developed and implemented the Comfort Curriculum, a comprehensive developmentally centred approach to optimise comfort in the cardiac ICU. Our primary aim was to explore the association between implementation of the Comfort Curriculum and cumulative opioid dose in neonates undergoing cardiac surgery in the first 5 days after cardiac surgery. Our secondary aim was to evaluate the association between the implementation of the Comfort Curriculum and pain scores in neonates in the first 5 days after cardiac surgery.
Materials and methods
Development of comfort curriculum
The cardiac ICU at Children’s National Hospital is a 26-bed unit limited to critically ill children and young adults with heart disease, with an estimated 300 annual admissions. An interdisciplinary team from the cardiac ICU developed the Comfort Curriculum which includes three components: (1) a Bedside Non-Pharmacologic Intervention Resource, (2) an Infant and Toddler Holding Guideline, and (3) a standardised nurse-driven Pain and Sedation Pathway.
The non-pharmacologic support reference table is an evidenced-based resourceReference Baudesson de Chanville, Brevaut-Malaty, Garbi, Tosello, Baumstarck and Gire18,Reference Lisanti, Vittner, Medoff-Cooper, Fogel, Wernovsky and Butler19 developed by cardiac ICU nurses, physicians, and a child life specialist. The table includes practical interventions for both the nurse and patient caregivers to provide comfort for infants using the five senses: vision, hearing, taste, touch, and smell. The Infant and Toddler Holding Guidelines were developed after gaining consensus on the criteria for patient holding from surgeons, intensivists, physical and occupational therapists, and nurses. The guidelines include agreed-upon criteria and detailed processes for holding infants and toddlers after cardiac surgery given the monitoring and intravenous lines, tubes, and surgical drains which must be managed for safe holding. For patients who are not deemed to be safe for holding, parents are encouraged to provide positive touch, gentle supportive touch that promotes comfort and bonding at the bedside. The third component of the Comfort Curriculum is a standardised nurse-driven Pain and Sedation Pathway developed by the cardiac ICU pharmacist and cardiac ICU nurse practitioner with input from cardiac anaesthesiologists, nurses, and intensivists. This nurse-driven pathway protocolises pharmacologic management of pain and agitation for patients of all ages in the cardiac ICU. The pain pathway includes two arms: Fast track and Complex. All patients receive acetaminophen around the clock for the first 48 hours. Fast track patients also receive an infusion of dexmedetomidine for the first post-operative night, ketorolac for 48 hours around the clock, in addition to as-needed morphine and/or oxycodone. The Complex pathway has three arms: Acute, Maintenance, and Recovery. Complex Acute is utilised during the acute post-operative period to achieve steady pain control with the ability to escalate the infusion based on the number of boluses needed. Fentanyl and dexmedetomidine infusions are used in the Complex Acute phase. Complex Maintenance phase delineates the assessment of procedural and non-procedural boluses given in a 12-hour period and recommends active weaning if fewer the four non-procedural boluses were needed in a 12-hour period. Recovery phase provides guidance on transitioning off opioid infusion in preparation for extubation and includes recommendations for methadone if needed. Lorazepam is recommended for amnesia in patients over 4 years old who are on the Complex Acute phase.
Implementation of Comfort Curriculum
Prior to implementation, there was not a standardised approach to pain management. The full cardiac ICU team was educated about the Comfort Curriculum in the Fall of 2019 during mandatory nursing annual skills day presentations, bedside just-in-time teaching, and presentations to fellows, attendings, and pharmacists. The Comfort Curriculum was implemented in January 2020. Laminated copies of the Comfort Curriculum were placed at every bedside and uploaded to the cardiac ICU intranet page. During the first 8 weeks after implementation, feedback was solicited from staff by email, bedside rounding, and paper forms at the bedside. All feedback was collated, revisions were made, and the second version of the Comfort Curriculum was implemented in early March 2020. The Comfort Curriculum is now embedded in the nursing orientation and nursing annual skills days and included in fellow and attending education.
Inclusion and exclusions
Given that the three components of the CC are designed for the care of neonates, all cardiac surgical patients ≤30 days of age at the time of their first operation were included. Patients who required Extracorporeal Membrane Oxygenation (ECMO) or reinterventions within the 5 days of surgery were excluded.
Measurement of variables
Data collected from the first five post-operative days following neonatal cardiac surgery included the initial opioid infusion dose, maximum opioid infusion doses, and total cumulative opioid dose, including both the infusions and bolus doses. Opioid dosages were converted to morphine milligram equivalents (MME). Total cumulative dexmedetomidine doses were collected.
Pain scores using the FLACC scaleReference Merkel, Voepel-Lewis, Shayevitz and Malviya20 were documented a minimum of every 4 hours. The FLACC pain scale is scored from 0 to 10 and is based on five categories that are measured from 0 to 1 and include face, leg, activity, cry, and consolability. The FLACC scale is validated in non-verbal patients and those ages 2 months–7 years.Reference Merkel, Voepel-Lewis, Shayevitz and Malviya20 While not yet validated in neonates, this is the measure used in our unit for all infants regardless of age.
Surgical complexity was categorised by STAT classification as defined by the Society of Thoracic Surgeons.Reference Jacobs, Jacobs and Lacour-Gayet21 Patient demographics, surgical procedures, and ICU variables were collected from the medical record (see Table 1). Data collection included three cohorts: (1) the pre-implementation phase (May–October 2019), (2) the post-implementation phase (May–October 2020), and (3) the sustainability phase (May–October 2021). Study data were collected and managed using redcap electronic data capture tools hosted at Children’s National Hospital.Reference Harris, Taylor and Minor22,Reference Harris, Taylor, Thielke, Payne, Gonzalez and Conde23 This study was approved by the Children’s National Hospital Institutional Review Board.
Demographics
* Gestational age missing for one infant in the pre-implementation phase.
^ Significance test excluded those with unknown race and ethnicity.
^^ Multiple testing adjusted pair-wise comparisons yielded the following statistically significant difference: the pre-implementation versus sustainability phase (p = 0.040).
Statistics of primary and secondary results
The statistical analysis included all infants stratified by the three time periods described above. Normality of continuous variables was tested using the Shapiro–Wilk normality test and visual inspection of histograms within each time cohort. Few of the continuous variables were found to be sufficiently normally distributed; therefore, non-parametric tests were used throughout the analysis. Descriptive statistics were generated for each cohort and consist of medians and interquartile ranges (IQR) for continuous variables, and frequencies and percentages for categorical variables. Comparisons of categorical outcomes among the three cohorts used Chi-square tests or Fisher’s exact tests when the expected values were small. Comparisons of continuous outcomes among the three cohorts used a Kruskal–Wallis test. If the Kruskal–Wallis test was statistically significant (p ≤ 0.05), post hoc pair-wise comparisons were performed between the three cohorts using Wilcoxon rank-sum tests. The resulting p-values from these pair-wise tests were adjusted for multiple comparisons using the Sidak method to maintain a family-wise type I error rate of 5%. No additional adjustments for multiple testing were applied. A p-value of ≤0.05 was considered statistically significant for all comparisons, and all analyses were performed using STATA V17 (College Station, TX).
Results
Ninety-five patients were evaluated in the study, with 36 in the pre-implementation phase, 38 in the post-implementation, and 21 in the sustainability phase. Patient demographics and clinical and surgical characteristics were similar across all three cohorts, including race, ethnicity, frequency of genetic abnormalities, CHD class,Reference Clancy, McGaurn and Wernovsky24 STAT category,Reference Jacobs, Jacobs and Lacour-Gayet21 or frequency of delayed sternal closure. Although those in the sustainability phase were significantly younger at surgery than the pre-implementation cohort (8.5 days versus 5 days p = 0.040), their weights were not significantly different. There is no significant difference in age between those in the pre-implementation phase and post-implementation phase nor between the post-implementation phase and sustainability phase (see Tables 1 and 2).
Clinical and surgical characteristics
*Total opioids missing for one infant in the sustainability phase.
**Benzodiazepines were used infrequently in all phases; see text for details.
^Multiple testing adjusted pair-wise comparisons yielded the following statistically significant differences:.
Pre-implementation phase versus post-implementation phase (p = 0.005).
Pre-implementation phase versus sustainability phase (p = 0.035).
^^Multiple testing adjusted pair-wise comparisons yielded the following statistically significant difference:.
Pre-implementation phase versus post-implementation phase (p = 0.029).
^Duration of open chest missing for one post-Implementation phase infant. Duration calculated and compared for only those with a delayed sternal closure.
**Duration of mechanical ventilation missing for one infant in each phase.
Initial opioid dose
During the pre-implementation phase, the initial dose of continuous opioid infusion in the first 5 days post-operatively was significantly higher [median (IQR); 0.10 (0.06) MME/kg/hr] than the post-implementation phase [median (IQR); 0.06 (0.10) MME/kg/hr]; (p = 0.005) and the sustainability phase [median (IQR); 0.07 (0.05) MME/kg/hr; p = 0.035] (see Table 3).
Medication data
*Total opioids missing for one infant in the sustainability phase.
**Benzodiazepines were used infrequently in all phases; see text for details.
^Multiple testing adjusted pair-wise comparisons yielded the following statistically significant differences:.
Pre-implementation phase versus post-implementation phase (p = 0.005).
Pre-implementation phase versus sustainability phase (p = 0.035).
^^Multiple testing adjusted pair-wise comparisons yielded the following statistically significant differences:.
Pre-implementation phase versus sustainability phase (p = 0.046).
Maximum opioid infusion dose
Similarly, during the pre-implementation phase, the maximum opioid infusion dose in the first 5 days post-operatively was significantly higher [median (IQR); 0.14 (0.15) MME/kg/hr] than the post-implementation phase [median (IQR); 0.10 (0.12) MME/kg/hr; p = 0.029] and the sustainability phase, although it did not reach statistical significance [median (IQR); 0.10 (0.05) MME/kg/hr; p = 0.12] (see Table 3).
Total cumulative opioid dose
During the pre-implementation phase, the total cumulative opioid dose in the first 5 days post-operatively was higher [median (IQR); 11.08 (14.92) MME/kg in 5 days] than the post-implementation phase [median (IQR); 6.72 (12.01) MME/kg in 5 days] and sustainability phase [median (IQR); 6.21 (10.48) MME/kg in 5 days], although none of the differences reached statistical significance (see Table 3).
Dexmedetomidine doses
During the pre-implementation phase, total dexmedetomidine dose in the first 5 days post-operatively was significantly lower than the [median (IQR); 16.03 (43.97) microgram/kg in 5 days] than the post-implementation phase [median (IQR); 32.88 (49.63) micrograms/kg in 5 days, p = 0.0005] and the sustainability phase [median (IQR); 49.87 (34.23) micrograms/kg in 5 days, p = 0.046] (see Table 3).
Pain scores
The maximum pain scores in the sustainability phase were significantly lower than the pre-implementation group on post-operative day 1 [median (IQR); (0 (0) versus 0 (4), p = 0.026)] and on post-operative day 5 [median (IQR); (0 (3) versus 4 (6); p = 0.045)]. In addition, maximum pain scores in the sustainability phase were significantly lower than maximum pain scores in the post-implementation phase on day 1 [median (IQR); (0 (0) versus 4; p = 0.008)]. No significant differences were observed between cohorts for days 2, 3, and 4, nor were there significant differences in comparisons of median pain scores (see Figures 1 and 2).
Maximum pain scores.
Median pain scores.
Discussion
In the first 5 days after neonatal cardiac surgery, we found a significant decrease in the initial and maximum opioid doses and a signal towards a reduction in total opioid dose after implementation of the Comfort Curriculum, including a standardised sedation pathway, bedside non-pharmacoligic reference, and holding guideline. Additionally, these reductions were sustained during a 1-year follow-up. The reduction in initial opioid infusion doses suggests adherence to the pain and sedation pathway. While not statistically significant, there was a trend towards a reduction in the total cumulative opioid dose, with a nearly 50% reduction between the pre-implementation phase and both post-implementation and sustainability phases. Total opioid dose was lowest in the sustainability group, despite the higher rate of delayed sternal closure in that group (67%) compared to the pre-implantation group (36%) and post-implementation group (39%). Dexmedetomidine dosing was significantly higher in both post-implementation groups compared to the pre-implementation group. Prior to the implementation of the Comfort Curriculum, the use of dexmedetomidine varied depending on provider preference; however, after review of the literarature,Reference Kiski, Malec and Schmidt25,Reference Liu, Bian, Liu, Zang, Gu and Chen26 standardised use of dexmedetomidine was added to the pain and sedation pathway. The higher dexmedetomidine dosing may reflect adherence to the pain and sedation pathway and indicate opioid-sparing benefits. Despite the reduction of opioid doses, pain scores were not significantly higher in any of the phases of the study; in fact, on post-operative days 1 and 5, the maximum pain scores were significantly lower compared to our previous practice.
It has been shown that protocolised management of analgesic and sedative medications has effectively reduced opioid and benzodiazepine exposure.Reference Donnellan, Sawyer, Peach, Staveski, Nelson and Pratap27–Reference Smith, Besunder and Betters29 Non-pharmacologic support of comfort has been well described in neonatal intensive care literature, but less studied in the paediatric cardiac intensive care setting. Lisanti et al developed the “Individualized Family-Centered Developmental Care” model for patients in the cardiac ICU,Reference Lisanti, Vittner, Medoff-Cooper, Fogel, Wernovsky and Butler19 a model with specific interventions to promote maternal bonding and family engagement, to optimise environmental factors, and to involve physical, occupational, and speech therapists early in the hospitalisation. This model informed the development of our educational and resource materials for non-pharmacologic support portions of the Comfort Curriculum. Elevating the role of the parent or caregiver at the bedside by involving them in comfort management has several potential benefits; it allows the parent to buffer the noxious stress of the environment, and it gives parents a role at the bedside, thus building their identity as parents and their connection to their infant.Reference Lisanti, Vittner, Medoff-Cooper, Fogel, Wernovsky and Butler19,Reference Lisanti, Kumar, Quinn, Chittams, Medoff-Cooper and Demianczyk30 Recently, Elhoff and colleagues reported significantly improved haemodynamics and oxygen delivery measured during one of the key components of our Comfort Curriculum, parental holding.Reference Elhoff, Acosta and Flores31
Adaptation to noxious and stressful stimuli early in life is necessary to maintain homeostasis; however, prolongation of these responses can lead to dysregulated stress response and symptoms of traumatic stress, anxiety, cognitive deficits, and neuromotor delays later in life.Reference Tortora, Severino and Di Biase7,Reference Boggini, Pozzoli and Schiavolin8,Reference Victoria and Murphy11,Reference Taddio, Katz, Ilersich and Koren32 Additionally, untreated or undertreated pain in the post-operative cardiac surgery patient can lead to poorly tolerated haemodynamic consequences, including tachycardia, hypoxaemia, hypertension, systemic blood pressure instability, and bleeding.Reference Pollak and Serraf33 The mainstays of sedation and pain management in the neonate after cardiac surgery are opioids. Animal and human studies indicate that risks associated with opioid exposure in the developing brain include apoptosis in microglial and neuronal cells and altered neuroplasticity in post-synaptic sites of the limbic system.Reference Beltrán-Campos, Silva-Vera, García-Campos and Díaz-Cintra9,Reference Attarian, Tran, Moore, Stanton, Meyer and Moore10 These findings can have long-term consequences that impact learning, motivation, reward, and memory.Reference Beltrán-Campos, Silva-Vera, García-Campos and Díaz-Cintra9,Reference Attarian, Tran, Moore, Stanton, Meyer and Moore10 Infants in the cardiac ICU often undergo surgical procedures and require days or weeks of mechanical ventilation, analgesia, and sedative medications, placing them at higher risk of impaired brain development. Both pain and medications to treat pain or agitation are detrimental to the developing brain. Expanding our pain management practice to involve parents and include non-pharmacologic interventions may be an opportunity to mitigate both the risk of pain and exposure to unnecessary doses of pain medication.
Limitations
While the results of the Comfort Curriculum show a signal towards lower cumulative opioid doses and less pain, there are several limitations worth noting. This is a single-centre study that examined neonates in the post-operative period; therefore, the results are not generalisable to other patient populations or time periods in the cardiac ICU. We implemented all three components of the Comfort Curriculum at the same time, making it difficult to discern if one of the components was more impactful than others. Our Electronic Medical Record (EMR) does not have a standard place to document holding or other non-pharmacologic measures, and these interventions are not typically documented. This was a limitation to evaluating the adherence to the non-pharmacologic practices. In a retrospective study, it is challenging to measure other aspects of care or standard practice that may also affect the primary outcomes being studied. Indeed, the implementation of the Comfort Curriculum coincided with the beginning of the COVID-19 pandemic and visitor restrictions was in place perhaps impacting the availability of families to hold infants at the bedside. We know that visitor guidelines were less restrictive during the sustainability phase, and interestingly, we noted significantly lower median maximum pain scores on post-operative day 1 in the sustainability phase compared to the post-implementation phase [1 (4) versus 0 (0); p = 0.008]. To our knowledge, there were no major changes to the anaesthetic approach between the cohorts; however, individual anaesthetic practices were not evaluated. The number of neonates in the sustainability cohorts was significantly lower (N = 21) compared to the pre-implantation cohort (N = 36) and the post-implementation cohort (N = 38). We do not have an explanation other than normal variation in neonatal admissions, and to our knowledge there were no programmatic differences in the Heart Center during this time.
Conclusion
The results of this preliminary evaluation of a comprehensive developmentally centred approach to optimise comfort in neonates following cardiac surgery show a trend towards reduced cumulative opioid exposure without an increase in pain scores. Development and implementation of the Comfort Curriculum required multidisciplinary expertise and the willingness of the bedside team to adopt non-pharmacological measures to optimise comfort. The sustained effects over 1 year from the implementation phase to the sustainability phase are encouraging. Future research involving larger samples are needed to test the effects of a holistic, developmentally focused approach to reduce opioid exposure, increase family bonding, and improve comfort in neonatal cardiac surgical patients.
Acknowledgements
Dr. Yves Dudekem, Dr. Richard Jonas, Dr. Nina Deutsch, Dr. Ricardo Munoz, Prof. Dr. Rolandaxt-Fliedner, and Dr. Can Yerebakan for their input to the project and support from the leadership level.
Financial support
None.
Competing interests
The author(s) declare none.
Ethical standard
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national guidelines on human experimentation (please name) and with the Helsinki Declaration of 1975, as revised in 2008, and have been approved by the Children’s National Institutional Review Board. The authors assert that all procedures contributing to this work comply with the ethical standards and have been approved by the Children’s National Institutional Review Board.
