Cover Page The handle https://hdl.handle.net/1887/3178040

Cover Page

The handle

https://hdl.handle.net/1887/3178040

holds various files of this Leiden

University dissertation.

Author: Roeleveld. P.P.

Title: Matters of the heart: clinical studies in paediatric cardiac intensive care

Issue Date: 2021-05-19

This is the time.

This is the Place.

So we look for the future.

Land of Confusion

— Genesis

Discussion and

future perspectives

Discussion

and future perspectives

DISCUSSION

The research in this thesis was driven by clinical questions that arose during the care of children following heart surgery. The thesis is divided into three sections based on three important aspects of paediatric cardiac intensive care. There are of course many more important aspects as briefly eluded to in chapter 1, the introduction to this thesis, that are not part of this thesis and also deserve further exploration and investigation as much of that what we do is not supported by empirical evidence (yet).(1, 2) Therefore, the approach to postoperative cardiac intensive care can differ quite significantly between different centres and even between intensivists within the same centre. For instance, an inotrope which is standard care in once centre may be considered harmful in another or an anticoagulation test which is the norm for one doctor can be disregarded by another. With all the best intentions we, the intensive care team (doctors and nurses), supported by cardiologists, cardiothoracic surgeons, paediatric infectious disease specialists, and neurologists base our approach on the limited evidence available and on our own combined clinical experience. And in the meantime, while treating our patients, we try and answer some of those clinical questions and contribute in some small or larger way to the available evidence. (See chapter 2)

SECTION II:

SUPPORTING THE CIRCULATION FOLLOWING

PAEDIATRIC HEART SURGERY

Inotropes and low cardiac output syndrome

In section two we discussed several modalities that are often implemented in supporting the circulation following paediatric heart surgery. In chapter 3 we reviewed the available literature for the use of inotropes following paediatric heart surgery and studied the international use of inotropes via an international survey. The literature at the time failed to show a positive effect on important clinical outcome measures, such as mortality, duration of mechanical ventilation, or PICU length of stay. We found there was clear practice variation and, despite the lack of compelling evidence, Milrinone was being used by the vast majority (97%) of caregivers for the prevention of postoperative low cardiac output syndrome (LCOS). For the treatment of LCOS there was more diversity in inotrope usage where often a combination of inotropes (milrinone, adrenaline, dopamine, and/or dobutamine) were described. We suggested that randomized trials comparing the influence of different inotropes in specific disease states on clinically important outcome measures should preferably be performed in the future. However, we also realized that it would be difficult to convincingly show benefit of one inotrope over another because of the many other factors influencing outcomes.

One very interesting paper by DeWitt at al we did not discuss in chapter three, because it was an animal study, described the effects of commonly used inotropes on myocardial function under constant ventricular loading conditions.(3) The authors created an isolated working heart model of animal myocardium which maintained constant left atrial pressure and aortic blood pressure to identify load-independent differences between inotropic medications. Of the catecholamines, dobutamine increased cardiac output, contractility, and diastolic performance more than clinically equivalent doses of norepinephrine (second most potent), dopamine, or epinephrine (P<0.001). The use of triiodothyronine and milrinone was not associated with

significant changes in cardiac output, contractility or diastolic function, either alone or added to a baseline catecholamine infusion. Results are shown in figure 1 and 2. Milrinone has a well-established role in the treatment of heart failure. It decreases systemic vascular resistance and increases cardiac output in patients after cardiac surgery.(4) These data by DeWitt et al, in rodents, could suggest that the clinical increase in cardiac output by milrinone is far more attributable to afterload reduction than to an actual increase in contractility, lusitropy or heart rate.(3) Of course, the question remains how these in vitro results translate to children with congenital heart disease after cardiac surgery.

Since publication of our survey several studies have been published investigating the possible pharmacological interventions in the prevention and treatment of LCOS in children:

Compared to placebo, oral Triiodothyronine supplementation was shown to lower the occurrence of LCOS and also shorten time to extubation from 47 to 32 hours (median) in children undergoing congenital heart operations on cardiopulmonary bypass.(5, 6) In another randomized trial oral thyroxin also reduced duration of mechanical ventilation, mean ICU length of stay (LOS) and mean hospital LOS in children undergoing complex congenital heart surgery (e.g. TGA, truncus, TAPVD, Fallot).(7)

The prophylactic use of Levosimendan compared to placebo in 187 children under the age of 48 months undergoing cardiac surgery was shown to be safe, but failed to provide significant benefit in reducing LCOS or 90-day mortality.(8) The effect of Levosimendan and Milrinone on echocardiographic strain, hemodynamic variables, and clinical outcomes were found to be comparable in a prospective double-blind clinical drug trial in children between 1 and 12 months of age operated for VSD, AVSD or tetralogy of Fallot.(9) Neither drug could prevent a substantial decrease in left ventricular longitudinal strain of approximately one-third for the left ventricle and 50% for the right ventricle on the first post-operative day. Levosimendan compared with milrinone also did not reduce the occurrence rate of acute kidney injury in the same study population.(10)

Figure 2: Eff ects on cardiac output. Percentage in change from baseline in cardiac output as

categorized by inotrope and dose. Within Milrinone group, end points were measured with milrinone alone (solid black lines) and in addition to a baseline of dopamine 5 ug/kg/min (dashed black lines) and epinephrine 0.1 ug/kg/min (dashed gray lines). Reprinted with permission.

Milrinone was compared to Dobutamine in a randomized trial of 50 children (median age 1.2 years) undergoing open-heart surgery for congenital cardiac lesions in Switzerland.(11) Both drugs were well tolerated and safe. Effi cacy was defi ned as the need for additional vasoactive support which was approximately in two-thirds of patients and similar for both drugs. There was also no diff erence in clinically relevant outcome measures. The authors concluded that both drugs were equally eff ective in preventing LCOS. However, because neither drug was compared to placebo, they might also have been equally ineff ective. Because the study was designed as a pilot study no power calculation was performed and results need to be interpreted with caution.

A prospective study of the administrative American Pediatric Health Information System database (PHIS) of patients who underwent congenital heart surgery showed that between 2004 and 2015 the use of vasoactive agents (epinephrine, norepinephrine, dopamine, dobutamine, milrinone, and vasopressin) had decreased and only milrinone was found to be associated with decreased inpatient mortality. (12) Dobutamine had the greatest decrease in use from 19% in 2004 to 2% in 2015. Epinephrine was used in 73% of admissions, norepinephrine in 3%, dobutamine in 7%, milrinone in 81%, and vasopressin in 4%. Vasopressin is not available in the

Figure 1: Eff ects on systolic function (A) and stroke volume (B). Percentage of change in dP/

dtmax from an immediately preceding baseline of each dose of inotrope measured. Within Milrinone group, end points were measured with milrinone alone (solid black lines) and in addition to a baseline of dopamine 5 ug/kg/min (dashed black lines) and epinephrine 0.1 ug/kg/ min (dashed gray lines). *p < 0.05, **p < 0.01, n = 10 replicates per group. Error = SE. Reprinted and adapted with permission from DeWitt et al.

Netherlands.(13) The overall decrease in the use of vasoactive agents is likely due to improvements in perfusion and anaesthesia strategies in the perioperative period combined with a possible to a change in perceptions of what vasoactive agents are more optimal.(12) The association of Milrinone with decreased mortality is interesting, but from these data the effect of severity of disease cannot be put into perspective and should therefore be viewed with caution. As the authors themselves comment, their findings highlight that important questions still remain with regards to the actual indications for vasoactive support and which vasoactive agent(s) to use in each specific situation. I agree with them that designing studies to answers these questions would be challenging and would probably raise more questions than clinically relevant answers.

In ninety children with severe pulmonary hypertension (pulmonary artery pressures above 50 mmHg) undergoing open-heart surgery three different doses of milrinone (0.375 ug/kg/min, 0.5 ug/kg/min, 0.75 ug/kg/min) were compared in a prospective randomized trial.(14) In all three groups additional inotropic support was required (dopamine and/or epinephrine) because of hypotension. In the high milrinone group (0.75 ug/kg/min) significantly higher doses of additional inotropes were required compared to the two other milrinone doses. However, duration of MV and PICU LOS were comparable. These data show that higher doses of milrinone cause more vasodilatation with hypotension and concomitant higher requirement for vasopressors without positively (or negatively) influencing the clinically relevant outcome measures such as duration of MV and PICU LOS in children with significant pulmonary hypertension. The authors conclude that a low-dose (0.375 ug/kg/min) was equally effective as higher doses and that low-dose infusions are more beneficial in avoiding adverse events and decreasing inotrope requirements without affecting the duration of MV and ICU stay. But again, because milrinone was not compared to placebo, we do not know if milrinone was at all effective in improving outcomes. In conclusion, it is still unknown if milrinone (or any other inotrope) has a clear beneficial effect in preventing and/or treating LCOS post congenital heart surgery. It has not been proven or disproven yet.

Paradoxical hypertension

In chapter 4 we described the international practice variability regarding the treatment of paradoxical hypertension following surgical correction of coarctation of the aorta (CoA) in children. Nitroprusside was the first drug of choice for acute blood pressure control in two-thirds of the respondents. The literature review only showed very limited evidence regarding treatment of paradoxical hypertension. No treatment was found to be superior to another and most often multiple drugs were used to control blood pressure. We are currently performing a retrospective analysis of our experience

with labetalol in 46 children following CoA-repair. Labetalol treatment had not been published in the literature before we started this study. Preliminary data show that labetalol was used in 36% of children (46/129) following surgical repair of CoA and that treatment goals were reached after approximately 3 hours with a median dose of 1.0 mg/kg/hr [0.6 – 1.25]. Labetalol was effective in controlling blood pressure without the need for multiple drug regimens. In 4.5% (2/46) labetalol had to be switched to another hypertensive drug due to bronchoconstriction (1 patient) and bradycardia (1 patient). Labetalol seems to be safe and effective in controlling blood pressure following CoA repair.

During our retrospective analysis one prospective observational study was published using labetalol in 15 children following CoA-repair.(15) All 15 patients received labetalol for blood pressure control. The authors maintained a maximum dose of 20 ug/kg/min (1.2 mg/kg/hr) of labetalol which is similar to the median dose we used. However, they added nitroprusside in 12/15 (80%) of their patients to control blood pressure which is in sharp contrast with our population where we were able to control blood pressure with labetalol alone.

What blood pressure to aim for?

For the future we should focus on crucial gap in our current knowledge, which is that we do not actually know what blood pressure to aim for. Should we maintain the systolic blood pressure under p95 or p99 for age? And what actually is the (upper limit of) normal blood pressure in postoperative children and should we aim for systolic blood pressure or the mean arterial blood pressure? (16) We have to realize why we aim to control blood pressure: it is to reduce acute complications such as rupture of the surgical anastomosis and the occurrence of mesenteric arteritis, or the development of chronic hypertension. Therefore, to find the best strategy in controlling blood pressure, we should base this on the clinically relevant outcomes and not on blood pressure alone.

Neonatal cardiac ECMO

In chapter 5 we presented a review of neonatal cardiac extracorporeal membrane oxygenation (ECMO). ECMO is an invaluable tool for neonates with therapy resistant circulatory failure, especially in a cardiac surgical centre. The most difficult aspect of ECMO is selecting the right patient at the right time. Especially in ECMO it is important that there is a high ‘suspicion’ that the underlying problem leading to circulatory compromise is amenable to treatment. And again, as with inotropes, specific treatment needs to be tailored to the specific patient with his/her specific congenital heart disease. Special importance is given to the essential pillars of neonatal cardiac ECMO support: 1) Finding the cause of circulatory compromise with cardiac catherization,

2) providing adequate systemic blood flow, and 3) adequately decompressing the heart. We end the chapter by suggesting multiple areas of research for children with cardiac disease on ECMO as many challenges are still unclear. One important area, not limited to cardiac neonatal ECMO, is the anticoagulation. We have made one small contribution towards this important and difficult topic in showing that an anti-Xa-based anticoagulation strategy, when compared to a time-based strategy was associated with fewer bleeding events and mortality rate, without an increase in thrombotic events. This meta-analysis was the topic of chapter 12. Anticoagulation will remain a topic of future research in our department as also discussed in chapters 10 to 12 of this thesis.

Another important future focus should be pharmacology in relation to ECMO as drug dosing to optimize pharmacotherapy during ECMO is a major challenge.(17) Age (developmental pharmacology), critical illness (inflammation, capillary leak, hypoxia/ asphyxia, multi-organ dysfunction syndrome), drug characteristics (hydrophilic, lipophilic, protein binding), possible renal replacement modalities, and the ECMO circuit (sequestration) all have a dynamic effect on pivotal pharmacological processes such as volume of distribution and clearance of drugs. Therapeutic drug monitoring is essential where possible and further research into pharmacotherapy is highly needed.(17) At the moment, in a prospective observational study, we are analysing our Levosimendan data in children with and without ECMO. We have gathered levels of Levosimendan and its two active metabolites and are in the process of making a pharmacokinetic model which we aim will contribute to dosing regimens of Levosimendan in children with and without ECMO.

ECMO and univentricular heart disease

In chapter 6 we studied a very specific group of 44 patients supported with ECMO using the Extracorporeal Life Support Organization (ELSO) database. We found that neonates supported with ECMO following the hybrid procedure for hypoplastic left heart syndrome only had 16% survival which is half that of neonates supported with ECMO following the Norwood procedure. We were not able to identify predictors of mortality. One of the limitations of our study was that we were not able to calculate the incidence of ECMO support following the hybrid procedure given the nature of the ELSO database. After our study, one more study was published in this very specific group of neonates. In this retrospective, single-centre review of all children who underwent the hybrid procedure only two out of 181 patients (1.1%) required ECMO-support. Both children died. The authors concluded that mortality in patients who underwent ECMO after the hybrid procedure was higher than reported for the Norwood procedure, but that the incidence was also significantly lower than in children following the Norwood procedure (approximately 10-20%).(18-22). At the time of writing this thesis, no neurodevelopmental follow-up studies have been published for children who survived ECMO following the hybrid procedure for single-ventricle heart

disease. Focus of future ECMO research should also include long-term outcomes, not just hospital survival.

SECTION III:

RESPIRATORY INFECTIONS IN PAEDIATRIC CARDIAC INTENSIVE CARE

Ventilator associated pneumonia

In section three we discussed the impact of respiratory infections in paediatric cardiac critical care. In chapter 7 we investigated ventilator-associated pneumonia (VAP) in children after cardiac surgery and found a higher mean VAP rate than reported in the general paediatric intensive care population. VAP was mainly, but not exclusively, caused by Gram-negative bacteria. Children with VAP had a prolonged need for mechanical ventilation and a longer length of stay in PICU. At the time diagnosis of VAP was difficult as precise definitions were missing and there is still controversy over definition, treatment and prevention of VAP.(23, 24)

We have not performed further VAP studies, and since our publication only few studies by other groups have been published in the literature. Our paper was included in a meta-analysis by He et al. of VAP and cardiac surgery in 2014. Of the 11 included studies only two, including ours, concerned children which indicates the paucity of available evidence.(25) A second meta-analysis of VAP in the entire PICU population included 9 papers.(26) In total only 213 patients with VAP were included of 4564 patients (5%). All 9 studies were single-centre studies, 7 were prospective and 2 were retrospective studies. Only two studies specifically investigated patients after cardiac surgery which were the same two studies as included by the meta-analysis by He et al.(27, 28) In PICU patients, genetic syndrome, reintubation or self extubation, steroids, bloodstream infection, prior antibiotic therapy and bronchoscopy were identified as risk factors for VAP.(26) The authors advocate preventative measures and further research.

In 2015, Li et al showed that VAP is a risk factor for prolonged mechanical ventilation in children after correction of tetralogy of Fallot, corroborating our findings.(29) In a retrospective cohort study, Turcotte et al identified health care-associated infections (HAI) in 6% of their children following cardiac surgery.(30) Most infections were bacteraemia and central line associated bloodstream infections and only 10% of HAI were VAP. The causative Gram-negative organisms for VAP were not susceptible to perioperative prophylaxis with Cefazolin which is also used in our institution. Mechanical ventilation, postoperative transfusion of blood products, postoperative steroid use, and continuation of antibiotic prophylaxis longer than 48 hours after surgery were associated with HAIs, but no sub analysis of VAP was performed. A

retrospective from Japan identified 13.5% HAIs in their children after cardiac surgery and 11% of those were VAP.(31) The authors identified delayed sternal closure and the use of Dopamine as a risk factor for the development of all HAIs. Again, no sub analysis was made for patients with VAP. Dopamine is known to modulate immune responses and may result in the inhibition of cytokine and chemokine production, inhibition of neutrophil chemotaxis, and disturbance of T-cell proliferation.(32) In children with septic shock dopamine has been shown to have a higher mortality when compared to noradrenaline.(33)

Ventilator care bundles impact positively on the incidence of VAP in critically ill neonates and children. Common VAP prevention bundles include elements such as VAP education, hand hygiene, head of the bed elevation, protocolized mouth care, endotracheal tube suctioning practice, peptic ulcer prophylaxis, feeding protocols, frequency of ventilator circuit changes, and sedation interruptions.(23, 34, 35) But the frequency of postoperative infections remains high.(31) And, the variations in bundle elements and insufficient valid evidence necessitates further research in the area to validate the findings and to ensure standardization of clinical practice.(36)

The impact of Rhinovirus infections

In chapter 8 we present our RISK study protocol: Rhinovirus Infections after cardiac Surgery in Kids. We wanted to study the impact of rhinovirus (RhV) after cardiac surgery for two reasons: first, if children developed a postoperative infection in PICU, parents almost always told us that their child had had a recent cold and second because we then very often found RhV. We wanted to know if it would be worthwhile to test all preoperative children for RhV to diminish postoperative infections and concomitant prolonged PICU admissions. We discussed the results in chapter 9 and found that RhV was not associated with more postoperative infections and did not prolong PICU length of stay (LOS). We therefore concluded that our findings do not support the use of routine testing of respiratory viruses in asymptomatic children admitted for elective cardiac surgery, nor when combined with a parental questionnaire focused on signs and symptoms of a respiratory tract infection in the weeks leading up to the operation. Very unexpectedly we actually found that pre-operative, subclinical, RhV was associated with fewer postoperative infections and shorter PICU LOS. As we discussed in chapter 9, we could not explain this from our data and it deserves further investigation.

Fifteen percent of our patients did develop signs a postoperative infection and had a significantly prolonged duration of mechanical ventilation, PICU LOS and hospital LOS. Future research should be aimed at reducing postoperative infections because of the significant impact on postoperative course. As identifying subclinical viral carriers pre-operatively, does not seem to be able to positively influence

postoperative outcomes, we should focus on other risk-factors of postoperative infection such as care bundles (VAP, central line associated bloodstream infections, urinary tract infections) and further study the influence of cardiopulmonary bypass on immunoparalysis. Immunophenotyping patients after CPB has been shown to predict the risk of developing postoperative infections.(37) The next step is not only to predict an increased risk but to try and mitigate that risk.

The data we gathered during our RISK study also gives us a chance to try and answer several more questions such as the impact of preoperative steroids and/or endotracheal tube size on extubation failure, the influence of fluid balance on post-operative course, or the influence of blood products on outcome.

SECTION IV:

PHARMACOLOGY IN PAEDIATRIC CARDIAC INTENSIVE CARE

Low molecular weight heparin in critically ill children

In section four we have discussed the pharmacology of tinzaparin, a low molecular weight heparin (LMWH) used in our centre to prevent and/or treat thromboembolisms. In chapter 10 we studied the anti-Xa levels of critically ill children who received tinzaparin in therapeutic doses because reports in the literature emerged of other LMWHs that required higher dosing in critically ill children compared to non-critically ill children. Tinzaparin had not been investigated at that time, therefore we retrospectively analysed our anti-Xa levels and tinzaparin doses in 46 dosing episodes and found that it took an average of almost 9 doses to reach therapeutic levels. Also, the tinzaparin dose to reach therapeutic levels was higher than the advised dose. A major limitation of that study was that anti-Xa testing was not available daily at that time in our institution. One of our suggestions was to perform the first anti-Xa level after the first tinzaparin dose and then after every dose increase (or decrease) until target levels had been reached.

We subsequently changed our therapeutic drug monitoring (TDM) accordingly and evaluated that in chapter 11. We did not yet increase our starting doses. In that follow-up study we included 56 dosing episodes and showed that we could reach target levels quicker with more frequent monitoring indicating that TDM is effective. However, it still took a mean of almost 5 doses to reach therapeutic levels and again the final doses needed to reach therapeutic levels was significantly higher than the advised starting doses (10-30%). Especially children with oedema, which is frequent in these critically ill children, required significantly higher doses compared to children without oedema. Currently we are making a pharmacokinetic model of the combined data from both studies. The results will help us in dosing tinzaparin aimed at currently advised

anti-Xa levels. However, the main limitation of both studies from a clinical standpoint is that it is actually unknown which anti-Xa levels to aim for as current targets have been derived from adult studies. We have discussed this in chapter 12 and have no answer (yet) at the time of writing this thesis. As suggested in chapter 12, only a prospective study would be able to identify clinically important targets (prevention and reduction of thromboembolic events) and related anti-Xa levels.

Heparin monitoring in paediatric ECMO patients

In chapter 12 we also discuss monitoring of anticoagulation therapy, but now in paediatric ECMO patients who are anticoagulated with unfractionated heparin. In this systematic review and meta-analysis, we compared time-based versus anti-Xa-based anticoagulation monitoring strategies and concluded that an anti-Xa-anti-Xa-based anticoagulation strategy was associated with fewer bleeding events and lower mortality rate, without an increase in thrombotic events. These results suggest to use anti-Xa when titrating unfractionated heparin in paediatric ECMO patients and not time-based tests such as the APTT. It is, of course, not as clear cut as that. Anticoagulation therapy in children is notoriously difficult, especially in critically ill children as discussed before. We should not rely on a single test (at the moment) to evaluate the entire clotting state of the patient while influenced by, but not limited to, critical illness, cardiopulmonary bypass, ECMO-circuit, and anti-coagulation therapy and other medications. Monitoring of anticoagulation should be based on at least two tests as advocated by ELSO.(38) Our meta-analysis helps support the anti-Xa in becoming the norm for monitoring heparin therapy, also in ECMO patients. In 2017 we changed our clinical anticoagulation protocol and now guide unfractionated heparin therapy based on anti-Xa instead of APTT. We still perform frequent APTTs and PTs, as well as platelet count and fibrinogen, as they all provide essential information about the coagulation state of the patient. Also, occasional thromboelastography is performed. At the moment of writing this thesis we are retrospectively evaluating our own experience with anti-Xa verses APTT-based dosing in ECMO patients. The next step would be to compare different anti-Xa targets for clinically important outcomes such as bleeding, circuit thrombosis, patient thrombosis (e.g. cerebral infarct), mortality, ECMO duration, hospital LOS and long-term neurodevelopmental and neurocognitive outcomes. This will require a large multicentre study.

SECTION V:

DISCUSSION AND FUTURE PERSPECTIVES

As discussed before, some questions are answered but many more lie ahead. There are still innumerous clinical challenges beyond our comprehension and therefore we should strive to find new knowledge in an attempt to progress scientifically. One major

challenge is that it is often unknown what targets to aim for: which blood pressure or which anti-Xa level will lead to the best outcomes for this child in this specific situation?

One of the major problems in paediatric research in general is the lack of

sufficient patient numbers and the lack of diagnostic gold standards (VAP, LCOS, anticoagulation). To find children with the exact same diagnosis in who we want to investigate the effect of one single intervention (for instance milrinone versus dobutamine or APTT versus anti-Xa) while all other aspects (for instance mechanical ventilation, fluid treatment, and sedation management, etc) remain comparable is almost impossible. This is especially true in children with congenital heart disease as within diagnostic groups (for instance pulmonary atresia, tetralogy of Fallot, tricuspid atresia, etc, etc.) many concomitant defects can exist that can have an effect on the surgical and postoperative approach. For instance, milrinone might be the perfect choice for a child with a ‘simple’ peri membranous VSD, but may in theory even be harmful to a child with a VSD, pulmonary stenosis and right ventricular hypertrophy. In the latter milrinone may reduce preload when an increased preload is warranted leading to increased fluid challenges and fluid overload, which may then in some children contribute to pleural effusions, prolonged mechanical ventilation, VAP and a protracted postoperative course in the cardiac intensive care unit.

Currently larger international studies in paediatrics and paediatric intensive care are being performed and we have to rely on these multicentre studies initiated by international research networks to find some of the answers.(1, 2) Smaller studies, albeit single-centre and/or retrospective studies such as presented in this thesis can help identify some of the problems we face and point towards a possible answer or direction for further research. And until we find true evidence, the cardiac intensivist still needs to tailor the postoperative approach to the specific patient with his or her specific cardiac lesion and operation by using the available pathophysiological and pharmacological knowledge, their clinical experience, sound judgment and common sense.

Being a good doctor, to me, means that you are constantly wondering if you really understand what is happening and if you are doing the best you can in treating the patient, the child, in front of you. If you do not have the answer, go and look for it. Be amazed. Wonder. Search. And most importantly, be critical of what you think you know.

12. Loomba RS, Flores S. Use of vasoactive agents in postoperative pediatric cardiac patients: Insights from a national database. Congenit Heart Dis. 2019;14(6):1176-84.

13. Roeleveld PP. Vasopressin Is No Inotrope. World J Pediatr Congenit Heart Surg. 2018;9(4):480.

14. Barnwal NK, Umbarkar SR, Sarkar MS, Dias RJ. Randomized comparative study of intravenous infusion of three different fixed doses of milrinone in pediatric patients with pulmonary hypertension undergoing open heart surgery. Ann Card Anaesth. 2017;20(3):318-22.

15. Charlton GA, Ladd DR, Friesen RM, Friesen RH. Labetalol Infusion Attenuates Paradoxical Hypertension and Decreases Plasma Renin Activity After Repair of Coarctation of the Aorta in Children. J Cardiothorac Vasc Anesth. 2020.

16. Roberts JS, Yanay O, Barry D. Age-Based Percentiles of Measured Mean Arterial Pressure in Pediatric Patients in a Hospital Setting. Pediatr Crit Care Med. 2020;21(9):e759-e68.

17. Raffaeli G, Pokorna P, Allegaert K, Mosca F, Cavallaro G, Wildschut ED, et al. Drug Disposition and Pharmacotherapy in Neonatal ECMO: From Fragmented Data to Integrated Knowledge. Front Pediatr. 2019;7:360.

18. Mitchell EA, Gomez D, Joy BF, Fernandez RP, Cheatham JP, Galantowicz M, et al. ECMO: Incidence and Outcomes of Patients Undergoing the Hybrid Procedure. Congenit Heart Dis. 2016;11(2):169-74.

19. Mascio CE, Austin EH, 3rd, Jacobs JP, Jacobs ML, Wallace AS, He X, et al. Perioperative mechanical circulatory support in children: an analysis of the Society of Thoracic Surgeons Congenital Heart Surgery Database. J Thorac Cardiovasc Surg. 2014;147(2):658-64: discussion 64-5.

20. Kogon BE, Kanter K, Alsoufi B, Maher K, Oster ME. Outcomes and hospital costs associated with the Norwood operation: beyond morbidity and mortality. Cardiol Young. 2015;25(5):853-9.

21. Ohye RG, Sleeper LA, Mahony L, Newburger JW, Pearson GD, Lu M, et al.

Comparison of shunt types in the Norwood procedure for single-ventricle lesions. N Engl J Med. 2010;362(21):1980-92.

22. Sherwin ED, Gauvreau K, Scheurer MA, Rycus PT, Salvin JW, Almodovar MC, et al. Extracorporeal membrane oxygenation after stage 1 palliation for hypoplastic left heart syndrome. J Thorac Cardiovasc Surg. 2012;144(6):1337-43.

23. Chang I, Schibler A. Ventilator Associated Pneumonia in Children. Paediatr Respir Rev. 2016;20:10-6.

24. Galal YS, Youssef MR, Ibrahiem SK. Ventilator-Associated Pneumonia: Incidence, Risk Factors and Outcome in Paediatric Intensive Care Units at Cairo University Hospital. J Clin Diagn Res. 2016;10(6):SC06-11.

REFERENCES

1. Zimmerman JJ, Anand KJ, Meert KL, Willson DF, Newth CJ, Harrison R, et al. Research as a Standard of Care in the PICU. Pediatr Crit Care Med. 2016;17(1):e13-21.

2. Peters MJ, Argent A, Festa M, Leteurtre S, Piva J, Thompson A, et al. The intensive care medicine clinical research agenda in paediatrics. Intensive Care Med. 2017;43(9):1210-24.

3. DeWitt ES, Black KJ, Thiagarajan RR, DiNardo JA, Colan SD, McGowan FX, et al. Effects of commonly used inotropes on myocardial function and oxygen consumption under constant ventricular loading conditions. J Appl Physiol (1985). 2016;121(1):7-14.

4. Hoffman TM, Wernovsky G, Atz AM, Bailey JM, Akbary A, Kocsis JF, et al. Prophylactic intravenous use of milrinone after cardiac operation in pediatrics (PRIMACORP) study. Prophylactic Intravenous Use of Milrinone After Cardiac Operation in Pediatrics. Am Heart J. 2002;143(1):15-21.

5. Marwali EM, Boom CE, Budiwardhana N, Fakhri D, Roebiono PS, Santoso A, et al. Oral Triiodothyronine for Infants and Children Undergoing Cardiopulmonary Bypass. Ann Thorac Surg. 2017;104(2):688-95.

6. Marwali EM, Caesa P, Darmaputri S, Sani AA, Roebiono PS, Fakhri D, et al. Oral Triiodothyronine Supplementation Decreases Low Cardiac Output Syndrome After Pediatric Cardiac Surgery. Pediatr Cardiol. 2019;40(6):1238-46.

7. Talwar S, Bhoje A, Khadagawat R, Chaturvedi P, Sreenivas V, Makhija N, et al. Oral thyroxin supplementation in infants undergoing cardiac surgery: A double-blind placebo-controlled randomized clinical trial. J Thorac Cardiovasc Surg. 2018;156(3):1209-17 e3.

8. Wang A, Cui C, Fan Y, Zi J, Zhang J, Wang G, et al. Prophylactic use of

levosimendan in pediatric patients undergoing cardiac surgery: a prospective randomized controlled trial. Crit Care. 2019;23(1):428.

9. Thorlacius EM, Wahlander H, Ojala T, Ylanen K, Keski-Nisula J, Synnergren M, et al. Levosimendan Versus Milrinone for Inotropic Support in Pediatric Cardiac Surgery: Results From a Randomized Trial. J Cardiothorac Vasc Anesth. 2020;34(8):2072-80.

10. Thorlacius EM, Suominen PK, Wahlander H, Keski-Nisula J, Vistnes M, Ricksten SE, et al. The Effect of Levosimendan Versus Milrinone on the Occurrence Rate of Acute Kidney Injury Following Congenital Heart Surgery in Infants: A Randomized Clinical Trial. Pediatr Crit Care Med. 2019;20(10):947-56.

11. Cavigelli-Brunner A, Hug MI, Dave H, Baenziger O, Buerki C, Bettex D, et al. Prevention of Low Cardiac Output Syndrome After Pediatric Cardiac Surgery: A Double-Blind Randomized Clinical Pilot Study Comparing Dobutamine and Milrinone. Pediatr Crit Care Med. 2018;19(7):619-25.

25. He S, Chen B, Li W, Yan J, Chen L, Wang X, et al. Ventilator-associated pneumonia after cardiac surgery: a meta-analysis and systematic review. J Thorac

Cardiovasc Surg. 2014;148(6):3148-55 e1-5.

26. Liu B, Li SQ, Zhang SM, Xu P, Zhang X, Zhang YH, et al. Risk factors of ventilator-associated pneumonia in pediatric intensive care unit: a systematic review and meta-analysis. J Thorac Dis. 2013;5(4):525-31.

27. Roeleveld PP, Guijt D, Kuijper EJ, Hazekamp MG, de Wilde RB, de Jonge E. Ventilator-associated pneumonia in children after cardiac surgery in The Netherlands. Intensive Care Med. 2011;37(10):1656-63.

28. Tang CW, Liu PY, Huang YF, Pan JY, Lee SS, Hsieh KS, et al. Ventilator-associated pneumonia after pediatric cardiac surgery in southern Taiwan. J Microbiol Immunol Infect. 2009;42(5):413-9.

29. Li S, Zhang Y, Li S, Wang X, Zhang R, Lu Z, et al. Risk Factors Associated with Prolonged Mechanical Ventilation after Corrective Surgery for Tetralogy of Fallot. Congenit Heart Dis. 2015;10(3):254-62.

30. Turcotte RF, Brozovich A, Corda R, Demmer RT, Biagas KV, Mangino D, et al. Health care-associated infections in children after cardiac surgery. Pediatr Cardiol. 2014;35(8):1448-55.

31. Hatachi T, Tachibana K, Inata Y, Tominaga Y, Hirano A, Kyogoku M, et al. Risk Factors for Healthcare-Associated Infections After Pediatric Cardiac Surgery. Pediatr Crit Care Med. 2018;19(3):237-44.

32. Beck G, Brinkkoetter P, Hanusch C, Schulte J, van Ackern K, van der Woude FJ, et al. Clinical review: immunomodulatory effects of dopamine in general inflammation. Crit Care. 2004;8(6):485-91.

33. Ventura AM, Shieh HH, Bousso A, Goes PF, de Cassia FOFI, de Souza DC, et al. Double-Blind Prospective Randomized Controlled Trial of Dopamine Versus Epinephrine as First-Line Vasoactive Drugs in Pediatric Septic Shock. Crit Care Med. 2015;43(11):2292-302.

34. Eom JS, Lee MS, Chun HK, Choi HJ, Jung SY, Kim YS, et al. The impact of a ventilator bundle on preventing ventilator-associated pneumonia: a multicenter study. Am J Infect Control. 2014;42(1):34-7.

35. Taylor RS, Shekerdemian LS. Avoidance of Hospital-Acquired Infections in Pediatric Cardiac Surgical Patients. Pediatr Crit Care Med. 2016;17(8 Suppl 1):S279-86.

36. Niedzwiecka T, Patton D, Walsh S, Moore Z, O’Connor T, Nugent L. What are the effects of care bundles on the incidence of ventilator-associated pneumonia in paediatric and neonatal intensive care units? A systematic review. J Spec Pediatr Nurs. 2019;24(4):e12264.

37. Cornell TT, Sun L, Hall MW, Gurney JG, Ashbrook MJ, Ohye RG, et al. Clinical implications and molecular mechanisms of immunoparalysis after cardiopulmonary bypass. J Thorac Cardiovasc Surg. 2012;143(5):1160-6 e1.

38. Laurance Lequier GA, Omar Al-Ibrahim, Melania Bembea, Dan Brodie, Tom Brogan, Shannon Buckvold, Louis Chicoine, Steve Conrad, David Cooper, Heidi Dalton, Jason Frischer, Bill Harris, Robert Mazor, Matthew Paden, Natalie Rintoul, Lindsay Ryerson, Phil Spinella, Jun Teruya, Annie Winkler, Trisha Wong, M. Patricia Massicotte. ELSO anticoagulation guideline Ann Arbor, MI, USA: The extracorporeal life support organization; 2014 [Available from: https://www.elso. org/Portals/0/Files/elsoanticoagulationguideline8-2014-table-contents.pdf.