Abstract
Objective
To describe factors associated with platelet transfusion during pediatric extracorporeal membrane oxygenation (ECMO), and the relationships between platelet transfusion, complications and mortality.
Design
Secondary analysis of data collected prospectively by the Collaborative Pediatric Critical Care Research Network (CPCCRN) between December 2012 and September 2014.
Setting
Eight CPCCRN-affiliated hospitals.
Patients
Age <19 years and treated with ECMO.
Interventions
None.
Measurements and Main Results
Of 511 children, 496 (97.1%) received at least one platelet transfusion during ECMO. Neonatal age, venoarterial ECMO, and various acute and chronic diagnoses were associated with increased average daily platelet transfusion volume (mL/kg body weight). On multivariable analysis, average daily platelet transfusion volume was independently associated with mortality (per 1 ml/kg, odds ratio 1.05, CI 1.03, 1.08, p<0.001) whereas average daily platelet count was not (per 1×109/L up to 115×109/L, odds ratio 1.00, CI 0.98, 1.01, p=0.49). Variables independently associated with increased daily bleeding risk included increased platelet transfusion volume on the previous ECMO day, a primary cardiac indication for ECMO, adolescent age, and an acute diagnosis of congenital cardiovascular disease. Variables independently associated with increased daily thrombotic risk included increased platelet transfusion volume on the previous ECMO day, and VA ECMO. Variables independently associated with decreased daily thrombotic risk included full-term neonatal age, and an acute diagnosis of airway abnormality.
Conclusions
Platelet transfusion was common in this multisite pediatric ECMO cohort. Platelet transfusion volume was associated with increased risk of mortality, bleeding and thrombosis.
Keywords: Extracorporeal Membrane Oxygenation, platelets, child, infant, neonate
INTRODUCTION
Platelet transfusions are common among critically ill children receiving extracorporeal membrane oxygenation (ECMO). Optimal transfusion thresholds are not well established, and the clinical factors, complications, and mortality associated with platelet transfusion during ECMO are uncertain. A recent multicenter study in a general population of critically ill children found that most platelet transfusions were prescribed prophylactically to those who were not bleeding (1). The same study also found an association between platelet dose and mortality. A single center study in a similar cohort of children reported the most common indication for platelet transfusion was prophylaxis for thrombocytopenia, and that mortality was high among those transfused and associated with disease severity (2). Many of these platelet transfusions, at least 30%, were given to children on ECMO.
Quantitative and qualitative platelet dysfunction has been described during ECMO and is associated with bleeding and mortality (3,4). Platelet transfusion also has risks including febrile reactions, transfusion-related acute lung injury, infection, thromboembolism, and multiple organ dysfunctions (5–8). The association between platelet transfusion and poor outcome in critically ill children may be related to the multiple pathologic processes that result in thrombocytopenia and lead to platelet replenishment (1,9,10). During ECMO, platelet consumption triggered by the ECMO circuit may lead to platelet transfusion resulting in multiple organ dysfunctions from microthrombi formation and immune dysregulation (9,10).
Guidelines to direct platelet transfusion are mostly not evidence-based but rather rely on expert opinion (11,12). Clinical decisions regarding platelet transfusion during ECMO are likely multifactorial with various child and ECMO characteristics playing a role. Prospective, multicenter data are needed to gain more generalizable knowledge about platelet transfusion during ECMO. Our objectives are to describe factors associated with platelet transfusion during pediatric ECMO and to determine the relationship between platelet transfusion, complications, and mortality.
METHODS
Design and Setting
This study was a secondary analysis of the Bleeding and Thrombosis during ECMO (BATE) study (13) which described the incidence of bleeding and thrombosis during neonatal and pediatric ECMO. The BATE study collected prospective observational data at eight CPCCRN-affiliated hospitals between December 2012 and September 2014. The Institutional Review Boards for each hospital and the Data Coordinating Center at the University of Utah approved the study with waiver of parental permission. All treatments provided were dependent on the clinical team and not dictated by study protocol.
Study Subjects
All children <19 years old treated with ECMO in a neonatal, pediatric or cardiac intensive care unit (ICU) were included in the BATE study (13). Only the initial ECMO course was included for children who required multiple courses of ECMO. Of 514 children enrolled in BATE, three had missing data for body weight and were excluded from this analysis due to inability to determine transfusion volume in mL/kg. Thus, 511 children are included in this report.
Data Collection
Pre-ECMO data included age, history of prematurity, acute and chronic diagnoses, Pediatric Risk of Mortality III score (14) and clinical site. Age was categorized as neonate ≤30 days, infant >30 days to ≤12 months, child >1 year to ≤12 years, and adolescent >12 years to <19 years. Prematurity was <37 weeks gestational age at birth and collected for neonates only.
ECMO data included primary indication for ECMO, placement on ECMO directly from cardiopulmonary bypass, mode of ECMO, location of ECMO within the hospital, average daily heparin dose, and average daily platelet and red cell transfusion volumes per kg body weight. Primary indication for ECMO was categorized as respiratory, cardiac or extracorporeal cardiopulmonary resuscitation (ECPR). Mode of ECMO was venoarterial (VA) or venovenous (VV). VV ECMO that was converted to VA was categorized as VA ECMO. Location of ECMO was categorized as neonatal, pediatric or cardiac ICU.
Laboratory data included arterial blood gases, blood lactate, and complete blood count with platelet count. Baseline laboratory values were obtained closest and prior to ECMO initiation; daily laboratory values were obtained closest to 7 AM on each ECMO day. Outcomes included complications during ECMO and hospital mortality. Complications included bleeding events, thrombotic events, neurologic events, hepatic dysfunction, and renal failure. Bleeding events were defined as blood loss requiring a red cell transfusion or intracranial hemorrhage. Thrombotic events included intracranial infarction, limb ischemia, pulmonary embolus, intracardiac thrombus, aorto-pulmonary shunt thrombus, other sites of thrombosis, and circuit thrombosis requiring replacement of a circuit component. Neurologic events included seizures, intracranial hemorrhage or infarction, and brain death. Hepatic dysfunction was defined as INR >2. Renal failure was defined as creatinine >2 mg/dL (>176.8 μmol/L) or use of in-line hemofiltration or other form of continuous renal replacement therapy (CRRT).
Statistical Analysis
Clinical data were summarized by average daily platelet transfusion volume using counts and percentages for categorical variables, and medians and first and third quartiles for continuous variables. Associations between clinical variables and average daily platelet transfusion volume were evaluated with the Cochran-Armitage trend test for binary variables, the Kruskal-Wallis test for other nominal variables, and the Jonckheere-Terpstra test for ordinal variables. These tests were chosen to account for the ordinal nature of average daily platelet transfusion volume.
The unadjusted association of average daily platelet transfusion volume and average daily platelet count with mortality was assessed with univariable logistic regression and natural cubic splines. Average daily platelet transfusion volume was determined to have a roughly linear association with the log odds of mortality and was therefore considered appropriate for inclusion in subsequent models as a linear (continuous) predictor (Figure 1). Average daily platelet count, however, was determined to have a roughly linear association with the log odds of mortality up to a platelet count of approximately 115 ×109/L, after which there was little evidence of any change in the risk of mortality with increasing platelet count (Figure 2). Subsequent modeling of mortality treated average daily platelet count as a linear (continuous) predictor only up to 115 ×109/L, after which further increases did not affect the modeled risk of mortality. This was accomplished by truncating the measurement of average daily platelet count to 115 ×109/L.
Figure 1.
Unadjusted effect of average daily platelet transfusion volume on mortality risk.
Figure 2.
Unadjusted effect of average daily platelet count on mortality risk.
The association between average daily platelet count and in-hospital mortality was assessed with univariable and multivariable logistic regression models. Variables at least marginally associated with mortality in univariable modeling (p <0.10) that were known for at least 90% of the cohort were considered candidate predictors for a multivariable model. The final multivariable model was created using a bi-directional stepwise selection process with criteria of p <0.10 and p <0.05 to enter and stay in the model, respectively.
The association between platelet count on the previous day with daily risk of bleeding was similarly assessed with day-level univariable and multivariable logistic regression models analogous to the models of mortality. An autoregressive covariance structure of order 1 and robust error estimates were used to account for correlation between different study days on the same subject. In particular, this accounts for a higher correlation between study days that are close together but relatively lower correlation between study days that are far apart. The association between platelet transfusion volume on the previous day with daily risk of thrombosis was similarly assessed with day-level univariable and multivariable logistic regression models analogous to the models of daily bleeding.
The decrease in platelet count after ECMO initiation was characterized by whether a platelet transfusion was administered on the day of ECMO initiation and by age. Only transfusions provided after ECMO initiation were recorded in the study, and platelet counts were only collected as obtained by standard of care. The decrease in platelets is summarized based on the last pre-ECMO and first on-ECMO platelet counts, and only subjects with measurements within a 12-hour period on the first day of ECMO are included in this summary. Platelet goals are summarized by institution, unit, and mode of ECMO using means and standard deviations. Analyses were performed using SAS 9.4 (SAS Institute; Cary, NC).
RESULTS
Of 511 children, 496 (97.1%) received at least one platelet transfusion during ECMO. Neonatal age group, VA ECMO, location of ECMO care in the neonatal ICU, acute diagnosis of congenital diaphragmatic hernia or persistent pulmonary hypertension of the newborn, and chronic diagnosis of congenital anomaly or chromosomal defect were associated with increased average daily platelet transfusion volume (Table 1). Acute diagnosis of acquired cardiovascular disease, pneumonia or bronchiolitis, respiratory distress/failure, and chronic diagnosis of neurologic condition were associated with decreased average daily platelet transfusion volume.
Table 1.
Clinical Characteristics by Average Daily Platelet Transfusion Volume
| Average Daily Platelet Transfusion Volume (mL/kg Body Weight)a |
||||
|---|---|---|---|---|
| Characteristic | < 9 (N = 187) | 9 - < 17 (N = 158) | ≥ 17 (N = 166) | P-value |
| Age | <.001b | |||
| Pre-Term Neonate | 4 (2.1%) | 18 (11.4%) | 28 (16.9%) | |
| Full-Term Neonate | 50 (26.7%) | 81 (51.3%) | 86 (51.8%) | |
| Infant | 55 (29.4%) | 29 (18.4%) | 38 (22.9%) | |
| Child | 44 (23.5%) | 23 (14.6%) | 12 (7.2%) | |
| Adolescent | 34 (18.2%) | 7 (4.4%) | 2 (1.2%) | |
| Primary ECMO Indication | 0.811c | |||
| Respiratory | 84 (44.9%) | 75 (47.5%) | 77 (46.4%) | |
| Cardiac | 76 (40.6%) | 61 (38.6%) | 69 (41.6%) | |
| ECPR | 27 (14.4%) | 22 (13.9%) | 20 (12.0%) | |
| Mode of ECMO | 0.048d | |||
| VA | 148 (79.1%) | 137 (86.7%) | 144 (86.7%) | |
| VV | 39 (20.9%) | 21 (13.3%) | 22 (13.3%) | |
| Location of ECMO Care | <.001c | |||
| PICU | 54 (28.9%) | 27 (17.1%) | 21 (12.7%) | |
| NICU | 23 (12.3%) | 65 (41.1%) | 68 (41.0%) | |
| CICU | 110 (58.8%) | 66 (41.8%) | 77 (46.4%) | |
| Baseline PRISM III | 8.0 [3.0, 12.0] | 9.0 [5.0, 13.0] | 9.0 [4.0, 14.0] | 0.117b |
| Acute Diagnoses | ||||
| Airway Abnormality | 4 (2.1%) | 2 (1.3%) | 6 (3.6%) | 0.380d |
| Immune Dysfunction | 17 (9.1%) | 11 (7.0%) | 8 (4.8%) | 0.117d |
| Cardiac Arrest | 21 (11.2%) | 17 (10.8%) | 10 (6.0%) | 0.099d |
| Cardiovascular Disease (Acquired) | 35 (18.7%) | 15 (9.5%) | 14 (8.4%) | 0.003d |
| Cardiovascular Disease (Arrhythmia) | 5 (2.7%) | 3 (1.9%) | 10 (6.0%) | 0.097d |
| Cardiovascular Disease (Congenital) | 70 (37.4%) | 54 (34.2%) | 69 (41.6%) | 0.446d |
| Hypoxic/Anoxic Injury | 3 (1.6%) | 7 (4.4%) | 7 (4.2%) | 0.163d |
| Gastrointestinal Disorder | 8 (4.3%) | 4 (2.5%) | 11 (6.6%) | 0.311d |
| Pertussis or Sepsis | 29 (15.5%) | 33 (20.9%) | 29 (17.5%) | 0.602d |
| Pneumonia or Bronchiolitis | 14 (7.5%) | 6 (3.8%) | 4 (2.4%) | 0.023d |
| Shock (non-septic) | 5 (2.7%) | 4 (2.5%) | 5 (3.0%) | 0.851d |
| Respiratory Distress/Failure | 79 (42.2%) | 44 (27.8%) | 47 (28.3%) | 0.005d |
| Neurologic Condition | 9 (4.8%) | 4 (2.5%) | 3 (1.8%) | 0.102d |
| Meconium Aspiration Syndrome | 11 (5.9%) | 17 (10.8%) | 19 (11.4%) | 0.067d |
| Congenital Diaphragmatic Hernia | 3 (1.6%) | 24 (15.2%) | 29 (17.5%) | <.001d |
| Persistent Pulmonary Hypertension of the Newborn | 16 (8.6%) | 42 (26.6%) | 32 (19.3%) | 0.006d |
| Chronic Diagnoses | ||||
| Chronic Lung Disease | 8 (4.3%) | 4 (2.5%) | 5 (3.0%) | 0.494d |
| Immune Dysfunction | 8 (4.3%) | 5 (3.2%) | 3 (1.8%) | 0.184d |
| Congenital Anomaly or Chromosomal Defect | 29 (15.5%) | 38 (24.1%) | 51 (30.7%) | <.001d |
| Neurologic Condition | 18 (9.6%) | 4 (2.5%) | 4 (2.4%) | 0.002d |
| Cardiovascular Disease (Congenital) | 31 (16.6%) | 36 (22.8%) | 29 (17.5%) | 0.791d |
ECMO = extracorporeal membrane oxygenation; VA = veno-arterial; VV = veno-venous; ECPR = extracorporeal cardiopulmonary resuscitation; PICU = pediatric intensive care unit; NICU = neonatal intensive care unit; CICU = cardiac intensive care unit; PRISM = Pediatric Risk of Mortality score.
Categories of average daily platelet transfusion volume are based on tertiles of platelet transfusion volume within this study cohort.
Jonckheere-Terpstra test.
Kruskal-Wallis test.
Cochran-Armitage trend test.
Univariable and multivariable logistic models of hospital mortality are shown in Supplemental Digital Content 1 and Table 2, respectively. On multivariable analysis, variables independently associated with increased mortality included increased average daily platelet transfusion volume, higher minimum lactate within 48 hours of ECMO initiation, ECPR, renal failure, new neurologic event on ECMO, pre-term neonatal age, and a diagnosis of congenital diaphragmatic hernia. Variables independently associated with decreased mortality included diagnosis of meconium aspiration syndrome and average daily heparin dose.
Table 2.
Multivariable Logistic Model of Mortality
| In-Hospital Mortality |
||
|---|---|---|
| Characteristic | Odds ratio (95% CI) | P-value |
| Average Daily Platelet Count (per 109/L up to 115×109/L) | 1.00 (0.98, 1.01) | 0.487 |
| Average Daily Platelet Transfusion Volume (1 mL/kg) | 1.05 (1.03, 1.08) | <.001 |
| Lowest Lactate in 48 hours Post-ECMO Initiation (mmol/L) | 1.38 (1.19, 1.66) | <.001 |
| Average Daily Heparin dose (units/kg/min) | 0.26 (0.07, 0.91) | 0.032 |
| Meconium Aspiration Syndrome | 0.15 (0.02, 0.63) | 0.007 |
| Renal Failure | 1.92 (1.15, 3.21) | 0.012 |
| New Neurologic Event | 4.05 (2.48, 6.73) | <.001 |
| Primary ECMO Indication | 0.027 | |
| Respiratory | Reference | |
| Cardiac | 2.08 (1.17, 3.76) | |
| ECPR | 2.34 (1.06, 5.24) | |
| Age | 0.002 | |
| Pre-Term Neonate | 3.97 (1.72, 9.53) | |
| Full-Term Neonate | Reference | |
| Infant | 2.50 (1.36, 4.68) | |
| Child | 2.22 (1.08, 4.60) | |
| Adolescent | 2.60 (1.06, 6.35) | |
| Congenital Diaphragmatic Hernia | 3.63 (1.56, 8.63) | 0.003 |
ECMO = extracorporeal membrane oxygenation; ECPR = extracorporeal cardiopulmonary resuscitation
The reported effect of average daily platelet count on mortality is for each additional 1×109 cells/L up to a count of 115×109/L, beyond which the odds of mortality is assumed constant.
Univariable and multivariable logistic models of daily bleeding are shown in Supplemental Digital Content 2 and Table 3, respectively. On multivariable analysis, variables independently associated with increased daily bleeding risk included increased platelet transfusion volume on the previous ECMO day, a primary cardiac indication for ECMO, adolescent age, and an acute diagnosis of congenital cardiovascular disease. Platelet count on the previous day was not associated with daily bleeding risk.
Table 3.
Multivariable Logistic Model of Daily Bleeding
| Bleeding Event |
||
|---|---|---|
| Characteristic | Odds ratio (95% CI) | P-value |
| Platelet Transfusion Volume on the Previous Day (1 mL/kg) | 1.01 (1.00, 1.01) | <.001 |
| Primary ECMO Indication | 0.033 | |
| Respiratory | Reference | |
| Cardiac | 1.72 (1.15, 2.58) | |
| ECPR | 1.42 (0.84, 2.38) | |
| Age | 0.003 | |
| Pre-Term Neonate | 1.39 (0.91, 2.13) | |
| Full-Term Neonate | Reference | |
| Infant | 1.18 (0.81, 1.73) | |
| Child | 1.74 (1.11, 2.71) | |
| Adolescent | 3.58 (2.17, 5.92) | |
| Cardiovascular Disease (Congenital) | 1.54 (1.04, 2.28) | 0.037 |
ECMO = extracorporeal membrane oxygenation; ECPR = extracorporeal cardiopulmonary resuscitation.
Univariable and multivariable logistic models of daily thrombosis are shown in Supplemental Digital Content 3 and Table 4, respectively. On multivariable analysis, variables independently associated with increased daily thrombotic risk included platelet transfusion volume on the previous ECMO day, and VA ECMO. Variables independently associated with decreased daily thrombotic risk included full-term neonatal age, and having an acute diagnosis of airway abnormality.
Table 4.
Multivariable Logistic Model of Daily Thrombosis
| Thrombotic Event |
||
|---|---|---|
| Characteristic | Odds ratio (95% CI) | P-value |
| Platelet Transfusion Volume on the Previous Day (1 mL/kg) | 1.01 (1.00, 1.02) | 0.010 |
| Mode of ECMO | <.001 | |
| VA | Reference | |
| VV | 0.40 (0.26, 0.64) | |
| Age | 0.032 | |
| Pre-Term Neonate | 1.08 (0.64, 1.81) | |
| Full-Term Neonate | Reference | |
| Infant | 2.09 (1.39, 3.13) | |
| Child | 1.62 (0.94, 2.78) | |
| Adolescent | 1.37 (0.78, 2.40) | |
| Airway Abnormality | 0.24 (0.09, 0.67) | 0.007 |
ECMO = extracorporeal membrane oxygenation; VA = veno-arterial; VV = veno-venous
Of the 511 children in the study, 287 (56%) had a platelet count recorded on the first day within 12-hours after ECMO initiation. Children who received platelet transfusion on the day of ECMO initiation had lower pre-ECMO platelet counts and lower first post-ECMO platelet counts than those who did not receive platelet transfusion on the day of ECMO initiation (Supplemental Digital Content 4). Neonates had a larger decline in platelet count on ECMO initiation than non-neonates (Supplemental Digital Content 5). Platelet goals ranged across centers from 67±23 ×109/L to 113±22 ×109/L for VA ECMO, and 50 ×109/L to 100±0 ×109/L for VV ECMO (Supplemental Digital Content 6).
DISCUSSION
Nearly all children in our multicenter ECMO cohort received platelet transfusions. Factors related to increased average daily platelet transfusion volume included age, mode and location of ECMO, and various acute and chronic diagnoses. Average daily platelet transfusion volume was independently associated with mortality. Platelet transfusion volume on the previous ECMO day was independently associated with daily risk of bleeding and thrombosis, however platelet count on the previous day was not.
In our ECMO cohort, neonates received a higher average daily platelet transfusion volume than older children. Thrombocytopenia is well described in the general neonatal ICU population (15–19). Additionally, neonatal platelets are hyporeactive in vitro in response to multiple agonists. Both low platelet count and hyporeactivity are more pronounced in pre-term than full-term neonates. Due to these concerns, early research suggested targeting a higher platelet count to prevent bleeding in neonates receiving ECMO (20). However, more recent work has shown that healthy full-term neonates display no clinical evidence of abnormal hemostasis and suggests that neonatal platelet hyporeactivity is part of balanced neonatal hemostatic system. Target platelet count was about 100 ×109/L in the neonatal ICUs included in our ECMO cohort, slightly higher than targets set in some of our pediatric and cardiac ICUs (Supplemental Digital Content 6). Previous studies have shown that most platelet transfusions administered to critically ill children are given prophylactically to meet a predefined threshold (1,2); however, the reasons clinicians transfused platelets were not captured in the BATE study. Location of ECMO in a neonatal ICU, and diagnoses such as congenital diaphragmatic hernia and persistent pulmonary hypertension of the newborn were also associated with increased average daily platelet transfusion volume likely due to their common occurrence in neonates. VA ECMO was associated with increased average daily platelet transfusion volume. In a prior study, neonates on VA ECMO received more platelet transfusions that those on VV ECMO (21). Mechanical flow differences between VA and VV ECMO, differences in disease severity and diagnosis, and other unmeasured factors were suggested explanations for the difference in platelet transfusions.
Interestingly, while average daily platelet transfusion volume was an independent predictor of mortality, average daily platelet count was not. Our findings suggest that platelet transfusion practice should not be based on a specific target level alone, but should also consider the potential harms associated with platelet transfusion including immune, inflammatory and hemodynamic consequences. A recent randomized trial in premature neonates with severe thrombocytopenia compared target platelet transfusion thresholds of 50 ×109/L versus 25 ×109/L and found a greater proportion of neonates exposed to platelet transfusion and higher rate of death or major bleeding in the 50 ×109/L target group (22). In our study, both average daily platelet transfusion volume and average daily red cell transfusion volume were associated with mortality on univariable analysis. However, on multivariable analysis, average daily platelet transfusion volume remained a significant predictor of mortality whereas average daily red cell transfusion volume did not.
Our findings suggest that platelet transfusion volume is associated with complications during ECMO. As noted in the BATE study, while the cumulative rates of bleeding and thrombosis were related to the duration of ECMO, the daily occurrence rates remained steady for the duration of ECMO (13). Consequently, it is unlikely that the duration of ECMO was a major contributor to daily complication rates. Platelet transfusion volume on the previous ECMO day was independently associated with daily bleeding events suggesting that children receiving platelets had ongoing bleeding and/or platelet consumption. Children with a primary cardiac indication for ECMO, an acute diagnosis of congenital cardiovascular disease, or adolescent age had increased bleeding risk. Postoperative cardiac surgical patients have increased bleeding risk due to surgical incisions and platelet dysfunction from cardiopulmonary bypass (23,24) which is amplified with systemic anticoagulation on ECMO. Bleeding during ECMO is multifactorial and related to increased fibrinolysis, loss of large von Willebrand factor multimers, and decreased platelet count and function (25–27). Although platelets have a key role in normal hemostasis, transfusing platelets to increase the platelet count during ECMO may not lead to the desired clinical outcome.
In our study, platelet transfusion volume on the previous day was independently associated with daily thrombotic events. Platelet activation occurs when blood is exposed to the ECMO circuit leading to increased clot burden (3,28). Recent studies have shown that activated platelets promote thrombosis by release of platelet-derived microparticles (PMPs) and that PMPs are generated by ECMO systems (29,30). Due to the observational nature of our study, we cannot differentiate whether ongoing thrombus formation and its associated platelet consumption led to decreased platelet count and subsequent transfusion, or whether platelet transfusion led to thrombus formation. Both processes are feasible and may occur simultaneously. We know from the BATE study that 37.5% of children had a thrombotic event. Of these, 12.8% were patient-related and 31.1% were circuit-related. Unfortunately, the relationship between platelet transfusion volume and site/type of thrombosis was not evaluated (13).
The mechanisms driving platelet consumption and activation are important to define in order to target therapies on ECMO. Children with ventricular assist devices routinely receive antiplatelet agents but children on ECMO do not; such agents may increase bleeding risk during ECMO without substantially reducing thrombosis (31). Others report using modified surfaces that create local release of nitric oxide to inhibit platelets that directly interface with the biomaterial surface while the remaining platelets maintain normal functionality (32–34). Further research regarding mechanisms of platelet dysfunction including activation, consumption and oxidative stress may help guide therapeutic interventions to reduce rates of thrombosis during ECMO (35).
Our findings demonstrate most children had a substantial reduction in platelet count after ECMO initiation. Children transfused platelets on the day of ECMO initiation had lower platelet counts pre-ECMO and after ECMO initiation. While neonates had higher platelet counts pre-ECMO, they also had a greater decline in platelets after ECMO initiation than older children. The decline in platelets observed in our neonates (56%) was even larger than a previous report (42%) (36). These findings may help bedside clinicians anticipate the magnitude of decline in platelet count prior to laboratory sampling. Anecdotal data suggests that including platelets in the ECMO prime for neonates prevents the large decline in platelet count seen on initiation. Whether this practice prevents bleeding, thrombosis or less platelet transfusion requires further study.
Strengths of this study include the multicenter design and daily prospective collection of data. Limitations include recording platelet counts obtained closest to 7 AM rather than all values. We used platelet counts and transfusion volumes from the previous day to evaluate associations with daily bleeding and thrombosis events. We did this because the exact timing of platelet counts and transfusions, and bleeding and thrombosis events, were not recorded in the BATE study. Only the day of the bleeding or thrombosis event was recorded, not the time. Platelet counts and transfusion volumes from the same day as the bleeding or thrombosis event could potentially have occurred after the event. This was avoided by using the platelet count and transfusion volume from the previous day in our analyses. Indications for transfusing platelets were not recorded in the BATE study. Thus, the number of children that received a platelet transfusion for bleeding or solely because their count was below a predefined threshold is unknown, as is the extent that predefined thresholds may have influenced our results. The lack of data on the reason for platelet transfusion is a major limitation of our study. Information regarding platelet function was also not available. Platelet transfusion volume may be a surrogate marker for severity of illness. Although we found no association between baseline PRISM scores and bleeding or thrombotic events, severity of illness could potentially be an explanatory factor in the relationship between transfusion volume and outcomes. Concomitant CRRT may cause thrombocytopenia, unfortunately the various forms of CRRT used in this study were not collected and their associations with outcome could not be assessed. Renal failure, which included the use of CRRT in its definition, was associated with mortality. Importantly, this is an observational study and the associations observed do not infer causation.
CONCLUSIONS
Our findings suggest that nearly all children undergoing ECMO receive platelet transfusion. Platelet transfusion volume is associated with increased mortality and daily bleeding and thrombotic risk. Future investigations should include measures of platelet function to determine whether platelet count is the optimal parameter to follow clinically or whether other indices should be used in decision making for platelet transfusion.
Supplementary Material
Supplemental Digital Content 6.
Title: Platelet Transfusion Threshold by Institution, Location, and Mode of ECMO
Description: This table shows the platelet transfusion threshold by institution, location of ECMO within the hospital, and mode of ECMO.
Supplemental Digital Content 1.
Title: Univariable Logistic Models of Hospital Mortality
Description: The associations between clinical characteristics and hospital mortality assessed with univariable logistic regression.
Supplemental Digital Content 2.
Title: Univariable Logistic Models of Daily Bleeding
Description: The associations between clinical characteristics and daily bleeding risk assessed with univariable logistic regression.
Supplemental Digital Content 3.
Title: Univariable Logistic Models of Daily Thrombosis
Description: The associations between clinical characteristics and daily thrombosis risk assessed with univariable logistic regression.
Supplemental Digital Content 4.
Title: Decline in Platelet Count by Transfusion Status on Day of ECMO Initiation
Description: This table shows the last platelet count pre-ECMO and the first platelet count after ECMO initiation for children who were transfused platelets compared to those who were not.
Supplemental Digital Content 5.
Title: Decline in Platelet Count by Age on Day of ECMO Initiation
Description: This table shows the last platelet count pre-ECMO and the first platelet count after ECMO initiation for neonates compared to older children.
ACKNOWLEDGEMENTS
The authors wish to acknowledge the contributions of the following research coordinators and data coordinating center staff: Stephanie Bisping, BSN, RN, CCRP, Alecia Peterson, BS and Jeri Burr, MS, RN-BC, CCRC from University of Utah; Mary Ann DiLiberto, BS, RN, CCRC and Carol Ann Twelves, BS, RN from The Children’s Hospital of Philadelphia; Jean Reardon, MA, BSN, RN and Elyse Tomanio, BSN, RN from Children’s National Medical Center; Aimee Labell, MS, RN from Phoenix Children’s Hospital; Margaret Villa, RN and Jeni Kwok, JD from Children’s Hospital Los Angeles; Mary Ann Nyc, BS from UCLA Mattel Children’s Hospital; Ann Pawluszka, BSN, RN and Melanie Lulic, BS from Children’s Hospital of Michigan; Monica S. Weber, RN, BSN, CCRP and Lauren Conlin, BSN, RN, CCRP from University of Michigan; Alan C. Abraham, BA, CCRC from University of Pittsburgh Medical Center. The authors also wish to acknowledge the contributions of Robert Tamburro, MD and Tammara Jenkins, MSN, RN from the Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health.
Conflicts of Interest and Source of Funding: This work was supported by the following cooperative agreements from the Eunice Kennedy Shriver National Institute of Child Health and Human Development, National Institutes of Health, Department of Health and Human Services: U10HD050096, U10HD049981, U10HD049983, U10HD050012, U10HD063108, U10HD063106, U10HD063114, and U01HD049934. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Copyright form disclosure: Drs. Dalton, Newth, and Carcillo’s institution received funding from the National Institute of Child Health and Human Development. Dr. Dalton disclosed off label product use of ECMO, and she received funding from Innovative ECMO Concepts. Drs. Dalton, Reeder, Zuppa, Shanley, Newth, Pollack, Wessel, Carcillo, Harrison, Dean, and Meert received support for article research from the National Institutes of Health (NIH). Drs. Reeder, Zuppa, Shanley, Pollack, Wessel, Harrison, Dean, Meert’s institution received funding from the NIH. Dr. Shanley received funding from PAS, Inc, IPRF, and Springer. Dr. Newth received funding from Philips Research North America (consulting). The remaining authors have disclosed that they do not have any potential conflicts of interest.
REFERENCES
- 1.Nellis ME, Karam O, Mauer E, et al. : Platelet transfusion practices in critically ill children. Crit Care Med 2018; 46:1309–1317 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Saini A, West AN, Harrell C, et al. : Platelet transfusion in the PICU: Does disease severity matter? Pediatr Crit Care Med 2018; 19:e472–e478 [DOI] [PubMed] [Google Scholar]
- 3.Saini A, Spinella PC: Management of anticoagulation and hemostasis for pediatric extracorporeal membrane oxygenation. Clin Lab Med 2014; 655–673 [DOI] [PubMed] [Google Scholar]
- 4.Saini A, Hartman ME, Gage BF, et al. : Incidence of platelet dysfunction by thromboelastography-platelet mapping in children supported with ECMO: A pilot retrospective study. Front Pediatr 2016;3:116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Khorana AA, Francis CW, Blumberg N, et al. : Blood transfusions, thrombosis, and mortality in hospitalized patients with cancer. Arch Intern Med 2008; 168:2377–2381 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Blumberg N, Heal JM, Liesveld JL, et al. : Platelet transfusion and survival in adults with acute leukemia. Leukemia 2008; 22:631–635 [DOI] [PubMed] [Google Scholar]
- 7.Refaai MA, Phipps RP, Spinelli SL, et al. : Platelet transfusions: Impact on hemostasis, thrombosis, inflammation and clinical outcomes. Thromb Res 2011; 127:287–291 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Sahler J, Grimshaw K, Spinelli SL, et al. : Platelet storage and transfusions: New concerns associated with an old therapy. Drug Discov Today Dis Mech 2011; 8:e9–e14 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Du Pont-Thibodeau G, Tucci M, Robitaille N, et al. : Platelet transfusions in pediatric intensive care. Pediatr Crit Care Med 2016; 17:e420–e429 [DOI] [PubMed] [Google Scholar]
- 10.Bochsen L, Johansson PI, Kristensen AT, et al. : The influence of platelets, plasma and red blood cells on functional haemostatic assays. Blood Coagul Fibrinolysis 2011; 22:167–175 [DOI] [PubMed] [Google Scholar]
- 11.Josephson C, Meyer E: Neonatal and pediatric transfusion practice In: AABB Technical Manual. Fung MK(Ed). 18th Edition. Bethesda, MD, AABB, 2014, pp 571–597 [Google Scholar]
- 12.Kaufman RM, Djulbegovic B, Gernsheimer T, et al. : AABB: Platelet transfusion: A clinical practice guideline from the AABB. Ann Intern Med 2015; 162:205–213 [DOI] [PubMed] [Google Scholar]
- 13.Dalton HJ, Reeder R, Garcia-Filion P, et al. : Factors associated with bleeding and thrombosis in children receiving extracorporeal membrane oxygenation. Am J Respir Crit Care Med 2017; 196:762–771 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Pollack MM, Patel KM, Ruttimann UE: PRISM III: an updated Pediatric Risk of Mortality score. Crit Care Med 1996; 24:743–752 [DOI] [PubMed] [Google Scholar]
- 15.Carr R, Kelly AM, Williamson LM: Neonatal thrombocytopenia and platelet transfusion: a UK perspective. Neonatology 2015; 107:1–7 [DOI] [PubMed] [Google Scholar]
- 16.Beiner ME, Simchen MJ, Sivan E, et al. : Risk factors for neonatal thrombocytopenia in preterm infants. Am J Perinatol 2003; 20:49–54 [DOI] [PubMed] [Google Scholar]
- 17.Baer VL, Lambert DK, Henry E, et al. : Severe thrombocytopenia in the NICU. Pediatrics 2009; 124:e1095–e1100 [DOI] [PubMed] [Google Scholar]
- 18.Ferrer-Marin F, Stanworth S, Josephson C, et al. : Distinct differences in platelet production and function between neonates and adults: Implications for platelet transfusion practice. Transfusion 2013; 53:2814–2821 [DOI] [PubMed] [Google Scholar]
- 19.Sola-Visner MC: Platelets in the neonatal period: Developmental differences in platelet production, function, and hemostasis and the potential impact of therapies. Hematology Am Soc Hematol Educ Program 2012;2012:506–511 [DOI] [PubMed] [Google Scholar]
- 20.Stallion A, Cofer BR, Rafferty JA, et al. : The significant relationship between platelet count and hemorrhagic complications on ECMO. Perfusion 1994; 9:265–269 [DOI] [PubMed] [Google Scholar]
- 21.Chevuru SC, Sola MC, Theriaque DW, et al. : Multicenter analysis of platelet transfusion usage among neonates on extracorporeal membrane oxygenation. Pediatrics 2002; 109:e89. [DOI] [PubMed] [Google Scholar]
- 22.Curley A, Stanworth SJ, Willoughby K, et al. : Randomized trial of platelet-transfusion thresholds in neonates. N Eng J Med 2019; 380:242–251 [DOI] [PubMed] [Google Scholar]
- 23.Brunetti MA, Gaynor JW, Retzloff LB, et al. : Characteristics, risk factors, and outcomes of extracorporeal use in pediatric cardiac ICUs: A report from the pediatric cardiac critical care consortium registry. Pediatr Crit Care Med 2018; 19:544–552 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Ranucci M, Pistuddi V, Di Dedda U, et al. : Platelet function after cardiac surgery and its association with severe postoperative bleeding; the PLATFORM study. Platelets 2018. October 26:1–7 doi: 10.1080/09537104.2018.1535706 [Epub ahead of print] [DOI] [PubMed] [Google Scholar]
- 25.Heilman C, Geisen U, Beyersdorf F: Acquired von Willebrand syndrome in patients with extracorporeal life support (ECLS). Intensive Care Med 2012; 38:62–68 [DOI] [PubMed] [Google Scholar]
- 26.McVeen RV, Lorch V, Carroll RC, et al. : Changes in fibrinolytic factors in newborns during extracorporeal membrane oxygenation. Am J Hematol 1991; 38:254–255 [DOI] [PubMed] [Google Scholar]
- 27.Mutlak H, Reyher C, Meybohm P, et al. : Multiple electrode aggregometry for the assessment of acquired platelet dysfunction during extracorporeal membrane oxygenation. J Thorac Cardiovasc Surg 2015; 63:21–22 [DOI] [PubMed] [Google Scholar]
- 28.Annich G: Extracorporeal life support: The precarious balance of hemostastis. J Thromb Haemost 2015; 13 (Suppl 1):S336–S342 [DOI] [PubMed] [Google Scholar]
- 29.Nomura S, Ozaki Y, Ikeda Y: Function and role of microparticles in various clinical settings. Thromb Res 2008; 123:8–23 [DOI] [PubMed] [Google Scholar]
- 30.Meyer AD, Gelfond JAL, Wiles AA, et al. : Platelet-derived microparticles generated by neonatal extracorporeal membrane oxygenation systems. ASAIO J 2015; 61:37–42 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Adachi I, Kostousov V, Hensch L, et al. : Management of hemostasis for pediatric patients on ventricular-assist devices. Semin Thromb Hemost 2018; 44:30–37 [DOI] [PubMed] [Google Scholar]
- 32.Annich GM, Meinhardt JP, Mowery KA, et al. : Reduced platelet activation and thrombosis in extracorporeal circuits coated with nitric oxide release polymers. Crit Care Med 2000; 28:915–920 [DOI] [PubMed] [Google Scholar]
- 33.Batchelor MM, Reoma SL, Fleser PS, et al. : More lipophilic dialkyldiamine-based diazeniumdiolates: synthesis, characterization, and application in preparing thromboresistant nitric oxide release polymeric coatings. J Med Chem 2003; 46:5153–5161 [DOI] [PubMed] [Google Scholar]
- 34.Amoako KA, Archangeli C, Handa H, et al. : Thromboresistance characterization of extruded nitric oxide-releasing silicone catheters. ASAIO J 2012; 58:238–246 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Cheung PY, Sawicki G, Salas E, et al. : The mechanisms of platelet dysfunction during extracorporeal membrane oxygenation in critically ill neonates. Crit Care Med 2000; 28: 2584–2590 [DOI] [PubMed] [Google Scholar]
- 36.Robinson TM, Kickler TS, Walker LK, et al. : Effect of extracorporeal membrane oxygenation on platelets in newborns. Crit Care Med 1993; 7:1029–1034 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental Digital Content 6.
Title: Platelet Transfusion Threshold by Institution, Location, and Mode of ECMO
Description: This table shows the platelet transfusion threshold by institution, location of ECMO within the hospital, and mode of ECMO.
Supplemental Digital Content 1.
Title: Univariable Logistic Models of Hospital Mortality
Description: The associations between clinical characteristics and hospital mortality assessed with univariable logistic regression.
Supplemental Digital Content 2.
Title: Univariable Logistic Models of Daily Bleeding
Description: The associations between clinical characteristics and daily bleeding risk assessed with univariable logistic regression.
Supplemental Digital Content 3.
Title: Univariable Logistic Models of Daily Thrombosis
Description: The associations between clinical characteristics and daily thrombosis risk assessed with univariable logistic regression.
Supplemental Digital Content 4.
Title: Decline in Platelet Count by Transfusion Status on Day of ECMO Initiation
Description: This table shows the last platelet count pre-ECMO and the first platelet count after ECMO initiation for children who were transfused platelets compared to those who were not.
Supplemental Digital Content 5.
Title: Decline in Platelet Count by Age on Day of ECMO Initiation
Description: This table shows the last platelet count pre-ECMO and the first platelet count after ECMO initiation for neonates compared to older children.