524264 research-article2014
PRF0010.1177/0267659114524264PerfusionAtallah et al.
Original paper
Evaluation of the activated clotting time and activated partial thromboplastin time for the monitoring of heparin in adult extracorporeal membrane oxygenation patients
Perfusion 2014, Vol. 29(5) 456–461 © The Author(s) 2014 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0267659114524264 prf.sagepub.com
S Atallah,1,2 M Liebl,1 K Fitousis,1 F Bostan1 and F Masud3
Abstract Introduction: Historically, the activated clotting time (ACT) has been the preferred monitoring test of the heparin effect in extracorporeal membrane oxygenation (ECMO) patients. However, few adult studies have evaluated its correlation to the heparin dose or other monitoring tests, such as the activated partial thromboplastin time (aPTT). This retrospective study sought to evaluate the correlation between the heparin dose and these monitoring tests. Methods: Patients administered a heparin drip during ECMO were included in this study. The primary endpoints were the correlation between heparin dose and ACT or aPTT and the relationship between paired ACT and aPTT samples. Results: Forty-six patients met the criteria for study inclusion. A better correlation was observed for heparin dose and aPTT (Pearson product-moment correlation coefficient (r) = 0.43 – 0.54) versus ACT (r = 0.11 – 0.14). Among the paired sample data, ACT values did not differ significantly between Groups two (aPTT 60 – 75 seconds) and three (aPTT >75 seconds). Conclusion: The heparin dose correlated better with aPTT relative to ACT and, thus, may be considered a more effective tool for the dosing of heparin in adult ECMO patients. Paired ACT and aPTT sample data suggested a poor relationship between these two anticoagulant monitoring tests. Keywords partial thromboplastin time; activated clotting time; extracorporeal membrane oxygenation; heparin; anticoagulation
Introduction Exposure of the non-biologic extracorporeal membrane oxygenation (ECMO) device to blood and cellular components results in both inflammatory and thrombotic responses.1 Consequently, therapeutic anticoagulation is required in order to reduce the risk of systemic thromboembolism and clotting within the ECMO circuitry. One must also consider that ECMO results in the consumption of various components vital to hemostasis, including platelets and clotting factors. Therefore, these patients are also at a significantly increased risk of hemorrhagic complications. Unfractionated heparin has typically been considered the anticoagulant of choice, given its short half-life and its reversibility with protamine, if necessary.2 The effect of heparin can be measured using a variety of tests, including the activated clotting time (ACT), activated partial thromboplastin time (aPTT), antifactor Xa level and thromboelastography. Despite the ACT being
the most common method of monitoring the heparin effect in ECMO, studies in neonatal patients have dem-
1Houston
Methodist Hospital, Department of Pharmacy, Houston, TX, USA 2Scripps Mercy Hospital, Department of Pharmacy, San Diego, CA, USA 3Methodist DeBakey Heart and Vascular Center, Houston Methodist Hospital, Houston, Texas, USA Corresponding author: Steven Atallah Scripps Mercy Hospital Department of Pharmacy 4077 Fifth Avenue MER 52 San Diego CA 92103 USA. Email: [email protected]
Downloaded from prf.sagepub.com at GEORGIAN COURT UNIV on February 21, 2015
457
Atallah et al.
onstrated a poor correlation with other monitoring methods, including the aPTT and anti-Xa.3,4 Unfortunately, there is a paucity of literature evaluating the correlation between different methods of anticoagulation monitoring in adult ECMO patients. In addition, the studies have not evaluated the correlation between the heparin dose and ACT or aPTT in adult ECMO patients. This retrospective study sought to evaluate the correlation between the heparin dose and either the ACT or aPTT. In addition, this study assessed the relationship between paired ACT and aPTT samples in ECMO patients.
Methods The retrospective study was conducted at Houston Methodist Hospital, a 948 bed academic tertiary referral hospital located in The Texas Medical Center at its Houston Methodist DeBakey Heart and Vascular Center and was granted expedited status from the institutional review board. Houston Methodist Hospital Perfusion Service provided a list of all adult ECMO patients admitted to Houston Methodist from July 1, 2011 to August 31, 2012, who were eligible for study investigation. During the study period, the heparin was dose adjusted based on the ACT value, which was typically obtained every 1 to 3 hours. Heparin was derived from porcine intestinal mucosa and standardized for use as an anticoagulant. ACT values were measured using the Hemochron Response point-of-care system (International Technidyne Corporation, Edison, NJ, USA) at the patient bedside and most commonly targeted a range of 140-180 seconds. APTTs were obtained based on physician preference and were sent to the central laboratory for measurement. The aPTT reagent used was the STA-PTT Automate 5 (Diagnostica Stago Inc., Parsippany, NJ, USA). Patients administered a heparin drip while on ECMO and in whom ACT values were used to titrate the anticoagulant dose were included in this study. Exclusion criteria consisted of the following: patients administered heparin for indications other than the prevention of ECMO-related clotting (i.e., venous thromboembolism, valve replacement, etc.), patients administered anticoagulant medications other than heparin for the prevention of ECMO-related clotting and patients not initiated on a heparin drip while undergoing ECMO. The primary endpoints evaluated in this study included: the correlation between the heparin dose and either ACT or aPTT and the relationship between paired ACT and aPTT samples. Secondary endpoints included a subgroup analysis, assessing the effect of numerous factors on heparin dose and ACT or aPTT correlation. The association between demographic and clinical variables and the average number of packed red blood cells
(pRBCs) transfused per day was also evaluated as a secondary endpoint. The correlation between the heparin dose and the ACT or aPTT was assessed through the use of the Pearson product-moment correlation coefficient. Based on the half-life of heparin and the expected time until attainment of steady state, ACT or aPTT values were only included if the heparin rate had not been adjusted for at least six hours prior to sample collection. These criteria would allow for a more accurate determination of the relationship between the heparin dose and the ACT or aPTT. For the paired samples, data were divided into three groups based on the level of anticoagulation as measured using the laboratory aPTT: Group 1 (aPTT less than 60 seconds), Group 2 (aPTT 60 to 75 seconds) and Group 3 (aPTT greater than 75 seconds). Group 2 was defined as the therapeutic aPTT, with Groups 1 and 3 classified as the sub-therapeutic and supra-therapeutic aPTT groups. The Pearson product-moment correlation coefficient was also calculated for the paired sample data. As a result of the ACT being a point-of-care test and the aPTT being a central laboratory test, samples were not typically measured at the same time. Thus, in this study, paired samples were defined as samples collected within 20 minutes of each other. Samples were not included if the heparin rate had been adjusted between the collection of the ACT and aPTT. Multiple variables, including mean ACT from all available ACT values, age, platelet count and type of ECMO, were also evaluated to determine if they were predictors of the average number of pRBCs transfused in a day. The dosing weight (DW) was defined as the actual bodyweight if the actual/ideal bodyweight was less than 1.2. For patients in whom the actual/ideal bodyweight was greater than 1.2, a calculated dosing weight was used where DW equals the ideal bodyweight (IBW) + 0.4(actual – ideal bodyweight).The ideal bodyweight was defined as 45.5 + [(2.3 x height (inches) – 60)] for females and 50 + [(2.3 x (height (inches) – 60)] for males.
Statistical Analysis Statistical analyses were conducted using the Minitab 16 software (Minitab Inc., State College, PA, USA). Continuous variables were analyzed using the two sample t-test. The correlation between two continuous variables was assessed via the Pearson product-moment correlation coefficient. Comparisons of correlation coefficients were performed using the Fisher r-to-z transformation.5 Data involving multiple groups were evaluated via the Tukey test. A p-value of less than 0.05 was considered statistically significant.
Downloaded from prf.sagepub.com at GEORGIAN COURT UNIV on February 21, 2015
458
Perfusion 29(5)
Results A total of 65 ECMO patients were identified during the defined study period. Forty-six patients met the study inclusion criteria. Lack of administration of a heparin drip was the primary reason patients (n=15) were excluded from this study. Two patients were excluded because they were administered bivalirudin instead of heparin. Electronic medical records could not be accessed for two patients, resulting in the exclusion of these patients, as well. A description of the patient enrollment is summarized in Figure 1 and baseline characteristics of the study population are summarized in Table 1. The mean duration of ECMO was 11 days, with a balanced distribution between venoarterial (VA) and venovenous (VV) ECMO. Few patients (n = 3) had a documented neurological insult, defined as either cerebral infarct or intracranial hemorrhage, while on ECMO (Table 2). The mortality rate at discharge was 87%. To assess the association between the ACT and heparin rate, 2301 ACTs were identified among 46 adult ECMO patients. The Pearson product-moment correlation coefficient ranged from 0.11 to 0.14 (Table 3). The intra-subject correlation coefficient for ACT and heparin rate was 0.14, indicating little to no correlation between the heparin dose and this method of anticoagulation monitoring (Figure 2). Correlation coefficient values ranged from 0.43 to 0.55 for the aPTT and heparin rate (Figure 3). Correlation was moderately increased when normalized for weight. The intra-subject correlation coefficient was not calculated, given the small aPTT sample size for each subject. Various factors were evaluated to determine their correlation with heparin dose and ACT or aPTT (Table 4). The only statistically significant difference observed was between VA and VV ECMO patients in the ACT group. Within the aPTT group, a trend towards a higher correlation coefficient was observed for patients undergoing VA ECMO, no continuous renal replacement therapy (CRRT) and an international normalized ratio (INR) less than 1.5. Between the paired ACT and aPTT data, a total of 143 paired samples were identified among 39 patients. A statistically significant difference was observed among the aPTTs for each specified group (Table 5). A statistically significant difference was not detected for the ACT values in Groups 2 and 3. The Pearson product-moment correlation coefficient for the paired ACT and aPTT data was 0.41. Neither age greater than 65 nor ACT greater than 180 seconds were associated with an increased requirement for pRBC transfusions (Table 6). Patients undergoing VA ECMO relative to VV ECMO were transfused a significantly greater number of pRBCs per day (Table 6). Patients with platelet counts greater than 50,000/μL
Figure 1. Inclusion and exclusion criteria. Table 1. Patient characteristics. Total number of patients, n Age, years (mean ± SD) Gender, % Male Female Weight, kg (mean ± SD) Actual Bodyweight Ideal Bodyweight Dosing Weight Reasons for ECMO initiation, % Cardiac Respiratory Both Baseline hemoglobin, grams/deciliter (mean ± SD) Baseline platelet count, x103/ microliter (mean ± SD)
46 56 ± 15 67 33 86.7 ± 23.1 64.9 ± 11.3 73.8 ± 14.1 45 45 10 10.3 ± 1.8 170 ± 99
Table 2. Patient outcomes. Duration of ECMO, days (mean ± SD) Days on venoarterial (VA) ECMO, % Days on venovenous (VV) ECMO, % Cerebral infarct during ECMO, n (%) Intracranial hemorrhage during ECMO, n (%) Mortality rate at discharge, n (%) Heparin dose, units/kg/hour DW (mean ± SD) ACT, seconds (mean ± SD) aPTT, seconds (mean ± SD)
11 ± 14.6 50.4 49.6 1 (2.2) 2 (4.3) 40 (87) 10.3 ± 5.2 164 ± 20.5 61 ± 29.2
ECMO: extracorporeal membrane oxygenation, DW: dosing weight, ACT: activated clotting time, aPTT: activated partial thromboplastin time.
also had a statistically significant increase in pRBC transfusions compared to those with a platelet count less than 50,000/μL (Table 6).
Downloaded from prf.sagepub.com at GEORGIAN COURT UNIV on February 21, 2015
459
Atallah et al. Table 3. Pearson product-moment correlation coefficient for heparin rate and either ACT or aPTT. Heparin rate
Correlation coefficient for ACT and heparin rate
Correlation coefficient for aPTT and heparin rate
Heparin rate (units/hr) Heparin rate (units/kg/hr) IBW Heparin rate (units/kg/hr) DW Heparin rate (units/kg/hr) ABW
0.11 0.12 0.14 0.14
0.43 0.52 0.55 0.54
IBW: ideal bodyweight, DW: dosing weight, ABW: actual bodyweight.
Figure 2. Scatterplot of ACT and heparin rate (units/kg/hr) using dosing weight.
Figure 3. Scatterplot of aPTT and heparin rate (units/kg/hr) using dosing weight.
Discussion Historically, the ACT has typically been considered the standard monitoring method for patients undergoing ECMO. In order to measure the ACT, whole blood is mixed with an activator, such as celite or kaolin, to provide a global functional test of hemostasis. Some factors that may affect ACT values include platelet activation, hemodilution of hemostatic factors, activation of the hemostatic system and hypothermia.6 The aPTT is one of most common tests used to monitor the effect of heparin in non-extracorporeal life support patients. The aPTT is a plasma test which evaluates the intrinsic and common coagulation pathways.7 In order to perform an aPTT test, surface activator, diluted phospholipid and calcium are added to citrated
plasma.8 The assay then measures the time it takes for a fibrin clot to form. Factors known to influence the aPTT include diurnal variation9,10, high concentration of citrate in the collection tube11, clotting factor deficiencies8,12 and consumptive coagulopathy. Several studies have compared the ACT to aPTT in adult non-extracorporeal life support patients.13–15 These studies have suggested that the aPTT correlates better with heparin concentration than the ACT. In addition, one study by Waele et al.13 demonstrated that the ACT was unable to differentiate between low and therapeutic levels of anticoagulation according to the aPTT. Until this study, no published data has compared the ACT to the aPTT for the monitoring of the heparin effect in adult ECMO patients. Our retrospective study revealed that the ACT appears to be an unreliable tool for monitoring anticoagulation in adult ECMO patients. While the aPTT moderately correlated with the heparin dose, a positive linear relationship with the ACT and heparin dose could not be established. Most institutions utilize a dosing nomogram to target their specified ACT or aPTT range. Thus, it is necessary that the monitoring tool exhibit a dose-response curve if such a strategy is to be used. In addition to the limited dose-response observed with the ACT, the relationship between the paired aPTT and ACT samples was poor. ACT values did not differ significantly between those who had therapeutic and supratherapeutic aPTTs. Thus, if the aPTT is considered a more reliable anticoagulant monitoring tool, these patients may have an increased risk of bleeding, as defined by a supratherapeutic aPTT, which would not be recognized by the “therapeutic” ACT value. Unfortunately, studies have not evaluated which anticoagulant monitoring tool is a better predictor of bleeding in adult ECMO patients. Our paired sample size data was not considered large enough to warrant an appropriate evaluation of this outcome. Larger, prospective studies will need to be conducted in order to validate whether the aPTT or ACT is a better predictor of bleeding in adult ECMO patients as well as the optimal target range. Univariate analysis revealed a trend towards improved correlation between aPTT and heparin dose
Downloaded from prf.sagepub.com at GEORGIAN COURT UNIV on February 21, 2015
460
Perfusion 29(5)
Table 4. Pearson product-moment correlation coefficient among subgroups.
VV ECMO VA ECMO CRRT No CRRT INR ≥ 1.5 INR < 1.5
Correlation between Heparin Rate and ACT
# of Samples
Correlation between Heparin Rate and a PTT
# of Samples
0.04 0.22a 0.09 0.17 Not studied Not studied
856 1374 1234 1065 Not studied Not studied
0.51 0.57 0.51 0.56 0.53 0.60
124 104 98 137 140 94
VV: venevenous, VA: venoarterial, CRRT: continuous renal replacement therapy, INR: international normalized ratio. ap 180 seconds Platelet count < 50,000/μL Platelet count > 50,000/μL
# of Samples
Mean # of pRBCs transfused/day
95% Confidence Interval for difference
155 304 223 222 320 54 168 289
1.41 1.76 2.23 1.13 1.47 1.46 1.32 1.84
−0.75 – 0.067
0.100
0.61 – 1.39
PRF0010.1177/0267659114524264PerfusionAtallah et al.
Original paper
Evaluation of the activated clotting time and activated partial thromboplastin time for the monitoring of heparin in adult extracorporeal membrane oxygenation patients
Perfusion 2014, Vol. 29(5) 456–461 © The Author(s) 2014 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0267659114524264 prf.sagepub.com
S Atallah,1,2 M Liebl,1 K Fitousis,1 F Bostan1 and F Masud3
Abstract Introduction: Historically, the activated clotting time (ACT) has been the preferred monitoring test of the heparin effect in extracorporeal membrane oxygenation (ECMO) patients. However, few adult studies have evaluated its correlation to the heparin dose or other monitoring tests, such as the activated partial thromboplastin time (aPTT). This retrospective study sought to evaluate the correlation between the heparin dose and these monitoring tests. Methods: Patients administered a heparin drip during ECMO were included in this study. The primary endpoints were the correlation between heparin dose and ACT or aPTT and the relationship between paired ACT and aPTT samples. Results: Forty-six patients met the criteria for study inclusion. A better correlation was observed for heparin dose and aPTT (Pearson product-moment correlation coefficient (r) = 0.43 – 0.54) versus ACT (r = 0.11 – 0.14). Among the paired sample data, ACT values did not differ significantly between Groups two (aPTT 60 – 75 seconds) and three (aPTT >75 seconds). Conclusion: The heparin dose correlated better with aPTT relative to ACT and, thus, may be considered a more effective tool for the dosing of heparin in adult ECMO patients. Paired ACT and aPTT sample data suggested a poor relationship between these two anticoagulant monitoring tests. Keywords partial thromboplastin time; activated clotting time; extracorporeal membrane oxygenation; heparin; anticoagulation
Introduction Exposure of the non-biologic extracorporeal membrane oxygenation (ECMO) device to blood and cellular components results in both inflammatory and thrombotic responses.1 Consequently, therapeutic anticoagulation is required in order to reduce the risk of systemic thromboembolism and clotting within the ECMO circuitry. One must also consider that ECMO results in the consumption of various components vital to hemostasis, including platelets and clotting factors. Therefore, these patients are also at a significantly increased risk of hemorrhagic complications. Unfractionated heparin has typically been considered the anticoagulant of choice, given its short half-life and its reversibility with protamine, if necessary.2 The effect of heparin can be measured using a variety of tests, including the activated clotting time (ACT), activated partial thromboplastin time (aPTT), antifactor Xa level and thromboelastography. Despite the ACT being
the most common method of monitoring the heparin effect in ECMO, studies in neonatal patients have dem-
1Houston
Methodist Hospital, Department of Pharmacy, Houston, TX, USA 2Scripps Mercy Hospital, Department of Pharmacy, San Diego, CA, USA 3Methodist DeBakey Heart and Vascular Center, Houston Methodist Hospital, Houston, Texas, USA Corresponding author: Steven Atallah Scripps Mercy Hospital Department of Pharmacy 4077 Fifth Avenue MER 52 San Diego CA 92103 USA. Email: [email protected]
Downloaded from prf.sagepub.com at GEORGIAN COURT UNIV on February 21, 2015
457
Atallah et al.
onstrated a poor correlation with other monitoring methods, including the aPTT and anti-Xa.3,4 Unfortunately, there is a paucity of literature evaluating the correlation between different methods of anticoagulation monitoring in adult ECMO patients. In addition, the studies have not evaluated the correlation between the heparin dose and ACT or aPTT in adult ECMO patients. This retrospective study sought to evaluate the correlation between the heparin dose and either the ACT or aPTT. In addition, this study assessed the relationship between paired ACT and aPTT samples in ECMO patients.
Methods The retrospective study was conducted at Houston Methodist Hospital, a 948 bed academic tertiary referral hospital located in The Texas Medical Center at its Houston Methodist DeBakey Heart and Vascular Center and was granted expedited status from the institutional review board. Houston Methodist Hospital Perfusion Service provided a list of all adult ECMO patients admitted to Houston Methodist from July 1, 2011 to August 31, 2012, who were eligible for study investigation. During the study period, the heparin was dose adjusted based on the ACT value, which was typically obtained every 1 to 3 hours. Heparin was derived from porcine intestinal mucosa and standardized for use as an anticoagulant. ACT values were measured using the Hemochron Response point-of-care system (International Technidyne Corporation, Edison, NJ, USA) at the patient bedside and most commonly targeted a range of 140-180 seconds. APTTs were obtained based on physician preference and were sent to the central laboratory for measurement. The aPTT reagent used was the STA-PTT Automate 5 (Diagnostica Stago Inc., Parsippany, NJ, USA). Patients administered a heparin drip while on ECMO and in whom ACT values were used to titrate the anticoagulant dose were included in this study. Exclusion criteria consisted of the following: patients administered heparin for indications other than the prevention of ECMO-related clotting (i.e., venous thromboembolism, valve replacement, etc.), patients administered anticoagulant medications other than heparin for the prevention of ECMO-related clotting and patients not initiated on a heparin drip while undergoing ECMO. The primary endpoints evaluated in this study included: the correlation between the heparin dose and either ACT or aPTT and the relationship between paired ACT and aPTT samples. Secondary endpoints included a subgroup analysis, assessing the effect of numerous factors on heparin dose and ACT or aPTT correlation. The association between demographic and clinical variables and the average number of packed red blood cells
(pRBCs) transfused per day was also evaluated as a secondary endpoint. The correlation between the heparin dose and the ACT or aPTT was assessed through the use of the Pearson product-moment correlation coefficient. Based on the half-life of heparin and the expected time until attainment of steady state, ACT or aPTT values were only included if the heparin rate had not been adjusted for at least six hours prior to sample collection. These criteria would allow for a more accurate determination of the relationship between the heparin dose and the ACT or aPTT. For the paired samples, data were divided into three groups based on the level of anticoagulation as measured using the laboratory aPTT: Group 1 (aPTT less than 60 seconds), Group 2 (aPTT 60 to 75 seconds) and Group 3 (aPTT greater than 75 seconds). Group 2 was defined as the therapeutic aPTT, with Groups 1 and 3 classified as the sub-therapeutic and supra-therapeutic aPTT groups. The Pearson product-moment correlation coefficient was also calculated for the paired sample data. As a result of the ACT being a point-of-care test and the aPTT being a central laboratory test, samples were not typically measured at the same time. Thus, in this study, paired samples were defined as samples collected within 20 minutes of each other. Samples were not included if the heparin rate had been adjusted between the collection of the ACT and aPTT. Multiple variables, including mean ACT from all available ACT values, age, platelet count and type of ECMO, were also evaluated to determine if they were predictors of the average number of pRBCs transfused in a day. The dosing weight (DW) was defined as the actual bodyweight if the actual/ideal bodyweight was less than 1.2. For patients in whom the actual/ideal bodyweight was greater than 1.2, a calculated dosing weight was used where DW equals the ideal bodyweight (IBW) + 0.4(actual – ideal bodyweight).The ideal bodyweight was defined as 45.5 + [(2.3 x height (inches) – 60)] for females and 50 + [(2.3 x (height (inches) – 60)] for males.
Statistical Analysis Statistical analyses were conducted using the Minitab 16 software (Minitab Inc., State College, PA, USA). Continuous variables were analyzed using the two sample t-test. The correlation between two continuous variables was assessed via the Pearson product-moment correlation coefficient. Comparisons of correlation coefficients were performed using the Fisher r-to-z transformation.5 Data involving multiple groups were evaluated via the Tukey test. A p-value of less than 0.05 was considered statistically significant.
Downloaded from prf.sagepub.com at GEORGIAN COURT UNIV on February 21, 2015
458
Perfusion 29(5)
Results A total of 65 ECMO patients were identified during the defined study period. Forty-six patients met the study inclusion criteria. Lack of administration of a heparin drip was the primary reason patients (n=15) were excluded from this study. Two patients were excluded because they were administered bivalirudin instead of heparin. Electronic medical records could not be accessed for two patients, resulting in the exclusion of these patients, as well. A description of the patient enrollment is summarized in Figure 1 and baseline characteristics of the study population are summarized in Table 1. The mean duration of ECMO was 11 days, with a balanced distribution between venoarterial (VA) and venovenous (VV) ECMO. Few patients (n = 3) had a documented neurological insult, defined as either cerebral infarct or intracranial hemorrhage, while on ECMO (Table 2). The mortality rate at discharge was 87%. To assess the association between the ACT and heparin rate, 2301 ACTs were identified among 46 adult ECMO patients. The Pearson product-moment correlation coefficient ranged from 0.11 to 0.14 (Table 3). The intra-subject correlation coefficient for ACT and heparin rate was 0.14, indicating little to no correlation between the heparin dose and this method of anticoagulation monitoring (Figure 2). Correlation coefficient values ranged from 0.43 to 0.55 for the aPTT and heparin rate (Figure 3). Correlation was moderately increased when normalized for weight. The intra-subject correlation coefficient was not calculated, given the small aPTT sample size for each subject. Various factors were evaluated to determine their correlation with heparin dose and ACT or aPTT (Table 4). The only statistically significant difference observed was between VA and VV ECMO patients in the ACT group. Within the aPTT group, a trend towards a higher correlation coefficient was observed for patients undergoing VA ECMO, no continuous renal replacement therapy (CRRT) and an international normalized ratio (INR) less than 1.5. Between the paired ACT and aPTT data, a total of 143 paired samples were identified among 39 patients. A statistically significant difference was observed among the aPTTs for each specified group (Table 5). A statistically significant difference was not detected for the ACT values in Groups 2 and 3. The Pearson product-moment correlation coefficient for the paired ACT and aPTT data was 0.41. Neither age greater than 65 nor ACT greater than 180 seconds were associated with an increased requirement for pRBC transfusions (Table 6). Patients undergoing VA ECMO relative to VV ECMO were transfused a significantly greater number of pRBCs per day (Table 6). Patients with platelet counts greater than 50,000/μL
Figure 1. Inclusion and exclusion criteria. Table 1. Patient characteristics. Total number of patients, n Age, years (mean ± SD) Gender, % Male Female Weight, kg (mean ± SD) Actual Bodyweight Ideal Bodyweight Dosing Weight Reasons for ECMO initiation, % Cardiac Respiratory Both Baseline hemoglobin, grams/deciliter (mean ± SD) Baseline platelet count, x103/ microliter (mean ± SD)
46 56 ± 15 67 33 86.7 ± 23.1 64.9 ± 11.3 73.8 ± 14.1 45 45 10 10.3 ± 1.8 170 ± 99
Table 2. Patient outcomes. Duration of ECMO, days (mean ± SD) Days on venoarterial (VA) ECMO, % Days on venovenous (VV) ECMO, % Cerebral infarct during ECMO, n (%) Intracranial hemorrhage during ECMO, n (%) Mortality rate at discharge, n (%) Heparin dose, units/kg/hour DW (mean ± SD) ACT, seconds (mean ± SD) aPTT, seconds (mean ± SD)
11 ± 14.6 50.4 49.6 1 (2.2) 2 (4.3) 40 (87) 10.3 ± 5.2 164 ± 20.5 61 ± 29.2
ECMO: extracorporeal membrane oxygenation, DW: dosing weight, ACT: activated clotting time, aPTT: activated partial thromboplastin time.
also had a statistically significant increase in pRBC transfusions compared to those with a platelet count less than 50,000/μL (Table 6).
Downloaded from prf.sagepub.com at GEORGIAN COURT UNIV on February 21, 2015
459
Atallah et al. Table 3. Pearson product-moment correlation coefficient for heparin rate and either ACT or aPTT. Heparin rate
Correlation coefficient for ACT and heparin rate
Correlation coefficient for aPTT and heparin rate
Heparin rate (units/hr) Heparin rate (units/kg/hr) IBW Heparin rate (units/kg/hr) DW Heparin rate (units/kg/hr) ABW
0.11 0.12 0.14 0.14
0.43 0.52 0.55 0.54
IBW: ideal bodyweight, DW: dosing weight, ABW: actual bodyweight.
Figure 2. Scatterplot of ACT and heparin rate (units/kg/hr) using dosing weight.
Figure 3. Scatterplot of aPTT and heparin rate (units/kg/hr) using dosing weight.
Discussion Historically, the ACT has typically been considered the standard monitoring method for patients undergoing ECMO. In order to measure the ACT, whole blood is mixed with an activator, such as celite or kaolin, to provide a global functional test of hemostasis. Some factors that may affect ACT values include platelet activation, hemodilution of hemostatic factors, activation of the hemostatic system and hypothermia.6 The aPTT is one of most common tests used to monitor the effect of heparin in non-extracorporeal life support patients. The aPTT is a plasma test which evaluates the intrinsic and common coagulation pathways.7 In order to perform an aPTT test, surface activator, diluted phospholipid and calcium are added to citrated
plasma.8 The assay then measures the time it takes for a fibrin clot to form. Factors known to influence the aPTT include diurnal variation9,10, high concentration of citrate in the collection tube11, clotting factor deficiencies8,12 and consumptive coagulopathy. Several studies have compared the ACT to aPTT in adult non-extracorporeal life support patients.13–15 These studies have suggested that the aPTT correlates better with heparin concentration than the ACT. In addition, one study by Waele et al.13 demonstrated that the ACT was unable to differentiate between low and therapeutic levels of anticoagulation according to the aPTT. Until this study, no published data has compared the ACT to the aPTT for the monitoring of the heparin effect in adult ECMO patients. Our retrospective study revealed that the ACT appears to be an unreliable tool for monitoring anticoagulation in adult ECMO patients. While the aPTT moderately correlated with the heparin dose, a positive linear relationship with the ACT and heparin dose could not be established. Most institutions utilize a dosing nomogram to target their specified ACT or aPTT range. Thus, it is necessary that the monitoring tool exhibit a dose-response curve if such a strategy is to be used. In addition to the limited dose-response observed with the ACT, the relationship between the paired aPTT and ACT samples was poor. ACT values did not differ significantly between those who had therapeutic and supratherapeutic aPTTs. Thus, if the aPTT is considered a more reliable anticoagulant monitoring tool, these patients may have an increased risk of bleeding, as defined by a supratherapeutic aPTT, which would not be recognized by the “therapeutic” ACT value. Unfortunately, studies have not evaluated which anticoagulant monitoring tool is a better predictor of bleeding in adult ECMO patients. Our paired sample size data was not considered large enough to warrant an appropriate evaluation of this outcome. Larger, prospective studies will need to be conducted in order to validate whether the aPTT or ACT is a better predictor of bleeding in adult ECMO patients as well as the optimal target range. Univariate analysis revealed a trend towards improved correlation between aPTT and heparin dose
Downloaded from prf.sagepub.com at GEORGIAN COURT UNIV on February 21, 2015
460
Perfusion 29(5)
Table 4. Pearson product-moment correlation coefficient among subgroups.
VV ECMO VA ECMO CRRT No CRRT INR ≥ 1.5 INR < 1.5
Correlation between Heparin Rate and ACT
# of Samples
Correlation between Heparin Rate and a PTT
# of Samples
0.04 0.22a 0.09 0.17 Not studied Not studied
856 1374 1234 1065 Not studied Not studied
0.51 0.57 0.51 0.56 0.53 0.60
124 104 98 137 140 94
VV: venevenous, VA: venoarterial, CRRT: continuous renal replacement therapy, INR: international normalized ratio. ap 180 seconds Platelet count < 50,000/μL Platelet count > 50,000/μL
# of Samples
Mean # of pRBCs transfused/day
95% Confidence Interval for difference
155 304 223 222 320 54 168 289
1.41 1.76 2.23 1.13 1.47 1.46 1.32 1.84
−0.75 – 0.067
0.100
0.61 – 1.39
