ABSTRACT
BACKGROUND: This study aims to compare lactate and central venous blood gas in the prediction of outcome in pediatric venoarterial mode extracorporeal membrane oxygenation (V-A ECMO). METHOD: This was a retrospective observational study conducted on patients undergoing V-A ECMO care in the pediatric intensive care unit of a tertiary medical center in Taiwan. Patients under 18 years of age undergoing V-A ECMO from January 2009 to April 2019 were included in this study. RESULTS: This study consisted of 47 children who received V-A ECMO with an overall weaning rate of 66.0%. The mean age was 5.5 years and mean ECMO duration was 11.6 days. Successful weaning group had significantly lower lactate levels at initial (58.7 ± 47.0 mg/dL vs. 108.0 ± 55.3 mg/dL, p = 0.003), 0-12 h (37.8 ± 29.0 mg/dL vs. 83.5 ± 60.0 mg/dL, p Z 0.001), and 12-24 h (29.4 ± 26.9 mg/dL vs. 69.1 ± 59.1 mg/dL, p = 0.003) after ECMO initiation; however, the central venous blood gas including pH, HCO3, CO2, base excess (BE), and O2 saturation showed no significant difference. The favorable outcome group had significantly lower lactate levels at 0-12 h (32.8 ± 26.3 mg/dL vs. 71.3 ± 53.3 mg/dL, p = 0.005), and 12-24 h (20.7 ± 10.2 mg/dL vs. 61.9 ± 53.5 mg/dL, p = 0.002); however, the HCO3 levels (26.2 ± 4.5 mmol/L vs. 22.9 ± 6.8 mmol/L, p = 0.042) and BE (2.2 ± 5.4 vs. 2.2 ± 8.5, p = 0.047) were significantly higher at 12-24 h. In multivariate logistic regression, 12-24 h lactate value was an independent factor for unfavorable outcomes (p = 0.015, odds ratio [OR] = 1.1) with the best cut-off value of 48.6 mg/dL (sensitivity 48%, specificity 100%). CONCLUSION: Lactate has better outcome prediction than central venous blood gas in pediatric V-A ECMO. The lactate value 12-24 h after ECMO initiation was an independent factor for unfavorable outcomes.
Subject(s)
Extracorporeal Membrane Oxygenation , Adolescent , Carbon Dioxide , Child , Child, Preschool , Extracorporeal Membrane Oxygenation/adverse effects , Humans , Intensive Care Units, Pediatric , Lactic Acid , Retrospective Studies , Treatment OutcomeABSTRACT
Objective To study the correlation between ion concentrations of arterial and central venous blood in blood gas anal‐ysis .Methods This prospective study was performed in 72 patients with a central venous catheter and an artery catheter .A central venous blood sample and an artery blood sample were obtained at the same time after 10 minutes′anesthesia induction .Correlation analysis of the ion concentrations such as Na+ ,K+ ,Ca2+ between those in arterial and central venous blood were performed .The ion concentrations were determined by using GEM Premier 3000 Blood Gas Analysis System .Results The concentration value of Na+ in artery blood was (138 .81 ± 2 .40)mmol/L and in central venous blood was (139 .90 ± 2 .39)mmol/L ,there was significant difference between them(P0 .01) .The values of Ca2+ in artery blood was (1 .30 ± 0 .04)mmol/L and in central venous blood was (1 .32 ± 0 .05)mmol/L ,which were significantly different(P<0 .01) .Conclusion There is significant difference between Na+ ,Ca2+ concentrations in arterial blood gas analysis and those in central venous blood gas analysis ,central venous blood is relatively higher than artery blood ,while no statistically significant difference is observed for K+concentration .
ABSTRACT
Antecedentes: El procedimiento de tomar las muestras de gases arteriales puede ser técnicamente difícil y tiene limitaciones y complicaciones. La colocación de un catéter venoso central es un procedimiento obligado en la unidad de cuidados intensivos (UCI). La toma de muestras de sangre venosa central es mucho más fácil y rápida, y menos complicada. Contexto: Correlación y concordancia entre los valores de pH, PO2, PCO2 y HCC3 en sangre arterial y venosa central de pacientes con ventilación mecánica. Objetivo: El objetivo era evaluar la correlación y la concordancia entre los valores de pH, PO2, PCO2 y HCC3 en sangre arterial y venosa central e inferir si los valores de los gases venosos centrales podían reemplazar la lectura de los gases arteriales. Lugar y diseño: UCI y estudio prospectivo de tipo observacional. Materiales y métodos: Se tomaron en total 100 muestras (50 gases arteriales y 50 gases venosos centrales) de 50 pacientes adultos con distintos procesos patológicos ingresados en la UCI de un hospital universitario. Se ingresaron en el estudio los pacientes normotensos y normotérmicos que requirieron ventilación mecánica y tenían arteria radial palpable. No se repitió ningún paciente en el estudio. En cada paciente se tomó primero la muestra de sangre arterial de la arteria radial y 2 min después se tomó la muestra de sangre venosa central. Se evaluaron la correlación y la concordancia entre los valores de pH, PO2, PCO2 y HCO3 - en sangre arterial y venosa central. Análisis estadístico utilizado: Correlación de Pearson y análisis de Bland Altman. Resultados: Se encontró una correlación significativa entre los valores venosos y arteriales de pH, PO2, PCO2 y HCO3 (r pH = 0,88, p< 0,001; rP02 = 0,358, p<0,05; rPC0(2) = 0,470, p< 0,001 y rHC03 = 0,714, p< 0,001). Se derivaron ecuaciones de regresión para predecir los valores en sangre arterial a partir de los valores en sangre venosa: pH arterial = 0,879 xpH venoso central +114,4 (constante), R² = 0,128, PC0(2) arterial = 0,429 x P0(2) venoso central + 24.627 (constante), R² = 0,2205 y HC0(3) arterial = 1.045 x HC0(3) venoso central+ 3.402 (constante), R² = 0,5101. La diferencia media de la lectura arterial menos la venosa para pH, P0(2), PC0(2), y bicarbonato fue de 0,053 ± 0,014, 56,04 ± 15,74, 2,20 ± 4,4 y 4,30 ± 1,64, respectivamente. Las gráficas de Bland-Altmanpara la concordancia de los valores de pH, P0(2), PC0(2) y bicarbonato mostraron límites de concordancia del 95% de -0,04 a 0,146, -52,51 a 164,59, -26,61 a 31,01 y -7,0 a 15,6, respectivamente. Conclusiones: Hubo una buena correlación entre los valores arteriales y venosos centrales de pH, P0(2), PC0(2) y HCO3 -. Sin embargo, solamente el valor venoso central del pH puede reemplazar el valor arterial del pH.
Background: The procedure for arterial blood sampling can be technically difficult with various limitations and complications. Central venous catheterization is a mandatory procedure in ICU. The sampling of central venous blood is much easier, quicker and less complicated. Context: Correlation and agreement between arterial and central venous blood pH, PO2, PCO2 and HCO3 − values of mechanically ventilated patients. Aims: Aim was to evaluate the correlation and agreement between arterial and central venous blood pH, PO2, PCO2 and HCO3 − values and infer whether central venous blood gas values could replace arterial blood gas values. Settings and design: Intensive Care Unit and prospective observational study. Materials and methods: A total of 100 samples (50 arterial blood gas and 50 central venous blood gas) were collected from 50 adult patients with varied disease process admitted in a medical college ICU. Patients requiring mechanical ventilation having palpable radial artery, normotensive and normothermic were enrolled for the study. No patient was repeated for the study. Arterial blood was collected from radial artery and within 2 minutes central venous blood was withdrawn from the same patient. Correlation and agreement were evaluated between arterial and central venous pH, PO2, PCO2 and HCO3 −. Statistical analysis used: Pearson's correlation and Bland-Altman analysis. Results: The pH, PO2, PCO2 and HCO3 − of CVBG correlated significantly with arterial values (r pH = 0.88, P < 0.001; rPO2 = 0.358, P < 0.05; rPCO2 = 0.470, P < 0.001 and rHCO3 = 0.714, P < 0.001). Regression equations were derived to predict arterial blood values from venous blood values as follows: arterial pH= 0.879×central venous pH+ 0.9422 (constant), arterial PO2 = 0.421×central venous PO2 + 114.4 (constant), R² = 0.128, arterial PCO2 = 0.429×central venous PO2 + 24.627 (constant), R² = 0.2205 and arterial HCO3 = 1.045×central venous HCO3 + 3.402 (constant), R² = 0.5101. The mean arterial minus venous difference for pH, PO2, PCO2, and bicarbonatewas 0.053±0.014, 56.04±15.74, 2.20±4.4 and 4.30±1.64 respectively. Bland-Altman plots for agreement of pH, PO2, PCO2, and bicarbonate showed 95% limits of agreement of −0.04 to 0.146, −52.51 to 164.59, −26.61 to 31.01 and −7.0 to 15.6, respectively. Conclusions: The arterial pH, PO2, PCO2 and HCO3 – values correlatedwell with central venous values. However, only the arterial pH value can replace the central venous pH value.
