Performance of the epoc Blood Analysis System
Performance of the epoc Blood Analysis System
The epoc system consists of test cards containing the sensors, a wireless card reader, and a personal data assistant (PDA) for data analysis. The test cards use selective electrode potentiometry to measure pH, pCO2, sodium, potassium, and ionized calcium. Glucose, pO2, and lactate are measured amperometrically. Hematocrit is determined conductometrically. Bicarbonate, total carbon dioxide, base excess, oxygen saturation, and hemoglobin are calculated.
The epoc card reader is a battery-powered portable device with a barcode scanner. Insertion of a test card into the reader releases calibration fluid over the sensors. After completion of a 190-second calibration cycle, the user has seven minutes to apply a patient sample to the card. The reader then sends the sensor data via Bluetooth to the PDA. The PDA is a handheld computer that calculates analytical values from the data sent by the reader. The host displays the results and sends them via Wi-Fi to the hospital network.
Data were collected at the Columbia University Medical Center (CUMC) campus of New York-Presbyterian Hospital (New York, NY), a tertiary care academic medical center with more than 1,000 adult and pediatric inpatient beds. Active transplant, cardiac, and general surgical services; multiple intensive care units (ICUs); and two emergency departments require fast turnaround times for blood gas results.
In 1995, the CUMC introduced i-STAT POC devices for measuring blood gases, electrolytes, and hemoglobin in adult and pediatric ICUs, operating rooms, and the pediatric emergency department. Initially, two types of instruments were used: 140 blood analysis modules (BAMs) built into patient monitors in the adult ICUs and 52 portable i-STAT clinical analyzers (PCAs) used in the pediatric emergency department, pediatric ICUs, and the operating rooms. Over time, the hospital discontinued using BAMs, and by 2008, only the portable PCAs were in use. Departments that did not use i-STAT instrumentation sent samples requiring blood gas analysis to the centralized core laboratory, where they were analyzed with Nova CCX instruments (Nova Biomedical, Waltham, MA).
In 2010, the hospital replaced the i-STAT instrumentation with epoc blood analysis systems. In addition, epoc instruments were deployed in several locations that were not previously using i-STAT devices for blood gas measurements. A total of 156 epoc systems were deployed over approximately three months. Each system consisted of a reader and a PDA. The vendor and the POC service staff trained more than 800 nurses, 300 physicians, 70 respiratory therapists, and 20 perfusionists. Training sessions lasted approximately one hour and included hands-on demonstrations of the instruments. Initial competency was assessed by a written quiz and direct observation of testing using quality control (QC) material. Competency was reassessed six months after initial training using direct observation of patient testing combined with an online review course and quiz.
Precision was assessed by analyzing three levels of aqueous QC material (Eurotrol BGEM control fluids [Eurotrol, Burlington, MA] and Mission Blood Gas and Electrolytes Control Calibration Verification Set [Diamond Diagnostics, Holliston, MA]) on 10 consecutive days. The precision data shown in this article are based on 37 readers. Three levels of QC material were run on each reader every day for 10 consecutive days, giving 370 measurements for each level of QC material. The means, standard deviations, and coefficients of variation were then calculated with Excel 2008 (Microsoft, Redmond, WA).
Linearity studies were performed on all 156 instruments with five levels of commercially available linearity material (RNA Medical, Devens, MA), with three repetitions of each level. Acceptability criteria were a slope between 0.9 and 1.1 and a recovery between 90% and 110%.
Correlation studies were performed on 15 randomly selected epoc systems, representing 10% of the analyzers purchased. Residual sample volume from patient samples collected for clinical analysis on Nova CCX instruments in the core laboratory was used for these studies. Samples collected in lithium heparin syringes were of arterial, mixed venous, and venous types. Approximately 40 patient samples were analyzed on each of the 15 epoc systems and on two i-STAT instruments; results from the 15 epoc systems were compared with those from the i-STAT analyzers and Nova CCX instruments.
After implementation of the epoc blood gas analyzer, the following quality assurance (QA) data were collected with the Electronic Data Manager (Alere) and analyzed in Excel 2008 (Microsoft): number of tests performed, number of tests reported in the electronic medical record, test card failure rate, and error codes associated with test card failures. The number of instruments replaced during the first year of use was tracked manually. Comparable QA data for i-STAT had been collected previously with an i-STAT Central Data Station. Data for the core laboratory were obtained with the laboratory information systems (Mysis, Tucson, AZ, and Cerner Millennium, North Kansas City, MO).
Materials and Methods
epoc Blood Analysis System
The epoc system consists of test cards containing the sensors, a wireless card reader, and a personal data assistant (PDA) for data analysis. The test cards use selective electrode potentiometry to measure pH, pCO2, sodium, potassium, and ionized calcium. Glucose, pO2, and lactate are measured amperometrically. Hematocrit is determined conductometrically. Bicarbonate, total carbon dioxide, base excess, oxygen saturation, and hemoglobin are calculated.
The epoc card reader is a battery-powered portable device with a barcode scanner. Insertion of a test card into the reader releases calibration fluid over the sensors. After completion of a 190-second calibration cycle, the user has seven minutes to apply a patient sample to the card. The reader then sends the sensor data via Bluetooth to the PDA. The PDA is a handheld computer that calculates analytical values from the data sent by the reader. The host displays the results and sends them via Wi-Fi to the hospital network.
Evaluation Site
Data were collected at the Columbia University Medical Center (CUMC) campus of New York-Presbyterian Hospital (New York, NY), a tertiary care academic medical center with more than 1,000 adult and pediatric inpatient beds. Active transplant, cardiac, and general surgical services; multiple intensive care units (ICUs); and two emergency departments require fast turnaround times for blood gas results.
In 1995, the CUMC introduced i-STAT POC devices for measuring blood gases, electrolytes, and hemoglobin in adult and pediatric ICUs, operating rooms, and the pediatric emergency department. Initially, two types of instruments were used: 140 blood analysis modules (BAMs) built into patient monitors in the adult ICUs and 52 portable i-STAT clinical analyzers (PCAs) used in the pediatric emergency department, pediatric ICUs, and the operating rooms. Over time, the hospital discontinued using BAMs, and by 2008, only the portable PCAs were in use. Departments that did not use i-STAT instrumentation sent samples requiring blood gas analysis to the centralized core laboratory, where they were analyzed with Nova CCX instruments (Nova Biomedical, Waltham, MA).
In 2010, the hospital replaced the i-STAT instrumentation with epoc blood analysis systems. In addition, epoc instruments were deployed in several locations that were not previously using i-STAT devices for blood gas measurements. A total of 156 epoc systems were deployed over approximately three months. Each system consisted of a reader and a PDA. The vendor and the POC service staff trained more than 800 nurses, 300 physicians, 70 respiratory therapists, and 20 perfusionists. Training sessions lasted approximately one hour and included hands-on demonstrations of the instruments. Initial competency was assessed by a written quiz and direct observation of testing using quality control (QC) material. Competency was reassessed six months after initial training using direct observation of patient testing combined with an online review course and quiz.
Evaluation Protocol
Precision was assessed by analyzing three levels of aqueous QC material (Eurotrol BGEM control fluids [Eurotrol, Burlington, MA] and Mission Blood Gas and Electrolytes Control Calibration Verification Set [Diamond Diagnostics, Holliston, MA]) on 10 consecutive days. The precision data shown in this article are based on 37 readers. Three levels of QC material were run on each reader every day for 10 consecutive days, giving 370 measurements for each level of QC material. The means, standard deviations, and coefficients of variation were then calculated with Excel 2008 (Microsoft, Redmond, WA).
Linearity studies were performed on all 156 instruments with five levels of commercially available linearity material (RNA Medical, Devens, MA), with three repetitions of each level. Acceptability criteria were a slope between 0.9 and 1.1 and a recovery between 90% and 110%.
Correlation studies were performed on 15 randomly selected epoc systems, representing 10% of the analyzers purchased. Residual sample volume from patient samples collected for clinical analysis on Nova CCX instruments in the core laboratory was used for these studies. Samples collected in lithium heparin syringes were of arterial, mixed venous, and venous types. Approximately 40 patient samples were analyzed on each of the 15 epoc systems and on two i-STAT instruments; results from the 15 epoc systems were compared with those from the i-STAT analyzers and Nova CCX instruments.
Quality Assurance Data
After implementation of the epoc blood gas analyzer, the following quality assurance (QA) data were collected with the Electronic Data Manager (Alere) and analyzed in Excel 2008 (Microsoft): number of tests performed, number of tests reported in the electronic medical record, test card failure rate, and error codes associated with test card failures. The number of instruments replaced during the first year of use was tracked manually. Comparable QA data for i-STAT had been collected previously with an i-STAT Central Data Station. Data for the core laboratory were obtained with the laboratory information systems (Mysis, Tucson, AZ, and Cerner Millennium, North Kansas City, MO).