The Radiometer ABL300 blood gas analyzer

A modern blood gas analyzer such as the ABL300 directly accepts samples from syringes or capillary tubing. The ABL300 measures pH, carbon dioxide tension, and oxygen tension on 85-μl samples by using specific electrodes, and estimates hemoglobin concentration from the optical density of a nonhemolyzed sample. From these it calculates bicarbonate, standard bicarbonate, total carbon dioxide, base excess, standard base excess, and oxygen saturation and content. The electrodes are automatically calibrated by pumping bicarbonate/phosphate buffer solutions, gas-equilibrated within the thermostatic chamber, into the measuring cuvettes. Samples are preheated in metal tubing before being pumped automatically into the measuring chamber. An open liquid junction to a half-saturated potassium chloride solution is renewed with each sample to complete the pH circuit. The electrodes are housed in a 37°C circulated-air thermostat, which also contains the equilibrators for the calibration solutions. Two known carbon dioxide mixtures are made from pure carbon dioxide by dilution with room air, which is compressed continuously, avoiding the use of premixed calibration gases. The computer determines the derived variables and can correct pH, carbon dioxide tension, and oxygen tension values to a user-defined body temperature. Corrections are made for predicted discrepancy between the measured and true value caused by the small sample size. The corrections depend on the difference in gas tension between sample and rinse solution.     

History of blood gas analysis. VII. Pulse oximetry

Pulse oximetry is based on a relatively new concept, using the pulsatile variations in optical density of tissues in the red and infrared wavelengths to compute arterial oxygen saturation without need for calibration. The method was invented in 1972 by Takuo Aoyagi, a bioengineer, while he was working on an ear densitometer for recording dye dilution curves. Susumu Nakajima, a surgeon, and his associates first tested the device in patients, reporting it in 1975. A competing device was introduced and also tested and described in Japan. William New and Jack Lloyd recognized the potential importance of pulse oximetry and developed interest among anesthesiologists and others concerned with critical care in the United States. Success brought patent litigation and much competition.     

Forehead pulse oximetry compared with finger pulse oximetry and arterial blood gas measurement

Usual monitoring sites for pulse oximetry involve the fingers, toes, ear lobe, and nasal septum. This study examined the performance of a forehead sensor compared with a finger sensor for the pulse oximeter and arterial blood gas (ABG) analysis. Ten healthy adult volunteers and 22 ventilator-dependent patients were studied. The arterial oxygen saturation detected by forehead pulse oximetry (SpO₂) correlated well with finger SpO₂ and arterial oxygen saturation (SaO₂) determined by arterial blood gas analysis in the healthy volunteers. Forehead SpO₂ in mechanically ventilated patients correlated well with finger SpO₂ and SaO₂ when heart rate detected by pulse oximeter differed less than 10% from apical heart rate. Factors that caused a difference in oximeter-detected heart rate and apical heart rate were extensive tissue edema, head movement, and difficulty securing good tape placement. This suggests that when signal strength is weak, causing poor pulse rate detection, there will also be problems associated with accurate SpO₂. The forehead pulse oximeter sensor works well on healthy, well-oxygenated volunteers. Difficulty was experienced when applying and using the sensor on critically ill patients. The reliability of the forehead pulse oximeter sensor has not been established at low saturations.     

History of blood gas analysis. VI. Oximetry

Oximetry, the measurement of hemoglobin oxygen saturation in either blood or tissue, depends on the Lambert-Beer relationship between light transmission and optical density. Shortly after Bunsen and Kirchhoff invented the spectrometer in 1860, the oxygen transport function of hemoglobin was demonstrated by Stokes and Hoppe-Seyler, who showed color changes produced by aeration of hemoglobin solutions. In 1932 in Göttingen, Germany, Nicolai optically recorded the in vivo oxygen consumption of a hand after circulatory occlusion. Kramer showed that the Lambert-Beer law applied to hemoglobin solutions and approximately to whole blood, and measured saturation by the transmission of red light through unopened arteries. Matthes in Leipzig, Germany, built the first apparatus to measure ear oxygen saturation and introduced a second wavelength (green or infrared) insensitive to saturation to compensate for blood volume and tissue pigments. Millikan built a light-weight car “oximeter” during World War II to train pilots for military aviation. Wood added a pneumatic cuff to obtain a bloodless zero. Brinkman and Zijlstra in Groningen, The Netherlands, showed that red light reflected from the forehead could be used to measure oxygen saturation. Zijlstra initiated cuvette and catheter reflection oximetry. Instrumentation Laboratory used multiple wavelengths to measure blood carboxyhemoglobin and methemoglobin is cuvette oximeters. Shaw devised an eight-wavelength ear oximeter. Nakajima and coworkers invented the pulse oximeter, which avoids the need for calibration with only two wavelengths by responding only to the pulsatile changes in transmitted red and infrared light. Lübbers developed catheter tip and cuvette fiberoptic sensors for oxygen tension, carbon dioxide tension, and pH.     

History of blood gas analysis. V. Oxygen measurement

The first biologic use of a platinum cathode for oxygen monitoring was reported in 1938 by Blinks and Skow, who were studying photosynthesis. Their report led to the tissue oxygen studies of Davies, Brink, and Bronk. Clark, by covering cathode and anode with a polyethylene membrane, changed the polarographic cathode from a sensor of oxygenavailability by diffusion to a measure of oxygentension (Po₂) in the solution and thereby facilitated an enormous expansion of the study of the respiratory physiology of blood oxygen after 1956. Clark's electrode led to the development of the present commercial blood gas systems that measure pH, carbon dioxide tension (Pco₂), and Po₂ and calculate many derived variables. Variations on Clark's electrode were designed for in vivo catheter-tip recording; gas phase oxygen monitoring; determining oxygen content of blood by releasing hemoglobin-bound oxygen and measuring Po₂; and determining oxygen consumption in cell cultures (thus replacing Warburg manometry). By reducing the cathode diameter, Staub and others eliminated the need for stirring the blood samples. Concurrent research with amperometric or polarographic oxygen measurement led Hersch to develop the means of determining oxygen content by coulometry in large cells that consumed all the injected oxygen. Methods of applying noninsulating, but protein impermeable, membranes to cathodes and of recessing cathodes into glass permitted measurement of Po₂ in tissues and fluids with microelectrodes.     

Progress in the development of a fluorescent intravascular blood gas system in man

In vitro and in vivo animal studies have shown accurate measurements of arterial blood pH (pHa), carbon dioxide tension (PaCO₂), and oxygen tension (PaO₂) with small intravascular fluorescent probes. Initial human clinical studies showed unexplained intermittent large drops in sensor oxygen tension (PiO₂). Normal volunteers were studied to elucidate this problem. In the first part of this study, the probe and cannula were manipulated and the probe configuration and its position within the cannula were varied. The decreases in PiO₂ were judged to be primarily due to the sensor touching the arterial wall. Retraction of the sensor tip within the cannula eliminated the problem. In the second part of this study, the accuracy of the retracted probe was evaluated in 4 subjects who breathed varying fractions of inspired oxygen and carbon dioxide. The arterial ranges achieved were 7.20 to 7.59 for pH, 22 to 70 mm Hg for PaCO₂, and 46 to 633 mm Hg for PaO₂. Linear regression of 48 paired sensor (i) versus arterial values showed pHi = 0.896 pHa + 0.773 (r = 0.98, SEE = 0.017); PiCO₂ = 1.05 PaCO₂-1.33 (r = 0.98, SEE = 2.4 mm Hg); and PiO₂ = 1.09 PaO₂-20.6 (r = 0.99, SEE = 21.2 mm Hg). Bias (defined as the mean differences between sensor and arterial values) and precision (SD of differences) were, respectively, -0.003 and 0.02 tor pHi, 0.77 and 2.44 mm Hg for PiCO₂, and -2.9 and 25.4 mm Hg for PiO2. The mean in vivo 90% response times for step changes in inspired gas were 2.64, 3.88, and 2.60 minutes, respectively, for pHi, PiCO₂, and PiO₂.     

The evaluation of a new intravascular blood gas monitoring system in the pig

Objective. Our objective was to investigate the accuracy of a new intravascular blood gas sensor, the Paratrend 7 (P7) (Biomedical Sensors Ltd, Pfizer Hospital Products Group, High Wycombe, England) in a porcine model.Methods. A total of 12 sensors were inserted into 10 animals under total intravenous anesthesia. Changes in blood gas chemistry were produced over a wide range by manipulating the inspired oxygen and carbon dioxide concentrations and by adjustments in minute ventilation. Blood gas samples (BGA) were taken and analyzed during periods of stability; the results obtained were compared with the readings from the intravascular sensor.Results. A total of 292 blood gas samples were taken and analyzed for pHa, Paco ₂, andPo ₂; the results were compared with the readings from the intravascular sensor. Correlation coefficients ofr=0.98 forPco ₂ andr=0.99 for Po ₂ were obtained. Analysis of bias and precision as mean±SD of the difference (P7 — BGA) gave the following results: pH bias=–0.03, precision=±0.04;Pco ₂ bias = 0.65 mm Hg, precision=±3.1 mm Hg; andPo ₂ bias=–6.50 mm Hg, precision=±0.6 mm Hg. No problems with clot formation on the sensor were seen, and the sensors did not appear to show the wall effect seen with other systems.Conclusions. The results obtained were well within the requirements for a clinically useful blood gas monitoring system.     

The theoretical effect of carboxyhemoglobin on the pulse oximeter

The relationship between arterial oxygen saturation as measured by the pulse oximeter (SpO₂) and the fractional arterial oxygen saturation (SaO₂) in the presence and absence of carboxyhemoglobin (COHb) has been derived according to the theory of absorption spectroscopy. We find that our theoretically derived correction equation is similar to that found in the technical literature of Nellcor. However, the correction equations presented by Barker and Tremper and the technical literature of Ohmeda differ substantially from our equation when sufficient quantities of reduced hemoglobin are present and the fractional COHb saturation (SaCO) is high. Our approximated equation, derived from the Lambert-Beer law, is SaO₂=SpO₂(1–0.932 SaCO)+0.032 SaCO. The equation of Barker and Tremper is SaO₂=SpO₂–0.9 SaCO. The Nellcor equation is SaO₂=SpO₂(1–SaCO).     

Transcutaneous Pco2 and Po2: A multicenter study of accuracy

A multicentcr study used 756 samples from 251 patients in 12 institutions to compare arterial (PaO₂, PaCO₂) with transcutaneous (PsO₂, PsCO₂) oxygen and carbon dioxide tensions, measured usually at 44°C. Of these samples, 336 were obtained from 116 neonates, 27 from 25 children with cystic fibrosis, and 140 from 40 patients under general anesthesia. Ninety-one patients were between 4 weeks and 18 years of age, 32 were between 18 and 60 years, and 12 were over 60. The ratio of transcutaneous to arterial P(s/a)CO₂ was 1.01 ±0.11 with PaCO₂ less than 30 mm Hg, increasing to 1.04 ±0.08 at PaCO₂ greater than 40 mm Hg. Mean bias and its standard deviation (PsCO₂ — PaCO₂) were + 1.3 ± 3.9 mm Hg in the entire group, + 1.8 ± 4.2 mm Hg in neonates (NS). Bias was +0.2 ± 2.7 mm Hg when PaCO₂ was less than 30 mm Hg (N = 175, NS), 1.0 ± 3.4 with 30 < PaCO₂ < 40 (n = 329,p < 0.001), and +2.04 ± 4.00 mm Hg with 40 < PaCO₂ < 70 (n = 229,p < 0.001). These data suggest that, using transcutaneous PCO₂ monitors with inbuilt temperature correction of 4.5%/‡C, the skin metabolic offset should be set to 6 mm Hg. The linear regression was PsCO₂ =1.052(PaCO₂)-0.56, Sy·x = 3.92, R = 0.929 (n = 756); and PsCO₂ = 1.09(PaCO₂)-1.57, Sy·x = 4.17, R = 0.928 in neonates (n = 336). The use of vasopressors and vasodilators had no significant effect on bias or its standard deviation or on regression slope and intercept (n = 78). In cystic fibrosis patients, bias and standard deviation were 0.0 ± 1.7 mm Hg (n = 27). Under anesthesia, PsCO₂ = 1.07PaCO₂-1.58, with bias and standard deviation = 0.6 ± 3.5 (n = 140). For oxygen, at PaO₂ ≤ 80 the ratio P(s/a)O₂ = 1.05 ± 0.16 in nconates and 0.93 ± 0.21 in older patients, but when PaO₂ > 80, P(s/a)O₂ fell to 0.88 ± 0.18 in neonates and 0.74 ± 0.21 in older patients. The errors were significantly greater (p < 0.001) in older patients than in neonates above but not below 80 mm Hg, and within both groups errors were significantly greater above than below 80 mm Hg.     

AVL COMPACT-3血气分析仪试剂的研制与应用

目的研制AVL COMPACT-3血气分析仪配套应用试剂。方法研制AVL COMP ACT-3血气分析仪配套试剂,并与进口试剂进行比较分析。结果研制试剂与进口试剂的测定结果各组无显着性差异(P>0.05),具有较好的相关性(γ≥0.99),精密度高,稳定性好,临床应用效果满意。结论研制试剂经济实用、稳定可靠、质优价廉,可替代原装进口试剂用于AVL COMPACT-3血气分析仪。     

Accuracy and precision of a new,portable, handheld blood gas analyzer,the IRMA®

Objective. The accuracy and precision of the new IRMA® (Immediate Response Mobile Analysis System, Diametrics, Inc.®, St. Paul, MN) handheld blood gas analyzer was compared with that of two benchtop blood gas analyzers. The IRMA consists of a notebook-sized machine and disposable cartridges, each containing a pH, a CO₂ and an O₂ electrode, and provides bedside (point-of-care) blood gas analysis.Methods. A total of 172 samples (arterial and mined venous) were obtained from 25 informed, consenting patients undergoing cardiopulmonary bypass. The pH, PCO₂ and PO₂ of each sample was determined on four blood gas analyzers: NOVA Statlabs Profile 5 (NOVA Biomedical, Waltham, MA), the ABL-50 (Radiometer, West Lake, OH), and two IRMA machines. Linear regression and bias ± precision were determined, comparing each of the analyzers with the NOVA.Results. All three machines showed a similar, high degree of correlation with the NOVA for pH, PCO₂, and PO₂. The bias and precision of the IRMA machines compared with the NOVA was similar to that of the ABL compared with the NOVA for pH (NOVA:ABL –0.005 ± 0.011; NOVA: IRMA 1 = 0.0026 ± 0.025; NOVA: IRMA 2 = 0.0021 ± 0.025), for PCO₂ (NOVA:ABL = –1.4 ± 1.3 mmHg; NOVA: IRMA 1 = –1.3 ± 1.9 mmHg; NOVA: IRMA 2 = –1.2 ± 2.1 mmHg) and PO₂ (NOVA:ABL = 3.6 ± 21.1 mmHg; NOVA: IRMA 1 = 3.4 = 19.9 mmHg; NOVA: IRMA 2 = 6.3 ± 20.9 mmHg). The bias found for pH, PCO₂, and PO₂ was not affected by extremes of temperature (range 25.5–40°C) or hematocrit (range 11–44%) for any machine.Conclusions. The new technology incorporated in the IRMA blood gas analyzer provides results with an accuracy that is similar to that of benchtop analyzers, but with all of the advantages of point-of-care analysis.     

Accuracy and precision of a new, portable, handheld blood gas analyzer, the IRMA?

Objective. The accuracy and precision of the new IRMA? (Immediate Response Mobile Analysis System, Diametrics, Inc.?, St. Paul, MN) handheld blood gas analyzer was compared with that of two benchtop blood gas analyzers. The IRMA consists of a notebook-sized machine and disposable cartridges, each containing a pH, a CO₂ and an O₂ electrode, and provides bedside (point-of-care) blood gas analysis.Methods. A total of 172 samples (arterial and mined venous) were obtained from 25 informed, consenting patients undergoing cardiopulmonary bypass. The pH, PCO₂ and PO₂ of each sample was determined on four blood gas analyzers: NOVA Statlabs Profile 5 (NOVA Biomedical, Waltham, MA), the ABL-50 (Radiometer, West Lake, OH), and two IRMA machines. Linear regression and bias ± precision were determined, comparing each of the analyzers with the NOVA.Results. All three machines showed a similar, high degree of correlation with the NOVA for pH, PCO₂, and PO₂. The bias and precision of the IRMA machines compared with the NOVA was similar to that of the ABL compared with the NOVA for pH (NOVA:ABL −0.005 ± 0.011; NOVA: IRMA 1 = 0.0026 ± 0.025; NOVA: IRMA 2 = 0.0021 ± 0.025), for PCO₂ (NOVA:ABL = −1.4 ± 1.3 mmHg; NOVA: IRMA 1 = −1.3 ± 1.9 mmHg; NOVA: IRMA 2 = −1.2 ± 2.1 mmHg) and PO₂ (NOVA:ABL = 3.6 ± 21.1 mmHg; NOVA: IRMA 1 = 3.4 = 19.9 mmHg; NOVA: IRMA 2 = 6.3 ± 20.9 mmHg). The bias found for pH, PCO₂, and PO₂ was not affected by extremes of temperature (range 25.5–40°C) or hematocrit (range 11–44%) for any machine.Conclusions. The new technology incorporated in the IRMA blood gas analyzer provides results with an accuracy that is similar to that of benchtop analyzers, but with all of the advantages of point-of-care analysis.     

The effects of bronchodilator-inhaler aerosol propellants on respiratory gas monitors

Spurious readings from a mass spectrometer have been reported following the administration of aerosol bronchodilators. We quantified the response of various respiratory gas analyzers to the aerosol propellant of albuterol inhalant (Proventil). The mass spectrometer systems tested, two Advantage systems, a SARA system, and a Model 6000 Ohmeda system, all displayed artifactual readings in response to the albuterol propellant. Each metered dose of the Proventil brand of albuterol contains 4 ml of Freon 11 (trichloromonofluoromethane) and 11 ml of Freon 12 (dichlorodifluoromethane). The concentration of propellant was expressed in doses/L, where each liter of gas contains 0.4 vol % of Freon 11 and 1.1 vol % of Freon 12 per dose. In proportion to the concentration of albuterol propellant, the two Advantage systems showed substantial readings of isoflurane (%) when no isoflurane was present (13% and 16% per dose/L) and reduced readings of enflurane (–8% and –10% per dose/L) and carbon dioxide (CO₂) (–3 and +5 mm Hg per dose/L). The SARA system showed substantial CO₂ readings when no CO₂ was present (5 mm Hg per dose/L) and displayed small enflurane readings (0.1% per dose/L) when no enflurane was present. The Model 6000 unit showed CO₂ readings when no CO₂ was present (5 mm Hg per dose/L). Neither the Raman spectrometer, the infrared spectrometers, nor the piezoadsorptive analyzer we tested showed an artifactual effect of albuterol propellant on any of its readings. Simulation and clinical tests demonstrated that a single dose of albuterol propellant into a breathing circuit at the onset of inspiration resulted in concentrations of 0.8 and 0.3 dose/L, respectively. The phenomenon may be clinically useful, by allowing the anesthetist to verify the uptake of an inhalant into a patient.     

The effects of bronchodilator-inhaler aerosol propellants on respiratory gas monitors

Spurious readings from a mass spectrometer have been reported following the administration of aerosol bronchodilators. We quantified the response of various respiratory gas analyzers to the aerosol propellant of albuterol inhalant (Proventil). The mass spectrometer systems tested, two Advantage systems, a SARA system, and a Model 6000 Ohmeda system, all displayed artifactual readings in response to the albuterol propellant. Each metered dose of the Proventil brand of albuterol contains 4 ml of Freon 11 (trichloromonofluoromethane) and 11 ml of Freon 12 (dichlorodifluoromethane). The concentration of propellant was expressed in doses/L, where each liter of gas contains 0.4 vol % of Freon 11 and 1.1 vol % of Freon 12 per dose. In proportion to the concentration of albuterol propellant, the two Advantage systems showed substantial readings of isoflurane (%) when no isoflurane was present (13% and 16% per dose/L) and reduced readings of enflurane (−8% and −10% per dose/L) and carbon dioxide (CO₂) (−3 and +5 mm Hg per dose/L). The SARA system showed substantial CO₂ readings when no CO₂ was present (5 mm Hg per dose/L) and displayed small enflurane readings (0.1% per dose/L) when no enflurane was present. The Model 6000 unit showed CO₂ readings when no CO₂ was present (5 mm Hg per dose/L). Neither the Raman spectrometer, the infrared spectrometers, nor the piezoadsorptive analyzer we tested showed an artifactual effect of albuterol propellant on any of its readings. Simulation and clinical tests demonstrated that a single dose of albuterol propellant into a breathing circuit at the onset of inspiration resulted in concentrations of 0.8 and 0.3 dose/L, respectively. The phenomenon may be clinically useful, by allowing the anesthetist to verify the uptake of an inhalant into a patient.     

Clinical assessment of a flow-through fluorometric blood gas monitor

We performed an observational study to evaluate a flow-through fluorometric instrument (Gas-STAT) that continuously measures the carbon dioxide tension (PCO ₂), oxygen tension (PO ₂), and pH of blood in the cardiopulmonary bypass circuit. Setup and calibration of the instrument typically required 20 minutes. During bypass, 129 blood samples were drawn from 16 patients for comparison with conventional measurements obtained with a blood gas machine. Data for each variable, within each sensor, were analyzed by linear regression. The ranges of the standard errors of the estimate were 0.7 to 4.2 mm Hg forPCO ₂, 18.3 to 78.7 mm Hg for the highPO ₂ range, 1.4 to 7.1 mm Hg for the lowPO ₂ range, and 0.008 to 0.049 for pH. The regression lines differed from the identity line (P<0.05) in at least one variable in most patients, and large deviations from the line of identity in both slope and intercept were common. Among 58 sensors evaluated, failures occurred in 5 (2.9%) of the 174 optodes, and minor leakage occurred in 2 (3.4%) of the flow-through cells. We conclude that although this flow-through fluorometric instrument is an adequate monitor of trends in blood gases during cardiopulmonary bypass, it is not accurate enough to supplant conventional laboratory measurements.Supported in part by grant HL-30881 from the National Institutes of Health.Presented in part at the National Meeting of the American Association for Clinical Chemistry, Atlanta, GA, July 1985, and at the Annual Meeting of the American Society of Anesthesiologists, San Francisco, CA, Oct 1985.The authors thank Dr Brenda Townes for supervising the bypass outcome study of which this was a component; our surgical colleagues, Drs Margaret Allen, Peter McKeown, and Gregory Misbach for allowing us to study their patients; cardiopulmonary perfusionists Gary Tarter, David Anderson, Roland Alberto, and Debora Bley; and the technologists of the Department of Laboratory Medicine.Gas-STAT and disposables were provided by Cardiovascular Devices, Inc.     

Comparison of the sphygmooscillographic method with the direct and auscultatory methods of measuring blood pressure

Objective. The objective of this study was to compare blood pressure (BP) measured by the sphygmooscillographic method with that measured by the direct and auscultatory methods.Methods. In 15 adult patients undergoing cardiac surgery, blood pressure was measured by the sphygmooscillographic and direct methods simultaneously on the same upper extremity. In another group of 86 children and 11 adults, blood pressure was measured by the sphygmooscillographic and auscultatory methods simultaneously, with one cuff. For the sphygmooscillographic measurement, we used sphygmomanometer-S, which measures blood pressure on the basis of the amplitude height (oscillometric) and the morphology (sphygmographic) of pulse waves recorded by a transducer placed in the cuff.Results. The systolic and diastolic blood pressure measured by the sphygmooscillographic method were both 2 mm Hg higher than those from the direct method; the mean blood pressure was 0.6 mm Hg higher. These differences were not significant. Compared with the auscultatory method, sphygmooscillographic systolic values were higher by 7 mm Hg, while diastolic values were lower by 9 mm Hg. These differences were significant.Conclusions. Blood pressure measurements obtained by the sphygmooscillographic method correlate well with the direct method for measuring blood pressure in children and adults; but, they do not correlate well with the auscultatory method.     

血气分析在婴幼儿急性感染性腹泻诊疗中的临床应用

目的探讨血气分析在婴幼儿急性感染性腹泻诊疗中的临床应用价值。方法应用美国MEDICA Easy BloodGas血气分析仪及配套试剂包,对急性感染性腹泻婴幼儿进行治疗前和治疗后12小时的动脉血血气分析并作比较。结果急性感染性腹泻的患儿大都存在明显的代谢性酸中毒。治疗前和治疗后12小时的动脉血血气分析比较有显著性差异。结论血气分析对临床医生及时掌握和纠正急性感染性腹泻婴幼儿的酸碱平衡紊乱有着十分重要的指导意义。     

Simple and noninvasive indicator of pulmonary gas exchange impairment using pulse oximetry

We postulated that the fractional inspired oxygen concentration (FiO₂) required to achieve a certain value of arterial oxygen saturation (SaO₂) can be used as an indicator of pulmonary gas exchange impairment in patients during mechanical ventilation. We tested this hypothesis in 20 patients. By reducingFiO₂ in increments of 10 vol% of capacity while monitoring SaO₂ with pulse oximetry, we could determineFi ₉₈,Fi ₉₇,Fi ₉₆, andFi ₉₅; that is, the yields 98, 97, 96, and 95% SaO₂, respectively. On the basis of our data, we choseFi ₉₈ as the most appropriate index, as an SaO₂ of 97% or below could not be achieved even with a lowFiO₂ in some of the patients. To test the significance of the newly proposed index, we comparedFi ₉₈ with the alveolar-arterial oxygen tension difference, P(A–a)O₂, and with the respiratory index, which are routinely used elsewhere. The correlation betweenFi ₉₈ and P(A–a)O₂ was excellent: P(A–a)O₂=490.5*Fi ₉₈+117.2 with a correlation coefficient of 0.906 (P<0.01).Fi ₉₈ also correlated significantly with the respiratory index: respiratory index=4.354*Fi ₉₈–0.776 (r=0.889,P<0.01).We conclude thatFi ₉₈ may be used as a simple index for the rough estimation of pulmonary gas exchange impairment without the need for invasive procedures. However, further studies are needed to confirm the validity of our method in hemodynamically unstable patients or when other brands of pulse oximeters are used.     

In vitro assessment of a flow-through fluorometric blood gas monitor

The Gas-STAT blood gas monitor uses fluorometric techniques to continuously monitor blood gas tensions and acid-base status in the extracorporeal perfusion circuit during cardiac surgery. We evaluated the in vitro performance of this instrument by using a tonometry loop to simulate the clinical environment and to provide controlled gas tensions and pH in the circulating fluid. In this article we report the in vitro study in which 35 Gas-STAT blood gas sensors were used to assess the precision, stability, response time, and specificity of the instrument and to confirm the sterile integrity of its flow-through cells. The blood gas monitor exhibited precision values for pH, carbon dioxide tension (PCO ₂), and oxygen tension (PO ₂) of 0.1%, 1.3%, and 1.0%, respectively; stabilities were 0.002 units/h for pH, 0.5 mm Hg/h forPCO ₂, and 1.4 mm Hg/h forPO ₂; time constants (, a response to within 1/e of a new gas tension, 63%) were 81 seconds forPCO ₂ and 72 seconds forPO ₂. No significant interference was detected in in vitro tests of 30 drugs and metabolites typically encountered during cardiac surgery. Bacterial challenge of the flow-through cell membranes showed that they provide an effective barrier isolating the sensors from contaminants in the fluid path. Our quality control consisted of measurement of a midrange gas standard as an unknown immediately following sensor calibration; this simple program is proposed as a complement to the manufacturer's operating procedures. This monitor is shown to be a valuable indicator of trends in blood gas tensions and acid-base levels in circulating fluids; however, as reported in a companion article, in vivo its accuracy in the flow-through environment does not match its calibration performance and is insufficient to supplant traditional analyses of discrete samples for absolute measurements.Supported in part by grant HL-30881 from the National Institutes of Health.Presented in part at the National Meeting of the American Association of Clinical Chemistry, Atlanta, GA, Jul 1985, and at the Annual Meeting of the American Society of Anesthesiologists, San Francisco, CA, Oct 1985.The authors thank Michael Nessly, Dept of Anesthesiology, for helpful discussions; Brenda Wagner for technical assistance; Scott Coble, Applied Mathematics graduate student, for response-time analysis; Linda Johnson and Gertrude Schmidt, Microbiology Division, for valuable assistance in the sterility assessment; and Del Landicho for computational expertise.Gas-STAT and disposables were provided by Cardiovascular Devices, Inc.     

全自动生化分析仪与血气分析仪电解质测定结果比较

目的研究全自动生化分析仪与血气分析仪电解质测定结果是否存在差异。方法两台仪器分别进行批内、批间精密度实验后,对62例住院患者同时采集肝素抗凝动脉血和无抗凝动脉血,采用Omin-C型血气分析仪分析动脉血电解质,未抗凝动脉血离心分离出血清在强生Vitros-350全自动干化学分析仪上测定电解质,实验数据用SPSS11.0进行配对样本t检验。结果两台仪器批内、批间精密度实验结果符合CLIA′88允许误差要求,但血气分析仪测定的钾、钠、氯值均低于Vitros-350测定结果,差异有统计学意义(P<0.01)。结论血气分析仪检测标本为肝素抗凝全血,与传统生化分析有显著差异,有必要建立适合血气分析仪的电解质检测生物参考区间。