Evaluation of factors affecting relationship between transcutaneous PO2 and probe temperature.

Several studies on transcutaneous O2 probes have shown that the transcutaneous PO2 increases to approximately 80% of the arterial PO2 when the probe is heated to 44 degrees C. It is not known whether this result reflects near-complete thermic arterialization or rather other factors such as the temperature-linked right shift of the hemoglobin O2-binding curve. In many clinical applications of transcutaneous probes the use of 44 degrees C is a major disadvantage because of the risk of skin burns. The development of new probes operating at lower temperatures is hampered by the lack of data on the temperature dependence of the factors influencing the relationship between the transcutaneous PO2 and the probe temperature. The present study attempts to estimate the temperature dependence of 1) the degree of arterialization of the blood in the skin capillaries, 2) the PO2 difference across the epidermis caused by the diffusion gradient and the epidermal O2 consumption, and 3) the arteriovenous saturation difference over the skin capillaries. The estimation is based on simultaneously measured transcutaneous PO2, PCO2, and argon partial pressure (PAr) values at seven different probe temperatures. The transcutaneous PCO2 is assumed equal to the mean capillary PCO2, which is used to calculate the mean capillary PO2 by the aid of a skin model. The O2 diffusion gradient is estimated from the transcutaneous PAr, and the PO2 difference caused by the epidermal O2 consumption is set equal to the difference between the mean capillary and transcutaneous PO2 less the partial pressure difference caused by the diffusion gradient. The degree of arterialization was found to be 53% at 38 degrees C and 65% at 44 degrees C. The partial pressure difference caused by the epidermal O2 consumption decreased from 33 Torr at 38 degrees C to 6 Torr at 44 degrees C. The PO2 difference across the epidermis caused by the diffusion gradient was 7 Torr at 38 degrees C and 5 Torr at 44 degrees C. The arteriovenous saturation difference fell from 31% at 38 degrees C to 12% at 44 degrees C.

Error due to dead-space admixture in single-breath method for estimation of pulmonary blood flow.

The single-breath method of Kim et al. (J. Appl. Physiol. 21: 1338-1344, 1966) for the estimation of pulmonary blood flow is based on a single-alveolus lung model for which an analytical relationship has been established between the kinetic behavior of the alveolar O2 and CO2 tensions and the pulmonary blood flow. The analysis is based on the assumption that the dead-space contribution to the expirate is negligible after expiration of a predefined volume. We have examined the influence of this assumption on the estimation of pulmonary blood flow by computer simulation in a lung model that incorporates deadspace contribution to the expirate. Data on the fractional contribution of the dead space to the expired gas were obtained from Tsunoda et al.'s study (J. Appl. Physiol. 32: 644-649, 1972) on the emptying pattern of normal adult lungs. The results show that failure to take account of the dead-space contribution can cause an underestimation in the pulmonary blood flow of greater than 30%. The error can be reduced by ignoring the first part of the expiration but only at the cost of an increase in the sensitivity of the single-breath method to measurement noise. This property of the system is demonstrated experimentally. The error due to dead-space admixture depends on the total volume of dead-space gas, the measurement noise, the pulmonary blood flow, and the emptying characteristics of the dead-space compartment during expiration. In normal subjects it is possible to optimize the experimental design so that the systematic error is less than 5% and the coefficient of variation is less than 10%.(ABSTRACT TRUNCATED AT 250 WORDS)

Estimation of blood flow distribution in skeletal muscle from inert gas washout.

A new method is evaluated for the estimation of blood flow-to-volume distribution in skeletal muscle from inert gas washout kinetics. Acetylene washout from the isolated, blood-perfused canine gracilis muscle was measured continuously with a blood gas catheter in combination with a mass spectrometer. The washout curves were transformed to flow-to-volume ratio distributions by means of a 50-compartment model. The algorithm fits the expression for the washout curve derived from the model by a least-squares method with enforced smoothing. The algorithm was evaluated using computer simulations in which artificial washout curves were generated by a multicompartment model with a known flow distribution. A wide range of given flow distributions could be recovered from the simulated data. The data were also analyzed using a linear programming technique. Analysis of the experimental data with the least-squares method showed that there is considerable heterogeneity in the distribution of perfusion in resting gracilis muscle. The distribution is characterized by at least two modes and a single compartment with a very low perfusion-to-volume ratio. Experimental noise made it impossible to obtain feasible flow distributions by means of linear programming.

A modified photo- and magnetoacoustic multigas analyzer applied in gas exchange measurements.

The feasibility of replacing a conventional mass spectrometer (MS) with a specially modified multicomponent (O2, CO2, Freon 22, and SF6) acoustic infrared and paramagnetic (IR/PM) gas analyzer in inert gas-rebreathing and metabolic gas exchange measurements has been investigated. Rebreathing variables were determined simultaneously with the MS and IR/PM analyzers in duplicate measurements at rest and during submaximal exercise in 10 subjects. The differences (means +/- SD, IR/PM - MS) were 0.028 +/- 0.048 liters [functional residual capacity (FRC)], 0.18 +/- 0.38 l/min [cardiac output (Qc)], -0.006 +/- 0.030 l/min [O2 consumption (VO2)], and -33 +/- 108 ml [combined lung tissue and capillary blood volume (Vti,c)]. The coefficients of variation on repeated estimates were 5.8% (FRC), 5.4% (Qc), 6.2% (VO2), and 17% (Vti,c) with the IR/PM analyzer and 5.9% (FRC), 4.2% (Qc), 5.0% (VO2), and 9.8% (Vti,c) with the MS. The differences (IR/PM - MS) obtained in mixed-expirate measurements were -0.006 +/- 0.020 l/min (VO2) and 0.020 +/- 0.021 l/min (CO2 production). Breath-by-breath estimates of VO2 and CO2 production with the IR/PM analyzer were, on average, 2.4 and 4.4% higher than the MS estimates, respectively. Our results demonstrate that the IR/PM gas analyzer, when appropriately modified, can substitute for a complex MS in a variety of noninvasive pulmonary gas exchange measurements.

Precision and accuracy of a noninvasive inert gas washin method for determination of cardiac output in men.

Stout et al. (J. Appl. Physiol. 38:913-918, 1975) suggested an open-circuit multibreath (MB) inert gas method for determining pulmonary capillary blood flow (Qc) in anesthetized dogs receiving artificial ventilation. In the present work we investigated the accuracy and reproducibility of the MB method in nine healthy human subjects at spontaneous ventilation, and we compared the MB method with the inert gas rebreathing (RB) method. Qc was calculated at rest and during exercise at 50 or 100 W, and experimental errors were evaluated in computer simulations of a two-alveoli lung model. The calculated mean Qc values of the MB method were 5.56 +/- 1.23 (SD), 10.02 +/- 0.78, and 13.2 +/- 0.84 l/min, and the mean difference (MB Qc - RB Qc) was not significantly different (P > 0.05). The variation in Qc of the MB method was found to be significantly larger than that in Qc of the RB method (P < 0.01). Random measurement errors and uneven distribution of ventilation contributed to the experimental errors. We conclude that the MB method is inferior to the RB method but that the MB method may be useful under exercise conditions.

In vivo quantitation of carbonic anhydrase and band 3 protein contributions to pulmonary gas exchange.

The contributions to pulmonary gas exchange of red blood cell (RBC) membrane band 3 protein HCO3(-)-Cl- exchange and carbonic anhydrase- (CA) catalyzed HCO3- dehydration have never been determined directly in the whole animal. We utilized an experimental and model approach to measure these by analysis of phase III exhaled CO2 and O2 profiles in anesthetized dogs. In this method, we inhibit RBC membrane band 3 protein and cytoplasmic CA in RBCs passing the pulmonary capillaries and lung vascular luminal membrane-bound CA during a single ventilatory cycle. This is achieved with appropriately timed right atrial infusions of 4,4'-dinitrostilbene-2,2'-disulfonate (DNDS) to inhibit band 3 protein, ethoxzolamide (a lipophilic CA inhibitor with rapid membrane penetrance) to inhibit RBC and lung tissue CA, and benzolamide (an extremely hydrophilic CA inhibitor with virtually no penetrance into RBC cytoplasm) to inhibit only lung vascular luminal membrane CA. DNDS caused a 15% reduction in CO2 production (VCO2) without any change in O2 consumption (VO2). The addition of benzolamide to DNDS did not cause any further decrease in VCO2. Inhibition of RBC CA by ethoxzolamide caused a 67% reduction in VCO2 and a 11.5% reduction in VO2. Inhibition of lung vascular CA by benzolamide alone caused no statistically significant changes in either VCO2 or VO2. These results are in general agreement with in vitro data and model calculations. The only exceptions are the higher than predicted effect of RBC CA inhibition on VO2 (Bohr effect) and the lack of any contribution to CO2 transfer in the dog by lung vascular CA with access to plasma as a possible consequence of an endogenous plasma CA inhibitor.

Contribution of continuing gas exchange to phase III exhaled PCO2 and PO2 profiles.

Changes in PCO2 and PO2 during expiration have been ascribed to simultaneous gas exchange, but other factors such as ventilation-perfusion inhomogeneity in combination with sequential emptying may also contribute. An experimental and model approach was used to study the relationship between gas exchange and changes in expired PCO2 and PO2 in anesthetized dogs during prolonged high tidal volume expirations. Changes in PCO2 and PO2 were quantified by taking the area bounded by the sloping exhalation curve and a line drawn horizontally from a point where the Fowler dead space plus 250 ml had been expired. This procedure is similar to using the slope of the exhalation curve but it circumvents problems caused by nonlinearity of the PCO2 and PO2 curves. The gas exchange components of the CO2 and O2 areas were calculated using a single-alveolus lung model whose input parameters were measured in connection with each prolonged expiration. The relationship between changes in experimental CO2 areas caused by sudden reductions in mixed venous PCO2 (produced by right atrial infusions of NaOH) and those calculated by the model was also studied. In seven dogs, calculated CO2 and O2 areas were 13% higher and 25% lower than the respective experimental areas, but interindividual variations were large. Changes in experimental CO2 areas caused by step changes in mixed venous PCO2 were almost identical to changes in the calculated areas. We conclude that the changes in PCO2 and PO2 during expiration cannot be explained solely by gas exchange. However, the single-alveolus lung model accurately predicts changes in the CO2 exhalation curve caused by alterations in the alveolar CO2 flow.

Effects of variation in systemic blood pressure on intraosseous pressure, PO2, and PCO2.

Several regulatory mechanisms have been shown to influence the intraosseous circulation. The influence of general hypovolemia on bone circulation and possible regulatory effects were investigated by recording the intraosseous pressure and PO2 and PCO2 continuously by mass spectrometry in rabbits. Hypovolemia was induced by repeated bleedings. The intraosseous and arterial pressures were found to be linearly related. The intraosseous PO2 already decreased after the first step of bleeding and decreased more than 50% of the initial value after an average blood loss of 40 ml. The intraosseous PCO2 showed an opposite pattern. The experimental PO2 vs. arterial pressure curves were similar to curves obtained by a computer simulation that assumes the blood flow to be proportional to the intraosseous pressure.

Transcephalic electrical impedance provides a means for quantifying pulsatile cerebral blood volume changes following head-up tilt.

Transcephalic electrical impedance (delta Z) was used to assess pulsatile cerebral blood volume changes while tilting nine premature (30-34 weeks) infants 20 degrees head up. High-frequency (1.50-4.00 Hz) delta Z variability decreased 27% while heart rate did not show any change. We would like to suggest that the variability of transcephalic electrical impedance analysed with spectral estimation seems to provide a means for quantifying alterations in cerebral circulation.

Cerebral circulation assessed by transcephalic electrical impedance during the first day of life--a potential predictor of outcome?

Pulsatile changes in intracranial blood volume (transcephalic electrical impedance, delta Z), arterial blood pressure (aBP) and respiration were studied during the first day after birth in 42 neonates with a birth weight of 1040-3850 g and gestational age of 26-36 weeks. The neurological outcome was assessed at 1 year of age to study the predictive ability of delta Z. delta Z, ECG, respirogram and direct aBP were recorded at 8-h intervals. Outcome was adverse in seven infants of whom two died from severe peri-intraventricular haemorrhage. PCO2 was higher (6.2 kPa) in the infants with adverse outcome than in those infants with favourable outcome (5.0 kPa) (P = 0.004). Blood glucose (4.5 vs. 3.3 mmol/l, P = 0.030) and first day administration of fluid (80 vs. 63 ml/kg/day, P = 0.003) behaved, respectively. Of the infants receiving dopamine therapy, 60% had adverse outcome while only 11% of those not receiving dopamine had adverse outcome (P = 0.016). Of the infants with high diastolic blood pressure levels, 32% had adverse outcome, while none with low diastolic blood pressure levels had adverse outcome (P = 0.031). Spectral analysis was used to examine signal variabilities in the frequency domain. During the first 24 h of life, the variabilities of aBP and respiration were equal in all the infants. The high-frequency delta Z signal variability (1.50-4.00 Hz, heart rate level) was found to be lower in the infants with adverse outcome (330 units) than in the infants with favourable outcome (1280 units, P = 0.017). The low delta Z variability allowed us to assume that there is a decrease of pulsatile cerebral blood flow (CBF) in the infants with adverse outcome. We speculate that this may result from the 'no reflow phenomenon', increased tissue pressure due to ischaemia and/or PIVH, the 'brain sparing effect' or constriction of main cerebral arteries due to increased pressure support or metabolic factors (PCO2, glucose). We believe that transcephalic impedance provides a potential cot-side method for monitoring cerebral circulation in the neonatal period with an ability to predict outcome.

Increased amplitude modulation of continuous respiration precedes sudden infant death syndrome -detection by spectral estimation of respirogram.

The immaturity of the control of the autonomic nervous system has been suggested as one of the key factors in the pathophysiology of sudden infant death syndrome (SIDS). Therefore, the attenuated control of respiration may also cause more slow oscillatory breathing among infants at risk of SIDS. In this study, patterns of respiratory activity (RAV) and heart rate variability (HRV) were examined in Medilog-records prospectively obtained from 22 tape recordings made on 16 babies subsequently suffering from SIDS and from 22 matched control babies. A total of 248 signal segments, 120 s in duration, representing the state of regular breathing were visually selected for further analysis. The digitised signal sets were detrended, Fast-Fourier-transformed and autospectra as well as cross-spectra for the HRV and HRV were computed. The RAV and HRV were examined at two spectral bands: (1) a low frequency (LF) band 0.03-0.17 Hz (1.8-10 cycles/min) and (2) a high frequency (HF) band 0.3-1.3 Hz (18-90 cycles/min). Different parameters of each band were tested in the spectral analysis of cardiorespiratory control. The LF/HF-ratio of the spectral peak area of the respiratory activity and the LF/HF-ratio of the spectral band area of the respiratory activity were greater in the SIDS group when compared to the controls. No significant intergroup differences were found in the parameters of HRV, or the cross-spectral parameters. Interestingly, the technique appeared helpful in displaying that the victims of SIDS had a significantly greater amount of slow oscillation in the continuous respiratory signal (1.05+/-1.89 vs. 0.41+/-0.57, P=0.02). In the victims of SIDS the respiratory control system seems to be less stable and cause more slow oscillatory breathing and this can be detected using spectral analysis of respiratory activity even during breathing that visually seems to be regular.

Adder bites. A report of 68 cases.

BACKGROUND AND AIMS: The adder (Vipera Berus) is the only venomous snake that exists naturally in Scandinavia. The aim of this study is to estimate the severity of adder bites, to form a general picture of symptoms of bites and to find out how effective the present treatment methods of adder bites are. MATERIAL AND METHODS: 68 true adder bites treated in Turku University Hospital during the years 1995-2000 were reviewed retrospectively. RESULTS: There were no deaths in this material. A bite caused severe symptoms to 10% of the patients. The symptoms were moderate in 21%, mild in 34% and minor in 35% of the cases. Children under 10 years were the proportionally biggest age group and severe poisonings were most frequent among small children. Rapidly progressive oedema, gastrointestinal symptoms, hypotension and early leucocytosis were signs of more severe poisonings. Antivenom therapy with specific ovine Fab antivenom proved to be an effective and safe treatment in severe poisonings. CONCLUSIONS: An adder bite may also cause severe symptoms for adults. All patients should be observed at least few hours after the bite and parental fluid therapy should be started at an early stage. In the treatment of severe poisonings an antivenom therapy should be considered. Rapidly progressive symptoms and early leucocytosis may serve as a warning signal for higher probability of severe reactions.

Comparison of the effects of antenatal magnesium sulphate and ritodrine exposure on circulatory adaptation in preterm infants.

We examined the effects of maternal magnesium sulphate (MgSO4) and ritodrine treatments on the autonomic cardiovascular control in preterm neonates with respiratory distress syndrome during the first 2 days of life. Serial measurements of heart rate (HR), blood pressure (BP) and respirogram were performed during the first 2 days of life in 28 preterm infants below 33 weeks of gestation with antenatal exposure to MgSO4 (n = 13) or ritodrine (n = 15), and in 12 nonexposed preterm controls. Spectral analysis was used for the quantification of HR and BP variability. Although antenatal MgSO4 exposure had no effect on HR or the systolic, diastolic or mean BP, it was associated with significant decreased beat-to-beat changes in BP. In contrast, ritodrine exposure had no consistent effects on the autonomic cardiovascular control during the first 2 days of life. Our data suggest that maternal MgSO4 treatment decreases the neonatal high frequency changes in BP. This early vascular stabilizing effect of antenatal MgSO4 exposure may contribute to a lowered risk of cerebral vascular catastrophes, in the vulnerable areas of the brain, among the preterm infants with respiratory distress syndrome.

Transcephalic electrical impedance in the study of cerebral circulation in a juvenile pig model.

Transcephalic electrical impedance offers a technique for non-invasive, cot-side monitoring of neonatal cerebral circulation but the exact nature of the signal is somewhat ambiguous. The impedance signal is examined in an animal project where the ventilator settings are adjusted (20 min-1-10 min-1-40 min-1 for 10 min periods each) to produce circulatory changes. Six juvenile pigs are intubated, and ECG, arterial blood pressure, carotid flow (CF) by electromagnetic flowmeter and impedance are continuously monitored and stored on analogue tape. Cardiac output by thermodilution, blood oxygen (pO2) and carbon dioxide (pCO2) tensions are measured. ECG is converted to heart rate, mean blood pressure is integrated, and the high-frequency (1.50-4.00 Hz) component of the impedance signal delta Z is computed using autoregressive spectral estimation. Stroke volume, peripheral vascular resistance (PVR) and cerebral vascular resistance (CVR) are calculated. pCO2 and CF increase and pO2 decreases during hypoventilation. CF correlates positively with cardiac output, stroke volume, delta Z and pCO2, and negatively with pO2 and CVR. delta Z correlates positively with heart rate and cardiac output, and negatively with PVR and CVR. It is concluded that the impedance signal is related to the amount of blood transmitted to the brain by every beat of the heart, depending on the changes in both the systemic circulation and the cerebral vascular compliance.

Multivariate autoregressive modelling combined with transcephalic electrical impedance: method to relate neonatal systemic circulation and respiration to cerebral circulation.

We studied the pulsatile component of cerebral circulation with transcephalic electrical impedance (delta Z) in six preterm newborns, three of whom had severe cerebral bleeding, peri-intraventricular haemorrhage (PIVH). The transcephalic electrical impedance delta Z signal, ECG, arterial blood pressure, (aBP) and respirogram were recorded on analogue magnetic tape for 30 min. Artefact-free stationary segments (lasting for 2 min) of the four signals were digitised. A digital multivariate autoregressive (MAR) model was used to study frequency-specific variability in the signals and to quantify interrelations between the variabilities of delta Z, HR, aBP and respiratory signals. MAR modelling describes a system where all the signals simultaneously explain each other. The inherent variability of delta Z was lower and the influences of respiration and aBP on delta Z significantly greater in infants with severe PIVH than in controls. These changes were observed at high frequencies corresponding to respiration and heart rate. This may be interpreted as a marker of pressure passivism in the cerebral circulation following PIVH. We conclude that in preterm babies the application of MAR modelling, together with transcephalic impedance, may be a new, helpful and quantitative method for the study of simultaneous interrelations between variables of cerebral and systemic circulations and respiration.

Beta-adrenergic control and inter-relationships between heart rate and blood pressure in neonatal lambs.

Our aim was to develop a signal analysis method for revealing interrelationships between heart rate and blood pressure and for displaying the influence of autonomic nervous control on these signals in a chronic lamb model. A chronically instrumented neonatal lamb model was made to record ECG and direct arterial blood pressure (N = 15). Continuous two-minute recordings of blood pressure (BP) and ECG were digitised. The instantaneous heart rate signal (IHR) was derived from the ECG. The IHR and BP signals were bandpass filtered. Autospectra, cross-spectra, coherence spectra and phase spectra for the signals were computed to study the relative magnitudes and inter-relationships of the cardiovascular signals under normal conditions and during beta-adrenergic blockade. It was noted that both in the BP and IHR there were oscillations at the frequency of less than 0.1 Hz and also at the respiratory rate around 0.6 Hz. Beta-blockade reduced the oscillations of the IHR in less than 30-day-old lambs. It did not affect the coherence spectra or the phase lag between the signals. During quiet sleep the variability of blood pressure was decreased. In over-30-day-old-lambs the beta-blockade did not affect the variabilities of the cardiovascular parameters. These findings indicate that in neonatal lambs the sympathetic control system is a major regulator of cardiovascular interactions.

Nonlinear curve fitting improves noise sensitivity of the single-breath method for estimation of cardiac output.

In a recent theoretical study, Srinivasan (10) presented a new nonlinear curve fitting algorithm for computing pulmonary blood flow by the single-breath method. The results obtained with the new algorithm showed a considerably lower sensitivity of the estimated pulmonary blood flow to experimental noise than with the linear curve fitting technique reported previously by Grønlund (5). In the present study, we have compared the linear and the nonlinear algorithms using computer simulations and experimental single-breath data. The results showed only a small difference between the two methods. Further, we showed that the difference between the noise sensitivities obtained by Grønlund (5) and Srinivasan (10) can be explained almost entirely by a difference between the models used to simulate experimental noise.

Single-breath method for estimating pulmonary blood flow: data reduction based on nonlinear curve fitting.

The single-breath method of Kim et al. for noninvasive estimation of pulmonary blood flow utilizes the PCO2 vs. PO2 curve recorded during a prolonged expiration. The calculation of pulmonary blood flow is complicated by the nonlinearity of the expression for the PCO2 vs. PO2 curve. In previous studies of the single-breath method, this problem has been solved by using a data reduction procedure which includes transformation of the nonlinear expression into a linear plot. The disadvantage of this data reduction procedure is that the transformation utilizes the first derivative of the PCO2 vs. PO2 curve, which is very sensitive to experimental noise. In the present study, we have examined an alternative data reduction procedure in which the nonlinear expression was fitted directly to the experimental curve. At an expiration length of 15 s, the two data reduction procedures gave almost identical standard deviations of 0.57 L.min-1 on repeated determinations of the pulmonary blood flow. However, the linear data reduction procedure was much more sensitive to omission of the PCO2 vs. PO2 curve segments obtained in the last part of the prolonged expiration.

High resolution ECG changes among Finnish pilots.

BACKGROUND: Abundant appearance of late potentials registered by high resolution ECG (HR-ECG) has been shown to predict a risk for cardiac arrhythmia. The aim of this study was to evaluate the appearance of normal and abnormal late potentials in a group of Finnish private and commercial pilots. METHODS: Standard 12-lead ECG and 40-250 Hz band-pass filtered HR-ECG were recorded in 168 healthy male pilots. The following parameters were estimated: the root-mean-square voltage in the terminal 40 ms (RMS40), the total filtered QRS duration (FQRSD) and the duration of terminal high frequency low amplitude signal (HFLAD) of less than 40 microV in the filtered QRS. RESULTS: Age (38 +/- 12 yr) and height (179 +/- 6 cm) had a significant (p < 0.005) and independent association with the total filtered QRS duration. No such correlations were observed for RMS40 or HFLAD. Mean +/- SD for RMS40 was 35.3 +/- 15.4 microV, for FQRSD 114.6 +/- 6.8 ms and for HFLAD 31.0 +/- 7.9 ms. Smokers tended to have longer late potentials (HFLAD) than non-smokers. Fasting blood glucose had a significant (p < 0.05) positive correlation with HFLAD. Serum cholesterol level had no correlation with the HR-ECG parameters. Age and height of the subject were associated with the total QRS duration as follows: FQRSD (ms) = -9.7* log age (yr) + 0.2*height (cm) + 94.6.

Improved method to estimate oxygen consumption, diffusing capacity and blood flow of synovial membrane.

The joint cartilage is depending on the oxygen diffusing from the capillaries of the synovial membrane through the synovial tissue and synovial fluid. In the present study we describe a new method to calculate the diffusing capacity (DO2), oxygen consumption (VO2) and blood flow of the synovial membrane. The principle of the method is to perfuse the joint cavity with two saline solutions, one with high and one with low oxygen and nitrogen partial pressures and to measure the oxygen and nitrogen partial pressures in the perfusate flowing out of the joint. Using a model of gas exchange between the joint and the membrane a set of equations was derived expressing the relationship between the blood flow, diffusing capacity and oxygen consumption and the oxygen and nitrogen flowing to and from the joint in the two different perfusion situations. In 12 rabbit knee joints we found a blood flow of 0.388 ml/min (SEM 0.027), VO2 of 0.495 microliters/min (SEM 0.196) and DO2 of 0.024 microliter/min/Torr SEM 0.003 (mean).