Median PaO2 Research Articles

Interleucina-8 en el lavado broncoalveolar del donante y su impacto en la disfunción primaria del injerto en el trasplante bipulmonar

Background and Objective The purpose of this study was to determine concentrations of interleukin 8 (IL-8) in the bronchoalveolar lavage (BAL) fluid from donor lungs and assess the role of IL-8 levels in the development of primary graft failure. Patients and Method Twenty patients who received a double lung transplant were studied. A series of data, including BAL fluid concentrations of IL-8, were collected for the donors. Data collected for the recipients included arterial blood gases after 6, 24, and 48 hours, and intubation time. Patients with a ratio of PaO 2 to the fraction of inspired oxygen (FiO 2) of less than 300 during the first 48 hours were diagnosed with primary graft failure. IL-8 levels were determined by enzyme-linked immunosorbent assay. Associations between the donor variables and IL-8 concentrations were evaluated using the Spearman rank correlation coefficient (ρ) and the Mann-Whitney test for categorical and continuous variables, respectively. Logistic regression was used for multivariate analysis. Results Fifteen of the 20 donors were men. The cause of brain death was trauma in 9 donors, 7 were smokers, 13 required inotropic support, and pathogens were isolated in the BAL fluid of 18. The median age was 35 years (interquartile range [IQR], 23.5–51.25 y), the median ventilation time was 1 day (IQR, 1–2 d), the median PaO 2/FiO 2 was 459.5 (IQR, 427–510.25), and the median IL-8 concentration in BAL fluid was 49.01 ng/L (IQR, 7.86–94.05 ng/mL). Ten of the recipients were men and the median age was 48.43 years (IQR, 25.4–56.81 y). The median ischemic time was 210 minutes (IQR, 176.25–228.75 min) for the first lung and 300 minutes (IQR, 273.75–333.73 min) for the second lung. The median PaO 2/FiO 2 ratio for the implant at 6, 14, and 48 hours was 329 (IQR, 190.25–435), 363.5 (IQR, 249–434.75), and 370.5 (IQR, 243.25–418.25), respectively. The median intubation time was 39.5 hours (IQR, 19.25–68.5 h) and the correlation with IL-8 values was positive: higher IL-8 concentrations in BAL fluid correlated with longer ventilation times (Spearman rank correlation, P=.007; ρ=0.583). Five patients developed primary graft failure; IL-8 concentrations were significantly higher in these patients than in those whose grafts did not fail (Mann-Whitney test, P=.003). Conclusion High IL-8 concentrations in donor BAL fluid lead to longer ventilation time in the recipients and favor the development of primary graft failure after lung transplant.

Interleukin 8 Concentrations in Donor Bronchoalveolar Lavage: Impact on Primary Graft Failure in Double Lung Transplant

Background and Objective The purpose of this study was to determine concentrations of interleukin 8 (IL-8) in the bronchoalveolar lavage (BAL) fluid from donor lungs and assess the role of IL-8 levels in the development of primary graft failure. Patients and Methods Twenty patients who received a double lung transplant were studied. A series of data, including BAL fluid concentrations of IL-8, were collected for the donors. Data collected for the recipients included arterial blood gases after 6, 24, and 48 hours, and intubation time. Patients with a ratio of PaO 2 to the fraction of inspired oxygen (FiO 2) of less than 300 during the first 48 hours were diagnosed with primary graft failure. IL-8 levels were determined by enzyme-linked immunosorbent assay. Associations between the donor variables and IL-8 concentrations were evaluated using the Spearman rank correlation coefficient (ρ) and the Mann-Whitney test for categorical and continuous variables, respectively. Logistic regression was used for multivariate analysis. Results Fifteen of the 20 donors were men. The cause of brain death was trauma in 9 donors, 7 were smokers, 13 required inotropic support, and pathogens were isolated in the BAL fluid of 18. The median age was 35 years (interquartile range [IQR], 23.5-51.25 y), the median ventilation time was 1 day (IQR, 1-2 d), the median PaO 2/FiO 2 was 459.5 (IQR, 427-510.25), and the median IL-8 concentration in BAL fluid was 49.01 ng/L (IQR, 7.86-94.05 ng/mL). Ten of the recipients were men and the median age was 48.43 years (IQR, 25.4-56.81 y). The median ischemic time was 210 minutes (IQR, 176.25-228.75 min) for the first lung and 300 minutes (IQR, 273.75-333.73 min) for the second lung. The median PaO 2/FiO 2 ratio for the implant at 6, 14, and 48 hours was 329 (IQR, 190.25-435), 363.5 (IQR, 249-434.75), and 370.5 (IQR, 243.25-418.25), respectively. The median intubation time was 39.5 hours (IQR, 19.25-68.5 h) and the correlation with IL-8 values was positive: higher IL-8 concentrations in BAL fluid correlated with longer ventilation times (Spearman rank correlation, P=.007; ρ=0.583). Five patients developed primary graft failure; IL-8 concentrations were significantly higher in these patients than in those whose grafts did not fail (Mann-Whitney test, P=.003). Conclusion High IL-8 concentrations in donor BAL fluid lead to longer ventilation time in the recipients and favor the development of primary graft failure after lung transplant.

Outcomes of Patients with Haematological Malignancies Admitted to Either High Dependency Unit (HDU) or Intensive Care Unit (ICU) with or without Haemopoeitic Stem Cell Transplantation (HSCT): Analysis of Prognostic Factors for Survival

Background: Patients with haematological malignancies (HM) receive intensive chemotherapy regimens including HSCT rendering them immunocompromised and susceptible to complications such as, cardio-pulmonary toxicity, hepatic and renal failure, GVHD, sepsis and haemorrhage. Patients (pts) with HM may require critical care (CC) admission. Many studies have looked at the ICU setting for these pts but few have evaluated the use of non invasive cardio-pulmonary support on HDU prior to intensive care escalation or invasive intervention. The Haemato-Oncology Unit at Sheffield Teaching Hospitals NHS Foundation Trust is a tertiary referral centre, serving a population of 1.5 million. Critical care facilities are divided into HDU and ICU. Levels of care provided on HDU include non-invasive respiratory (non-invasive ventilation or continuous positive airway pressure) and cardiovascular (vasopressor/inotropic) support, but not mechanical ventilation (MV) or renal replacement therapy (RRT), both of which are provided on ICU. We evaluated all admissions to HDU and ICU for pts with HM in order to better understand the prognostic markers for this cohort of pts.Subjects and Methods: We retrospectively analysed consecutive HM (including HSCT) admissions to HDU or ICU over a 66 month period, between September 2002 and March 2008. We analysed predictors of mortality including age, white cell count (WBC), Acute Physiology, Age, and Chronic Health Evaluation II (APACHE II) score, PaO2:FiO2 ratio as well as requirement for respiratory or cardiovascular support or RRT. Unit survival, hospital survival and overall survival at the time of study (OS) were evaluated.Results: One hundred and sixty two pts (n=107 HDU, n=55 ICU) were admitted to CC during the study period comprising n=191 admissions (n=128 HDU, n=22 HDU transferred to ICU, n=63 ICU). Of the HDU admissions n=38 pts received HSCT, and of the ICU admissions n=28 pts received HSCT. The median age for the entire cohort was 59 yrs (range 16–83 yrs). The median age for the HSCT and non HSCT pts was 50 yrs (range 17–68 yrs) and 63 yrs (range 16–83 yrs) respectively. Of the n=162 pts, n=99 (61%) were male, n=63 (39%) female. Types of HM varied, AML n=48 (25%), Myeloma n=40 (21%), NHL n=36 (19%), MDS n=16 (8%), ALL n=16 (6%), CLL n=12 (6%), CML n=9 (5%), others n=13 (7%). Overall median WBC was 3.6×109/l (range 0.0–746×109/l), median WBC for HSCT and non HSCT admissions was 3.1×109/l (range 0.1–31.6×109/l) and 4.1×109/l(range 0.0–746×109/l) respectively. The overall mean APACHE II score for all HDU admissions was 16.2 (SD ± 5.9) and 23.7(SD±7.2) for all ICU admissions. The overall median PaO2:FiO2 ratio for all HDU admissions was 15 kPa and 12 kPa for all ICU admissions. In total n=61 (32%) admissions required respiratory support (n=35(18%) MV, n=26 (14%) non invasive respiratory support), n=91 (48%) received cardiovascular support while n=25 (13%) required RRT. For all HDU admissions, unit survival, hospital survival and OS was 81%, 55% and 27% respectively compared to 52%, 44% and 13% for all ICU admissions. For HDU admissions escalated to ICU, unit survival, hospital survival and OS was 32%, 32% and 9% respectively. For non HSCT pts admitted to HDU, unit survival, hospital survival and OS was 79%, 54% and 26% respectively, compared to 84%, 55% and 29% for HSCT pts. For non HSCT pts admitted to ICU, unit survival, hospital survival and OS was 53%, 44% and 14% respectively, compared to 54%, 43% and 14% for HSCT pts. An APACHE II score of >30 was associated with a unit survival of 44%. No pts with an APACHE II score >36 survived. Unit survival for admissions requiring no organ support was 87%, one organ support 79%, two organ support 22% and three organ support 58%. Unit survival for pts requiring MV was 23%.Conclusions: Our study demonstrates that MV, requirement for two organ support, admission APACHE II score >30 are all poor prognostic factors for pts with HM admitted to CC. Interestingly of the n=12 pts requiring MV, cardiovascular support and RRT n=7 (58%) survived to ICU discharge, demonstrating that selected pts requiring multi-organ support may have a better prognosis. Requirement for single organ support was associated with better outcome. Our study demonstrates that the majority of HM pts requiring CC can be appropriately managed in the HDU setting. We aim to develop a predictive scoring system for HM pts requiring CC using the evaluated prognostic factors.

Prone Positioning and Acute Respiratory Distress Syndrome After Cardiac Surgery: A Feasibility Study

To determine the feasibility, safety, and efficacy on PaO(2)/F(I)O(2) ratio of prone positioning (PP) for acute respiratory distress syndrome (ARDS) after cardiac surgery. Retrospective review of information entered prospectively in the authors' database. A private community nonteaching hospital. Sixteen patients who developed ARDS after cardiac surgery from January 2004 through June 2005. PP to improve oxygenation. After a median duration of 18 (range, 14-27) hours in PP, PaO(2)/F(I)O(2) improved in 14 (87.5%) patients. For the entire population, median PaO(2)/F(I)O(2) rose from 87 (range, 56-161) before PP to 194 (range, 94-460; p < 0.05) after it. After supine repositioning (SR), PaO(2)/F(I)O(2) declined to 146 (range, 72-320; not significant). PaO(2)/F(I)O(2) at the end of PP and 1 day after SR were comparable, respectively, 194 (range, 94-460) and 184 (range, 105-342). No severe complication was associated with PP, but 5 patients developed pressure sores and 2 others had superficial sternal wound infections. Intensive care unit mortality of 37.5% reflected the number of organ failure(s); there were no deaths with 2 failures, and 60% with > or = 3 organ failures died (p = 0.03). Mortality rates were comparable regardless of whether patients were PaO(2)/F(I)O(2) responders or their PaCO(2) decreased by > or = 1 mmHg. PP to treat ARDS after cardiac surgery is feasible, safe, and can efficiently improve oxygenation. Measures to prevent pressure sores are mandatory.

Selective Recruitment Maneuvers for Lobar Atelectasis: Effects on Lung Function and Central Hemodynamics: An Experimental Study in Pigs

We investigated whether selective lung recruitment of a lobar collapse would improve oxygenation and lung volume as well as a general (global) lung recruitment maneuver, with fewer circulatory side effects. In 10 ventilated, anesthetized pigs, a bronchial blocker was inserted in the right lower lobe, which was selectively lavaged to create a dense lobar collapse. The pigs were randomized into two orders of lung recruitment maneuvers (40 cm H2O airway pressure for 30 s): either a selective lung recruitment maneuver (using the inner lumen of the bronchial blocker) followed by a general lung recruitment maneuver, or vice versa. Median end-expiratory lung volume and median Pao2 increased significantly by approximately 100 mL and 16 kPa, respectively, with no significant differences between the two recruitment methods. There were no circulatory changes during the selective lung recruitment maneuver, but during the general lung recruitment maneuver, mean arterial blood pressure decreased significantly by 36 (21, 41) mm Hg (median, 25th and 75th percentiles), cardiac output by 2.1 (1.6, 2.5) L/min and left ventricular end-diastolic area by 4.4 (3.5, 4.5) cm2. In conclusion, a selective recruitment maneuver improved lung function similar to a general lung recruitment maneuver but without any circulatory side effects.

Liver transplantation for hepatopulmonary syndrome: a ten-year experience in Paris, France.

Although the possibility of reversing hepatopulmonary syndrome (HPS) after liver transplantation is now well established, the proportion of patients in whom reversibility is observed and the time to resolution of HPS remain uncertain. We analyzed the outcome of all adult patients with HPS who underwent orthotopic liver transplantation in all the liver transplant centers in Paris, during a 10-year period. Twenty-three adult patients (median age, 47 years; range, 14-64) underwent transplantation in four institutions. Median PaO(2) was 52 (range, 32-67) mm Hg and median alveolar-arterial oxygen gradient was 66 mm Hg. When patients were breathing 100% O(2), median PaO(2) was 310 (range, 74-663) mm Hg. Median isotopic shunt ratio was 33% (range, 0-80%). The overall mortality during the study period was 30.5% (7/23). Perioperative mortality was 8.5%, whereas late mortality was 22%. None of the preoperative characteristics of HPS (isotopic shunt ratio, PaO(2) on room air or on 100% oxygen) was associated with overall postoperative mortality. Of the 21 patients surviving the perioperative period (median follow-up, 17 months; range, 0.5-72), a decrease in alveolar-arterial oxygen gradient of at least 5 mm Hg and at least 10 mm Hg was observed in 21 of 21 and in 18 of 21 patients, respectively, with great variations in the time of improvement. The threshold of 70 mm Hg was reached in 15 patients. The lower the preoperative PaO(2), the longer the time to reach this point. Our data strongly support the role of orthotopic liver transplantation in adult patients with HPS, regardless of its severity.

Optimizing peroperative compliance with PEEP during upper abdominal surgery: effects on perioperative oxygenation and complications in patients without preoperative cardiopulmonary dysfunction

Background and objective Late postoperative hypoxaemia after upper abdominal surgery is common even among cardiopulmonary healthy patients. Atelectasis may develop after intubation and persist into or reveal a disposition for atelectasis in the postoperative period. Positive end-expiratory pressure (PEEP) eliminates peroperative atelectasis but the effect on perioperative oxygenation is controversial. This study evaluated the effect of peroperative PEEP optimized pulmonary compliance on perioperative oxygenation and complications. Methods Forty patients assessed by electrocardiography, spirometry, functional residual capacity and diffusion capacity were randomly assigned to receive positive end-expiratory pressure (PEEP) or zero end-expiratory pressure (ZEEP) during surgery. PaO2, SPO2 and complications in the postoperative period were evaluated without knowledge of peroperative PEEP or ZEEP application. Results Peroperative arterial oxygenation improved for all patients receiving PEEP, mean 2.1 kPa (0.7–3.5 kPa). There was no difference in postoperative median PaO2 between the groups. The differences in the incidence of late prolonged postoperative hypoxaemia and complications were 25% (−5% to 55%) and −1% (−31% to 29%) between the ZEEP and the PEEP group, but were not statistically significant.

“Alveolar Recruitment Strategy” Improves Arterial Oxygenation During General Anaesthesia

Abnormalities in gas exchange during general anaesthesia are caused partly by atelectasis. Inspiratory pressures of approximately 40 cm H2O are required to fully re-expand healthy but collapsed alveoli. However, without PEEP these re-expanded alveoli tend to collapse again. We hypothesized that an initial increase in pressure would open collapsed alveoli; if this inspiratory recruitment is combined with sufficient end-expiratory pressure, alveoli will remain open during general anaesthesia. We tested the effect of an 'alveolar recruitment strategy' on arterial oxygenation and lung mechanics in a prospective, controlled study of 30 ASA II or III patients aged more than 60 yr allocated to one of three groups. Group ZEEP received no PEEP. The second group received an initial control period without PEEP, and then PEEP 5 cm H2O was applied. The third group received an increase in PEEP and tidal volumes until a PEEP of 15 cm H2O and a tidal volume of 18 ml kg-1 or a peak inspiratory pressure of 40 cm H2O was reached. PEEP 5 cm H2O was then maintained. There was a significant increase in median PaO2 values obtained at baseline (20.4 kPa) and those obtained after the recruitment manoeuvre (24.4 kPa) at 40 min. This latter value was also significantly higher than PaO2 measured in the PEEP (16.2 kPa) and ZEEP (18.7 kPa) groups. Application of PEEP also had a significant effect on oxygenation; no such intra-group difference was observed in the ZEEP group. No complications occurred. We conclude that during general anaesthesia, the alveolar recruitment strategy was an efficient way to improve arterial oxygenation.

‘Alveolar recruitment strategy’ improves arterial oxygenation during general anaesthesia

Abnormalities in gas exchange during general anaesthesia are caused partly by atelectasis. Inspiratory pressures of approximately 40 cm H2O are required to fully re-expand healthy but collapsed alveoli. However, without PEEP these re-expanded alveoli tend to collapse again. We hypothesized that an initial increase in pressure would open collapsed alveoli; if this inspiratory recruitment is combined with sufficient end-expiratory pressure, alveoli will remain open during general anaesthesia. We tested the effect of an 'alveolar recruitment strategy' on arterial oxygenation and lung mechanics in a prospective, controlled study of 30 ASA II or III patients aged more than 60 yr allocated to one of three groups. Group ZEEP received no PEEP. The second group received an initial control period without PEEP, and then PEEP 5 cm H2O was applied. The third group received an increase in PEEP and tidal volumes until a PEEP of 15 cm H2O and a tidal volume of 18 ml kg-1 or a peak inspiratory pressure of 40 cm H2O was reached. PEEP 5 cm H2O was then maintained. There was a significant increase in median PaO2 values obtained at baseline (20.4 kPa) and those obtained after the recruitment manoeuvre (24.4 kPa) at 40 min. This latter value was also significantly higher than PaO2 measured in the PEEP (16.2 kPa) and ZEEP (18.7 kPa) groups. Application of PEEP also had a significant effect on oxygenation; no such intra-group difference was observed in the ZEEP group. No complications occurred. We conclude that during general anaesthesia, the alveolar recruitment strategy was an efficient way to improve arterial oxygenation.

Helicobacter pylori infection and chronic gastric acid hyposecretion

BACKGROUND & AIMS: We have identified a subgroup of Helicobacter pylori- infected subjects with low or absent gastric acid output. The aim of this study was to document the morphological and functional abnormalities in these subjects and to assess the effect of eradicating the infection. METHODS: The 16 hypochlorhydric subjects (6 men) had a mean age of 55 years (range, 36-79 years). They underwent a 14C-urea breath test, H. pylori serology, fasting gastrin, gastric autoantibodies, gastroscopy with antral and body biopsies, and measurement of peak acid output to pentagastrin (PAO(PG)). Their histology was compared with that of age- and sex-matched duodenal ulcer and nonulcer dyspepsia patients (16 each). H. pylori infection was eradicated in the hypochlorhydric subjects, and the investigations were repeated 6 months later. RESULTS: Compared with controls, the hypochlorhydric subjects had less dense H. pylori colonization, body- predominant colonization and gastritis, and increased prevalence of body atrophy and intestinal metaplasia. Median PAO(PG) before eradication in the hypochlorhydric subjects was 1.1 mmol/h and increased to 12.6 mmol/h after eradication (P < 0.001), with no significant change in body atrophy or intestinal metaplasia. CONCLUSIONS: In some subjects, chronic H. pylori infection produces a body-predominant gastritis and profound suppression of gastric acid secretion that is partially reversible with eradication therapy. (Gastroenterology 1997 Jul;113(1):15-24)

Ventilation-perfusion inequality in nocturnal hypoxaemia due to chronic obstructive lung disease (COLD).

Nocturnal hypoxaemia is often noted in COLD patients with a daytime PaO2 above 8.0 kPa. It has been assumed that ventilation-perfusion inequality contributes to nocturnal hypoxaemia. 10 patients with advanced COLD [median FEV1 0.73 (range 0.50-1.32)l], but without daytime hypoxaemia [median PaO2 8.35 (range 8.0-12.2) kPa] were investigated with regard to possible nocturnal hypoxaemia using polysomnography. Daytime lung function was assessed by spirometry and carbon monoxide diffusion capacity (DLCO). Daytime ventilation-perfusion (VA/Q) relationships were measured by the multiple inert gas elimination technique. Dispersion of perfusion and ventilation distributions was increased [log SDQ 1.01 (range 0.80-1.35) and log SDV 0.91 (range 0.69-1.86) resp.]. Around 8% of the ventilation was directed towards high VA/Q areas (10 < VA/Q < 100). All subjects reached all sleep stages, and all but one had a nadir nocturnal oxygen saturation (SaO2) of below 90%. Their median lowest nocturnal SaO2 was 84.0 (range 70-93)% and their mean oxygen saturation in the course of desaturation episodes (MminSaO2) was 86.4 (range 83.6-91.5)%. An increased mean VA/Q ratio of ventilation distribution was associated with a reduced DLCO. Increased nocturnal episodes of wakefulness and of stage I sleep correlated with increased dead space ventilation and dispersion of the ventilation distribution. Patients with deep nocturnal desaturations had a low mean VA/Q ratio of the perfusion distribution (Q mean) (r = 0.87, P < 0.01) and increased perfusion of inferior VA/Q areas (0.1 VA/Q < 0.3). Low MminSaO2 was associated with low morning PaO2 and a low Q mean. COLD patient with solely nocturnal hypoxaemia have a high degree of pulmonary hyperinflation and emphysema. Increased sleep disruption is associated with more severe small airway disease. Increased perfusion of sparsely ventilated areas is associated with more pronounced nocturnal desaturations.

Effect of Alcohol on Acute Ventilatory Adaptation to Mild Hypoxia at Moderate Altitude

To evaluate the influence of alcohol on acute adaptation to mild hypoxia at moderate altitude. Randomized, double-blind, placebo-controlled crossover trial. University clinic and mountaineering resort at altitudes of 171 m and 3000 m, respectively, in the Austrian Alps. 10 healthy male alpinists, 22 to 24 years of age. Single dose of 50 g of alcohol or placebo at altitudes of 171 m and 3000 m. Arterial oxygen pressure (Pao2) and arterial carbon dioxide pressure (Paco2) before and 1 hour after consumption of alcohol or placebo. At an altitude of 171 m, 50 g of alcohol caused no statistical change in Pao2 and Paco2 (median Pao2, 91.5 compared with 90.5 mm Hg [P = 0.89]; median Paco2, 37.5 compared with 36.0 mm Hg [P = 0.41]). At an altitude of 3000 m, the median Pao2 decreased from 69.0 to 64.0 mm Hg, a median decrease in the paired difference of 4.0 mm Hg (95.1% CI, 1.5 to 6.5 mm Hg; P < 0.01), and the median Paco2 increased from 32.5 to 34.0 mm Hg, a median increase in the paired difference of 3.0 mm Hg (95.1% CI, 2.0 to 4.0 mm Hg; P < 0.01) 1 hour after drinking 50 g of alcohol. Placebo did not influence Pao2 or Paco2 at either altitude. Alcohol inhibits the initial stages of adequate acute ventilatory adaptation to mild hypoxia at moderate altitude. Caution in the use of alcoholic beverages at moderate altitude is therefore necessary.

87 VARIABLES INFLUENCING THE LEVEL OF PLASMA LACTATE

It has been suggested that plasma lactate (L) as an indicator of hypoxemia may identify serious anaemia requiring transfusion. One might expect anaemia to exacerbate the effect of hypoxemia with increased L resulting. The influence of haemoglobin (Hb) (range:3.2-9.6 g/100ml), glucose, weight, PaO2 and blood pressure on L (range:2.4-9.3 mmol/l) was studied in 31 piglets aged 12-60h. The animals were anaesthetised with halothane, intubated and sampled under stable conditions. 17 of the animals were sampled again after a period of severe hypoxemia (median PaO2=3.3kPa). Regression analysis were performed showing Hb to be the best predictor of L at base line conditions, (log L=1.1-0.1 Hb p=0.0001 Adj.R2=0.5). Hb did not however influence the amount of increase in L (ΔL) (range:10.6-28.9) during severe hypoxia. None of the recorded variables alone or in combinations were able to predict ΔL with p<0.05.

Nasal intermittent positive pressure ventilation in acute exacerbations of chronic obstructive pulmonary disease — a preliminary study

Ten patients (two male) suffering from acute exacerbations of long-standing chronic obstructive pulmonary disease and admitted in hypoxic, hypercapnic respiratory failure were treated with Nasal Intermittent Positive Pressure Ventilation (NIPPV) plus supplemental oxygen, on a general medical ward. The median (range) pH on admission was 7·30 (7·20–7·35), the median age was 67 years (47–77) with an FEV 1 (percent of predicted) of 30 (17–39). On admission the median arterial oxygen tension ( PaO 2) was 4·71 kPa (3·45–6·26) on air, and the carbon dioxide tension ( PaCO 2) was 7·68 kPa (6·85–9·83). With controlled oxygen therapy there was no significant improvement in PaO 2, but the median PaCO 2 increased significantly to 9·75 kPa (7·04–11·70) ( P < 0·05). By using NIPPV with supplemental oxygen it was possible to significantly improve the median PaO 2 to 11·25 kPa (6·70–26·90) ( P < 0·01) without worsening PaCO 2 levels (8·96 kPa; 6·85–13·10). NIPPV was applied by a senior, respiratory physiotherapist and used intermittently depending on patient tolerance and clinical response. The median total time on NIPPV was 27 h, delivered over 1–5 days. One patient found the mask difficult to tolerate beyond a short period of time. NIPPV was well accepted on a general ward by nursing staff. Three patients later died with progressive hypercapnia, despite an initial response; with one of these patients also receiving intubation and mechanical ventilation. A further patient also received intubation and mechanical ventilation and was eventually discharged. NIPPV plus supplemental oxygen offers a method to correct hypoxaemia on a general medical ward without worsening hypercapnia for acute on chronic, hypoxic, hypercapnic respiratory failure, and warrants further investigation.

Maintenance of Oxygenation during One-Lung Ventilation

The aim of this study was to evaluate the effect on oxygenation of intermittent inflation with oxygen of the collapsed lung during one-lung ventilation (OLV). Sixteen patients were studied during pulmonary surgery. Balanced anesthesia with nitrous oxide and an inspired oxygen fraction of 0.5 was used. The control group (N = 8) had a median PaO2 of 19.2 (range 11.2-30.2) kPa before OLV, and 10.2 (8.2-16.0) kPa after 9 minutes of OLV without further reduction in PaO2 for another 10 minutes. In the treatment (inflation) group, the collapsed lung was manually inflated with 2 liters of oxygen and was then immediately allowed to collapse again. This procedure was repeated every 5 minutes during OLV. PaO2 increased more than 4 kPa following each inflation in seven patients. In the eighth, PaO2 remained high throughout OLV. Although PaO2 decreased between inflations, it never reached the level observed in controls during 19 minutes of OLV.