Ch. Hörmann

The prone position in ARDS. A successful therapeutic strategy

Zusammenfassung. Im Rahmen einer prospektiven Studie wurden an 7 polytraumatisierten Intensivpatienten im akuten Lungenversagen die Auswirkungen eines 12-stündlichen Lagewechsels zwischen Bauch- und Rückenlage auf Gasaustausch und auf die im Thorax CT nachgewiesenen Atelektasen untersucht. Bei allen Patienten kann eine signifikante Verbesserung sowie eine dauerhafte Stabilisierung des Gasaustauschs nachgewiesen werden. Parallel dazu kommt es zu einer signifikanten Abnahme des Shunt. Im Kontrollthorax-CT nach mehrtägigem Wechsel zwischen Bauch- und Rückenlagerung zeigt sich gegenüber der initialen CT-Untersuchung eine deutliche Verringerung der Atelektasen. Durch konsequente Lagerung in Bauch- bzw. Rückenlage konnte bei den untersuchten Patienten eine dauerhafte Verbesserung und Stabilisierung der Oxygenierung erzielt werden, die auch nach Ende der Lagerungsmanöver bis zur Beendigung des Weaning (2,6 Tage nach Beendigung des Lagewechsels) bestehen blieb. Dafür dürfte einerseits eine Verbesserung des Ventilations-/Perfusionsverhältnisses in den nach Umlagerung in Bauchlage basal liegenden Lungenarealen verantwortlich sein. Andererseits kommt es in den nach Lagewechsel oben liegenden, dys- und atelektatischen Lungenarealen zu einer Rekrutierung von zuvor kollabierten Alveolen.Abstract. As early as 1974, Brian advocated the prone position for ventilated patients [5]. He suggested that this position might enhance ventilation of the dorsal parts of the lungs, thereby improving oxygenation. These considerations have been confirmed by several experimental and clinical studies [1, 6, 13, 17]. Better secretion removal, decreased intrapulmonary shunting, and an increased FRC are thought to be responsible for the observed improvement of oxygenation. However, the prone position never became very popular in the clinical treatment of the adult respiratory distress syndrome (ARDS). Routine performance of thoracic CT scans in ARDS patients demonstrated preferential distribution of pathological densities in the dependent lung areas. The prone position therefore could possibly benefit these patients, as shown by two recent studies [8, 11]. The aim of our study was to evaluate the influence of repeatedly turning the patient to the prone position on gas exchange and thoracic CT findings in multiple-trauma patients. Methods. Seven ventilated intensive care patients with severe ARDS (Murray Score >2.5 [9], Quotient >0.7 [4], mean airway pressure >18 cm H2O, thoracic CT scan showing dorsal atelectases) were included in the study. Patients were turned from the supine to the prone position at 12-h intervals using an air-cushion bed (Mediscus, Austria). Redistribution of dystelectatic or atelectatic dependent lung areas was verified by means of repeated thoracic CT scans (Figs. 1, 8). Results. The patients were intermittently turned for 6.5 ± 1.1 days. The course of gas exchange is shown in Figs. 2 and 3. Initially, improvement of the respiratory quotient could only be achieved during prone positioning, from the 2nd day in the supine position as well. Intrapulmonary shunting showed a similar trend (Figs. 4 and 5). No significant changes in cardiovascular parameters could be observed. Control thoracic CT scans showed uniform reduction of atelectases in dependent lung areas (Figs. 1 and 8). The inspiratory fraction of oxygen could be reduced significantly as of the 2nd day (Fig. 7). Constant levels of positive end-expiratory pressure and tidal volume were associated with decreasing mean and plateau airway pressures (Fig. 6). Discussion. Repeatedly turning the patient to the prone position produced long-lasting improvement of arterial oxygenation, which persists up to the end of the weaning process. This is in good accordance with other studies [1, 2, 6, 8, 11, 13, 17], however, this is the first study to report an observation period of more than 6 days of repeatedly turning the patient. These positive effects on gas exchange can be attributed to sudden improvement of the ventilation-perfusion ratio within the lung areas that become dependent after turning to the prone position. Due to reduced hydrostatic pressure [11] and relative hyperventilation [7, 16], previously collapsed alveoli are recruited in the lung areas that become non-dependent after turning to the prone position.

Comparison of different modes of high-frequency ventilation in surfactant-deficient rabbits

Various modes of high‐frequency ventilation (HFV) have been developed to avoid the disadvantages of conventional mechanical ventilation. In the present study, we examined the hypothesis that high‐frequency oscillation (HFO) is superior to high‐frequency positive pressure ventilation (HPPV) and combined high‐frequency ventilation (CHFV) in surfactant‐deficient rabbits. The aim of the ventilator strategy was to adjust the mean airway pressure to 2 cm above critical opening pressure of the inflation limb of the respiratory system pressure volume (P/V) curve, achieve a normal tidal volume (Vγ) (5 ml/kg body weight) and apply repeated sustained inflations. We studied the effect of these HFV modes on oxygenation, lung mechanics and lung histology in 15 New Zealand White rabbits during a 6‐hour experiment. Statistically, the HFO group demonstrated significantly better oxygenation (P < 0.05), lung mechanics (lung stability index: P < 0.05), and better lung tissue histology compared to the HPPV and CHFV groups. In contrast to the HPPV and CHFV groups, the P/V curves of the HFO group showed significant recovery over the 6‐hour period after lavage. The lungs of the HFO‐treated group had a more uniform distribution of alveoli and less overdistention than the HPPV group (P < 0.002), and less atelectasis than the CHFV group (P < 0.05). The HFO group had less lung injury than the CHFV groups (P < 0.01) and its lungs contained significantly less water than both other groups (P < 0.05). We conclude that the relationship between mean and end‐expiratory pressures impacts strongly on both oxygenation and the progression of injury during HFV at the same mean airway pressures. The HFO group showed less acute lung injury than the other ventilatory groups. Pediatr Pulmonol. 1996; 22:263–270.

Tidal volume, breathing frequency, and oxygen consumption at different pressure support levels in the early stage of weaning in patients without chronic obstructive polmonary disease

The objective of this study was to evaluate the influence of different PSV levels on VT, F, VO2 in the early weaning phase of patients without chronic obstructive pulmonary disease. These parameters were tested for the predictive power for the success of the weaning. Patients on SIMV were studied during the first weaning attempt with PSV. Depending on their ventilatory support demands after 24 h they were divided into responders (patients breathing on CPAP) and nonresponders (patients being on a more invasive ventilatory mode). 14 ICU patients without pre-existing pulmonary disease being ventilated for at least 3 days entered the study. 2 of them could be studied a second time after failing the first weaning attempt. Beside the level of ventilatory support no other changes (drugs, nutrition) were allowed. VO2, VT, F were measured by a computer controlled, metabolic unit connected to the expiratory port of a Siemens Servo Ventilator. In addition, airway pressures, arterial pressure and heart rate were recorded. The measurements were performed at PSV of 5, 10 and 20 cmH2O. Arterial bloodgases were drawn at the end of each 60 min lasting PS period. Responders and nonresponders could be separated by the response of VO2, VT and F to a change in PS 10 to PS 20 cmH2O. Patients who significantly increased VT and significantly decreased F did not show a significant change in these two parameters, but a significant increase in VO2 at PS 20 cmH2O could be observed. In patients offering a steady increase in VT and decrease in F up to high levels of Ps the contribution of respiratory muscles to the volume displacement seemed to be poor. Patients who could be weaned used their respiratory muscles to controll their VT at higher levels of PS.

Effects of normo- and hypocapnic nitrous-oxide-inhalation on cerebral blood flow velocity in patients with brain tumors.

Nitrous oxide (N2O) use during anesthesia for intracranial procedures has been a subject of controversy in the past. To date, the isolated influence of N2O on mean cerebral blood flow velocity in the middle cerebral artery (VMCA) has not been investigated during hypocapnia in patients with brain tumors. We compared VMCA during normocapnic (ETCO2: 40 mm Hg) and hypnocapnic (ETCO2: 25 mm Hg) inhalation of air and 50% nitrous oxide in oxygen N2O/O2 in eight patients with unilateral brain tumors on both the tumor side and the healthy side. Six patients completed the study. Mean VMCA increased during normocapnic inhalation of N2O/O2 (tumor side: 86 ± 16 cm sec-1; healthy side: 74 ± 17 cm sec-1) when compared with air (tumor side: 72 ± 18 cm sec-1; healthy side: 62 ± 14 cm sec-1, p < 0.01), whereas during hyperventilation VMCA decreased on both sides (p < 0.001). Mean VMCA values were quite similar during hypocapnic inhalation of 50% N2O/O2 (tumor side: 50 ± 12 cm sec-1; healthy side: 45 ± 13 cm sec-1) and air (tumor side: 51 ± 14 cm sec-1; healthy side: 45 ± 12 cm sec-1). The data of our study suggest that in patients with cerebral tumors the N,O-induced increase in mean VMCA can be completely reversed by hyperventilation.

Inhalation injury treated with extracorporeal CO2 elimination

A 38-year-old male was admitted to the intensive care unit with a full-thickness burn involving 30 per cent of his total body surface area (TBSA) and severe inhalation injury. Respiratory failure developed within 54 h and CO2 could not be eliminated, even by very invasive mechanical ventilation. Because of the patients age and the minor extent of the burned TBSA, we started extracorporeal CO2 elimination (ECCO2-R) and continued ECCO2-R for 30 days, when the patient was weaned from ECC. The clinical course during ECCO2-R was complicated by major bleeding from a thoracotomy tube, from the site of tangential excision and by four septic episodes. Lung biopsy was performed twice on day 29 (during ECCO2-R) and day 58 (after ECCO2-R) after admission and revealed bronchiolitis obliterans without tendency to recovery. The patient died of sepsis with multiorgan failure on day 81 after trauma.

Hypocapnia reverses the fentanyl-induced increase in cerebral blood flow velocity in awake humans.

Investigations on the effects of opioids on cerebrovascular dynamics have repeatedly demonstrated mild to moderate increases in cerebral blood flow velocity in the middle cerebral artery (CBFVMCA), cerebral blood flow, and cerebrospinal fluid pressure in humans and animals. However, the influence of hypocapnia on these fentanyl effects has not been investigated. We compared mean CBFVMCA during normo- and hypocapnia before and after administration of fentanyl (2.5 μg/kg i.v.) in 20 awake humans. During normocapnia (end-tidal carbon dioxide [ETCO2] 40 mmHg) fentanyl significantly increased mean CBFVMCA (60 ± 10 cm/s vs. 81 ± 12 cm/s [mean ± SD]; p < 0.01), whereas during hypocapnia (ETCO2 25 mmHg) mean CBFVMCA values were identical (40 ± 7 cm/s vs. 40 ± 7 cm/s) before and after fentanyl administration. These results confirm previous findings that administration of fentanyl increases CBFVMCA, but, more importantly, clearly indicate that hypocapnia reverses this potentially undesirable effect.

Influence of urapidil on cerebrospinal fluid pressure in humans with uncompromised intracranial compliance

ObjectiveDetermine the influence of urapidil on mean lumbar cerebrospinal fluid pressure (CSFP), mean arterial pressure (MAP), mean central venous pressure (CVP) and heart rate (HR) in awake humans without any evidence of cerebral or cardiovascular disease.DesignOpen, single-dose volunteer study.InterventionsCSFP was measured via a spinal needle after i.v. injection of a single dose of 0.2 mg kg−1 urapidil in six volunteers (2 female, 4 male).Measurements and resultsAfter administration of urapidil, CSFP increased from 7±1 mmHg to 10±1 mmHg (p<0.05), MAP decreased from 88±7 mmHg to 74±5 mmHg (p<0.05), CPP decreased from 81±7 mmHg to 64±5 mmHg (p<0.05) and CVP decreased from 0±1 mmHg to −3±1 mmHg (p<0.05).ConclusionOur data suggest that in humans with presumed normal intracranial compliance the administration of urapidil causes a small but statistically significant increase in CSFP due to a parallel decrease in MAP.

Low-dose sufentanil increases cerebrospinal fluid pressure in human volunteers.

Although sufentanil is frequently used in neuroanesthesia, the effect of the drug on intracranial pressure is still controversial. In our study, we used an invasive measurement technique to study the effects of 0.1 micrograms/kg-1 sufentanil on mean lumbar cerebrospinal fluid pressure (CSFP), mean arterial pressure (MAP), cerebral perfusion pressure (CPP), central venous pressure (CVP), heart rate (HR), and end-tidal dioxide (ETCO2) in five human volunteers. After i.v. injection of sufentanil, mean lumbar CSFP increased from 6 mm Hg to 12 mm Hg (p < 0.05), and mean CPP decreased from 92 mm Hg to 78 mm Hg (p < 0.05), whereas MAP, CVP, HR, and ETCO2 remained stable. The results of this study clearly show that even a low dose of sufentanil transiently increases lumbar CSFP in volunteers with uncompromised intracranial compliance.

Surfactant replacement therapy in acute respiratory distress syndrome from viral pneumonia

A modified natural surfactant was administered to a patient with life-threatening adult respiratory distress sydrome caused by viral pneumonia. Subsequently, there was a marked improvement in gas exchange. In order to assess the mechanism for improved oxygenation, computed tomography of the lungs was done. Quantitative analysis of the scans taken before and after surfactant administration indicates that improvement in gas exchange was largely due to the expansion of underinflated and collapsed lung areas. Although this is a single case report, it provides insight into the possible beneficial effect of instilled surfactant in severe respiratory distress from viral pneumonia.

Nitrous oxide and cerebral haemodynamics

Sir, The study by Strebel et al. (1) evaluates the influence of nitrous oxide on cerebral hemodynamics. Although they clearly showed that N,O is a potent cerebral vasodilator during normocapnia leading to an increase in cerebral blood flow velocity (CBFV), their conclusions and recommendations are somewhat misleading. When the authors write that “in patients with an elevated ICP or a reduced intracranial compliance, the possibility that nitrous oxide may be a potent cerebral vasodilator should be considered,” they forget that these patients, who were not the subject of their study, should be hyperventilated intraoperatively, while the cerebral healthy (!) patients examined by the authors were normoventilated. A recent study (2), however, was able to show in cerebral healthy volunteers that the increase in CBFV caused by 50% N,O in oxygen can be completely reversed by hyperventilation. In our opinion it is always problematic to attempt to draw direct conclusions for the treatment of patients from findings made in healthy individuals.