Over a 2-year period we identified 40 patients with Pao^Fio2 ratios below 150 mmHg (20 kPa) owing to either lung trauma or acute pneumonic lung disorders. Three-quarters showed maintained improvements in oxygenation with 'combination ventilation' (i.e. EHFO plus positive-pressure ventilation). Overall mortality was 46 per cent, a figure only marginally better than the quoted mortality of around 60 per cent for similar case-mix groups receiving optimal conventional ventilation alone. In most of the patients studied, optimal settings were a frequency of 90 cycles/min, an I:E ratio of 1:1, an inspiratory pressure of -35 to -45 cmH 2O, and an expiratory pressure of +5 cmH2O (Campbelland Nevin 1,99.3).
A similar study by a group at St George's Hospital, London, examined 'combination ventilation' in ARDS patients ( Taylor.eial 1994). They showed an initial improvement in 60 per cent of patients, but this was not sustained at 5 h. Overall mortality was 75 per cent.
The differences in outcome from these two initial assessments of 'combination ventilation' warrant an attempt at explanation. Although all patients in both studies met the entry criteria for extracorporeal support, there were major differences which may provide valuable information regarding future applications of the technique. In the St George's study, many of the patients had undergone extensive periods of conventional ventilation prior to transfer, whereas our study, albeit containing some patients (invariably pneumonias) who had received extensive conventional ventilatory support before entering the trial, consisted mainly of patients who commenced combination ventilation within 72 h of their initial insult. This 'rapid entry' was achieved by setting up a rapid response team which took the equipment to the primary hospital site. A total of 16 hospitals provided patients for the study over the 2-year period. Many patients showed a fluctuating Pao^Fio2 ratio during treatment with initial improvement followed by subsequent deterioration and a later recovery. These patients, many of whom were eventual survivors, would have been withdrawn from the St George's study after the measured deterioration.
Analysis of outcome related to the duration of the pre-EHFO period (Fig 2) strongly suggests that early treatment with combination ventilation is associated with a greater chance of survival than in patients who have received extensive periods of conventional ventilation before proceeding to combination therapy. There appears to be little advantage in using the technique in patients with ARDS secondary to a pneumonic process.
As with all new developments in intensive care practice, there is a need for a prospective randomized assessment to compare outcomes against conventional treatment. While this may be theoretically ideal, it presents many practical problems, not the least of which is the paucity of appropriate cases and the difficulty in standardizing case mix for associated multiple organ failure, as well as the duration and type of pre-entry conventional ventilation. Despite this, an assessment is under way, although several years may be required to produce any clinically significant results.
The Hayek oscillator is not without its problems. Despite considerable efforts by the manufacturers, it remains noisier than conventional ventilators. In addition, the continuous movement of the cuirass can easily cause skin burns. This complication can be avoided if adequate protection is taken by covering the skin contact points with a shock-absorbing protective material such as Granuflex. Although the majority of patients rapidly become accustomed to the oscillator, some spontaneously breathing patients may not tolerate the ventilatory mode. This problem can often be circumvented by starting oscillation at both low frequency and short span. There may be a requirement for sedation in some circumstances. Nevertheless, all the above problems should be overcome with sound sense and a sensitive approach to the feelings of both the patient and attendant nursing staff.
Was this article helpful?