A review of the clinical value of mechanical ventilators and extracorporeal membrane oxygenation (ECMO) equipment

Abstract

Acute healthcare providers operate large, diverse medical equipment inventories. Resources for managing these inventories is frequently scarce so must be prioritised such that maximum benefit is conferred per unit of expenditure. This review identifies publications which have discussed the clinical value conferred by mechanical ventilation (MV) and by extra corporeal membrane oxygenation (ECMO). Respectively, mechanical ventilators and ECMO units are necessary to deliver these therapies. Systematic searches for publications which discuss the clinical value conferred by MV and by ECMO were conducted. The identified articles included reviews, prospective studies, retrospective studies, and models. Most presented findings in terms of the cost-effectiveness ratio. The patient populations studied, and analytical methods used varied widely. The clinical value conferred by MV varied with dependencies on several factors including the age- and disease- profile of the patient population. It was not possible to infer these dependencies from the literature which exists for ECMO. More relevant literature existed for MV, the more mature technology, than did for ECMO. The ECMO literature also tended to be more recent and included more modelling studies and fewer prospective studies. The data extracted could inform estimates of the clinical value likely to be delivered by mechanical ventilators operated by a specific institution. Estimates for ECMO are likely to carry greater uncertainty than those for MV.

Keywords: Clinical Engineering; Extracorporeal membrane oxygenation; Mechanical ventilation; medical equipment management; prioritisation; value.

Fig. 1

Results of the literature search for MV and ECMO publications.

Fig. 2

MV QALE values from included studies. The label on the vertical axis shows the author's name, year of publication and a short descriptor of the patient cohort; “consecutive” indicates that the cohort consists of consecutive patients seen at one or more centres, “composite” indicates a modelled cohort not intended to represent a specifically defined patient cohort. The marker colour denotes the quality of the analysis, green denotes high quality, orange denotes moderate quality, and red denotes low quality. Error bars show the standard deviation, where available. Where more than one marker appears for a single category, this indicates that the article published multiple values (e.g., Hung et al. published separate values for patients cognitively able to complete the quality-of-life questionnaire and for those unable to do so), consult the supplemental material for further details. Data are presented by order of publication, with the earliest at the top of the plot. Acronyms: ARDS = acute respiratory distress syndrome, COPD = chronic obstructive pulmonary disease, LTOT = long-term oxygen therapy, PMV = prolonged mechanical ventilation, UTI = urinary tract infection.

Fig. 3

ECMO QALE values from included studies. The label on the vertical axis shows the author's name, year of publication and a short descriptor of the patient cohort; “consecutive” indicates that the cohort consists of consecutive patients seen at one or more centres, “composite” indicates a modelled cohort not intended to represent a specifically defined patient cohort. The plots on the left show the total QALE, the plots on the right show the incremental QALE with respect to the comparator technology, where available. The marker colour denotes the quality of the analysis, green denotes high quality, orange denotes moderate quality, and red denotes low quality. The error bars for the Gravesteijn et al. data show the credibility interval for the QALE estimate. Data are grouped by the different ECMO variants, the “ECMO (all variants)” data is from articles which presented results which included data from more than one ECMO variant. Within each group, data are presented by order of publication. The Agus et al. [40] data is omitted, as this group used a one-year horizon for their calculations, so their results are not comparable to those of other studies, which used lifetime horizons. Acronyms: ARDS = acute respiratory distress syndrome, ASY = asynchrony, CA = cardiac arrest, CS = cardiogenic shock, PEA = pulseless electrical activity, VF = ventricular fibrillation, VT = ventricular tachycardia.

Fig. 4

Plots of the dependence of QALE on age, extracted from tabulated data presented by Dewar et al. [65], Linko et al. [68] and Hung et al. [61]. The Hung et al. data is aggregated data for patients with chronic obstructive pulmonary disease (COPD), heart disease, intracranial or spinal injury, septicaemia or shock, or stroke. “Partial cognition” refers to patients cognitively able to complete a quality-of-life questionnaire, “poor cognition” refers to those unable to do so. Error bars show standard deviations, where available.