Mechanical ventilators for low- and middle-income countries: informing a context-specific and sustainable design

Abstract

Editor—The COVID-19 pandemic transiently raised the need for greater intensive care mechanical-ventilation capacity around the world. Whilst ICUs were under immense pressure globally, several high-income countries (HICs) did not ultimately require novel prototype ventilators, as standard ventilator supplies were sufficient to meet demand. In contrast, mechanical-ventilation capacity was, and largely remains, insufficient in the vast majority of low- and middle-income countries (LMICs), limiting access to critical care and surgery.1Baker T. Critical care in low-income countries.Trop Med Int Health. 2009; 14: 143-148Crossref PubMed Scopus (109) Google Scholar Standard ventilators used in HIC hospitals are unaffordable and unsustainable for hospitals with limited resources in LMICs. In response to the COVID-19 pandemic, some non-governmental organisations specifically began work on designing ventilators for LMICs. However, to our knowledge, no one has systematically surveyed a variety of LMIC sites to understand the contexts and challenges that need to be addressed in ventilator design. Here, we identify the key technical requirements for the design of a context-specific and sustainable mechanical ventilator for use across LMICs. We performed a cross-sectional survey aimed at providing an overview of mechanical-ventilation capacity and infrastructure in LMICs. The information was requested with reference to typical working patterns, not specifically to any surge in service demand, like that associated with the COVID-19 pandemic waves. Results from this survey and those available in the literature were critically considered to identify key technical parameters to support the design of a context-specific and sustainable mechanical ventilator. Ethical approval was not required for this survey because it did not require patient data. The study followed the Enhancing the Quality and Transparency of Health Research Network guidelines and standards for reporting qualitative research. The survey (Supplementary Appendix) consisted of 63 questions used to define respondents' profiles, their hospitals' characteristics, current mechanical-ventilation infrastructure and environment, and the estimated need for mechanical ventilation, associated maintenance and monitoring equipment. The survey was shared via e-mail to 16 anaesthesia and critical care national and international societies and healthcare institutions, and via social media, directly reaching out to a convenience sample of 81 LMICs as defined by the World Bank. Fifty respondents, mostly anaesthetists (45%) and intensivists (44%), with >6 yr experience (72%), from different large teaching and public hospitals (76%) contributed from 23 LMICs in Africa, Asia, and Central and South America. Characteristics of respondents and their institutions are presented in Supplementary Table S1, country-by-country details are listed in Supplementary Table S2, and the geographical distribution of responses is illustrated in Supplementary Figure S1. Supplementary Table S3 summarises the availability of power and gases. Loss of electricity was found to occur daily 19% (9/48), weekly 25% (12/48), less frequently 48% (23/48), or did not occur 8% (four/48). The majority of these outages (72% [31/44]) were reported to last <2 h, with back-up power from external fossil fuel generators and batteries reported to be supplied within 1 h in 91% (40/44). This electricity availability was similar to that reported in a survey of 10 hospitals in Tanzania.2Baker T. Lugazia E. Eriksen J. Mwafongo V. Irestedt L. Konrad D. Emergency and critical care services in Tanzania: a survey of ten hospitals.BMC Health Serv Res. 2013; 13: 1-9Crossref PubMed Scopus (74) Google Scholar As access to electricity varies within most LMICs, where larger hospitals typically have more reliable supply,3Adair T.H. Gay W.J. Montani J.P. Growth regulation of the vascular system: evidence for a metabolic hypothesis.Am J Physiol. 1990; 259: R393-R404PubMed Google Scholar the design of a context-specific mechanical ventilator needs to include an emergency battery to supply electricity, but our results suggest that the battery power may only be needed for a few hours (Table 1), limiting the cost for this relatively expensive component.Table 1Overview of key requirements for the design of a mechanical ventilator usable and sustainable in low- and middle-income countries. Bold text, requirements identified by this survey; italic text, requirements based on prior work,4US Food and Drug AdministrationCoronavirus disease 2019 (COVID-19) emergency use authorizations for medical devices.2020https://www.fda.gov/medical-devices/emergency-use-authorizations-medical-devices/coronavirus-disease-2019-covid-19-emergency-use-authorizations-medical-devicesDate accessed: January 19, 2022Google Scholar expert opinion, and on the authors' baseline experience of known needs. ∗For electricity international standards, see International Electrotechnical Commission guidelines: https://www.iec.ch/world-plugs; †98% (135/137) LMICs have either 220 or 230 V, AC, and 84% (115/137) LMICs have a frequency of 50 Hz; ‡listed in order from highest perceived importance from our survey (see Supplementary Fig. S6 for details); ¶apart from suction, all these monitoring priorities match the recent recommended standards of monitoring during anaesthesia and recovery;5Klein A.A. Meek T. Allcock E. et al.Recommendations for standards of monitoring during anaesthesia and recovery 2021.Anaesthesia. 2021; 76: 1212-1223Crossref PubMed Scopus (43) Google Scholar §calibration and testing systems need to consider environmental conversion factors (e. g. high altitude); ||tolerance error <10%; #30 ml for ventilation in neonates. AC, alternating current; CE, Conformité Européenne; DC, direct current; FDA, US Food and Drug Administration; LMICs, low- and middle-income countries; PEEP, positive end-expiratory pressure.ParameterRequirementElectricity supplyAC: 100–240 V; 50/60 Hz; DC: 12–30 V; battery operation >2 h∗†Compressed airAir compressor∗Oxygen supplyOxygen concentrator to deliver >10 L min−1∗ or green-labelled gas cylinder to support 2 h of operationElectricity and gas supplyMatch the national infrastructure (e. g. gas connectors, compatible gas pressures, electricity voltage and frequency, plugs, and sockets∗)Operating temperature15–35°COperating humidity20–70%Patient monitoring‡¶Pulse oximetry, noninvasive BP, electrocardiography, airway oxygen, pressure and flow, end-tidal carbon dioxide, and suctionInternet connectivityNetwork access for remote device monitoring (location, usage, self-calibration,§ and self-testing reports) and firmware updatesConsumables costs<US$ 50 per patient per daySupport: technical and trainingLocal or in country as a priority, long-term contract for remote support, built-in self-calibration system, product life cycleManufacture, assembly, and testingIn country as a priority, bench testing to meet regulatory standards; continuous operation >14 days for critical care applications; built-in self-testing system§Regulatory standardsCE marked or FDA approved (or similar)Modes of ventilationPressure and/or volume control; spontaneous breathing support if ventilation is required for >1 dayFraction of inspired oxygenUp to at least 80%Ventilatory frequency10–50 bpm; increments of 1 or 2||PEEP5–20 cm H2O, in 5 cm H2O steps or less||Inspired-to-expired ratio1:1 to 1:3Tidal volume50#–700 mL, in steps of 50 mL or less||Inspiratory pressure limits15–40 cm H2O, in 5 cm H2O steps or less||AlarmsGas or electricity supply failure, switched off, tidal volumes and pressures exceeded or not reachedFiltering and humidificationBacterial/viral filter; heat and moisture exchangerDisplay and ventilation monitoringSet and delivered inspiratory pressure, tidal volume, ventilatory frequency, PEEP, fraction of inspired oxygen, and ventilation mode Open table in a new tab Compressed air was always available in 35% of hospitals, either piped 68% (19/28) or via cylinders 29% (eight/28) (Supplementary Fig. S2). Oxygen was reported as always available in 74% of hospitals, at high pressure either piped 44% (20/46) or via cylinders 46% (21/46), or at low pressure via concentrators 11% (five/46). This oxygen availability was lower than observed in 99 Sri Lankan ICUs, where piped oxygen was available,6Haniffa R. De Silva A.P. Iddagoda S. et al.A cross-sectional survey of critical care services in Sri Lanka: a lower middle-income country.J Crit Care. 2014; 29: 764-768Crossref PubMed Scopus (34) Google Scholar and lower than in 10 hospitals in Tanzania.2Baker T. Lugazia E. Eriksen J. Mwafongo V. Irestedt L. Konrad D. Emergency and critical care services in Tanzania: a survey of ten hospitals.BMC Health Serv Res. 2013; 13: 1-9Crossref PubMed Scopus (74) Google Scholar Our result on oxygen availability matches a survey of 97 clinicians from public district and provincial hospitals in urban and rural areas in 19 countries throughout Africa, Asia, and South America, where 75% reported that oxygen was available most or all of the time7Ginsburg A.S. Van Cleve W.C. Thompson M.I.W. English M. Oxygen and pulse oximetry in childhood pneumonia: a survey of healthcare providers in resource-limited settings.J Trop Pediatr. 2012; 58: 389-393Crossref PubMed Scopus (51) Google Scholar and is similar to that reported from 97 Ugandan anaesthetists, where 63% always had access to oxygen.8Hodges S.C. Mijumbi C. Okello M. McCormick B.A. Walker I.A. Wilson I.H. Anaesthesia services in developing countries: defining the problems.Anaesthesia. 2007; 62: 4-11Crossref PubMed Scopus (245) Google Scholar Overall, our results were more conservative than or comparable with those obtained from a total of 293 respondents plus 10 hospitals across 22 LMICs, supporting that our findings are representative of a broader scenario. Considered together with evidence from the literature,2Baker T. Lugazia E. Eriksen J. Mwafongo V. Irestedt L. Konrad D. Emergency and critical care services in Tanzania: a survey of ten hospitals.BMC Health Serv Res. 2013; 13: 1-9Crossref PubMed Scopus (74) Google Scholar,6Haniffa R. De Silva A.P. Iddagoda S. et al.A cross-sectional survey of critical care services in Sri Lanka: a lower middle-income country.J Crit Care. 2014; 29: 764-768Crossref PubMed Scopus (34) Google Scholar,8Hodges S.C. Mijumbi C. Okello M. McCormick B.A. Walker I.A. Wilson I.H. Anaesthesia services in developing countries: defining the problems.Anaesthesia. 2007; 62: 4-11Crossref PubMed Scopus (245) Google Scholar our results on compressed air and oxygen availability suggest that the design of a mechanical ventilator for hospitals with limited resources may need to include an air compressor and either a gas cylinder or an oxygen concentrator (Table 1), where modular designs would allow integration of features as necessary. Most institutions (96%; 48/50) reported having emergency departments, operating theatres, and ICUs, and 4% (two/50) reported having only two of these three resources. Supplementary Figure S3 displays the number of ventilators that were functioning, non-functioning, and perceived to be required for each of these three departments. More than half of respondents (25/48) reported that ventilators were used 7 days a week, and 15 of these reported they were in use 24 h per day. ICUs had the greatest estimated requirement for ventilators with 28% (13/47) requiring more than 15, 40% (19/47) requiring from six to 15, and 32% (15/47) requiring from zero to five ventilators per Unit. Emergency departments and operating theatres required a relatively smaller number of ventilators, with 76% (35/46) and 50% (23/46) requiring between 0 and 5 per Unit, and only 9% (four/46) emergency departments and 22% (10/46) operating theatres requiring >10 ventilators. A lack of funds and technical maintenance support combined was the most frequent reason for non-functioning or not using ventilators, where a lack of funding was cited in 74% (34/46) of responses (Supplementary Fig. S4). The median reported minimum/maximum operating temperatures were 20/30°C, and humidity was 30/60%. The design needs to consider the different environmental working conditions, where temperature and humidity reported here were typically higher in LMICs than in HICs. Testing must be performed in a relevant range of temperature, humidity, and atmospheric pressure, depending on flow and pressure sensors used (Table 1). Detailed information on the frequency of anaesthesia mode is presented in Supplementary Figure S5, the perceived importance of monitoring techniques in Supplementary Figure S6, and the proposed sustainable costs of consumables/patient/day in Supplementary Figure S7. Further results are presented and discussed in the Supplementary Appendix. Table 1 summarises the key features to be considered when designing a mechanical ventilator to be useful and usable in LMICs, based on this survey's new and confirmatory findings, on evidence available from the literature,1Baker T. Critical care in low-income countries.Trop Med Int Health. 2009; 14: 143-148Crossref PubMed Scopus (109) Google Scholar, 2Baker T. Lugazia E. Eriksen J. Mwafongo V. Irestedt L. Konrad D. Emergency and critical care services in Tanzania: a survey of ten hospitals.BMC Health Serv Res. 2013; 13: 1-9Crossref PubMed Scopus (74) Google Scholar, 3Adair T.H. Gay W.J. Montani J.P. Growth regulation of the vascular system: evidence for a metabolic hypothesis.Am J Physiol. 1990; 259: R393-R404PubMed Google Scholar, 4US Food and Drug AdministrationCoronavirus disease 2019 (COVID-19) emergency use authorizations for medical devices.2020https://www.fda.gov/medical-devices/emergency-use-authorizations-medical-devices/coronavirus-disease-2019-covid-19-emergency-use-authorizations-medical-devicesDate accessed: January 19, 2022Google Scholar, 5Klein A.A. Meek T. Allcock E. et al.Recommendations for standards of monitoring during anaesthesia and recovery 2021.Anaesthesia. 2021; 76: 1212-1223Crossref PubMed Scopus (43) Google Scholar, 6Haniffa R. De Silva A.P. Iddagoda S. et al.A cross-sectional survey of critical care services in Sri Lanka: a lower middle-income country.J Crit Care. 2014; 29: 764-768Crossref PubMed Scopus (34) Google Scholar, 7Ginsburg A.S. Van Cleve W.C. Thompson M.I.W. English M. Oxygen and pulse oximetry in childhood pneumonia: a survey of healthcare providers in resource-limited settings.J Trop Pediatr. 2012; 58: 389-393Crossref PubMed Scopus (51) Google Scholar, 8Hodges S.C. Mijumbi C. Okello M. McCormick B.A. Walker I.A. Wilson I.H. Anaesthesia services in developing countries: defining the problems.Anaesthesia. 2007; 62: 4-11Crossref PubMed Scopus (245) Google Scholar, 9US Food and Drug AdministrationAppendix B: authorized ventilators, ventilator tubing connectors, and ventilator accessories.2021https://www.fda.gov/medical-devices/coronavirus-disease-2019-covid-19-emergency-use-authorizations-medical-devices/ventilators-and-ventilator-accessories-euas#appendixbDate accessed: January 19, 2022Google Scholar, 10Ginsburg A.S. Gerth-Guyette E. Mollis B. Gardner M. Chham S. Oxygen and pulse oximetry in childhood pneumonia: surveys of clinicians and student clinicians in Cambodia.Trop Med Int Health. 2014; 19: 537-544Crossref PubMed Scopus (11) Google Scholar on the authors' baseline experience of known needs, and on collaborators' expert opinion. The requirements listed in this table will not be fully applicable to the very wide variety of all LMIC sites. However, there will be similarities in resource-poor settings that are applicable to many sites. In this sense, the table considers the lower end of the resource spectrum and provides the ‘lowest common denominator’, which we think will capture some common issues of many sites. The trade-off between cost and sophistication is obvious, where in-country manufacture (or at least assembly) is needed for the design to be sustainable. The authors thank all respondents for their time and contribution, expert physicians and engineers who contributed to the preparation of the survey, Rafael Badell-Grau (OxVent, Abingdon, UK), Martin Dres (Sorbonne Université, Paris, France), Andrew D. Farmery (University of Oxford, Oxford, UK), and experts in healthcare in low- and middle-income countries for discussions and critical feedback, including Arturo Mendoza (British Embassy, Mexico City, Mexico), Bárbara Elena Montalvo Alvarez (Embassy of Cuba, UK), Benjin Joshua (Assist International), Caroline Murphy (King's College London, London, UK), Dylan Bould (University of Ottawa, Ottawa, ON, Canada), Laura Magee (King's College London), Lucio Luzzatto (Muhimbili University of Health and Allied Sciences, Dar es Salaam, Tanzania), Mark Ansermino (Centre for International Child Health and University of British Columbia, Vancouver, Canada), Masreshaw Demelash (Addis Ababa Institute of Technology, Addis Ababa, Ethiopia and Assist International, Ripon, California, USA), Paul Fenton (formerly professor of anaesthesia, College of Medicine, Malawi), Peter von Dadelszen (King's College London), Raoul Kamadjeu (Public Health Emergencies at UNICEF), and Robyn Meurant (NSF International and ACT-IVD). FF reports grants from the National Institute for Health Research (UK), the National Association of Academic Anaesthetists , and the Wellcome EPSRC Centre for Medical Engineering ; he is one of the volunteer directors of OxVent, a joint-venture social enterprise between Oxford University and King's College London for mechanical ventilation. All other authors declare no competing conflicts. The following is the Supplementary data to this article: Download .pdf (.43 MB) Help with pdf files Multimedia component 1