A Framework for Assessing Healthcare Facilities in Low-Resource Settings: Field Studies in Benin and Uganda

The aim of this paper is to present and validate a framework for assessing healthcare facilities in low-resource settings to collect evidence and inform policies on the harmonisation, regulation and contextualised design of medical devices. A literature review and focus groups with several experts of medical device design, clinical engineering, health technology assessment and management, allowed the creation of a protocol, comprising two parts: a semi-structured interview and electrical safety measures. Three hospitals were assessed in Benin and three in Uganda. All the health centres resulted to be facing typical challenges for low-resource settings, including the lack of funding, expertise, a well-established maintenance program, spare parts and consumables, and unreliable power supplies. As there is a paucity of information regarding low-resource settings, the proposed framework can be used by clinical or biomedical engineers to assess and thereby propose actions for improving the conditions of healthcare settings.


Introduction
United Nations' (UN) 2030 Agenda for Sustainable Development includes 17 Sustainable Development Goals (SDGs), which are an urgent global call for action [1]. SDG3, Good health and well-being, strives for ensuring healthy lives globally, emphasizing the access to quality essential healthcare services, which is currently inequitable at the expenses of low-and middle-income countries (LMICs) [1,2].
During the Sixtieth World Health Assembly, the World Health Organization (WHO) recognized the severe implications of the inappropriate provision of health technologies to support the quality of health in developing countries. The lack of appropriate technologies was recognized as a barrier to the achievement of the SDGs, specifically SDG3 with its specific target of Universal Health Coverage (UHC), which impacts on the guarantee of human rights [3].
Patients and healthcare providers in LMICs have limited access to medical devices (MDs) [4]. One barrier for the accessibility to medical technologies is the lack of funds and the high costs of imported MDs. However, if this had been the main and only problem, donations would have been a solution [5]. In this regard, WHO estimates that more than 80% of medical equipment in LMICs is donated [6], but only 10-30% of this is operating [7], due to the lack of spare parts [8], of specialized personnel [9,10], of a good maintenance and management systems [5,11] and for the harsh environmental conditions.
Other more impelling barriers derive from the fact that medical equipment is largely designed for and manufactured Licia Di Pietro and Davide Piaggio contributed equally to this work.

Electronic supplementary material
The online version of this article (https ://doi.org/10.1007/s4084 6-020-00546 -3) contains supplementary material, which is available to authorized users. in high-income countries (HIC), following good manufacturing practices, international standards and minimum requirements, optimized for well-structured facilities, resulting difficult to be implemented in low-resource settings (LRSs) [12]. Medical equipment is designed to operate in a clean, sterile, climate-controlled environment, with a reliable main power supply, a good maintenance system, a working supply chain and expert healthcare operators. However, in LMICs, the safe and efficient use of MDs is hindered by harsh environmental and climatic conditions (e.g., high temperatures, humidity and dust), unreliable main power supplies and poor infrastructures [13,14]. These conditions cause frequent failures and trigger a higher demand for spare parts, which are expensive and difficult to find, making the maintenance of MDs as problematic as their acquisition [15]. As regards local hospitals LMICs, the challenges are exacerbated by the underlying structures of the buildings [16][17][18][19][20][21] (see Online material [1][2][3]. Given the paucity of information on the conditions of LRSs, several authors have performed reviews, interviews and observations, aimed at assessing the adequateness and appropriateness of local hospitals to deliver surgical care [16,17], or intensive care [18][19][20][21] or to manage non-communicable diseases (NCDs) [22] such as hypertension [23]. In 2011, Hsia et al. [16] ran a survey based on the WHO list of essential surgical services, analysing hospitals in Kenya, Rwanda, Tanzania, Uganda and Ghana, focusing on different variables, including basic infrastructure, medicine storage capability and quality systems. The authors concluded that none of the analysed countries had the proper infrastructures for delivering surgical care. The same problems along with supply chain difficulties in terms of equipment and supplies, old infrastructures not big enough for the ever-growing population, a limited number of beds and monitored beds, were identified by Albutt et al. [17]. They also stressed the fact that the equipment, if present, is outdated and works intermittently, lacking maintenance. In some cases, the situation is exacerbated by the differences between rural and urban areas [21,22].
Accordingly, the authors of this paper decided to implement and test a systematic framework, consisting in semistructured interviews with qualified personnel and direct measurements campaigns, for the assessment of health centres and hospitals in LRSs, given the lack of a standard protocol and of tangible, quantifiable and comparable information regarding these settings.
This paper presents the framework and its application during two field studies performed in Uganda and Benin.

Two Countries of the Developing World: Uganda and Benin
The UN created the Human Development Index (HDI) to evaluate the development of a country, based on different factors, including economic growth, life expectancy at birth, education and the standards of living [24]. A developing country, or a LMIC, is a country with few resources, a low HDI compared to other countries, and a gross national income per capita (GNI) below 4035$. Uganda, an Eastern African country with a GNI of 620$, and Benin, a Western African country with GNI of 870$, are both ranked as low-income countries (LICs) and are among the 25 poorest countries in the world [25]. Table 1 summarises some relevant data for both countries in comparison with a typical HIC.

The Ugandan National Health System
Uganda is organized into four administrative regions (i.e., Northern, Eastern, Central and Western), which are further The multifaceted realities of the departments are difficult to map out and present differences, mainly between the north and the south of the country. In addition to the ordinary health structures, there are numerous vocational hospitals spread around the country [31].

Selection of Health Facilities
Benin and Uganda were selected as first destinations, because of our previous experiences and the networks. Local contacts were able to link us with the hospitals, where we performed our assessment.
The selected health structures were:

Questionnaire Preparation and Validation
The questionnaire was drafted during focus groups among experts of medical device design, contextualised design, health technology assessment and management and clinical engineering (see Fig. 1). During these focus groups, relevant literature was used and integrated by the experts' knowledge to define the different questions and sections of the questionnaire: in particular, the sections on the general characteristics of the facility, electrical access, human resources and facility environment were based on Refs. [32,33], those on medical electrical equipment on Refs. [34,35] and the one on patient data management on Ref. [31]. At first, the questionnaire was conceived to be online, but we realised that more relevant information could be captured if it was administered in person as a semi-structured interview (see Online material 4). The first draft of the questionnaire was circulated among the participants of the first focus group, for internal validation and possible corrections. Once the final version was ready, it was tested for further validation on an anaesthesiologist with working experience in hospitals in Sierra Leone (with Emergency).

Electrical Safety Measures
Electrical safety measures were carried out in accordance with IEC 62353-2015, using an electrical safety analyser (ESA620) by Fluke (see Online material 5). The MDs analysed were selected depending on availability at the health centres at the time of the assessment. Our inspection protocol included: • Voltage and earth check on the main power supply; • Visual inspections to check the integrity of the devices, the cables and the accessories; • Protective earth measurement; • Insulation resistance measurement; • Leakage currents measurement; • Functionality tests (e.g., ECG wave simulation).

Questionnaire
The final questionnaire (see Online Material 6) consisted of 71 questions, organised in 9 sections: Introduction and Authorisation, Personal Information, Facility Information, Facility general characteristics, Electrical Access, Human Resources, Facility Environment, Medical Electrical Equipment, and Database. The questions were of different type (e.g., Yes/No, multiple choice etc.) and part of them was quantitatively assessing some dimensions, others were assessing how some dimensions were perceived by the interviewee. The latter had six possible answers, based on a 5-step Likert-type scale, i.e., "Very low", "Low", "Middle", "High", "Very high", "Do not know".
The six interviewees, one per hospital, were all males (21-50 years old), had a college or university degree and were 4 biomedical engineers, 1 medical doctor and 1 nurse. All had been working in those roles for an average of 7 years.

General Information About the Facility
All the assessed hospitals are public and third-level structures (urban), but the one in Zinviè, which is private and a second-level structure (semi-urban). All the hospitals rely on piped water, but H1, which relies on tanker water (also available in H2, H3 and H6), or water from a well (also available in H4, H6). H1 also has structures for collecting rainwater. Only some of the structures have a functioning landline telephone (H2, H3, H5). Other basic facilities are: a mobile phone (H1, H2, H3, H5), a short wave radio (H2), a computer (all but H4), an internet service (H2, H3, H5, H6), an ambulance (all but H5). As regards the access to electricity, all the hospitals can rely on the central supply, most of them have a generator (H1, H2, H4, H5, H6) and some of them have a solar electric system (H1, H2, H6) (see Table 2 for further information). All the hospitals reported a certain degree of incompatibility among the local sockets and the plugs of donated MDs (see Fig. 2).

Human Resources
All the facilities reported a chronic lack of doctors (1 for more than 30 patients), clinical officers (1 for more than 30 patients), of nurses (with an average of 1 nurse per 20 patients), and laboratory technicians. In general, most of the interviewees judged this situation as "poor" or "very poor".

Facility Environment
In general, most of the facilities were judged by the interviewee as poorly insulated or distant from dust, smoke, and undue noise. A better situation was found regarding the insulation or distance from foul odours, the lighting, and the adequateness of ventilation (see Table 3). Similarly, the disposal of different kinds of waste (e.g., non-risk waste, infectious waste, sharps etc.) was mostly rated as "very good", apart from H1, for which it was rated as "very poor". All the hospitals' environments are cleaned 5 + times per week. Figure 3 shows the distribution of essential MDs and services within the six hospitals. The list is ordered from the least available piece of equipment to the most available one. Complex devices like colonoscopes, mammographs, CTscanners and X-Ray machines are a rare find, compared to blood pressure machines, thermometers, pulse-oximeters, scales, and patient monitors. All the hospitals have a biomedical engineering department, except H1 and H4. The approach to medical device maintenance varies depending on the facility. Most of the structures that have a biomedical engineering department are in charge of the preventive and the corrective maintenance (H2, H5, H6), with the exception of H3 that only follows corrective maintenance practices. However, the structures that do not have a biomedical engineering department (H1, H4) are only relying on on-call biomedical engineers/technicians, thus following only corrective maintenance practices.

Medical Electrical Equipment, Maintenance, and Management
The most recurrent problematics resulted to be the lack of funding, of essential medical device, of spare parts and consumables, of expertise. One of the hospitals (H1) also denounced the inexistence of a policy regulating the donation of MDs.
As regards the management, all the hospitals but one (H4) have a form of inventory: 3 hospitals have a paper-based one (H1, H2, H3) and 2 have or are transitioning to a computerised one (H5, H6). The most recorded information regards the year of manufacture, the type, the serial number, the year of acquisition, the class function, the date of service, the routine servicing, the reason for acquisition and the technical characteristics.

Electric Safety Measurements
We were able to perform these measurements in 4 hospitals (1 in Uganda, 3 in Benin). As regards the intensive care unit of H2, the inspected sockets were up to standards (voltage on the mains of 238 V; voltage between the neutral and the ground of 0.2 V). In H2, out of the five tested devices (i.e., 1 defibrillator, 3 patient monitors and 1 ECG), two of the patient monitors 1,2 did not pass visual inspection because they were lacking respectively the blood pressure cuff, the ECG cables and the power cable, and the ECG cable. Moreover, the ECG cables of the defibrillator 3 were not working, in fact, when inputting a simulated signal nothing was showing on the screen. All the other devices passed the inspection.  Fig. 2 The compatibility between the local sockets and the plugs of donated medical devices for Uganda and Benin (voltage on the mains of 240 V; voltage between the Neutral and the Ground of 2.6 V). H6 had functional sockets in the ambulatory, however the maintenance lab and two surgical theatres had no ground, with voltages between the neutral and the ground of 89.9-134.5 V. Throughout the different buildings we noticed that the grounding system was a common problem, along with the inversed polarity of the sockets. In H5, all the inspected departments (i.e., Biochemistry, Haematology, HIV lab, new equipment room) presented troubles with the earth system (with voltages between the Neutral and the Ground between 8 and 43 V), but the department of radiology. The report can be seen in Table 5.

Discussion and Conclusions
This paper introduces a framework to assess clinical locations, specifically in LMICs, through a semi-structured interview and electrical safety measures. Such a framework, tested in Uganda and Benin, is crucial for mapping out the different realities of healthcare locations of LRSs.
Many developing countries still lack access to MDs and equipment that are appropriate for their specific clinical needs because of poor regulatory controls [36]. More importantly, what they do have is often inappropriate due to a mismatch between working conditions and design constraints.  B  a  d  n  a  g  U  a  d  n  a  g  U  a  d  n  a  g  U  y  r  t  n  u  o  C  6  H  5  H  4  H  3  H  2  H  1  H  l  a  t  i  p  s  o  H  0  0  0  0  1  0  e  p  o  c  s  o  n  o  l  o  C  0  0  0  1  0  0  h  p  a  r  g  o  m  m  a  M  0  0  0  1  1  0  r  e  n  n  a  c  s  -T  C  0  0  0  2  1  0  e  p  o  c  s  o  r  t  s  a  G  1  1  0  1  0  0  e  r  t  n  e  c  n  o  i  t  a  m  i  n  a  e  r  t  n  a  f  n  I  1  1  1  1  3  1  e  n  i  h  c  a  M  y  a  R  -X  2  0  2  1  3  0  e  c  n  a  l  u  b  m  A  0  2  3  A  N  4  2  r  o  t  a  l  l  i  r  b  i  f  e  D  1  2  0  8  4  0  U  C  I  r  o  t  a  l  i  t  n  e  V  1  4  8  1  4  1  r  e  t  e  m  o  t  y  c  o  m  e  H  3  2  4  2  8   In Africa, for example, competent authorities, who control the quality and safety of MDs, often lack adequately trained staff for consultation [37] due to limited human capacity in Biomedical Engineering [38].
We believe that the problem of existing standards is that of generalism, or of a non-inclusiveness, which does not take into consideration all the specific realities and that, for this reason, is not reachable by them. This dilemma, however, can be addressed with an inductive method, which, by examining specific situations, can inform the writing of new regulations and standards that are more inclusive and universally applicable. It is only by studying and taking into account different contexts and types of users that the design of a safe and high-quality medical device for low-resource settings can be successful [39]. It is true that many challenges are common to these settings, but it is also true that each of them has its own identity and peculiarities. According to this, open-source and collaborative methods have the potential to improve the design of needs-based MD, offering specific solutions to problems not properly considered by current standards, oriented to well-structured healthcare environments. When properly deployed, the open design paradigm has the potential to increase access to medical technologies, reducing the management, maintenance, and repair costs due to the openaccess of device blueprints [40].
In this context, UBORA, the open Biomedical Engineering e-platform for collaborative design, can be effectively used to develop safe and effective MDs, and the relative spare parts [41], the analysis of the technical needs, the risk management process, legal aspects, safety criteria and performance data, fundamental for maintaining the compliancy of the repaired devices with the medical device regulation [42].
As shown by the results of the questionnaire, all the analysed facilities are inefficiently built, with poor isolation from foul odours, smog and undue noises, and unstable and unsafe power supplies. The scarcity of essential MDs, spare parts and consumables, together with a poor maintenance system, bolstered by the chronic lack of biomedical engineers, technicians and healthcare personnel (from nurses to specialised doctors), hinders the safe and efficient care of patients, impeding universal health coverage.
The electrical safety measurements confirmed the results of the questionnaire, highlighted the common issue of too high protective earths of some MDs, and, above all, clearly proved that in Benin there are problems with the grounding systems.
One limitation of the study is that it was not always possible to test MDs, because of the lack of authorisation or because they were all being used for the care of patients.
Nevertheless, the availability of the technical staff, who allowed the measurements and the interviews on the field, made us understand the urgency for the locals of the problems highlighted in this paper. Therefore, this study is an important starting point to frame the problem and present the framework which favoured a thorough investigation in the field and facilitated the subsequent processing of the data, showing itself a suitable, exportable and repeatable tool to offer an overview of healthcare locations in specific contexts. The study will be expanded to include more hospitals around Uganda, Benin and other LMICs in order to provide the basis for promoting awareness of the issues they face and towards global policy changes for health equity.
Acknowledgements The authors would like to acknowledge all the interviewees and all the locals who helped in making this possible. Special thanks go to Joshua Kiberu, Ambrose Ziwage, Marius Yabi, Funding DP and LP received support from the University of Warwick with two Warwick Impact Found grants supported by the EPSRC Impact Accelerator Award (EP/K503848/1 and EP/R511808/1). LDP, CDM and AA were supported the European Union's Horizon 2020 research and innovation programme under grant agreement N° 731053.
Availability of Data and Material Data related to this study are available upon request to the authors.

Compliance with Ethical Standards
Conflicts of interest The authors declare that they have no conflict of interest.

Ethical Approval
The study was performed in accordance with ethical approval REGO-2018-2283.

Consent to Participate
All the interviewees were asked and gave informed consent to participate.

Consent for Publication
All the interviewees were asked and gave informed consent to publication.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.