Building Interoperability Forward Versus Retrospectively Fitting it, an Approach for Developing Countries (Preprint)

Abstract

BACKGROUND The aim of the study was to conduct an in-depth review and assessment of implementation of ehealth systems implementation in Zimbabwe with objectives that include determining the status of implementation of ehealth interoperability in the country and to propose a framework for guiding implementation of ehealth interoperability there. The framework will provide guidance to developers as well as decision makers at the health ministry level in government. The expectation is that such a framework can also be of value in other developing country contexts, to their developers as well as decision makers at the health ministry level in their governments. The rationale for the study is that there are frameworks for implementation of ehealth interoperability, however these current frameworks the majority of them were developed within the context of developed country settings. Furthermore there are developments in the area of interoperability in developing country contexts such as the OpenHIE initiative that are very promising, which add to the relevance of this study, and which call for an update on the topic of interoperability in developing country contexts. The study employed a cross-sectional study design and qualitative methods, main data source were interviews of key informants. The key informant sources consisted of e-health experts from the Health Ministry in Zimbabwe, e-health systems developers from the public and private sectors, vendors of health IT systems and non-governmental organisations (NGOs) active in e-health implementation and support. Findings include evidence of much activity in respect of ehealth implementation in Zimbabwe, from point of care systems and electronic health records, to laboratory systems and pharmacy systems, as well as administrative systems. Development of these systems is funded by the government of Zimbabwe as well as its international donor funders. However the degree of implementation was not ranked as high, with a potential for greater and more implementation of ehealth to meet the country’s needs. A five point framework has been proposed from the results of the study to address some of the key barriers to strategies and approaches for ehealth interoperability implementation in Zimbabwe. A number of future research directions have also been delineated in the study. The study was therefore fruitful in its results on the topic of implementation of ehealth interoperability in Zimbabwe with some implications for other developing country contexts. OBJECTIVE The aim of the study was to develop a framework intended to provide guidance and strategies for implementation of e-health interoperability in developing country settings, using Zimbabwe as a case. The specific study objectives were: 1. Determine the current status of e-health implementation in Zimbabwe; 2. Determine the current status of e-health interoperability in Zimbabwe; 3. Determine the barriers and enablers in the implementation of e-health interoperability in a developing country context; 4. Develop a framework for implementation of e-health interoperability in a developing country context. METHODS This study pursued a qualitative approach. Face to face structured interviews were conducted with the ten participants. The average duration of each interview was one hour and all interviews were recorded on a voice recording device. The interview questions were centered on research objectives one, two and three. These are: “To determine the current status of e-health implementation in Zimbabwe”; “To determine the current status of e-health interoperability in Zimbabwe”; and “To record barriers and enablers in ehealth interoperability implementation in Zimbabwe”. The collected data was useful in informing the development of the framework for implementing e-health interoperability in a developing country context. Interviews were considered ideal for data collection since there was a need to get a deeper comprehension of the matters pertaining to e-health implementation and e-health interoperability in Zimbabwe. A professional transcriber was used to transcribe the interviews. The researchers then verified all the interview transcriptions before data analysis commenced. After analysis of the interviews, a document review was conducted. This was part of the triangulation process which is typical of qualitative research. The purpose of the document review was to corroborate and compare findings from the interviews. In this regard the researcher collected relevant documents from the interviewees and downloaded others from the internet. Documents reviewed comprised of national health planning documents, government acts, bills, policies and charters, e-health interoperability documents as well as reports on health information systems. RESULTS 4.1 Results 4.1.1 Current status of ehealth implementation in Zimbabwe Two major themes that emanated from the interviews were the types of Health Information Systems (HISs) and the maturity level of e-health systems implementation. Several health information systems were identified including Human Resources systems, Laboratory Information Systems, administrative systems, and systems for HIV/AIDS. Below is an example of a quote from respondents on the types of systems in Zimbabwe. “We have what is called ePMS, electronic Patient Management System, this system is currently managing HIV/AIDS data in the country. I think this is being supported in about 680 sites in the country”. – Respondent 1 A number of other systems were mentioned by name by the respondents including DHIS2 for routine health data capture, PACS for medical imaging, Laboratory Information System for laboratory services, and for administration purposes Public Management System and SAP Healthcare. Findings from the respondents suggested that the maturity level of e-health systems implementation was low. Thematic analysis showed that 80% of the respondents (Respondents 1, 2, 3, 4, 5, 6, 7 and 8) agreed that the maturity level of e-health systems implementation in Zimbabwe was low. “We are at the basic level of basic data entry and basic analytics, that is where we are”. – Respondent 1 “We're at 40% (4/10) maturity, being that we have a lot of systems, but we're not even using any of the systems to their maximum potential”. – Respondent 5 4.1.2 Current status of ehealth interoperability in Zimbabwe Thematic analysis revealed that the status of e-health interoperability in Zimbabwe was generally low. “Most of the systems are working in isolation, there is a need for the systems providers to work hand in hand, to have a proper interface”. – Participant 7 “There is no recognised architecture for health information exchange. I have been around since 2016… This EHR has got its own architecture… its’ own registries ...”. – Respondent 5 However, since the national EHR system was being developed locally (in Zimbabwe), some respondents expressed the status of e-health interoperability in terms of which health information system could interoperate with it. “EHR and DHIS-2 can communicate at this stage”. – Respondent 2 4.1.2.1 Technological Barriers Respondent 6 revealed that point to point integration was being used to connect systems. Point to point integration refers to the use of custom software to connect two applications and systems. It is simple and can be easily implemented. However, it is not scalable since its efficiency deteriorates as more applications and systems are integrated. “We currently don’t have e-health interoperability. We have point to point integration but not interoperability”. – Respondent 6 Responding to whether SAP Healthcare System could talk to DHIS, Respondent 7 said the following: “Interfacing is possible but presently they are not communicating … We are currently working on the interface of SAP and DHIS, it is work in progress”. – Respondent 7 4.1.2.2 Organisational Barriers Respondent 2 pointed to the challenges of unclear lines of command when he said the following: “For example, I have my co-Deputy Director, who is in charge of ICT, but when you then look at his area of operation, there is going to be infringement into other spaces and the other way around. If those things can be set up to say, where to start and there is harmonisation.” – Respondent 2 Participant 10 responding to the question regarding the need for additional training of university graduates in order to qualify them to develop interoperable e-health systems. We have classified this is an organisational concern, he said the following: “People coming out of university at the moment… will need quite a lot more knowledge to design interoperable solutions. I think both in terms of technologies like individual standards, even the functionality that are delivered by the components of a health information exchange, which is one of the major enablers of interoperability”. – Respondent 10 4.1.2.3 Terminology Barriers Fast Healthcare Interoperability Resources (FHIR) is a recent standard for exchanging healthcare information electronically [29]. FHIR (pronounced as “fire”) was developed by Health Level Seven (HL7), a standards development organisation for clinical data, and was developed after HL7 version 2 and HL7 version 3. This is what Respondent 4 said about it: “And then we talked about systems wanting to communicate with each other. The question would be how do they communicate? You can think HL7 or FHIR”. – Respondent 4 Respondent 9 elaborated further on the suitability of the FHIR standard for use in developing countries or in settings where there are no standards yet, stating the following: “HL7 FHIR has more potential in Africa and Asia than it does in places like Europe and the US,..because they have already put together these giant, monolithic systems that are already using things like CDA; and to get them to change to something like FHIR is going to take more effort than it would here where you have these emerging ecosystems that don't yet have standards in place.” – Respondent 9 Respondent 2 highlighted the need for cross country common standards and terminologies to ensure the continuity of care beyond borders, here is the quote: “There is no way you can talk of interoperability without the standards. We need to go for standards that cut across other countries, we don't need to go for standards that only work in Zimbabwe. For example, Namibia, Botswana, Zimbabwe and South Africa have relations, we have chronic patients on ART moving between these countries.” – Respondent 2 4.1.2.4 Legal and Regulatory Factors The thematic analysis revealed that 70% (7 out of 10) of the respondents mentioned legal and regulatory barriers. The sub-themes that emerged were: e-health strategy, governance, legal and regulatory issues, policy, legal statutes, professional bodies in healthcare and regulatory framework. “… we need a policy document, we need an e-health governance institution to actually regulate these issues of privacy, standards and ensuring that we meet the minimum interoperability requirements as defined by national and international standards”. – Respondent 1 Respondent 10 added that: “It's important to have that [legal framework], and to match that with training courses. And also have the compliance test or conformance testing. If you take those three pillars of it, you can have a very enabling environment for achieving interoperability, and you need all three of them.” – Respondent 10 Respondent 10 further proposed a maturity model for implementing e-health interoperability. “..And you need to implement it in a good way, like maybe a maturity model is a good way to do that. So, you start off just providing guidance, like this is the architecture and it would be good for everyone if we subscribe to this, and you get people voluntarily doing it. And then you provide connectathons and hackathons, to provide people with a forum to test compliance to the standards in the framework. And then slowly over time, you have training courses to help people build their systems to match it.” – Respondent 10 4.1.3 Interoperability Approaches In this study, participants advocated for a loosely coupled approach. Respondent 9 proposed a loosely coupled approach through the OpenHIE. “One of the key things with OpenHIE is that it provides an architecture, and is designed to be an architecture first and foremost. And from there, there are technologies, what are called reference technologies that can be used to address some of those architectural components and problems. So, the overall architecture is looking at some standardized workflows, and data standards”. – Respondent 9 Respondent 5 implied the OpenHIE framework when he stated the following: “I know people like Jembi, they have done this nice architecture, you are supposed to look at your entity as an enterprise…this is how things interoperate, these are your standards. For me, that would be the first thing”. – Respondent 5 4.1.4 Document Review and Nvivo Analysis The document review and Nvivo analysis was done as part of triangulation of the interview data. Their results delivered similar themes to those of the interview data from the key informants, and were also similar in respect of information systems implemented in Zimbabwe, as well as in classifying the level of maturity of ehealth implementation and interoperability. CONCLUSIONS Conclusions This case study on ehealth interoperability in Zimbabwe yielded much data. Findings include evidence of much activity in respect of ehealth implementation in Zimbabwe, development of these systems is funded by the government of Zimbabwe as well as its international donor funders. However the degree of implementation was not ranked as high, with a potential for greater and more implementation of ehealth to meet the country’s needs. The framework which has been proposed for implementation of ehealth interoperability proposes that in developing country contexts, implementation of interoperability can be done forward, alongside increasing and growing implementation of ehealth. In the framework various strategies are proposed for achieving this aim from measures such as conformance testing to leveraging on global public goods for health. The study has therefore been fruitful, its other contribution is that it factors in promising developments in ehealth interoperability in developing countries such as OpenHIE. There are a number of future research directions emanating from the study, such as validation of the proposed framework. Specific points from within the framework also constitute future research directions, such as conducting a study on conformance testing and hackathons in the context of ehealth interoperability in developing countries.