Facilitating health information exchange in low- and middle-income countries: conceptual considerations, stakeholders perspectives and deployment strategies illustrated through an in-depth case study of Pakistan
Background Health information exchange (HIE) may help healthcare professionals and policymakers make informed decisions to improve patient and population health outcomes. There is, however, limited uptake of HIE in many low- and middle-income countries (LMICs). While resource constraints are an obvious barrier to implementation of HIE, it is important to explore what other political, structural, technical, environmental, legal and cultural factors may be involved. In particular, it is necessary to understand associated barriers in relation to context-specific HIE processes and deployment strategies in LMICs with a view to discovering how these can be overcome. My home country Pakistan is currently struggling to implement HIE at scale and so I undertook a detailed investigation of these issues in the context of Pakistan to generate insights on how best to promote uptake of HIE in Pakistan and in LMICs more generally. Aims The concept of HIE is evolving both over time and by context. To gain a clearer understanding of this terrain, I began by identifying different definitions of HIE in the literature to understand how these had evolved and the underlying conceptual basis for these changes. Second, I sought to understand the barriers and facilitators to the implementation and adoption of HIE in LMICs. Building on this foundational work, I then sought to explore and understand in-depth stakeholders perspectives on the context of and deployment strategies for HIE in Pakistan with a view to also identifying potentially transferable lessons for LMICs. Methods I undertook a phased programme of work. Phase 1 was a scoping review of definitions, which involved systematically searching the published literature in five academic databases and grey literature using Google to identify published definitions of HIE and related terms. The searches covered the period from January 1900 to February 2014. The included definitions were thematically analysed. In Phase 2, to identify barriers and facilitators to HIE in LMICs, I conducted a systematic review and searched for published and on-going (conference papers and abstracts) qualitative, quantitative and mixed-method studies in 11 academic databases and looked for unpublished work through Google interface from January 1990 to July 2014. Eligible studies were critically appraised and then thematically analysed. Finally, in Phase 3 I conducted a case study of HIE in Pakistan. Data collection comprised of interviews of different healthcare stakeholders across Pakistan to explore attitudes to HIE, and barriers and facilitators to its deployment. I also collected evidence through observational field notes and by analysing key international, national and regional policy documents. I used a combination of deductive thematic analysis informed by the theory of Diffusion of Innovations in Health Service Organisations that highlighted attributes of the innovation, the behaviour of adopters, and the organisational and environmental influences necessary for the success of implementation; and a more inductive iterative thematic analysis approach that allowed new themes to evolve from the data. The findings from these three phases of work were then integrated to identify potentially transferable lessons for Pakistan and other LMICs. Results In Phase 1, a total of 268 unique definitions of HIE were identified and extracted: 103 from scientific databases and 165 from Google. Eleven attributes emerged from the analysis that characterised HIE into two over-riding concepts. One was the ‘process’ of electronic information transfer among various healthcare stakeholders and the other was the HIE ‘organisation’ responsible to oversee the legal and business issues of information transfer. The results of Phase 1 informed the eligibility criteria to conduct Phase 2, in which a total of 63 studies met the inclusion criteria. Low importance given to data informed decision making, corruption and insecurity, lack of training, lack of equipment and supplies, and lack of feedback were considered to be major challenges to implementing HIE in LMICs, but strong leadership and clear policy direction coupled with the financial support to acquire essential technology, provide training for staff, assessing the needs of individuals and data standardisation all promoted implementation. The results of Phases 1 and 2 informed the design and content of Phase 3, the Pakistan case study. The complete dataset comprised of 39 interviews from 43 participants (including two group interviews), field observations, and a range of local and national documents. Findings showed that HIE existed mainly in/among some hospitals in Pakistan, but in a patchy and fragmented form. The district health information system was responsible for electronically transferring statistical data of public health facilities from districts to national offices via provincial intermediaries. Many issues were attributed to the absence of effective HIE, from ‘delays in retrieving records’ to ‘the increase in antibiotic resistance’. Barriers and facilitators to HIE were similar to the findings in Phase 2, but new findings included problems perceived to be the result of devolution of health matters from the federal to provincial governments, the politicised behaviour of international organisations, healthcare providers’ resistance to recording consultations to avoid liability and poor documentation skills. Public pressure to adopt mobile technology frameworks was found to be a novel facilitator whereas sharing regional health information with international organisations was perceived by some participants as disadvantageous as there were concerns that it may have enhanced espionage activities in the region. Conclusions HIE needs to be considered in both organisational and process terms. Effective HIE is essential to the provision of high quality care and the efficient running of health systems. Structural, political and financial considerations are important barriers to promoting HIE in LMICs, however, strong leadership, vision and policy direction along with financial support can help to promote the implementation of HIE in LMICs. Similarly, the federal and provincial governments could play an important role in implementing HIE in Pakistan along with the support of international organisations by facilitating HIE processes at federal and provincial levels across Pakistan. This however seems unlikely for the foreseeable future. At a meso- and micro-level, HIE in Pakistan and other LMICs could be achieved through using leapfrog mobile technologies to facilitate care processes for local organisations and patients. Specifically, the study on Pakistan has highlighted that LMICs may achieve modest successes in HIE through use of patient held records and use of now ubiquitous mobile phone technology with some patient and organisational benefits, but scaling these benefits is dependent on the creation of national structures and strategies which are more difficult to achieve in the low advanced informatics skill and resource settings that characterise many LMICs.