Health Information Systems Workshop
Health Information Systems in Developing Countries Workshop
held on 28th September 2010 at LSE, London
Report prepared by Dr. Shirin Madon with assistance from rapporteur Silvia Masiero
A workshop was held to mark the conclusion of a three-year British Council DelPHE Academic Exchange project involving Indian Institute of Management, Bangalore, London School of Economics & Political Science, and Imperial College London. The aim of the workshop was to consolidate the key points that have emerged from the project and to provide a forum for discussion of critical issues regarding the implementation of Health Information Systems in Developing Countries.
The British Council DelPHE project was entitled ‘IT for Improving Public Health Systems in Rural Karnataka’. A summary of the main goals of the project are provided below:
- To identify the effectiveness of the primary health centre (PHC) as a nodal agency for collecting and using data on rural health for local planning and analysis. A goal was to compare government-run versus NGO-run PHCs.
- To extend the health management information system by developing capabilities based on epidemiological analysis for improving resource allocation and health systems planning.
- To develop capacity to use information tools effectively by designing and running training workshops for health staff at district and facility level.
- To evaluate the extent to which the formal system of monitoring PHC performance is able to account for the quality and relevance of basic health services to rural poor citizens.
Each of the three project participants focused on a different aspect of these overall project goals - Professor Krishna from IIMB on overall health policy in Karnataka, Dr. Shirin Madon from LSE on PHC accountability to community, Dr. Edwin Michael from Imperial on critical data issues for epidemiological analysis.
Following the workshop programme and list of participants, this report summarises the main contents and discussions that took place during the course of the day.
Workshop on Health Information Systems in Developing Countries
28th September 2010, LSE, New Academic Building Room 2.14
9.30 Arrival and Introduction to the Workshop
9.45 Health Information Systems in Developing Countries
Geoff Walsham, Judge Business School, University of Cambridge
10.30 Coffee Break
10.45 DelPHE project IT for Improving Primary Healthcare in Rural India
· S. Krishna – Overview of Health in Karnataka
· Shirin Madon – Improving accountability of primary healthcare delivery
· Edwin Michael – Critical data issues for epidemiological analysis
1.00 Sandwich Lunch (provided)
1.45 HMIS to Promote Good Governance and Prevent Corruption in PHC
Reinhard Huss, Nuffield Centre for Intl. Health, Leeds Institute of Health Sciences
2.30 Experience of HMIS in Two African Countries
Tolib Mirzoev, Nuffield Centre for Intl. Health, Leeds Institute of Health Sciences
3.15 Coffee Break
3.30 healthEnet: A proposal for an e-Health Platform based on Second Opinions
4.15 Roundup and Close of Workshop
Workshop on Health Information Systems in Developing Countries
28th September 2010
NEW ACADEMIC BUILDING 2.14
Diane Whitehouse, Business Partner, Castlegate Consultancy (policy consultancy)
Geoff Walsham, Professor of Management Studies (Information Systems) Judge Business School, University of Cambridge
S.Krishna, Professor of Information Systems, Indian Institute of Management, Bangalore, India
Geoff Walsham, Professor of Management Studies (Information Systems) at the Judge Business School, University of Cambridge. Geoff has extensive international experience teaching, researching and consulting in many countries in Europe, Africa, The Americas and Asia. He is one of the founding pioneers of the field of IT for Developing Countries and has written extensively on the subject.
This opening talk presented the various opportunities as well as the challenges of health information systems in developing countries. In the context of India, where health is recognised as a state government responsibility, various data management systems have been established in order to monitor the provision of services. Information usage within these systems is primarily aimed at data flowing upwards to higher levels of the health administration for accounting for monies spent and resources allocated, rather than providing an adequate mechanism for feedback and local usage of information by the implementers of healthcare programmes. An alternative paradigm suggested is that of ‘information for action’ which means making regular and widespread use of information in order to enable more effective action in pursuing health outcome (Stoops et al., 2003). A precursor to the achievement of information for action is the availability of complete, timely, accurate, and relevant data for decision-making at various levels of the health administration hierarchy. This is often lacking as data sets are subject to unintended mistakes, deliberate misreporting/underreporting, and gaps in data due to the sheer volume of data to be gathered by health fieldworkers. An example was given from earlier fieldwork of an ANM (auxiliary nurse and midwife) in Orissa who typically required three hours per day to manually systematise the data gathered during the day. While essential for improving healthcare, good quality information per se is not enough. We also need people at all levels in the health system who have the knowledge to use information effectively which calls into play various institutional and policy issues. Drawing on the case of the Health Information Systems Programme (HISP) which aims to provide a computerized information system for health planning and management first in South Africa then in other developing countries, experience shows that long-term commitments are necessary for institutionalizing and strengthening information systems, and providing support and incentive systems for motivating staff to use information effectively. In conclusion, health management information systems (HMIS) need to be viewed as only one part of heterogeneous networks composed of people, technology, standards, institutions, processes etc. One way of conceptualizing such networks is Actor-Network Theory (ANT) which provides scope for studying heterogeneous networks over extended time periods.
S. Krishna, Professor of Information Systems at Indian Institute of Management, Bangalore. He is involved in research on global software alliances and work management, as well as on health information systems in developing countries and has published books and articles on these topics. Krishna has been commissioned by the World Bank to prepare reports on the health information systems infrastructure in Karnataka.
Prof. Krishna in his presentation on IT and health care in Karnataka and more generally in India, contrasted technology use in private hospitals as against public health care systems in the country. Private hospitals in a city like Bangalore use the best technology, both IT and others, to provide healthcare which is comparable to best hospitals in the world. However, public health clinics and hospitals remain in a pathetic state. Quite often they are unable to render minimal services due to non availability of personnel, drugs, equipment and other infrastructure. Effective use of IT could help in improving healthcare delivery in many ways although in projects of the NRHM and in those derived from the HISP project, the focus has been on health statistics. Though statistical data is useful for certain kinds of decision making and actions for improving healthcare delivery, there is scope for use of IT in many other ways. In order to evaluate the relevance of data for healthcare delivery in public systems, Indian national health documents over the last twenty years were analyzed. It was noticed that of around twenty major issues of concern for national health policy, hardly three or four could benefit from accurate and complete health statistics. Programs like NRHM need to look beyond improvements in health data gathering. In particular, implementation and sustenance of systems for critical health applications needs to be improved. Prof. Krishna presented case studies of four health IT projects implemented through support from the World Bank. One of these could not be implemented even as a pilot and the other three, which were partially successful, could not be sustained by Karnataka State Government after the period of support from the World Bank ended and were totally non functional. As such, it is very important to focus on selection of IS projects and plans for sustenance as part of any IS project implementation. He mentioned that applications like surveillance of communicable diseases and drug inventory management could have significant impact in improving health care whereas projects like PDA based data collection are costly mistakes. Technologies like telemedicine can occasionally work well though these are not scalable and may not be useful widely. In terms of possibilities for the future, mobile phone based systems, if designed well, could have a significant impact because of the wide availability of mobile phones in small towns and rural areas in India.
Shirin Madon, Senior Lecturer in ICTs and Socio-Economic Development at the LSE. Shirin has several years of research experience in India studying the impact of ICTs for improving rural development planning and administration. More recently, she has been involved in studying the role of ICTs for governance reform both in terms of decision support for policy-makers, and in terms of improving the interface between government and citizens.
This presentation focused on the issue of accountability of primary healthcare delivery. While progress in the health sector is typically measured in terms of immunization coverage and family planning, the routine provision of rural health services remains poor with a high prevalence of major diseases (Roy et al., 2004; Gopinathreddy et al., 2006). Evidence shows that services either do not reach disadvantaged sections of the population or are not accessed by them. Yet, there is poor understanding of what is needed to make the PHC system more responsive to local needs. This state of affairs signals that an inadequate system exists to account for PHC performance. The relationship between accountability and performance monitoring depends on how we interpret ‘performance’. With pressures to increase cost effectiveness in healthcare provision, most performance monitoring systems focus exclusively on information about programme costs and work completed (managerial accountability). In India, this type of performance monitoring is carried out by various information systems implemented under the directive of the National Rural Health Mission (2005-2012). However, this type of accountability system has failed to respond to local needs and priorities of communities. A wider conceptualization of accountability was proposed that takes into account both managerial and socializing forms of accountability (Roberts, 1991) which provide the contextual information (sometimes in the form of soft data) needed to interpret why health indices are improving or deteriorating. Emergent processes introduced by the NRHM to improve accountability of primary healthcare delivery to citizens were described such as the Village Health and Sanitation Committees and Community Monitoring which aims to engage citizens in health-related activities over a period of several months leading to a recording of community preferences on nine parameters of service delivery. So far, results of the three rounds of community monitoring in Chamarajnagar district of Karnataka show that community perceptions about PHC service delivery are improving. However, further research is needed to interpret and link the results from the community monitoring exercise and the various social processes that underlie this exercise with service delivery parameters (health outputs) and improved health indicators (outcome).
Edwin Michael, Senior Lecturer in Infectious Disease Epidemiology at Imperial College London. Edwin focuses on modeling the transmission and control of tropical parasitic and infectious diseases. He has been involved in field-based projects in many developing countries to develop epidemiological tools for the control of these diseases including health information systems tools to support the implementation of nation-wide control programmes.
This presentation commenced by emphasizing the importance of data for a range of activities including informing policy, forecasting the future, monitoring programme implementation, and evaluating policy impact. Edwin focused on the need to address issues related to the supply and demand of data for epidemiological analysis. The importance of data analysis for public health management was emphasized in order to understand patterns and relative prevalence of specific health problems and to consequently devise appropriate measures to counteract them. Data are also needed to evaluate the efficiency of an existing health policy and to modify it in an informed manner. Most importantly, Edwin emphasized that enhancing data quality was essential in order to provide a measure of protection against interventions that were driven by donor policy priorities. With respect to statistics on health, countries were classified into four clusters (Scott, 2005). In vicious-circle countries, statistics are weak and therefore poorly used for the purpose of policymaking and analysis. In countries that are data supply-constrained, the state is willing to rely on statistics for policy purposes but as the statistics supplied are weak, they provide a poor basis for policy formulation. In data demand-constrained countries, statistics are good but there is poor or no willingness to utilize them for policy-making. In virtuous-circle countries, statistics are of good quality and are utilized for policy-making. India can be described as falling in the vicious-circle cluster as its statistics are weak and usage by government is poor. Two examples of strengthening the epidemiological component of healthcare provision were provided. The first was related to the estimation of tuberculosis (TB) incidence. Comparison across three databases – HISP, NIC and RNTCP – revealed huge differences in data. An example was given of Gulbarga district where HISP reports approximately 3 million TB cases, while NIC reports 16 million for the same time period. The second case was related to measuring the technical efficiency of the PHCs. Input-output technical efficiency analysis is an econometric method which has become a dominant approach for measuring performance in healthcare. Input proxies include number of staff, drugs and infrastructure. Output proxies include field visits, immunization coverage, deliveries, outpatients, community sensitization sessions. In Karnataka, the conduct of the technical efficiency exercise signaled problems related to scale efficiency, rather than to actual operational problems. Dr. Michael concluded that in India, statistics are poor on average. However, the major problem is caused by demand-side constraints. It seems that policymakers, instead of relying on the idea of information for action, do not really consider data in formulating their policies. As Professor Walsham pointed out, building a culture of data usage is of paramount importance in India.
Dr. Reinhard Huss, Director of MSc. International Public Health at The Nuffield Centre for International Health & Development, University of Leeds. Reinhard has worked for seven years as regional advisor for improving primary and secondary health services in Central African Republic and give years as a medical officer in Zimbabwe. He is currently coordinating a research project on governance and corruption in the health sector in Karnataka state.
Dr. Huss focuses on HMIS to promote good governance and prevent corruption in the health sector – a sector particularly prone to corruption because of uncertainty about future health, asymmetry of information between actors, and complexity of system (large number of actors and interaction) (Savedoff & Hussman, 2006). Lack of clear data on health management sharply reduces the traceability of decisions and actions to single individuals and breeds corrupt practices. Dealing with corruption can be ascribed to what Ostrom (1990) refers to as a collective action problem in which people may think that if they are the only ones acting in the name of the whole group they will only be able to extract a small share of benefits from the action, given that the overall benefit will be spread across the entire group. One way of constructively dealing with the problem of collective action lies in ‘co-production’ in which individuals of diverse organizations cooperate towards a common end in order to cross-fertilize each other with their insights and knowledge. Dr. Huss’s discussion of the Karnataka case study is based on force field analysis which involves the identification of the factors that influence healthcare delivery and of the forces that connect and create interaction between these factors at the field level. Two hospitals were compared in which factors determining corruption were seen to differ. In one case, low salaries were the main determinant of corruption, while in the other the transfer of corrupt staff played a dominant role. At the organizational level, positive and negative incentives seemed to be the most effective tool for reducing corruption – rewards to those who stick to the rules, sanctions to those who do not. Dr. Huss concluded that in order to fight corruption, a higher degree of power should be given to people at the local level so that demotivation which derives mainly from being bypassed by higher level decisions is reduced. HMIS can be a key factor for good governance of PHC if it provides oversight and transparency for service to managers and leaders, professionals and other staff members, and users in the health sector.
Dr. Tolib Mirzoev, Lecturer in International Health and Development at The Nuffield Centre for International Health & Development, University of Leeds. Originally from a clinical background, Tolib has developed an interest and expertise in health systems in lower middle income countries. Tolib’s research and teaching covers different aspects of health policy and systems research.
This presentation described a multiple study related to strengthening mental health information systems in Ghana and South Africa. Improving HIS for mental health is an inherently complex task and is grossly underfunded due to unpopularity of the subject and stigma associated with mental ill-health. The study presented was part of an action research project which initially involved carrying out a situational analysis in cooperation with respondents in the field, followed by devising, implementing and evaluating a therapeutic intervention with support of a HIS. The situational analysis revealed numerous complex information problems at the local level. In Zambia, information systems were so poor that the actual number of psychiatrists operating in the nation could not be known with any degree of certainty. The situation in Ghana and South Africa was somewhat better but gaps in the data were nonetheless profound. In Ghana, no management indicators existed for the mental health sector. Despite these problems, computerisation had been initiated in both nations at province and district levels. Therapeutic intervention was grounded not on the implementation of information systems per se, but on the practical and operational use of information. HIS were implemented at diverse levels. In Ghana, specialised hospitals were predominant whereas in South Africa but public facilities were addressed by the programme. Contextual factors have shaped the outcomes of the project through different channels. In South Africa, consultations with local provinces has absorbed considerable amounts of time whereas in Ghana the project was led just before political elections which resulted in willingness of politicians to engage with popular, technology-based projects. Implementation of mental HIS was conducted by task teams who worked on different parts of the process. A key role in this process was assigned to those who actually built up a database of patients, each with a personal identification number, in order to protect data security. Data protection issues were especially relevant in Ghana where, at the diagnostic stage of analysis, lack of personal data security was perceived as the predominant issue in mental health. Three main lessons to be learnt from the project were presented. First, capacity needs in mental healthcare impact different levels of action and HIS needs to be devised in order to address each level specifically. Second, political support and resource commitments are very important in implementing these projects, especially when an unpopular topic such as mental health is coupled with a popular toolkit such as ICTs. Third, a key issue when dealing with HIS on a specific health problem is to integrate the new system with the broader structure of HIS as operational in the country as a whole.
Dr. Pratap Kumar, Dr. Pratap Kumar, Health Economics Consultant and Social Entrepreneur. After qualifying as a doctor Pratap obtained a PhD in neuroscience. Thereafter, Pratap decided to apply his clinical and scientific experience to issues of access to healthcare in developing countries. He is founder of Health E-Net which is a social enterprise.
The presentation described Health E-Net which is an e-health platform currently being implemented based on the value of second opinions. The starting point for this project was recognition that in rural India poor people have extremely limited access to specialists when it comes to tackling a specific severe health problem. As argued by Prof. Krishna earlier, inequality in healthcare is very severe in India even within the same district or town. Moreover, whereas primary healthcare is, at least in theory, provided universally through PHCs, secondary healthcare is a privilege of the few since only better-off individuals can access specialised healthcare performed by skilled professionals. The next observation was that second opinions on existing health problems improves health outcomes and seeking such opinions is a universal behaviour irrespective of socioeconomic status. The purpose of the project, therefore, is to endow individuals in rural India with access to second opinions by specialised doctors over a medical data storage platform. While telemedicine applications may offer a solution, the barriers to scalability are very high preventing such applications to be rolled out to all areas. Health E-Net requires three key ingredients. First, access to existing records of medical data, which must digitised and stored in a secure database accessible only by the patient. Second, patient demand for second opinions, which already exists. Third, access to appropriate specialists who can provide specialist opinions on the medical records. The universal demand for second opinions allows for a self-sustaning financing model where the affluent can be charged for specialist second opinions while the same service is provided to the poor for free. The framework for Health e-Net does not envisage a complete electronic medical record solution or even a dial-a-doctor service. Rather, it combines elements of these two diverse toolkits and provides an innovation which can be tailored to the needs of the rural poor in India. Essentially, Health E-Net focuses on providing free access clinics in underserved rural areas to achieve the ultimate purpose of providing specialist second opinions to poor patients. The ultimate outcome of this project, carried out by a network of volunteer specialists, is to strengthen the health system and improve the quality of healthcare in developing countries.
Gopinathreddy, M.K., Jayalakshmi, K. and Goetz, A,. (2006) Politics of Pro-poor Reform in the Health Sector, Economic & Political Weekly, 41, 5, pp. 419-426.
Ostrom, E. (1990) Governing the commons: The evolution of institutions of collective action. Cambridge: Cambridge University Press.
Roberts, J. (1991) The Possibilities of Accountability, Accounting, Organization and Society, 16, 4, pp. 355-368.
Roy, T.K., Kulkarni, S., Vaidehi, Y. (2004) Social Inequalities in Health and Nutrition in Selected States, Economic & Political Weekly, 39, 7, pp. 677-683.
Savedoff, W.D. & Hussman, K. (2006) Why are health systems prone to corruption? In: Kotalik J, Rodriguez D (eds.) Global corruption report 2006 London: Pluto Press. 4-16.
Scott, C. (2005) Measuring up to the Measurement Problem: The role of statistics in evidence-based policy-making. PARIS 21 (Partnership in Statistics for Development in the 21st Century). Available at: www.paris21.org/documents/1509.pdf
Stoops, N., Williamson, L. and Braa, J. (2003) Using Health Information for Local Action: Facilitating organisational change in South Africa. In Krishna, S. and Madon, S. (editors) The Digital Challenge: Information Technology in the Development Context, Ashgate Publishing Limited, Aldershot, Hants., pp. 156-170.