Public Sector Health Information Systems

eGovernment for Development
eHealth Case Study No.1

A Health Information System for Indian Districts

Case Study Author

Dr. Zubeeda Banu Quraishy ( zubeeda@hotmail.com )

Application

In 2001, the State Government in Andhra Pradesh, India (in collaboration with the University of Oslo, Norway) began introducing new information systems to support health sector reforms in rural districts.

Application Description

The District Health Information System (DHIS) runs in a set of primary health centres in Chittoor district in Andhra Pradesh. At the end of each month, primary care staff report in to the designated health centre and enter into a PC health-related data gathered from the communities in which they operate. In addition, those staff also enter less transient data such census and population data that provides a baseline profile for communities in the district. The data can be used for a variety of health reporting and decision-making:

In all cases, the geographic information system capacity (i.e. mapping) is a key element of reporting and input to decisions. For example, it has identified poor planning of previous health centre location, showing places where centres are too close together (and thus underutilised), and also identifying underserved areas in the district.

The information system was first piloted in nine primary health centres in Kuppam constituency within Chittoor district during the whole of 2002. From February 2003, it was rolled out to a further 37 health centres (representing roughly half of all those in Chittoor district). In practice, the implementation team has been collating monthly data from all 84 health centres in the district (46 already computerised, and 38 planned computerisations). The plan is to roll-out the application in the other centres and districts of Andhra Pradesh state.

Role of ICT

The information system runs on basic PCs; the DHIS software is custom-written and open source. It has a facility to use Arcview formats to view data via a digital map, in geographic information system mode. There are plans to Web-enable the application so that monthly reports could be sent for collation via the Internet; currently reports are sent from health centres for collation at district headquarters via hard copy and floppy disk.

Application Drivers/Purpose

The main rational purpose for DHIS introduction was to improve the quality and range of informational inputs to health decision-making and, hence, to improve the quality of health decisions (covering planning, monitoring, evaluation, etc.). In a broader sense DHIS was introduced to support health sector reforms; specifically health care decentralisation and the better delivery of health services in rural areas.

In terms of stakeholder driving forces, the application was part of the State Chief Minister's high-profile political agenda of trying to use ICTs to improve both governance and delivery of public services. The State Department of Health wanted to improve health reporting since they had previously found it difficult to generate aggregated data reports from health centres. Health workers felt that their existing data work was repetitive and monotonous, yet that they were excluded from the health decision-making for which the data was used; they therefore wanted a more modern but decentralised approach to health information management and decision-making. Finally there were decision-makers in the district government and in external agencies with a health agenda, such as PATH-India and UNICEF, who wanted improved access to better quality health data in order to improve both their decision-making and their ability to report upwards and outwards to other authorities.

Stakeholders

The core stakeholders are the primary health care workers who gather and enter the data that is the foundation for the system, and the community members in whose service those workers act, seeking to improve their health. Other public sector stakeholders noted above include the Chief Minister of the state, senior public officials in the Department of Health, plus district-level officials. Secondary stakeholders include the team from the University of Oslo who provided facilitating inputs, the IT vendors who supplied the technology, and external agencies involved in health-related projects/programmes.

Health and the Poor

The DHIS does not have a direct impact on the poor. However, its data focus is poor communities, and it is directly used by those health workers who spend their lives addressing the health needs of those poor communities. One intention of the project has been to reduce the amount of time health workers spend filling forms, in order to release them to spend more time serving communities.

Impact: Costs and Benefits

Being open source, the DHIS is provided free of charge, together with training manuals. Funding has also been provided by the state and from overseas sources that enable training to be provided free of charge to relevant state, district and health centre staff. The state's Department of Health paid the required sums for the hardware infrastructure, for trainers, and for a local systems developer whose job was to help implement the system.

Benefits to date in the implementation have particularly been of the 'proof of concept' type, showing that the application can be made to work over a period of time, that hundreds of staff can be trained, and that support can be garnered from public officials for this type of e-health application. A key benefit is seen to be the orders issued by senior officials requiring staff to use the system, and threatening sanctions for those who do not. The system has also demonstrated a capacity to produce information that can be valuable for decision-making, and has also shown the value of a GIS-/map-type interface to help with a better locational understanding of health planning in the district.

Previously, it is estimated that health workers spent over 60% of their time on data collection, collation, entry and transmission. When it is used, the system cuts the amount of time primary health staff spend on data work, while still improving the quality of health data in the health system.

More subtly, the DHIS may deliver some changes in culture and values; for example, introducing a greater appreciation of the value of information within the health system, and potentially empowering health workers by giving them greater access to data and greater understanding of indicators and targets.

Evaluation: Failure or Success?

The DHIS programme is a rolling one in which implementations carried out to date are now being stabilised, and in which an ambitious set of further implementations are planned. It is therefore too early for a formal evaluation of the system. There have been evaluative inputs from various stakeholders during implementation that have been used in amendments to design; for example in relation to the content of training or to the customisation of reports to match local needs.

Enablers/Critical Success Factors

  1. Participatory and prototyping approach . The progress made with this e-health application can be put down in significant part to the participatory, prototyping approach used. Data sets and entry/output interfaces were designed and then demonstrated to, used by, and discussed with a variety of stakeholders - health staff in the field, district administrators and statistical officers. Training programme design was also amended a number of times in response to user feedback. Such an approach takes time - more than 12 months in total - but significantly improves the quality and acceptance of e-health applications.
  2. High-level support . A key success factor for e-health applications is support from high levels within the public sector - this support releases both physical and political resources that help ensure the application can be implemented. In this case, support came right from the very highest level in the state: the Chief Minister.
  3. Application flexibility . The DHIS software is designed to enable flexible customisation of interfaces, outputs and datasets to match particular local needs. This is a valuable asset for this type of e-health system, that must cover the needs of a rich variety of different stakeholders. Customisation is essential, and can be seen as an ongoing - not one-time - activity; to date, customisation of implemented systems still continues; for example, to provide new reports to match emergent management requirements.

Constraints/Challenges

  1. Bureaucratic politics . eHealth systems introduced in the public sector may need to tread carefully through the minefield of bureaucratic politics that is so often found in multi-institutional settings. In this case, there was support from the very top of the political tree, and also interest at the level of the district, perhaps partly due to the IT infrastructure that the project promised. However, some of the concerned state departments and commissioners at the mid-level were not supportive, partly because they had initiated their own health information systems project that could be seen as competing with DHIS, and partly because their permission had not initially been sought. Only after showing results of the pilot study, plus showing how there could be synergies rather than just competition between the two HIS projects, did support come from those stakeholders. As stated above, the emergence of this support is seen as one of the key successes of the project to date.
  2. Competing priorities . Of the state's 16 health performance indicators, sterilisation rates is the one that most concentrates the minds of all public servants, from the health workers upwards. Both those workers (who want to achieve their targets) and especially higher-level officials created a lot of fuss when requests to run DHIS training courses were made by the project team because they did not want staff diverted from their health work. This highlighted the tensions for e-health systems between longer-term gains from systemic change vs. the short-term need to deliver against health targets.
  3. Staff transfers . The public sector - particularly in India - is characterised by continuous transfers of staff. This has been a non-stop problem for the project, with staff at all levels - sub-centres, health centres and district - moving on. At the district's Medical and Health Office, four main staff left within a 16-month period. In a direct sense, there was the problem of loss of capacity since these staff had been trained to use and understand the e-health system. Even more difficult were the 'soft' issues caused by staff transfer: each time, the project team had to start all over from scratch building rapport, building interest and building motivation with the new staff member before they could think of then trying to build capacity.
  4. Infrastructural shortcomings . The DHIS project was piloted in Kuppam, the constituency of the Chief Minister. Even here, there were power supply problems with electricity only available for around five hours per day. Training programmes had to work around this. In some health centres, power was only reliable in the late afternoon, but this conflicted with the desire of health workers to be travelling back at their homes, leading to difficulties in making progress with training. In one case, team members set up in the nearby town - where most staff lived - and provided training in the evening.
  5. Vendor support issues . The hardware infrastructure for the project was procured through the government system, with a relatively high cost being paid partly to cover guarantees of continuous maintenance, repair and replacement of faulty equipment. In practice, though, within a few weeks of hardware installation problems arose with PC breakdown, UPS burn-out, or printer failure due to poor power supply, poor site preparation, or poor quality of initial equipment. Fault reporting to the vendor met with either no response or long time delays, and a requirement for continuous badgering in order to obtain any level of service.
  6. Institutionalising new practices . Even given all the challenges that were met, it still proved easier to install the computerised systems than it did to get staff to sustainably alter their working practices. A few weeks after training, staff began to revert to their old, manual ways of working, despite these being more time-consuming. The main reason emerging is that, under the new computerised system, staff are unable to manipulate the figures before they are submitted to higher authorities, as has been their norm.

Recommendations

  1. Adopt a participatory, prototyping approach . eHealth systems will work best where they take feedback from stakeholders at all levels during all phases in order to improve system quality and ownership.
  2. Consider social issues . Technology may lie at the heart of e-health systems, but it is essential to pay attention to the human/social issues during e-health implementation. If such issues are ignored then any success a system might have will be short-lived.
  3. Pay attention to all bureaucratic structures/players . It pays to involve all public sector stakeholders in an e-health system, even if you initially perceive them to be relatively weak. Failure to do so can store up problems for later.

Further Information

http://www.hispindia.org/

Case Details

Case Editor : Richard Heeks.
Author Data Sources/Role : Project Coordinator Role.
Outcome : Too Early to Evaluate
Region : South Asia. Start Date : 2001.Submission Date : January 2004.

Last updated on 19 October, 2008.
Please contact richard.heeks@manchester.ac.uk with comments and suggestions.