Principles for Effective Patient Information Systems

Over the coming years, many countries will be setting up information systems to better manage patient information. These systems are fundamentally different from the usual routine health information systems with their register and tally sheets producing aggregate information. An information system to collect individual patient information and manage it over the life of the patient will provide new perspectives on health and disease and new insights and opportunities to improve the health of individuals. It is worth considering a broad overview of what these systems should look like and how they should be implemented and used.

Define success locally
These systems collect individual patient information locally, of course, but what is often overlooked is the value of that information at the point of care to improve health decisions. Many times systems are designed to collect data to be sent to "higher levels" and the local analysis and use of the information is overlooked. When you combine good quality data about an individual patient with decision making tools and resources, you have an important opportunity to directly impact patient care and improve health.
It is also important to understand that "local" for a patient does not usually mean a single point of care. Patients often encounter different doctors, clinics, and ancillary services such as laboratories, radiologists, pharmacies and community health workers. Each of these encounters can benefit from access to good information on the person and in turn collects valuable information that should be part of the persons health record. Therefore, it is fundamentally important that individual patient information systems be designed to allow all relevant parties to access and update the record. This can be done using modern information and communication technologies focused on a personal health record.
In addition to individual care decisions, this local data also creates important insights into the local health environment. Analysis of this information can give important insights into the community access to care, coordination of care, and patterns of health and disease.
The bottom line on the "local" principle is that information systems should be designed to be accessed locally and not to send data off to a "higher authority", never to be seen again. Data extracts and aggregations as well as complete anonymized individual records in some cases can be valuable at higher levels for management, planning and policy as well as operational research but these are secondary uses. The systems should be designed around local use.

Improve what you have
Rather than embarking on a large project to design and install the "ultimate" computer system, it is much better to take a good look at the existing health information system and assess how to improve it. Often the current system is not working well and there many be computers and related resources that are not used well because of poor design, implementation or training.
Experience has shown that it is not wise to expect big and complicated things to somehow become small and simple. Large information technology projects are particularly prone to failure because people universally underestimate the complexity of large IT projects and overestimate the ability of the organization to change to adopt a new system.
Therefore, an incremental development approach is much preferable to large projects. If you start at the local level and look at work flow and information flow and work on improving this incrementally and continuously, it will give you a much better and more reliable result that tearing out everything and starting over.
Likewise, systems that are designed around relatively simple technology that is easily implemented and that can be readily scaled to larger numbers of users has a much better chance of success than trying to install large complex systems. In resource poor environments, these principles are even more important since the option of adding more people, computers, communications technology, and training is not available.
Even resource poor environments often have access to simple information and communications technology that can be easily implemented and scaled. Most countries have basic mobile phone communication technology available. The low bandwidth of even basic mobile systems (SMS and GPRS) is not an impediment to effective use. If you look at the massive success of Twitter, for example, it is based on very simple mobile technology of mobile phones and low bandwidth computers. It is easy to implement at low cost and the basic design scales to massive numbers of users easily. Each user of the system adds value and improves the functioning of the network logarithmically. It is possible to create valuable health information networks using these same principles. In fact, it is already being done.

Design patient information systems for the smallest unit of care
Unfortunately, most patient information systems are designed for large enterprises such as hospitals or large clinics and are designed to meet the perceived needs of these sites who have complex management needs for the most complex health problems. This results in very complex systems that are expensive to implement and maintain and unfortunately, do not have a good record of success.
In keeping with our principle of simplicity, it is much preferable to start with a minimal system that would be suitable for the smallest unit of care which would be an individual doctor, nurse or community health worker. One should also consider that the system should be designed to give the individual patient access to their own health data.
When you design such a minimal system you end up with a system that is easy to use, affordable, and easy to adopt. When you have a minimal data set of only the essential health information, you also minimize the necessary bandwidth and the complexity of the devices to enter and access the information. Therefore, it can easily use existing mobile communications technology. These systems can easily scale to larger number of users.

Focus on connectivity
Information is only valuable if it is used. It must be communicated to be used. This communication can be at a site (where a patient is seen at the same clinic), local (where results from a lab on the other side of town are available), regional (where a district office manages multiple clinics, or national (where health information is used for policy and planning).
In order to communicate information, it must be put into a form that has a common meaning for the sender and the recipient. This means that you have to have standards for data representation and semantics. In addition, you need to have a communications protocol that will allow the data to be sent and received without loss or corruption.
Having data and communication standards also allows you to have a flexible system design. You do not have to have the same hardware and software at every point in the system. If you think of the Internet and email as an example, there are multiple different devices from phones to terminals, to computers running hundreds of different software systems that can all send and receive standard email messages. You can do the same thing by defining health data standards and communication protocols.
This also avoids the necessity of building and maintaining large data repositories and multiple communication protocols. The only thing that is necessary is that the standards be defined. This fits in well with our simple system design that can be easily implemented and scaled at multiple sites on varied platforms.

Separate data from applications and the data transport layer
The success of the Internet is based on this fundamental principle of having a simple universal data transport layer that can reliably send and receive data. The data is represented in standard formats such as email or HTML (web) syntax that are universally recognized. There are thousands of applications which understand the data standards and transport protocols. These applications can be very simple or very complex and can run on simple technology such as mobile phones or complex computers. The applications themselves can be simple or complex. As long as an application can understand the data standard and the data transport protocol, it can communicate.
When you design a health information system to use data standards and define data transport protocols, you have everything that is required for others to build effective applications for all levels of use. These can connect patients, care providers, ancillary services, and health resources in compelling connected sustainable systems.

Sustainable information technology is driven by the desire to connect
Human beings seek connection. Health requires connections between patients, care providers, resource providers, and others. A system that facilitates these connections will be enthusiastically adopted and will be sustainable. In fact, it would be difficult to stop.
Maintaining and restoring health, preventing disease, and caring for others are among the most basic social activities. Unfortunately, our health systems and health information systems are not usually set up to facilitate these very basic human connections. Often due to reasons of power, control, and money health systems are set up to isolate patients, care providers and others who could and should be involved leading to fragmented, disconnected and isolated health services. Existing health information systems are unfortunately often designed to collect data, remove it and hide it in inaccessible data warehouses away from those who could most benefit.
Patient health records that use the principles we outline here can share data and connect the experience of patients, caregivers, doctors, and others. These will be adopted and utilized and as a result will be sustained.