Toolkits - programming made easy

Help key First DataBank Europe’s (FDBE) clinical decision support functionality checks information in the electronic patient record against medication about to be prescribed or dispensed and generates alert messages if potential clinical problems are found.

Given the complexity of the data that enables clinical decision support, how do system suppliers integrate this functionality into healthcare IT systems? FDBE’s implementation services director, Chris Gray, explains how FDBE best serves its client base.

How does FDBE help its customers to integrate the Multilex Drug Data File (Multilex DDF) and its clinical decision support?

We offer two main models of integration: ASCII data files and software toolkits, also known as application programming interfaces (APIs).

ASCII data files contain the raw drug data in tables which offers customers the maximum flexibility for development. These files enable customers to use their own data structures and select which items of data they wish to implement. However, customers who implement via the ASCII data files are responsible for the development and testing of the functions that use FDBE’s data.

For those customers who prefer not to undertake this level of development and testing, FDBE also provides API toolkits. These are essentially a layer of software between the data and the end user system, which support a range of functions from simple data access to complex processes such as drug interaction checking. Our toolkits shield our customers’ developers from complicated database structures and are fully regression tested, which makes the development process faster and more efficient for our customers.

How can FDBE be sure that its toolkits work across a wide range of systems?

We have over 20 years’ experience of implementing our clinical decision support into a range of systems across the entire healthcare spectrum. To ensure that we can support the vast array of Windows-based and platform independent systems, FDBE offers both COM and JAVA toolkits.

It is also important that we deliver APIs with the same relational database format that our customers will be using in their system. The formats we support include SQL Server 2000, Interbase 5.5 and Oracle 9i or 10g and we connect the APIs to the database using the recognised standards of ADO, ODBC or JDBC. This allows us to recommend and deliver the offering that best fits into each customer’s architecture.

How does the Implementation Services Team support FDBE’s customers across this range?

We work alongside our customers to assess their user requirements and gain a detailed understanding from a technical, functional and business perspective of what they are trying to achieve. We then base our solution on those requirements.

We are with our customers every step of the way from a series of initial meetings to a number of touch points in a development period. We conduct regular design reviews, from which we make recommendations from both a cosmetic and clinical perspective on areas for improvement.

How does FDBE assist its customers in an enterprise environment?

In recent years, we have seen a shift from local ‘site based’ implementations of clinical systems to enterprise environments, with many users accessing the system remotely at the same time.

The biggest difference in a remote environment is the potential workload that our APIs need to handle. We may have very large numbers of concurrent users accessing the toolkit at any one time and that creates extra performance demands. Within the JAVA toolkit, we conduct rigorous performance testing and we maintain a test system that simulates multiple requests. This makes it even more vital to understand the system supplier’s requirements and architecture, so that we can recommend and deliver the best solution.

What new developments is FDBE offering customers?

We offer a standard development pathway for both our ASCII data files and APIs. Once on these offerings, our customers can move up through that pathway to meet the current and medium-term needs of their business. Each year we invest a significant amount of time and resource in developing a major product release. This provides our customers with a base point for future developments.

Our latest product release, launched in July 2006, integrated NHS Connecting for Health’s standard terminologies, the NHS dm+d (Dictionary of Medicines and Devices) and SNOMED CT (Systemised Nomenclature of Medicine Clinical Terms). In fact, we were among the first adopters of these standard terminologies. The product release is being used by a number of our customers to support core elements of the national programme, namely the Electronic Prescription Service and the National Care Records Service.

During 2007, FDBE will be reducing the number of alerts generated by continuing to review editorial policies for decision support modules and will be tackling information overload by the increased provision of filtering for warnings and referential information.

FDBE will also be developing new methods to integrate the patient’s medical history so that we are able to use the patient’s context to generate more specific alerts.

From a broader perspective, we are seeing the globalisation of healthcare application providers. Drug information differs from country to country, yet system suppliers need a common implementation solution. To meet the challenges of implementing across regions, we have developed an international API for integrating drug content, which is particularly suitable for those looking for a solid foundation upon which to build a global solution.