National Health Service (NHS) England’s tech fund schemes had a focus on open source projects from the start. The first round (Safer Hospitals Safer Wards: Technology Fund) allocated £20m of its £260m total specifically for open source and this was rolled over into the second round (Integrated Digital Care Fund) due to time constraints. The second round also provided strong support for open source solutions, but the recent announcement that the £240m budget has been cut by £90m in order to divert funds towards solving the A&E crisis will reportedly have little effect on the number of open source projects benefitting from the tech fund. The reasoning behind this, and the advantages of open source software within healthcare, don’t always stand up to examination, however.
Is Open source really lower cost?
When looking at the licensing cost of software, this is certainly true. However, looking at the total cost of ownership is more telling. In the most likely scenario, a consultancy company will be employed to implement and support the software. In the case of a full-blown, big-bang Electronic Patient Record (EPR) -type system, this may be more cost-effective, but the bulk of the price of a small to medium enterprise software purchase is not the software license – there is a far greater percentage which the company allocates towards sales, marketing, functions which the consultancy company will need to shoulder at least as much as the commercial software vendor, as the consultancy isn’t associated with the product. For example, a colleague within the NHS once directly assessed two integration engines, one open source and one licensed. The support package of the open source option made the overall cost of ownership over 5 years significantly greater than purchasing the licensed option.
Is Open source software more customisable?
The customisability of a software package has less to do with its licensing model and more to do with its design, quality, the availability of resources and the business model around it. A piece of OSS which is badly designed with poor quality programming will cost more to customise than a commercial company will charge for the work to alter their proprietary code. Even an in-house team trying to customise a piece of OSS could struggle to compete against the customisability of the software being produced by the SMEs of today – where commercial platforms are being deliberately designed to provide hospitals with software they can customise to their exact requirements. The healthcare software industry can be exceptionally slow to respond to emerging trends within software and technology – take some of the traditional EMR user interfaces as a classic example. The ability of an organisation to incorporate new trends into its software is an issue which has many more factors than the licensing model, however. The size and ethos of the organisation, the maturity of the software, the language it is developed in, number of programmers working on it and, above all, the usefulness of incorporating an emerging trend all will have an effect.
A common accusation levelled at proprietary software vendors is that integration with other software is deliberately limited in order to lock clients into a closed system. While this can be true, there are again other factors which come into play – chiefly, the vendor’s strategy. Large software companies who provide packages for all specialties may be able to use an unwillingness to integrate as a tool to beat clients kicking and screaming towards further investments with them, but by providing multiple methods of interacting with a piece of software, a vendor provides potential clients with fewer barriers against buying their products.
Speed of reaction to bugs
Because some open source projects may have many people working on their code and bugs may be fixed quickly, but smaller projects will have far fewer coders working on them. Regardless of the project size, the programmers have little to lose by not responding within a set period of time and the quality of the code produced isn’t guaranteed. Conversely, a commercial organisation working within the bounds of ap service level agreement is contractually bound to fix bugs within a given timeframe and a failure to meet those terms is bad for business.
Many of the arguments in favour of open source software over its commercial counterparts are far more complicated than paid vs free. The license cost of open source software may save a hospital money in the short term, but the benefits of choosing innovative software from the right company can counterbalance the short term gains and leave a client far better off in the long run. It is certainly worth careful consideration and also raises the question as to whether healthcare organisations should concentrate on open data standards to create their efficiencies.