Outcomes-based commissioning has been around for a few years now and with the NHS under ever greater pressure for cost efficiency, many of the commissioners see it as the best or even only way they could meet the financial, quality and transformational challenges they face. At the same time, a push towards enhanced use of digital technologies is promising to relieve pressure on repetitive administrative tasks and offer better access to patients. Many of the digital solutions that are being implemented, however, are relatively new and lack robust, long-term data necessary to establish definitive evidence for their outcomes. Healthcare providers and commissioners are often reluctant to invest in digital innovation due to a perceived uncertainty, lacking the information necessary to make confident decisions.

Presented below are lessons we’ve learned through our work on digital innovation in healthcare as well as case studies that we recently presented at the Digital Design: Transforming Healthcare event at the Health Foundry.

Designing for outcomes – lessons learnt

• Describe the desired outcomes for patients and stakeholders. We do this by mapping stakeholders and their outcomes, connecting them into a Theory of Change – a logical model that describes what we aim to do and why it matters.

• Look for outcomes, not outputs. Meaningful impact might not be easy to measure, but measuring outputs is not a good substitute. A service that processes a higher number of patients is not necessarily better at improving people’s health.

• Gather evidence appropriate to the stage of innovation. In the early stages of innovation, this might mean gathering qualitative feedback from patients and clinicians to understand their underlying needs. Reliable quantitative evidence for innovation is more easily provided in later stages, by testing a functional prototype with a pilot group. See Nesta’s Standards of Evidence for further information about this topic.

• Keep metrics in mind from the start. It will be much easier to demonstrate that a service is causing an outcome if measurements are set up as early on as possible. Measurement points should be designed in a patient-friendly way – as an integrated part of the journey, rather than an unwanted nuisance.

• Get feedback in a real-world context. One of the most important elements of designing for outcomes is to keep a connection with the people and the environment we’re designing for. We do this by creating functional prototypes early and testing them in context with patients and clinical staff. This provides us with necessary guidance to iterate and improve our digital solutions.

Mapping outcomes for planned interventions at Guy’s and St Thomas’ hospital.

Finding robust evidence of outcomes for an established service

The first case study we presented at the Health Foundry event deals with the Social Inclusion, Hope, and Recovery Project (SHARP), an already established service that wanted to provide evidence of positive impact they’re having on the local community. Unboxed worked with the service team to formulate a Theory of Change that was used to structure areas of impact. This formed the basis for a thorough pre- and post- analysis of their interventions assessed against a randomised comparison group.

SHARP Impact Assessment presented by Sarah Josefsberg from South London and Maudsley NHS Foundation Trust and Boris Divjak from Unboxed (Link to slides)

How do you measure a service that doesn’t exist?

The second case study is related to early stages of development of new digital services. We demonstrate our outcomes-driven approach to digital design through our work with Guy’s & St Thomas’ hospital in an effort to improve their patient referral pathways.

Working toward better patient referrals with Guy’s & St Thomas’ hospital presented by Dr. Toby Garrood from Guy’s & St Thomas’ NHS Foundation Trust and Leon Odey-Knight and Martyn Evans from Unboxed.

The above case studies offer a quick look at how we’ve been trying to design new digital services for healthcare with a focus on outcomes. We will develop our approach further. In the meantime we welcome any feedback or thoughts around this topic!