What was the problem?
Now that people in Britain are free to choose their own hospital to treat non-urgent conditions, how can providers give them the information they need to make informed decisions?
Most people look for information on the Government website, NHS Choices, which 'scores' individual hospitals according to different aspects of quality. But many choices involve complex trade-offs between competing objectives. The challenge is first to understand how people make their choices, and then to present information in a way that allows them to identify and select the best-quality hospital to meet their particular needs.
In spring 2009 a large-scale research programme was designed to test different configurations (or 'choice architectures') of the NHS Choices website. Its explicit aim was to determine which configuration would be most effective in enabling patients to choose amongst competing hospitals for the treatment of their non-urgent health conditions.
Leading the research was ÐÓ°ÉÂÛ̳ Associate Professor of Behavioural Science Barbara Fasolo, an expert in the science of decision making and now Head of the ÐÓ°ÉÂÛ̳ Behavioural Research Lab. She worked closely with Elena Reutskaja, Associate Professor of Marketing at the University of Navarra's IESE Business School, and Drs. Anna Dixon and Tammy Boyce of the King's Fund, England's leading independent health charity. The King's Fund supported the research with a £27,000 grant.
The research team began by conducting seven focus groups to identify the type of information people wanted and the way in which it should be organised. Results were published and discussed with staff at NHS Choices. The findings from the focus groups were then fed into an extensive two-stage online experiment involving 1,200 UK citizens and launched in the summer of 2009. In the first stage, participants tested a simplified 'scorecard' summarising the performance of local hospitals, and were presented with different “nudges” or configurations that could help them to choose the best quality hospital. Three main nudges were compared, in addition to the configurations used by NHS Choices. One nudge was “feedback”, which provided participants information on the quality of the hospital they had chosen, another nudge prompted participants to think about what they wanted in a hospital, and the last nudge preselected the best hospital by default (but participants were also provided the opportunity to “opt out”). Six weeks later, participants completed a second part of the online experiment to gauge what they had learned about hospital choices.
The research found that people typically spent just a few minutes choosing hospitals online and could easily be overwhelmed by too much information. This indicates the need for websites to organise information in such a way that people are informed only about those aspects of care that matter the most to them.
Aggregate quality measures found favour with older people and those with lower levels of education, but these groups also found them confusing. The research confirmed that only highly numerate people could properly process numerical information about the quality of health care, especially complex statistical information such as hospital standardised mortality rates. It also found that participants were likely to become suspicious when told that data were unavailable or missing.
The research also revealed that people's preferences could change according to how providers presented information, and that no 'nudge' was effective for everyone - different groups of people valued different choice architectures.
What happened?
Impact on NHS Choices
In April 2009 the research team presented their interim report on the focus group findings to NHS Choices and several of their recommendations were immediately incorporated into the website. For instance, the research had identified which indicators should be given priority and proposed ways in which the format could be clarified and strengthened.
In May 2011, Anna Dixon of the King's Fund presented the team's overall research findings to the Clinical Information Advisory Group of NHS Choices. Further developments to the site followed, including simpler labels, grouping into domains and clearer definitions.
As a result of these changes, the NHS Choices website has received more than 27 million visits per month, as compared with 7 million per month when the research commenced.
Impact on other organisations
The researchers engaged directly with other organisations that provide information to patients. Discussions in September 2010 with the UK Care Quality Commission influenced the way it presents the outcomes of its inspections and risk quality profiles to the public. For example, the Commission switched to using simpler and more understandable messages and symbols to communicate information about its standards.
The King's Fund drew upon this research in a number of responses submitted to Government and other organisations in terms of presenting provider performance information to the public, including the Nuffield Trust's review of health and social care.
Interpreting and publicising hospital standardised mortality indicators was especially problematic. The King's Fund team members presented the relevant research findings to the steering group established by the National Health Service's Medical Director to review how this information should be made public. The study findings were directly cited in the steering group's report to the Department of Health.