Thanks for visiting Research Information.

You're trying to access an editorial feature that is only available to logged in, registered users of Research Information. Registering is completely free, so why not sign up with us?

By registering, as well as being able to browse all content on the site without further interruption, you'll also have the option to receive our magazine (multiple times a year) and our email newsletters.

Semantic searching can help patients with healthcare information

Share this on social media:

Topic tags: 

Boosting 'findability' is vital for medical professionals and researchers to uncover knowledge for the benefit of patients, writes Jeremy Bentley of Smartlogic

The understanding and interpretation of information, quickly and in context, is essential for clear communication. Nowhere is this more important than in the healthcare sector, where medical researchers and professionals need to be able to access accurate information – and, in many cases, a lot of it – often at very short notice. Unfortunately, the seemingly straightforward mantra of the modern professional – ‘just Google it’ – is not always applicable within the workplace, where internal information assets can be massive and hugely complex – and not at all easy to access.

A new study commissioned by Smartlogic and conducted by MindMetre Research demonstrates that, in too many cases, organisations are simply not meeting the ‘enterprise search’ expectations of their own managers and directors. Enterprise search, the process for quickly finding accurate information within organisations when it is needed, is vital to any business or service provider that relies on internal information resources to serve their clients – or patients in the case of frontline healthcare organisations – and ensure the smooth running of day-to-day operations.

The research, which took in the responses of more than 2,000 managers and directors in the USA, UK, Germany and France, showed that nearly two thirds of those surveyed (65 per cent) define a ‘good search’ as taking less than two minutes to find what they were looking for, but fewer than half (just 48 per cent) report being able to achieve that result in their own organisation – a 17 per cent gap between expectation and reality. Nearly 90 per cent of the respondents believe that taking four minutes or more to find the information they want does not constitute an acceptable search experience, and yet more than a quarter (27 per cent) say this is the case within their own organisations. The figures are very similar for the subset of the respondents who work in the healthcare sector.

In most industries, failure to find information from within the enterprise can slow operations, impact profits and frustrate employees, but in the medical sector the stakes can be much higher if knowledge about a condition, injury or disease cannot be recovered. When it comes to finding data to be used in diagnosis, patient care and the assessment of treatment options, being able to access information quickly can, quite literally, be a matter of life and death.

Finding information

Clearly, ‘findability’ – the ability to recover stored information – has become a real issue when medical professionals and researchers search their organisations’ information management systems. These in turn often tap into other sources of healthcare information such as studies by educational institutions, investigations by labs, government reports, journals, conference papers, hospital and health authority findings, charity databases and medical blogs.

To help medical professionals and researchers find what they are looking for and be confident they are getting all the necessary facts, it is fundamental they be able to look up information quickly and easily. This might be descriptions of symptoms, background on conditions, details of possible remedies, facts about medications, previous case histories, accounts of experimental trials, patient records, and other diagnostic, treatment and experimental resources.

Healthcare organisations and research labs need to take their information retrieval platforms beyond the sort of conventional search that simply takes the few terms inputted and, using closed algorithms, scans the entire index for these words. For traditional search to work effectively, the user’s understanding and use of the search term has to line up nearly exactly with the language of the author of the particular document with the information the searcher is seeking. Even if this works, the document that is needed by the searcher may be buried deep below page after page of other results.

Semantic searching

In order to provide more precise information retrieval, a medical organisation needs to apply the principles of ‘semantic search’, which seek to address searchers’ intent and the contextual meaning of the terms they are using. In developing a ‘semantic search’ capability for a healthcare information management system, an organisation must first put in place an ‘ontology’ – a ‘semantic model’ that encompasses a vocabulary of all the medical terms that would be used by clinical staff or researchers. Such a semantic model recognises the significance of keywords in certain contexts – for instance, MS meaning ‘multiple sclerosis’ rather than ‘Microsoft’ or ‘Marks and Spencer’. The ontology also embraces contextual relationships between the symptoms inputted, such as ‘headache’ and ‘meningitis’. When this is used to drive the search experience, it can eliminate ambiguity, direct the user and present the information in context.

The ontology is employed to implement automatic content classification. This is essential because the ‘tagging’ of documents within an information management system is usually applied manually, making it inconsistent and costly (where it has been done at all). The growing volume of information, human error and the differing standards and methodologies of the various groups, departments and individuals supplying and labelling content makes an automatic meta-tagging solution essential to delivering a consistent and satisfactory ‘find’ experience.

Software for providing automatic tagging scans each document, intelligently recognises key terms from the ontology, and tags the content with the right ‘labels’ for later retrieval during a search. In this way, all information on an information management system is automatically tagged with accurate, standardised and consistent labels – metadata – so that when a professional searches for medical information they only receive relevant, topical content in the correct context.

The National Health Service (NHS) in the UK has already implemented a semantic approach to its primary patient portal, NHS Choices. This helps patients to tap into its information resources online to get information of general healthcare, medical conditions and treatment options. Last November the health service reported that NHS Choices received more than 100 million visits over a 12-month period, while a study from Imperial College London found that a third of those logging onto www.nhs.uk decided against seeing a doctor afterwards – dramatically cutting the number of unnecessary doctor and hospital visits in the UK and saving the NHS an estimated £44 million in costs.

Using an accurate and efficient internal search platform is essential for any organisation. Failure to locate and utilise information assets effectively is irritating at best, costly at worst. Nowhere can this be more important than within the medical arena where the ability to pull up accurate information in an instant is, in the true sense of the word, critical.

Jeremy Bentley is chief executive of Smartlogic