With digital misinformation often making headlines, it’s easy to forget that this is a golden age of online health communication. The internet is flooded with websites on every imaginable medical condition, with lots of good content that enhances patient–doctor dynamics and improves disease management. It’s all at our fingertips, in front of our eyes, if we are able to navigate the maze of available information online.
When creating a disease information website, considerable time and effort is spent ensuring that the content is accurate and – in the case of patient resources – lay friendly. But what about the search experience for patients who are desperately seeking answers, particularly if they are undiagnosed, and know little or nothing about their medical condition?
Do people really search with the most effective key words when they are experiencing physical symptoms that they don’t understand? Early signs from research are that they don’t. The search experience for many people is frustrating as the results returned are not optimised to support the type of symptom based searches undiagnosed patients and their families undertake.
These patients and their families may have no idea about their condition, especially if they experience seemingly unconnected symptoms. Consequently, there is a risk that undiagnosed people might never find online resources that are designed to support them because their understandably vague searches take them though a confusing maze of information dead-ends.
Haphazard searches
When we first look online about something new, that we can’t explain clearly, our internet searches are rarely logical or precise. Instead, they are haphazard and exploratory.
Exploratory search behaviours are not necessarily bad. Broad terms and basic single words are used or typed in groups (long-tail searches), to filter results. Casting a wide net enables many thoughts to be processed, and sources compared, until content becomes relevant. But we might go down many rabbit holes before we reach meaningful results.
The simplest approach to searching online is look-up or ‘known-item’ searching, involving key words and a clear goal (e.g. a direct question). Patients/carers generally begin to take this more focused approach once they have sought medical help and have a possible diagnosis. By then they are learning the relevant medical language and the quality of their searching improves. But this was probably preceded by weeks or months when they were unable to find ideal content to answer their questions.
Although there is general research on social media search behaviours, few studies investigate how people look online for health information when they don’t know the right medical language. However, preliminary evidence from swii.ch health suggests that, especially for unusual symptoms, haphazard exploratory searches are common – especially before a diagnosis is confirmed. In the swii.ch research, people did not reach the most suitable websites unless they were aware of key medical terms.
For rare and ultra-rare diseases especially, the person seeking advice may be a parent or carer. Trying to verbalise unusual symptoms that haven’t been personally experienced is an added complication that may keep information out of reach.
Enhancing online resources to support patient searches
Optimising website content for searching is far more than ensuring that keywords are relevant to a search query. It’s a key area of search behaviour research and development.
When creating online resources targeted at patients/carers, research to investigate their information search behaviours is invaluable. As part of this, beta testing should always involve people with no previous knowledge of the subject matter, and should evaluate their search strategies, the language of their searches, resources identified and first impressions of website content and design.
Analysis of search behaviour not only uncovers the triggers to initiate a search and questions asked but also the sites that patients land on. In some cases, a symptom based search may return results that appear to provide the answers a patient is seeking. Satisfied with the information provided, patients may end their search at this point without connecting their symptoms to what the reality of their condition might be. This represents a significant opportunity to create a digital breadcrumb trail of information and links that can sign-post patients to the accurate and up-to-date information they need.
There are simple measures that any content developer can take, to make searches lay friendly. Images are as important as words, and good Alt-text descriptions are critical for accessibility and optimised searching. For medical content, Alt-text should describe graphics with basic phrases as well as key medical terms. Writing a perfect short Google description of a website (snippet) is also essential. People make quick decisions about whether to visit or dismiss a website based on its snippet, so the content should be simple, informative and engaging.
The bottom line
Indeed, the most powerful assets on websites aren’t the copy, the images, or the videos. The most powerful assets are the metadata and tags, and they deserve a great deal of thought during content generation. They alone might guide a patient from a state of haphazard confusion to empowerment, by leading them quickly to good content and effective support.
Useful resources
Exploratory search behaviours: (https://dl.acm.org/doi/abs/10.1145/3352683.3352687)
Fixing the broken patient search experience: https://www.reutersevents.com/pharma/patients-and-medical/fixing-broken-patient-search-experience
Online Health Information Seeking by Parents for Their Children: Systematic Review and Agenda for Further Research: https://www.jmir.org/2020/8/e19985/
Illness perception and information behaviour of patients with rare chronic disease: http://www.informationr.net/ir/21-1/paper707.html#.Y3JI2i-l30o
