Despite the health information revolution and health care consumerism that the Web has ignited, many decisions in medicine today are still made without reliable comparative information. The analytics methodology that can address patients' and physicians' needs for comparative information does exist. However, it is not consistently applied and it is not easily accessible at the point of care. Specifically, analytics on comparing hospital quality have primarily focused on experts and have made their way into professional reports that are sold for thousands of dollars to hospital administrators. For the most part, they have not yet been made usable or accessible to patients and their busy physicians.
Number Competent Patients: A Segment Ready for Analytics
One of the reasons commonly cited for the shortage of usable patient analytics tools is that patients lack number and technology literacy. However, there is evidence to the contrary. According to the 2007 Health Information National Trends Survey -- or HINTS -- between 50% and 78% of patients find medical statistics easy to understand, prefer numbers to words for expressing the chance of an event, can interpret numbers correctly, and depend on numbers or statistics to make decisions.
In terms of specific decisions, finding a health care provider is one of the most significant areas, with 34% of surveyed patients turning to the Web for comparative information on hospitals and physicians.
Where Patient Analytics Fall Short: Building Confidence
Patients' competence with numbers does not translate to confidence in using them to make decisions. For example, a study published in the journal Supportive Care in Cancer found that while 63% of cancer patients reported using the Internet to search for cancer information, only 13.3% of those patients said their decisions about treatments were affected or changed by the search and just 11.4% said their Internet use affected their choice in physicians.
A separate study warned that patients perceive analytics tools as:
- Overwhelming: While patients want numbers to make decisions, too many numbers become overwhelming. Fifty-three percent of surveyed patients "feel uncomfortable with health information that has a lot of numbers and statistics."
- Contradictory: Patients reported that multiple measures of comparison are confusing when performance is different across different measures. Currently, analytics tools do not reconcile what may seem contradictory to a non-expert patient.
- Confusing: Borrowing from consumer goods comparative information, patient analytics aggregate multiple numbers into star ratings in an attempt to make comparisons easier. However, this aggregation becomes confusing to patients who want to know where the stars or the aggregate numbers come from. It leads them to question the reliability and credibility of the information.
- Purely rational: Current analytics tools ignore patients' emotionally charged state and reliance on personal experience when making health care decisions. They present facts without guidance and without room for personalization.
- Burdensome: When analytics rely on patients to provide the data, use of the system drops. Patients are reluctant to enter information into a system. Only 15% use the Web for storing health information, and self-care or disease management systems have been a challenge for patients to adopt.
The Future: 'It's All About Me'
In the course of our NIH-funded research, we found that analytics should not be about the content but about the patient. In their "think out loud" sessions, patients who used an online guide to compare kidney transplant centers trusted and engaged with the system if they perceived it as "empathetic" and easy to use. Their comments expressed a "trust-engage-no-think-no-work" reaction to analytics tools consistent with studies on human-computer interaction. Some of the lessons learned include:
-
Help me trust you: Patients are more likely to trust information if they feel emotional and practical support that relieves some of the anxiety from their disease.
-
Engage me: Patients seek interactivity that presents information in response to a question and linked to a decision rather than in an article-like format.
-
Don't make me think: Patients need the big picture first, in the form of an aggregate summary followed by a drill-down functionality to help them understand where the summary comes from.
-
Don't make me work: Patients want personalization only for the key metrics critical to their decision and are willing to forgo it if it requires too much data entry on their part.
The opportunity for analytics to make a difference in patients' decisions lies in incorporating these "trust-engage-no-think-no-work" elements in the user interface design. At the emotional level, the user interface can employ aesthetics to acknowledge and honor one's personal experience. At the practical level, a user interface designed for usability can make content engaging and interactive rather than merely educational.