CCC 2014: Mobile Devices and Information Technology for Improving Care

December 10, 2014

While modern medicine is heavily dependent on sophisticated technologies to diagnose and treat disease, health care generally lags behind other sectors in adopting advances in information technology. A number of presentations at the 2014 Canadian Cardiovascular Congress looked at ways in which information technology and health care do or could intersect to improve patient care and health promotion.

The Value of Free Mobile Apps

There are more than 10,000 health and medical mobile apps available in the iTunes Store alone, but how appropriate and effective are these unregulated tools? According to one study, the answer is buyer beware, and you get what you pay for. A group from England’s King’s College London assessed 96 free hypertension and blood pressure apps from the Apple UK App Store. The apps, all targeted to the general public, were examined for how well they conformed to medical guidelines.

The researchers found that the apps “conformed poorly to recognized guidelines,” with 89 per cent providing poor information on data security. They concluded that they could not recommend patients use any of the apps in the group reviewed.

Text-Messaging Reminders for Medicine and Exercise

In two separate studies, researchers from Waterloo, Ont., looked at the use of text-messaging reminders to improve adherence to exercise programs and medications following a heart attack. Less than 50 per cent of patients keep to their exercise programs one year after a heart attack. For medications, up to 80 per cent of non-adherence may be due to forgetfulness. The researchers set up programs that generated four text messages throughout the day, reminding participants to exercise or to take their medication.

As the 12-month studies progressed, participants who did not receive text messages saw significant fall off, both in taking their medications and exercising as recommended. Those who received the texts closely followed their exercise programs. Similar results were found for the medication participants, with an especially strong benefit found for those expected to be at high risk for not staying with their regimens. While many apps are available for people to send themselves reminders, the researchers believe that reminders generated by a health care professional have greater impact because they are less easy to ignore.

The Medium Affects the Message

Another study had practical implications for the communication of health information. It looked at the popularity of various media types for learning about heart disease risk factors and whether the medium impacted awareness of those risk factors. Source media compared included the internet, print sources, TV and radio, health care providers, community agencies, and friends and family. The survey of 4,682 visitors to a Toronto-area urgent care clinic generated some surprising results. As expected, use of the internet increased steadily as the age of respondents decreased, but more than half of all respondents in all age groups still rely on print sources. Also striking was that print sources resulted in the greatest awareness of risk-factor information by a wide margin. Health care professionals fared only moderately well for generating awareness and TV and radio fared poorly.

Data Mining

Canada’s large provincially based health care systems generate vast sets of patient data that represent an under-utilized resource for improving the delivery of care. Researchers at the University of Ottawa Heart Institute developed an automated data-mining system to extract relevant information from more than 82,000 dictated and transcribed clinical notes on more than 30,000 patients. The goal was to extract data necessary to determine the rate at which the medical management of patients with both heart disease and diabetes was fully compliant with clinical diabetes guidelines.

The system parsed the language of the transcribed clinical notes and resolved the context of the information to identify targeted data points. These included blood pressure, glucose and cholesterol measures, as well diagnosis of diabetes and drug names and doses. One of the big challenges was to get the system to recognize the context in which information was presented so that accurate data could be extracted from the clinical notes. Richard Davies, MD, who presented the findings, explained that accuracy of the system was improved through manual data-quality checks and through the system’s own self-learning capacity.

The project demonstrated that it was possible to extract the desired information from the unstructured data sources using an automated system. Analysis of the data identified trends in adherence to care guidelines, revealing areas for improvement.