Routine data can be a helpful tool in healthcare, but for such information to have a real impact--especially in the creation of a learning health system--certain challenges must be addressed.
Unlike information collected for research purposes, routine data is analyzed to help in the delivery of care for patients, a commentary on the trend published BMJ Quality and Safety discusses. Sometimes this data doesn't always show the full picture of the patient's health; the authors--from the London-based Health Foundation--call this a "data shadow" because often there is information that may be missing from a patient's record that has direct influence on their health.
Routine data can include administrative information such as reimbursement and contracting; clinically generated data, which includes information collected by healthcare workers to provide diagnosis and treatment; and some patient generated data that can be asked for by a clinician or offered up by the patient.
However, information that may be left out, the authors say, includes symptoms and response to treatment that occurs outside the doctor's office; input from other groups like caregivers; and lifestyle information, such as whether the person lives alone, if they smoke and stressors such as personal relationships and financial security.
While much of the industry has been focused on the collection and use of big data, there's also a lot to say for the influence of "small data," Larry Stofko, executive vice president of the Innovation Institute at St. Joseph Health System in Orange County, California, wrote last fall in an article for InformationWeek. Stofko said billions of dollars could be saved if we can connect the dots between small data, which maximizes individual care, and big data, which uncovers solutions that can have a global impact.
To fix this problem, the researchers say that instead of increasing the amount of data collected, questions should be asked about which information is the best to choose from, where it should be collected, when it should be collected and how to collect it. The collection of such data also should be contingent upon patient preference, since some consumers may be OK submitting information electronically, while others may not.
In addition, while collection of data in healthcare settings is now easier than ever, making sense of all the information is going to be the industry's real challenge.
The authors add that data analytics should be applied to the information, but the healthcare system needs skilled analysts who can "adapt to local needs and bring their expertise in the design, analysis and interpretation of data."
"A learning healthcare system may address the challenges faced by our health systems, but for routinely collected data to be used optimally within such a system, simultaneous development is needed in several areas, including analytical methods, data linkage, information infrastructures and ways to understand how the data were generated," the authors say.
To learn more:
- read the paper