Advances in health information technology suggest the possibility of enabling end-users options to configure their interfaces with other members of the healthcare marketplace to provide a more personalized experience. Given the different patient preferences, diagnostic approaches and treatment options available for a given set of symptoms, using health information technology to customize the patient’s (or caregiver’s) experience might help improve understanding or suggest a next best course of action.
In 2014, Minvielle et al. suggested six key factors to mass-customization in health care. First, delivery providers could stratify patients into segments. Second, information technology could affect service delivery in multiple ways, including greater choice of providers or products, and greater personalization based on more complete recall of a patient’s prior choices. Third, healthcare workers can apply human skills to tailor a experience to more closely match a patient’s treatment plan with their preferences as well as monitor for unintended adverse effects. Fourth, empower patients with more self-management tools. Fifth, engaged patients could provide information about their preferences to further customize the specific healthcare experience. Finally, tracking the economic impact of the customization would be important, both positive (cost-savings from stopping unnecessary actions) and negative (higher costs to train employees to tailor specific healthcare experiences based on a patient’s preferences). In 2017, Sean Duffy suggested tracking patient behaviors to validate their true preferences and better customize a specific user’s experience.
Given my experience with health information technology, I believe thoughtful choices about what data elements to collect and then using that information to display specific components to suggest a change in knowledge, attitudes and/or behaviors would the most effective way to provide a customized experience for patients. When a patient enrolls in an online program, the intake process should include some assessment of the patient’s current health status and behaviors along with diagnostic and treatment preferences as well as their communication and data sharing preferences. Health systems and payors may be able to discern some of this information from their own medical records and claims information, respectively. The program could then offer the relevant self-management tools to help that member determine actions to best improve their health.
Once an online platform performs better than a general Internet search at providing tailored information to a patient, the odds that patient will use the online program for other health needs is likely to increase. An acute or chronic change in health prompt the patient to better understand what diagnostic approaches might be most appropriate for someone with their specific health profile. Unlike existing provider quality metrics that focus on adherence to national guidelines or specialty recommendations, an online program would need to define a relevant cohort that matches the patient’s health profile as well as suggest options based on that patient’s preferences around diagnostic certainty, convenience and anticipated cost. After a diagnosis is made, the online program could provide options about how best to manage that diagnosis using a similar understanding of the patient’s treatment preferences.
Healthcare entities interested in providing a mass-customization experience may need to consider how they might simultaneously use patient preference and historical treatment information to suggest next best steps without violating patient privacy. Today, most online platforms that provide mass-customization rely on users providing extensive personal information for algorithm training and optimization. Unfortunately, we in America have not had a strong record of protecting that personal information for “other” purposes. We as consumers have not made choices that support higher levels of privacy over convenience. Healthcare information with its implications on health insurance premiums, risk for social ostracization and associated effects on family members, may require a higher level of data security to maintain a patient’s confidence in sharing such information.
Health information technology driven by designers focused on collecting information to provide customized options for each patient could distinguish a health care provider or payer in the marketplace. Those platforms that are more explicit about how they use patient information including methods for patients to remove their information upon request may be more likely to increase long-term patient engagement regardless of where the patient receives their care or manages their health insurance premiums. I believe the American healthcare system could benefit from a deeper exploration in this space.