When it comes to patient marketing and the healthcare industry at large, privacy is of the utmost concern. As more and more health data gets generated each day, healthcare marketers need to be sure the solutions they’re using to reach patient audiences have been designed to protect their privacy and comply with HIPAA regulations.
In developing our Patient Modeled Audiences solution, DeepIntent partnered with Mirador Analytics to address these specific needs and ensure our clients can reach highly-qualified patient audiences in a transparent and privacy-compliant way. Mirador Analytics CEO Jamie Blackport joins us for an in-depth Q&A about how they’re working to lower risk and maximize utility in healthcare data.
1. What is Mirador Analytics’ mission?
We aim to reduce the bottleneck in expert determination while ensuring risk is assessed and managed appropriately. We want to create the optimal balance between patient privacy and data utility to create a system where there’s no need to compromise individual privacy to gain great healthcare insights.
2. Describe how Mirador Analytics is integrated into the DeepIntent Patient Modeled Audiences solution to ensure HIPAA compliance.
DeepIntent’s Patient Modeled Audiences offers a precise patient targeting solution for healthcare organizations. The system brings together industry-leading data and advanced analytics. While DeepIntent continues to develop their offering to give more accurate results, Mirador Analytics is helping ensure data remains de-identified throughout DeepIntent’s processes by assessing those processes and making risk-mitigating recommendations, thereby protecting patient privacy.
3. Can you describe the “Mirador Metric”?
The Mirador Metric is our core offering that runs through all of our products and measures the disclosure risk analysis of healthcare data in relation to de-identified data.
To do this, we work on an ongoing basis with our partners to gather, collate, and understand their data and processes so that we can perform our expert determination assessments. This ongoing model allows us to bring more speed, utility, and risk mitigation to our clients in a way that wouldn’t be feasible with one-off engagements.
4. What are some misconceptions about patient marketing and targeting in the pharmaceutical industry?
An ongoing concern of stakeholders is the potential involvement of individuals’ health data in patient marketing. Ad tech companies like DeepIntent are not seeking to pick out an individual’s information and target them; they are looking for patient trends on a wider scale as they look to reach those potentially in need of life-changing therapies. There’s a focus on the processes to ensure the risk is lowered as they seek wider insight to minimize the need for concern of individuals’ data use.
5. How do you minimize the risk of patient re-identification?
There is always going to be residual risk in data (HIPAA describes de-identified data as having no more than a very small risk). Knowing that and seeking to find and address this risk is the best way to minimize it. At Mirador Analytics, we use expert statistical analysis to evaluate the relationship between smaller groups in the dataset and the broader population, and the degree to which linkage can be achieved. The methods used to determine and justify the expert’s opinion are documented and recommendations are made on mitigative tactics to reduce risk. Having a documented report helps companies utilizing de-identified data understand their boundaries better.
6. When it comes to patient marketing, how can marketers target precise patient audiences while data remains de-identified?
Healthcare marketers are not typically targeting individuals based on their healthcare data; what they do is gain in-depth and valuable insights from healthcare data that shows precise trends in a wider population based on other determinants. It’s the precise patient trends that are valuable to pharma as well as their customers.
Targeting precise patient audiences wouldn’t benefit a wider population, who may have a higher likelihood of having a condition and needing help, whereas precise patient trends would allow healthcare marketers to reach these wider groups and potentially save lives.
7. As people become savvier about ad targeting and more states pass laws, like California’s CCPA or the Colorado Privacy Act, what challenges do you see this presenting healthcare marketers?
Effective personalization is key to connecting with patients and increasing interest and sales. However, if the type of ad targeting causes the patients to feel their privacy has been invaded, it will have the opposite effect. Regulations like California’s CCPA or the Colorado Privacy Act make people more conscious about their data privacy and give them the power to act if they feel compromised.
Multiple regulations can cause challenges in process consistency across regions, which can slow down implementation. Documentation of processes and assessment of risk helps highlight risk and what companies are doing to minimize it. This can help to build vital trust factors as entities have implemented necessary safeguards needed to protect the privacy of sensitive medical information.
These assessments can also support the navigation of regulation for companies working hand-in-hand with their legal parties to comply as necessary, and in some cases going beyond.
8. What shifts have you seen in healthcare marketing since the beginning of the COVID-19 pandemic?
The global pandemic has sparked an accelerated movement toward virtual health, including telemedicine and remote patient monitoring. In addition, the lockdowns and self-isolation have supported the emergence of the new culture of social media and data sharing, which encourages society to more willingly share personal information through fitness and health-related mobile applications.
Finally, non-profit and governmental organizations have established centralized medical data systems to support the research and development of new COVID-related therapies. All these movements have generated an expansion in real-world data, which ultimately create more and more opportunities for the research community and healthcare marketers to access the high utility datasets and deep patient marketing insights.
Increased data and increased accessibility means the need for further privacy and security measures, as bigger data can potentially result in bigger breaches and incidents.
9. How can healthcare data benefit society while still protecting the privacy rights and personal data of individuals?
Today, advances in data science allow for sophisticated analysis of increasingly large datasets. In addition, the emergence of electronic health records, the real-life character, and the sheer amount of data make health datasets an attractive resource for public health and biomedical research.
The increased access to big data simultaneously increases the risk of re-identification. Conservative risk mitigation methods, such as Safe Harbor, can result in poor data utility. Therefore, it is crucial to adopt risk mitigation methods like expert determination, which creates an outstanding balance between data utility and privacy. This risk evaluation method can be tailored to specific projects to generate data richness and provide accurate insight and replicability for research.
A bigger focus on what information is necessary for use cases versus the gathering of all data needs to happen. Focusing on this will naturally reduce risk as less information is shared, and also give potential for more granularity within the attributes that are important.
10. What’s next for Mirador Analytics?
We’ll continue further breaking down of the bottleneck to ensure risk-monitoring is more regular and more widely used within our industry. We just launched our expert-led automated risk software to provide customers with the potential for disclosure risk assessment of data provided in minutes. This new product is an effective tool for organizations that regularly make deliveries from their data layout and want to ensure that the data they share is confirmed to be de-identified without introducing undue time delays.
We’re already working on developing and implementing these solutions further and wider, with the aim to decrease analysis time so there’s more time to focus on risk. Healthcare data is growing and evolving at such a pace, we need to keep developing to ensure there’s infrastructure in place for ongoing risk assessment in the future.