Patient X and big data

Patient X first experiences disease symptoms at home and reaches for Google in search of an answer. He explores numerous websites and arrives at the doctor’s office with disparate sources of information that will have to be translated by the medical staff. Alternatively, he is a pro-active participant, wirelessly transmitting his health habits to a central provider and expecting information to be synthesized into easy-to-understand summaries by his doctor. Patient X, in essence, becomes a complex series of data points that will have to be carefully analyzed in order to develop an individualized treatment regimen. Digital data will have to be compared with evidence-based research and ongoing in-person medical evaluations in order to initiate a meaningful conversation. This is a tall order for Patient X’s physician.  Big data analysis may aid in accelerating and optimizing a management scheme. However, expectations will have to be managed on both sides. As Patient X’s love affair with digital self-management via his smartphone cools into a mature relationship, he needs to become aware of the merits of combining this data with larger sets of information and the security hurdles doctors and other healthcare stakeholders face in storing big data.

Big Data

Michael J. Fox and other celebrities have become symbols of diseases affecting countless patients across the globe. They have drawn attention to the research performed in academia and the private sector and focused the nation’s attention on illnesses that cannot be cured. These diseases require us to “maximize the number of minds working on the problem in order to hasten the pace of investigation and drug development” (Tim Scherer, CEO of The Michael J Fox Foundation for Parkinson’s Research).

Collaboration 2.0, i.e. enhancing traditional scientific collaborations with big data in order to gain a better understanding of any given disease, is becoming the new norm for Alzheimer’s and Parkinson’s disease states. The Parkinson’s Progression Marker’s Initiative (PPMI) has offered de-identified data from an initial group of 423 patients and 196 controls to approved investigators through an online database. Collecting this data and integrating the information with other sources may ultimately lead to “better care for individuals, better care for all, and greater value for dollars spent (The Institute of Medicine’s Clinical Effectiveness Research Innovation Collaborative [CERIC]) [1].” Interested readers are referred to Eric Larson’s JAMA article, “Building Trust in the Power of Big Data Research to Serve the Public Good [1],” for a review of opportunities, challenges and successes associated with the use of patient metadata.

Security Issues

While the Snowden affair continues to be debated, Patient X is also aware that his online information is being used in a more “benign” manner by marketers to provide targeted emails/ads reflecting his personal tastes. If Patient X belongs to the social-media-savvy younger generation, the thought of sharing medical information in order to find a cure for his disease may seem like a logical/necessary step. Creating new jobs by addressing the big data skills gap and crowdsourcing medical information may overshadow any hidden fears that Patient X has about revealing his health data to third parties. However, security breaches impacting more than 21 million patients’ medical records in the last three years, illustrate the magnitude of this problem. Penalties imposed by the US Department of Health and Human Services (such as the $1.7 million that Wellpoint, a healthcare giant, had to pay for inadvertently exposing patient data), may do little to address Patient X’s concerns.

In the absence of an all-encompassing “opt-out” button to sidestep the need to evolve digital privacy laws, Patient X has to ask the appropriate questions to his care team and other healthcare stakeholders. The team may be composed of healthcare providers, health information technology experts, patient navigators, caregivers and others who aggregate data streaming in from smartphones, wearable sensors and electronic medical records. Ideally, this information is then stored securely, analyzed and displayed in an easy-to-read format that will engage the doctor and patient.

Assume for a moment that Patient X’s data is being analyzed as part of larger datasets by human beings and artificial intelligence. The possibility for accidental leaks or security breaches are magnified. According to the Cloud Security AllianceSM, the top ten big data security and privacy challenges are:

1. Secure computations in distributed programming frameworks

2. Security best practices for non-relational data stores

3. Secure data storage and transactions logs

4. End-point input validation/filtering

5. Real-time security/compliance monitoring

6. Scalable and composable privacy-preserving data mining and analytics

7. Cryptographically enforced access control and secure communication

8. Granular access control

9. Granular audits

10. Data provenance

Security experts have clarified major aspects of the big data infrastructure attack surface (a detailed description of each challenge can be found here) and are currently working with healthcare stakeholders to mitigate these risks or provide alternative options.

Reference

1.  Larson, E.B., Building trust in the power of “big data” research to serve the public good. JAMA, 2013. 309(23): p. 2443-2444.