2016 predictions for health IT

No, Watson will not replace clinicians
By John Halamka
12:00 AM
John Halamka MD

As the year ends and we archive the accomplishments and challenges of 2015, it's time to think about the year ahead. Will innovative products and services be social, mobile, analytics, and cloud (SMAC)? Will wearables take off? Will clinicians be replaced by Watson? Here are my predictions:

1.  Population health will finally be defined and implemented. Recently I asked a number of clinicians to define population health. Although no one could define it, they were sure that their current EHR did not provide the desired functionality. To me, population health tools in 2016 will automatically aggregate data from multiple provider, payer and patient sources then create lists of patients with care gaps to be closed. This article in Harvard Business Review illustrates some of the functionality we've built at BIDMC in anticipation of 2016 needs.

2.  Security threats will increase. Two weeks ago, the Attorney General for National Security from the Department of Justice visited Boston to meet with a group of CIOs and CISOs, describing the escalating number and sophistication of cybersecurity attacks. He concluded that if a device is internet connected, it will be compromised. At BIDMC, we will continue to invest millions in security technology, rewrite many of our policies and invest in continuous security education for all our staff. Despite our best efforts, I cannot promise a breach-free year in 2016.

3.  The workflow of EHRs will be re-defined. In 12 minutes, can a clinician enter 200 structured data elements, manage 140 quality measures, be empathic, never commit malpractice and make eye contact with the patient? Nope, it's impossible. This Wall Street Journal piece illustrates the problem. The EHR must evolve from a fraud-prevention tool in a fee for service world to a team-based wellness tool supporting alternative payment models. I've told CMS that the ideal EHR will be a combination of Wikipedia (group authored notes) and Facebook (you'll have a wall of health related events).

4.  Email will gradually be replaced by groupware. Managing daily email is a burden with minimal rewards. Facebook has announced Facebook for Work to provide enhanced communication among teams, supported by enterprise grade security. I receive over 1500 emails a day and might declare email amnesty in 2016 (an out of office message declaring email to be an ineffective communication medium and suggesting that I will never respond).

5.  Market forces will be more potent than regulation. Meaningful use has accomplished its goals. MU is dead, long live MU. We need to move away from prescriptive regulations so complex that no one understands them. Instead, we need pay for performance based on outcomes, giving providers and industry the freedom to achieve these outcomes using whatever technology they feel appropriate.

6.  Apps will layer on top of transactional systems empowered by FHIR. Epic, Cerner, Meditech, Athena, and eClinicalWorks are all fine companies. However, will the next great app be authored by their staff? I'm guessing a better approach is crowdsourcing among clinicians that will result in value-added apps that connect to underlying EHRs via the protocols suggested in the Argonaut Project (FHIR/OAuth/REST). One of our clinicians has already authored a vendor neutral DICOM viewer for images, a patient controlled telehealth app for connecting home devices, and a secure clinical photography upload that bypasses the iPhone camera roll. That's the future.

7.  Infrastructure will be increasingly commoditized. In 2016, I will be moving select applications to Amazon and Google. They can offer a better/stronger/faster/cheaper service because of their scale than I can do myself. They are willing indemnify BIDMC and sign Business Associate Agreements. Why do I want the risk of operating multiple data centers myself for commodity services like web hosting?

8.  Less functionality with greater usability will shape purchasing decisions. Recently a clinician told me that EHR A has half the features of EHR B, therefore EHR A is twice as good! Remember Wordstar and Word Perfect? Try authoring an outline in the most modern version of Microsoft Office. Prepare to have your work destroyed by feature bloat in Office. Clinicians want usability, speed, and simplicity, not more features.

9.  The role of the CIO will evolve from provisioner/tech expert to service procurer and governance runner. From 1996-2001 I wrote many of the foundational applications of Beth Israel Deaconess. My education at Stanford, UCSF, UC Berkeley, Harvard, and MIT enabled me to innovate rapidly as a clinician, domain expert, and engineer. Today I do not write code and my role is to empower/enable talented people around me with funding, protected time, and political will. The CIO of 2016 will increasingly be an orchestra conductor and not a technology expert.

10.  The healthcare industry will realize that IT investments must rise for organizations to meet customer expectations, survive bundled payment reimbursement methods, and create decision support/big data wisdom. I often tell my stakeholders that scope, time and resources are tightly coupled. You cannot increase scope without increasing time or resources. As more automation is deemed critical for the needs of the business, IT budgets will be increased as a strategic imperative. There will be a tension; the CFO will want to increase capital budgets (purchasing of stuff) while the CIO will want to increase operating budget (purchasing of services and subscriptions to cloud functionality).

That's my top 10 list. And no, Watson will not replace clinicians, although Natural Language Processing is a technology to watch in 2016. Other companies will do it better than IBM.

Happy New Year!