Raju Chauhan attended the second-annual HealthImpact Southwest conference in Houston, Texas. In this article, he shares with us a session involving the creation of new patient-care systems and the improvement of dashboard designs for better diagnosis and treatment at point of care, leveraging clinical-support tools and advanced informatics:
A full day at the conference was filled with interesting topics confounding industry leaders. Good use was made of time because of the “No BS” style of conference leader Shahid Shah, “The Healthcare IT Guy.” The conference was formatted around information sharing with several industry leaders onstage. Then there was a follow-up with questions from the audience that brought up real-life issues surrounding the topics.
Samir Badlani, chief health information officer at Intermountain Healthcare; Arun Matthews, chief medical information officer at Medical Center Hospital; and Sean Benson, vice president, innovation, at Wolters Kluwer Health shared some thoughts.
I think tackling this issue is important, since doctors and other healthcare providers are already a strained resource in today’s world. We need systems that simplify their lives and present actionable information to them that can help in a patient’s care.
EHRs started as financial-tracking systems and have now grown into patient record-keeping repositories with a whole lot of patient health information. EHRs have now become clinical-transaction systems and can provide a wealth of information, but the question is whether the captured information is actionable or not.
For example, systems developed at the University of Chicago can predict that a patient will have a myocardial infarction in 31 hours, but there is nothing that a doctor can do to avoid it. Is this insight actionable and useful? On the other hand, the systems are able to help decide whether, based on outcomes, to utilize the latest technology or older, cheaper technology when making a purchase decision for a tonsillectomy. Incidentally, in this case it was found that older, cheaper blades used for the procedure versus newer technology had no significant outcome improvement – but had a cost difference of 10,000 percent ($0.06 vs. $600). This insight makes the decision-making process simple and useful.
Modern patient-care systems collect a whole lot of patient information, but the question is, “How do you get clinicians excited about data collected through various channels and analytical insights from such data”? Some thoughts:
- First and foremost, transparency is a must; be prepared to share details and processes by which data are synthesized.
- Include multiple physicians in the design and issue-solution processes. It is not sufficient to simply include a single physician/clinician on the team. Multiple experts should be consulted.
- Design systems to solve specific problems and demonstrate specific improvements.
- Educate patients on how the analytics and data help refine processes and improve patient care.
- Be cognizant of alert fatigue and attempt to reduce it, and avoid false positives that cause interruptions in the clinical workflow. Alerts should be based on a simple principle that the clinician “did not know and wants to know.”
- Keep physician/clinician workload in mind when proposing and designing new systems. The new systems should:
> be disruptive to benefit but not disrupt clinician workflow.
> help automate clinician workflow.
> automatically capture useful information as the clinician engages in patient care, e.g., moving in and out of the examination/hospital room and remotely transcribing voice notes captured in exam rooms.
> help automate the associated business processes, e.g., capture progress notes electronically as clinician dictates.
> determine what work can be offloaded by the doctors to other licensed and qualified resources like nurses and pharmacists.
- Include patients as part of the solution as well as demonstrate a benefit to patient care by leveraging analytics to share with the patient. For example, present relevant knowledge and educate the patient.
- Include data from devices directly in patient records by using device interoperability, so that it can be easily factored in by the clinicians. Today, such data may be available but not integrated. Doctors may be reading such devices and duplicating such information in EHRs, which can be error-prone
Additionally, it is important to improve the dashboards that present patient information to a doctor for diagnosis and treatment at the point of care. Things to consider in effective dashboard design are:
- It is important to validate the visual presentation of data with the clinicians.
- Include analytical groups to guide the design, development, and presentation of insights
- Provide next possible actions to take. Today, dashboards are designed as scorecards and focused on where the clinician may be lacking. Instead, it would help if the dashboard could show leading actionable measures. As an example, it would be helpful to know when a patient needs his next antibiotic dose as well as how many are waiting for their doses so that a nurse can plan out the work. An actionable leading analytics would help improve patient care.
- Present all relevant data to the doctors on a single screen.
In order to implement the recommendations mentioned above, digital health organizations are encouraged to take the following steps:
- Create systems that provide specific actionable guidance to healthcare providers.
- Include several healthcare providers when designing systems.
- Bring data from multiple sources to provide integrated information.
- Integrate devices to feed into resulting dashboards.
Obviously, the meaningful-use initiative has overwhelmed physicians with additional tasks in their day-to-day activities, and it is important to help doctors with compliance so that they can keep their focus on patient care.
Share your thoughts on presenting leading versus lagging analytical results on the dashboards.