Big Data Can Deepen or Dilute Caregiver-Patient Engagement

Big Data Evangelist, IBM

Bedside manner is something that some physicians have in spades and others totally lack. It’s not just a matter of personalities, etiquette and social pleasantries. Depending on the patient’s situation, it can make as much difference as medication, therapy and surgery in influencing whether patients stay well, avoid illness, or, if stricken, recover fully and rapidly.

keyboard-stethoscope.jpgBedside manner is in danger of being demoted to a second-tier concern in the age of data-driven medicine—or, at least, that’s what this recent article in The Atlantic is arguing. What jumped out at me is the following sentence: “Hospital interns spend only about 12 percent of their time interacting with patients. By contrast, they spend 40 percent of their time interacting with hospital information systems.”

Whether that’s an optimal allocation of physician time, considering metrics on healthcare outcomes in hospital settings, is an open issue. In the article, author Richard Gunderman doesn’t present any quantitative research that would indicate that outcomes improve, worsen, or stay the same as the amount of time that physicians spend perusing electronic medical records increases.

What he does is present a possibly apocryphal anecdote about an intern who trusted a record that erroneously described a patient he had never met as an amputee, rather than simply walk down the hall and meet the patient in person. He then sums up his thesis, without empirical corroboration, as “Often the presence of more information is symptomatic of decreased levels of communication and understanding.”

Let’s give today’s data-driven medical practitioners the benefit of the doubt on this matter. Not having any empirical research that would indicate otherwise, it’s safe to assume that most working physicians have found the right balance between data analytics and human interaction in their daily rounds. If healthcare outcomes haven’t suffered from expansion of data analytics applications in operational settings, even a significant drop in interpersonal communication may not be a critical issue.

And let’s also consider the possibility that interns may be going deep on analytics at the direction of the most experienced doctors who supervise them. This is a plausible scenario in environments where the established doctors may in fact have exquisite bedside manner and insist on handling the lion’s share of patient interaction.

Nevertheless, Gunderman’s concern should be heeded. Healthcare organizations can’t simply assume that expansion of data analytics in operational settings is always a positive development. Clearly, the need to attend to more medical record data, more diagnostic charts, more interactive visualizations, and so forth means that less time is available for other activities.

One of those critical activities is the heart of medicine: the need to physically examine, interact with, and treat the patient directly. In addition, in-person conversations are fundamental: they enable caregivers to gather a wide range of observational data that may not ever be recorded through standard tests, instrumentation and other formal means. All of this in-person observation enables caregivers to use all of their senses to glean “soft” data that may make all of the difference in diagnosis, monitoring and treatment.

So clearly, the line between data analysis and human communication isn’t hard and fast. The article fails to note that many real-world healthcare information systems provide operational guidance to doctors, nurses and others in their daily interactions with patients.

For example, just as call-center systems generate scripts that prompt agents to ask customers key questions about their requirements and help them probe for further detail relevant to trouble calls, many healthcare professionals have decision-support tools to assist in the detail-oriented gathering of relevant data and delivery of targeted advice. Indeed, when healthcare professionals are dealing with a steady stream of patients suffering from myriad maladies, these sorts of data-driven conversation-management applications can be a critical productivity aid with a substantially positive impact on outcomes.

No one’s arguing that big data can help doctors with their interpersonal skills. But big-data-driven healthcare applications can be an effective assistant, prompting physicians to probe for important details in their conversations with patients. Considering how busy, complex and stressful their jobs are, healthcare professionals should welcome big data analytics as a more powerful type of operational checklist.