"Big data" is one of the most promising developments in health care in a decade. It is already being used to identify at-risk populations and individuals who can be steered toward the right care, at the right time and place. And it is serving as a catalyst for new partnerships and collaborations, such as the one between Mayo Clinic and IBM that is putting the supercomputer, Watson, to work on efficiently creating customized treatment plans for patients.
There is much gold to be mined from health care's ocean of data, but there are reasons to be cautious. Information technology and its purveyors already have demonstrated a talent for pulling billions of dollars out of hospitals seemingly ill-equipped to judge IT viability, direct its deployment or require that it meet a minimal level of accountability for results. Big data and analytics could turn into another expensive IT stampede that kicks up a lot of dollars, dust and dubious results.
Of greater concern is what I'd describe as the McNamara effect. In the '60s, Robert McNamara moved from industry into government and eventually served as secretary of defense under former Presidents John F. Kennedy and Lyndon B. Johnson. He brought with him a band of whiz kids and a pile of confidence in the power of big data and analytics.
At the time, McNamara believed that victory in Vietnam was a numbers game. The key was to drop lots of bombs from high altitude, count body bags and wait for the inevitable collapse of enemy resistance. That is, of course, not quite the way things worked out. McNamara and his whiz kids were far removed from the human dynamics on the ground. Despite a commitment to win the hearts and minds of the Vietnamese, they lost both.
That's often the problem with big data, analytics and chasing metrics. They too often breed myopic overconfidence. The numbers, after all, never lie — except when they do. Numbers without context or interpretation can create a dangerous distance between key decision-makers and reality.
For example, analytics applied to patient population data could further distance health care leaders from the point where most of the real value will always be created in health care — between an individual patient and an individual caregiver. A population of patients could become a faceless herd, subject to economic experiments designed to steer them toward the lowest cost point and targeted utilization patterns.
Such efforts would presume, of course, that patient populations are steerable. Some would point to numbers indicating reduced utilization during the heyday of HMOs as evidence of steerability, but it is just as likely that recession and loss of employer health benefits were the real drivers. And despite efforts to declare slower increases in health care spending in the past five years a victory for legislators and regulators, the Centers for Medicare & Medicaid Services' Office of the Actuary recently attributed the slow rise to a sluggish economic recovery and sequestration. What big data means often depends on who's doing the analytics.
The Human Factor
Big investments in the quantification of big data may result in an underinvestment in much-needed qualitative inputs (the kind you get from careful and intentional observation of human behavior). An investment solely in big data is going to yield just that — a big amount of data. Without interpretation at the human level to add value, meaning and relevancy, data can never make the transformation into usable information.
It's the sort of insight McNamara's whiz kids might have gotten if they had had an opportunity to spend time in the London Underground during the Blitz. There they could have watched ordinary people in an exceedingly threatening situation harden into resistance and resolve.
It's been my experience that answers are most likely to be found at the intersection of the qualitative and the quantitative. If you've got a qualitative insight, it helps to get some quantitative validation. The reverse is also true. When you've reached a quantitative conclusion, it's best to be sure it works in practice. Although it's often hard to keep a dispassionate distance, you can learn a lot by just looking and listening. We need the qualitative perspectives of "big anthropology" to balance the quantitative insights of big data.
In their research, anthropologists traditionally have used qualitative approaches, which emphasize the way situational context impacts culture and behavior. Qualitative research tends to be flexible and interactive. As a result, it can generate important findings that may not have resulted from other methods of data collection and analysis.
Anthropological findings can develop without a predefined question or hypothesis. Conclusions are often emergent. They arise out of the observation of people in context. Ann Jordan, a professor at the University of North Texas and a practicing business anthropologist, caught the essence of the challenge: "Human behavior must always be studied in context; to find what consumers want, one must ask and observe them."
Jordan shares an observation from Francisco Aguilar, a business anthropologist who has emphasized that, "Anthropologists have the ability to begin their data-gathering as an unbounded inquiry with no preconceived notion of what is important. They have to discover the questions to ask as well as the answers."
Jordan describes the kind of qualitative perspectives an anthropological approach may uncover: "If the site has a parking lot, for example, I notice the size and percentage of occupancy of the lot. I notice if there are reserved parking spaces and, if so, for whom they are reserved. I notice the location of outdoor gathering areas, whether there might be picnic tables for lunch or an area where employees gather to smoke. From these observations, I have learned potentially useful information about the organization before I ever enter the building.
"Reserved parking spaces for executives and various levels of managers may give me some clues about the hierarchical nature of the organization. The degree of rigidity of the hierarchy may impact successful completion of some kinds of work. Amenities for employees, such as picnic tables, a designated smoking area or a basketball court, may indicate an attempt to provide workers with extra perks to improve the quality of their experience at the workplace and may impact the relationship between employees and executives," she adds.
Hospitals, by their nature, provide rich settings for generating anthropological questions. Consider sitting and observing in the waiting area of a large urban hospital emergency department. Then consider doing the same in a small rural hospital. Watch. Do patients and families have a hard time orienting themselves when they walk in the door? Why? Are they out of breath? Why? Does the triage nurse make eye contact? Why not? What do male patients do? What do female patients do? Today, family members can spend many hours, sometimes days, in a hospital room with a patient. How do they live?
Dispassionate external observation, known in anthropology as the "etic approach," can also help in nonhospital settings such as physician practices. Why does the receptionist have a frosted sliding glass window between her and the reception area? Why is there so little communication between front office staff and patient care staff?
Baby boomers sandwiched between aging parents and adult children are becoming increasingly sensitized to the anthropological dynamics of health care. They are spending more and more time sitting next to a hospital bed in what they too often experience as a confusing and foreign world inhabited by tribes of health care workers.
I had an introduction to this world when my youngest son Andrew spent three months in a children's hospital battling an immune disease that almost killed him. At that point, I'd already spent nearly two decades in health care, a lot of that as a consultant to hospitals. Still, I was a stranger to the forces that exist inside the vortex of a hospital and its numerous tribes.
There was the tribe of the "inside doctors." And the tribe of the "outside doctors." And there was the nursing tribe. And lots of other tribes. On occasion, the tension and dysfunction between the tribes was palpable. While Andrew experienced extraordinary competence and kindness, he also experienced mistakes. And when the mistakes occurred, tribe behavior often kicked in — most noticeably, silence in the face of an obvious error. And then tribe coalescence might be triggered as its members tightened around one of their own, endangered by the possibility of being blamed for whatever mistake had been made.
Then there were the parents, including the tribe of the dying. These were parents and grandparents of children so sick they had few prospects for leaving the hospital. And there was the passive-aggressiveness of the tribe of the vigilant parents, compelled to push and bargain, but not too hard lest they alienate the nurses and doctors their child's life depended on. The administrative managers' tribe was invisible.
Many years later, I remember waiting to watch an open-heart procedure at the Cleveland Clinic. The team assembled in the OR was boisterous and animated. A subculture in action. The team settled down to attentive silence when heart surgeon Toby Cosgrove walked in to perform the procedure. More observable tribe behavior.
Cosgrove is now CEO of the Cleveland Clinic. He was preceded by another heart surgeon, Fred Loop. During his tenure, Loop had held regular meetings that brought together the clinic's top physician and administrative leaders in sessions that could be described as semistructured at best. He likened them to the Israeli Knesset because only the most compelling ideas and presentations were likely to command the attention of the group. It was a Darwinian kind of survival, where ideas competed and only the strongest arguments were likely to prevail. This, too, was distinctive tribal behavior.
Anthropologists in Business
There is significant precedent for using an anthropological approach in contemporary organizations. Anthropologists have been deeply involved in a wide variety of business enterprises. Corporations, particularly those in the consumer product arena, make extensive use of anthropologists. They include giants like Intel, GM, Microsoft and Xerox.
Methods derived from anthropology have become a mainstay of market research into consumer needs and behavior. McDonald's is faced with declining sales, particularly among younger customers, and is described by some of its own executives as having lost relevance with consumers. In June of this year, it set up a "learning lab" at a McDonald's restaurant in California to better understand what people want and to experiment with customizable burgers.
Medical anthropologists have conducted studies that focus on improving communication and patient discharge at the Veterans Health Administration. Other studies have been conducted in hospital emergency departments related to throughput. One study explored the values, motivations and alliances that influence physicians' attitudes toward reporting incidents, suggesting that professional subcultures can prevent reporting but also provide opportunities for improvement.
Mayo Clinic has developed an innovation center that has conducted a number of anthropological studies. Anthropologists working with Mayo Clinic have generated research suggesting that patients tend to view the health care system as inflexible, intimidating and guarding information that should be shared. Patients also want to be known as a whole person, and they value most a relationship with a particular professional who is nurturing and easy to talk to.
A Mayo anthropologist used the entire town of Austin, Minn., as context for study. She found that the No. 1 reason people want to stay healthy is that someone depends on them, that churches can serve as resources for social outreach and that, although there is a wide array of social services agencies, they don't communicate with each other or the public.
When Mayo created its clinics in Jacksonville and Scottsdale, there were concerns about the best way to preserve and extend Mayo's defining cultural characteristics. In response, Mayo hired another corporate anthropologist. At the end of a six-week study in which she posed as a patient, observed families in waiting rooms and shadowed Mayo physicians and staff, the anthropologist presented 11 recommendations. All of them were implemented.
Mayo also has created a "living lab" in Charter House, a retirement community with 400 residents that is physically connected to Mayo Clinic in Rochester, Minn. The goal is to support "aging in place" so seniors can remain at home while staying healthy and independent. New services and technologies are designed, prototyped and piloted in the lab.
A Qualitative Perspective
Much ink is spent pointing to the need to reshape a hospital's culture. One problem with this is the presumption that a hospital has one prevailing culture when, in reality, there are multiple, arguably hundreds, of fluid subcultures at work. Leaders need to recognize this and target interventions to strategically important subcultures (the ED subculture or the OR subculture, for example). Rather than presuming to shift an organization's culture wholesale, they should focus on a few tribes at a time, applying Pareto's law, which suggests that 20 percent of the variables yield 80 percent of the impact.
Health care needs a class of professionals whose job is to dispassionately watch, report and recommend. Teams of such professionals could help to bridge the gap that already separates many health care leaders from the front lines where patients and caregivers interact. They could help to prevent the McNamara effect by providing qualitative perspectives to counter overreliance on quantitative perspectives generated by big data. Such teams would need to be sufficiently immersed in the organization to discern patterns, habits and conventions. They would need to be attuned to behaviors that manifest themselves as better care and worse care.
Because these professionals need familiarity with the complexities of health care, it might be best to draw from the ranks of nurses and midlevels who have training in anthropology. A risk in the approach, however, is that these observers may not be able to remain "culturally neutral" and, thus, may project their bias. Great care would be needed to maintain a balanced perspective by ensuring that teams remain sufficiently detached.
Team members also would require access to administrative and clinical leaders so they could share observations unfiltered. This would require a degree of independence from the organizations they serve and might be accomplished best by rotating teams through multiple hospitals. A multihospital system or network could provide an appropriate vehicle to sponsor these teams.
American caregivers, including physicians and nurses, may argue that they are sufficiently attuned to the needs of patients and have no need of outside perspectives. Research suggesting that patients experience hospitals as foreign and often hostile environments would argue otherwise.
A Lesson from the Military
In 2007, the U.S. Army launched human terrain teams, or HTTs. Each HTT generally consisted of five members with training in the social sciences and linguistics embedded with military units. Their job was to provide commanders and their staff with a cultural understanding of their operational environment.
The Army had come to realize that even after several years of fighting in Iraq and Afghanistan, military units on the ground had little cultural understanding of local populations and that effective engagement of the population is central to winning struggles with insurgents. In an initial pilot, five teams were deployed. These were well-received, and more teams were requested. By 2008, the Army had 30 teams on the ground and by 2010, the number of HTTs peaked at 45.
My son Andrew survived his hospital stay and grew up healthy. Today, he is involved in intelligence work on behalf of the Department of Defense. He is also an intelligence officer in the U.S. Navy Reserve. The work allows him to apply degrees in anthropology and business. It's also given him a perspective on the Army's use of HTTs. His observations are cautionary.
HTTs have been controversial. They've generated strong ethical concerns among professional anthropologists, including the American Anthropological Association. They remain concerned that a research field that dedicates itself, along the lines of the Hippocratic Oath, to do no harm to a culture is being used in warfare.
The HTTs also have detractors in the military and intelligence communities; much criticism is related to the effectiveness with which the program was initially deployed and managed. Some commanders feel that the HTTs, by merit of their civilian status, don't fully understand the strategic and tactical needs that drive their research while the intelligence community views their work as inferior and duplicative to existing methods of human intelligence.
Other commanders, including those involved in studying counterinsurgency and asymmetrical warfare, found HTTs and the information they produced to be very useful. There continues to be a recognition that the need for which HTTs were created remains as pressing as ever — insurgencies are best fought by engaging hearts and minds at ground level. Doing this requires a base understanding of what guides a particular culture's heart and the acceptance that "our way" may not be the same as "their way."
Though the HTT's use has diminished, their experiences, methods and results have, either through purposeful distribution or serendipitous realization, been exported out to the rest of the military. Cultural sensitization has become a driving force behind counterinsurgency operations in the global war on terror and now includes realized concepts such as expanded foreign internal defense and female engagement teams. These concepts have evolved from a need to understand cultural barriers in an environment in which lives may literally depend on it.
We may learn that the same is true when it comes to combating intractable barriers to improving the quality, accessibility and affordability of health care. Like HTTs, "care terrain teams" would need to be sensitive to the culture, obligations and stresses of those ultimately accountable for generating results. In health care, it is caregivers on the front line who produce the results that matter.
Dan Beckham is the president of The Beckham Co., a strategic consulting firm based in Bluffton, S.C. He is also a regular contributor to H&HN Daily.