Quantified Health: Empowering Healthcare for Better Outcomes

Big Data
Data Security
Customer Experience

Luiz Cieslak

We are experiencing  a transformation of behavior in our daily lives with access to more and more data points about our health. With more information about ourselves available now than ever before, the question that arises is: are we actually getting healthier?
When Victor was born, his parents Mateus and Lilian felt a lot of the same things most new parents feel—joy, pride, excitement, maybe a little fear...and definitely deep love for the newest member of their family. When Victor had his first open heart surgery on his 5th day, it was clear that medical care would be a big part of his life.
For Victor, who was born with a heart defect called Hypoplastic Left Heart Syndrome (HLHS), every little detail matters. Things that would be considered par for the course in a typical newborn, like having trouble sleeping or eating, slow weight gain, and even just general fussiness, can mean real trouble. Advice like “you’ll know what to do” and “follow your instincts” were replaced by “doctor knows best.” 
“The doctor is the authority in the hospital,” says Mateus, “A parent’s feeling is not really considered as much.”
Fortunately for Victor, his parents were not the type to leave the data to the doctors. As professionals in digital business, they were both well-versed and highly comfortable wrangling data and they started to track everything from feedings (how much, how often, what time, and what he ate) and sleep, to blood pressure, O2 saturation, and heart rate. “I used to track my own data,” says Lilian, “so it kind of came naturally to do it for Victor.” 

Health Data Tracking is a Growing Trend

Although remarkable, Victor’s parents are not unique. They are an example of a growing trend in health data tracking, sometimes called the “Quantified Self” movement, that takes advantage of advances in digital technology to track, store, report, and sometimes share health measurements in a form that is easy not only to capture but to analyze and visualize. 
With the introduction of the FitBit in 2008 and Apple Health and Google Fit in 2014, access to sophisticated data capture and analysis has exploded over the last decade. We surveyed over 300 healthcare consumers across a spectrum of ages and demographics and found that around 71% of people surveyed reported tracking some form of health data (e.g. steps, weight, calorie intake, blood pressure, etc.), and  around 37% tracked three or more different things. As you might imagine, the biggest adopters are in the 18–24 age group—those who have grown up with this technology—but the number of “trackers” doesn’t fall below 75% until age 61 and up. Even then over half of them still track something—even if it’s only one thing like steps— suggesting that this is not a fad confined to digital natives.

Data Supports Better Conversations with Health Care Teams

For Victor’s family, tracking the data and looking at it themselves was only the beginning. “I left my computer open at Victor’s bedside,” says Lilian, “When different nurses and doctors asked questions, I would just show them.” 
In our survey, we saw that the tendency to use data to improve conversations with doctors tracked with the amount of tracking done. Those who responded that they did use their data to have better discussions with their doctors tracked on average two to three different data points, while those who did not use the data tracked less excluding those who did not track data at all). In short, the more you track, the more likely you are to talk about your data with a healthcare professional. 
Mateus shared, “When you show the data the impact is there. They start to listen to you, and they have to act—we were one step ahead of the doctors.” Ownership of the data, its collection, and being able to discuss it level the playing field between patient and doctor. 
It’s worth noting here that it did not seem to matter specifically what people were tracking—no single factor, like steps, calories, blood pressure, etc. seemed to correlate nearly as strongly as how much they were tracking  and that they tracked anything at all. There were a few specific exceptions—people tracking data specific to a diagnosis, such as a person with diabetes tracking blood glucose, did have a strong correlation, for perhaps obvious reasons.

Quantified Health is Better Health

Of course, what matters ultimately is the answer to the question “Does all this tracking actually lead to better health?” The question of outcomes is the most important. 
Victor’s parents tell the story of how they were “encouraged” to keep Victor in the hospital rather than bring him home because the doctors were concerned about his lack of weight gain, a common issue with children in Victor’s condition. Mateus and Lilian were sure that Victor would do better at home, without the stresses and disruptions of the hospital environment. Again, they turned to the data. 
“Doctors wanted to keep him in the hospital to gain weight, but we were able to show the results were better at home. Showing the trends and the difference influenced the decision. We were discharged that day—it was so rewarding to know that we were making a difference.”

In our survey, we saw a definite correlation between tracking and outcomes. We asked people how they compared against a variety of stated health-related goals, things like getting more exercise, losing weight, better sleep, and quitting harmful behaviors like smoking. In general, regardless of tracking behavior, about 30% of people said they met or exceeded their goals. Those who did not track did worse at about 20%. Once people started tracking three or more data points, that number started to trend upward. Those who tracked between zero and two data points averaged 24.8%, meeting or exceeding goals. Those tracking three to five  things averaged 29.8%. Those tracking six or more? 56.3%. 

Success rates over number of factors tracked graph

Sharing Increases the Impact

We found a similar possible correlation with not just tracking, but sharing the data. People who shared what they tracked with one to two others had about a  25% success rate, whereas those who shared with three to four had around 41%, and of those sharing with five, 62.5% reported meeting or exceeding their goals. 

Success rates over sharing depths graph

Where are We Headed?

As powerful as this is at the individual level using self-tracked data, it’s only  the beginning. As in so many other areas of life, data is all around us and as its mass and volume increase, its power grows exponentially.
Perhaps the most apparent facet of this is aggregation—the now-familiar world of “big data.” For years, our online behavior has been tracked and stored (in the best case anonymously) and combined with that of other users to create vast oceans of data. This data, using machine learning, data visualization, and other large-scale techniques, can be converted into incredibly precise, individualized experiences—no two people have seen the same Amazon homepage in a decade. Netflix can guess pretty accurately not only which shows we will enjoy, but which cover art versions will be the most likely to convince us to watch them. The formula takes the relatively small data set of an individual, mixes it with the vast and impersonal data set of millions, and then—like magic—comes back to the individual with something that is uniquely tailored to them—and surprisingly often it fits perfectly.
Translating this to health and medicine is in some ways obvious, but the implications are mind-boggling. The same effect that today makes sure that we see ads, promotions, and opinions that are a striking match for us as individuals can transform generalized, condition-centric treatment into personal, human-centered therapy. Rather than prescribing what works for “most people” with a given disease or condition, every medication, every course of treatment—corrective as well as preventive—would be based on the individual and what will work for them. 

Treating the Person, not the Condition

When we go from condition-based treatment to individualized treatment, we are talking about a change that could be as dramatic as the shift from treating symptoms in isolation to treating root causes at the source. No two people have the same concert of anatomical, biochemical, and physiological features and no health condition exists in isolation. We are heading for a world where no two cancer patients, or people struggling with depression, or even just sufferers of the common cold, will take the same medicines or follow the same exercise routines or receive the same diet recommendations. Our medical and health care will become as unique to us as individuals as our Netflix queues. At the pace that data science and its related disciplines are advancing, this could be a real possibility in as little as three to five years.

From Individual Care to Societal Outcomes

As healthcare becomes more successful at the individual level through personalized medicine and therapy, the benefits will quickly extend beyond the person receiving treatment. Pervasive and persistent illnesses, failed treatments for conditions, such as substance abuse, with socioeconomic components have a huge impact at the community level in terms of quality of life as well as lost productivity and opportunity (a student whose education suffers due to a long-term illness, for instance). In these and other ways, better health outcomes for individuals in greater numbers translate into better social and economic outcomes for whole communities—and even nations.

What Will Need to Change?

In spite of—and in many ways because of—the success of data-based approaches in the marketplace of media, resistance is building and privacy advocates rightly worry about the misuse of this power to harm consumers. The Health Insurance Portability and Accountability Act, or HIPAA, passed in 1996 is only one of many state, federal, and international regulations that are in place to protect people from the misuse of their health data. Knowing what movies we like is one thing, but knowing about our serious health conditions, our mental health history, and conditions like chemical dependencies can be truly damaging if used with the wrong intentions—or even the right intentions with unintended consequences.

In the survey we conducted, we asked for respondents’ feelings about how difficult or easy it is for their doctors (and others who did have their interests in mind) to share data that would help achieve better care outcomes. 59.4% agreed or strongly agreed that it is too difficult for doctors and healthcare providers to share information.
Agree or Disagree: It is more difficult than it should be for all of my doctors and providers to share information about me with each other.

Shared info between doctors graph

Trust Factors

Along with regulation changes must come increased consumer trust. Doctors and pharmacists rank highly in this regard, but the insurance industry and government have significant work to do.

Trust ranking from most to least graph

Industry Behavior Must Change First

Clearly, the regulatory environment and the trust environment as a whole need to change, but that can’t happen in a vacuum. In order for regulations to change, the industry will need to show a much higher degree of responsibility and accountability. Genetic data aggregators will need to ensure that their samples are actually representative of the population as a whole, not just a narrow (and often privileged) segment in a few geographic locations. Insurers will need to show that they will not deny coverage not only pre-existing conditions but for highly likely future conditions. And everyone will need to prove without a doubt that the data is absolutely 100% safe from breaches and intrusions by bad actors. Regulation does need to change to allow the future we envision, but it cannot be at the expense of human beings and their quality of life.

On the Horizon

While there is good reason to view the future of health data with caution (and maybe even a bit of trepidation), the possibilities are extremely compelling if we can get there safely and responsibly. Rather than trying multiple treatments over months and even years hoping something will finally work, we will have a high confidence in finding the right therapy faster. Instead of patients and their doctors looking at a collection of symptoms and struggling to find a diagnosis, we will have answers in a fraction of the time with far fewer missed guesses. Far from being replaced or supplanted in this future, doctors and other health professionals will become more successful and achieve better outcomes more often for their patients. Quality of life will improve dramatically for all, not just the few and the fortunate—something we can all look forward to with the hope that we will overcome the barriers that hold us back.

Luiz Cieslak, VP, Digital Solutions, CI&T

Luiz Cieslak

VP, Digital Solutions, CI&T