Improving Healthcare Delivery: Data as Service Provider

Struggles with Healthcare Delivery in Real Time

Karen GrossWe have all experienced, or know someone who has experienced, problems with our healthcare delivery systems. To be sure, there are commonalities among the issues that arise but each patient’s situation has unique, personal features.  These differences can lead to vastly different outcomes, including those affecting family caregivers. 

Consider a couple general examples. There can be a range of medical errors, whether caused by physicians or other medical personnel, some largely inconsequential and others leading to devastating outcomes; there can be discontinuity of care with siloed or non-integrative providers and procedures where each sector of the medical profession is working to solve the body part problem in their limited sphere and coordination is hard to achieve; there can be the failures to compare healthcare data taken over time and in different locales, at least in part because the information is disaggregated; there is rudeness; there are delays; there are whole knowledge bases that go untapped because of information asymmetry producing lots of wasted time, exasperation and inability to arrive at optimal choices and evaluation of suitable solution.  

In short, as with any complex system involving people and problems, there are opportunities for failures; indeed, we could remark that success in healthcare outcomes occurs despite – not because – of the large delivery systems and insurance conundrum.

Think about a marine like Chris Ayers and his family.  His injuries in battle were severe, and they had consequences that extended far beyond the physical repair of his severely damaged leg; he had psychological issues caused by brain injury and a tough adjustment from military life to family life. Not only did Chris have trouble adjusting and navigating his condition but so did his wife and children.  Yet it took years to get coordinated services; it took falling deep into the well of pain to enable resolution of the myriad of injuries that plagued Chris and his family.  His situation, as we reflect on it in retrospect, begs for us to find better ways to help those who were injured serving our nation.

Then, ponder the situation of MaryAnne Sterling who is the co-founder of Connected Health Resources. She thoughtfully recounts her experience (and she is a trained healthcare professional) trying to deal at a distance with her aging mother who kept on falling down. The mother was moved from facility to facility, often without the daughter having time to assess both the injuries and the quality of the locale where her mother was located. Yes, the daughter tried through phone, email and in-person visits to stabilize her mother’s condition but at every turn, there was a new event, a new provider, a new location with little or no coordinated care; it was as if the wheel was being reinvented monthly, if not weakly.  And, the mother’s condition took a toll on her daughter and her daughter’s work and family.  At one point, the daughter observed, quite rightly sadly, that she had become a health educator by necessity not by desire. 

One Solution to Fragmented Care that Actually Exists – if We Can Embrace It

We have the capacity today to begin to use open-source data to provide personalized information and services to patients, their families and medical professionals that will facilitate quality healthcare delivery. We have the capacity to build a nationwide shared services digital infrastructure that could be used to improve healthcare outcomes. And, if done well, this system has the potential to be transported effectively into other arenas including education and social services. 

But, the task of getting these systems to work now and into the future is daunting (indeed, perhaps more daunting than it needs to be). To move from idea into action, we need a minimum of these three things:

  1. We need to change our collective perception of data and its usage in the 21st century so there is less fear of data and its use among all people; data are not and should never be in the sole purvey of the elite and the highly educated professional (whether in medicine or academia); 
  2. We need to understand how these open source, open access systems are constructed and what are their capacities and we need support (and not just economic support but leadership support) to highlight outcomes and the value proposition – the ROI;
  3. We need to educate consumers on the data available and its use, showing how it can be deployed with relative ease, protection of personal privacy and in ways that foster improved healthcare results.

Marc Wine

Data Scare Us

Historically, we have left the conversation on data to data experts.  We think of data as complex and outside the understanding of the average person (whether young or old).  We look at statisticians and quiver.  We think courses are statistics are for eggheads or math geniuses.  We abdicate the whole topic of data.

Why are we so willing to abdicate data decision-making and allow others to decide what data matter and what data need to be collected?  Part of the answer is, our lack of understanding as to what data can and should be used for and when.  Yet another aspect, a critical one too, is that we have done a poor job in our educational system about teaching children and adults about data – from pre-K onwards.

Even assuming we appreciate the need for data, we have surely not settled on what data are important to collect, what criteria should be used for data collection and what conclusions including public policy recommendations flow from available data. Now, we know that old adage: garbage in equals garbage out.  So, if we use bad or incomplete data, the conclusions we draw will be flawed because they are not based on fulsome information. We are also missing some data; omissions can stem from data unavailability. In other words, we want to know certain information but that information has never been gathered at all or has been gathered in a fashion that does not enable its entry into a database, at least not without considerable time, effort and money – as well as data cleansing for errors and flaws. 

Surely, virtually everyone is aware of the risks data present if individual or corporate privacy are not preserved.  We know from our own personal experience that revealing data on a form – say in a doctor’s office – leaves us feeling vulnerable (even assuming we answer all the questions accurately).  We have seen our financial data compromised when our credit cards are stolen or even our identity is absconded.  We have read about and perhaps experienced the wrath of hackers and then the myriad of efforts needed to restore systems.

These aren’t easy issues – on any front.  And of course, we haven’t addressed in detail foreign intervention and data hacking for a myriad of purposes including blackmail and interference with international commence and safety.

After Fear, What’s Next?

The way we focus on data is, in a sense, very traditional.  We see it as inanimate. We collect it; we aggregate it; we interpret it; we report on it; we make changes based on what is suggests.  That is how we have become so enamored of the phrase “Data Informed Decision-Making” across the disciplines.  In fact, decisions made in the absence of data are challenged – even if there are non-data reasons to effectuate a particular initiative.

So, what if we changed how we think about data and moved from a static model to an active and engaged model?  Indeed, this is consistent with how we are now thinking about education: moving from teachers/professors being sages on stages to guides on the sides so students can be more actively involved in the learning process.

Picture data as a service provider, meaning that individuals and entities could access the data to get information that would be helpful to them – accurate, private, useful information.  Ponder viewing data as a source for engagement where individuals or corporations ask for answers from the data and the data are programmed (with or without artificial intelligence) to provide answers or pathways for decisions (like decision-trees) or even alternative approaches.

This may be hard to picture in the abstract for many; so, consider these concrete examples: 

Suppose there is a database that contains the health information of millions of people – not far fetched at all and these data are already being collected.  Posit that the data points can be matched to you and your particular test results and conditions.  Then, you and your physician would be able to assess how others with your disease and test results had fared and which treatments had the best results. Suppose too that the data gave you the names of ongoing clinical trials to which you could link to see if enrollment was an option. 

Suppose too that the database has information (personal) about you – the person inquiring and could link you to other services you may need and for which you might be eligible – needed prescription drugs for the condition at a reasonable cost; support groups for this illness; key articles that discuss the condition and its effects. In other words, a robust database could not only provide information; it could provide answers.  And, it could anticipate needs and desires and offer meaningful approaches and suggestions. 

In this just described incident, the data are performing a service for an individual; they are answering questions and providing information in real time that is useful.  The “backdoor” informatics are a series of databases integrated together, along with artificial intelligence that can process which information you need and what added material might be of use to you.

In short, in this example, the individual is getting the data to work for him/her. The data are providing a service – more than an information service. The data are fulfilling a need and much like any other service providers, are meeting (one hopes) the person’s present needs.

And, suppose – to go just a tad outside the proverbial box – the data are assessing the inquiring patient’s health through blood pressure and answers to select questions. And suppose the database with its complex interfaces has a sense that the patient is at risk for suicide or opiate use – pick your dilemma.  And suppose the database could contact third party providers of assistance who could find the patient (whose location could be determined from the web and tracking mechanisms) and intervene.  This is suicide prevention without it being expressly sought or intervention in a potential drug overdose before there is irreparable harm.

This would yet another example of data as service provider – a role largely untapped at present but with enormous and realizable potential to change the lives of individuals and organizations.

A New Data Paradigm

If we move data from being static and handled by experts at each point to information that can be used to aid individuals in their daily lives – helping them become better educated, healthier, more fiscally stable and more civically engaged -- then data will cease being viewed as “egg-heady.”  Instead, data will be seen as active providers of services and as such, a welcomed and much needed part of the lives of individuals.

As a first but key step, we can change our healthcare and educational systems so that they promote this changed approach to data.  In healthcare, we need for individuals to learn the possible benefits data can provide and in education, among other things, we need to teach all of our children the amazing ways in which – in their lifetimes – data can animate decision-making.  In the absence of these two points, we will never have and sustain a learning health network.  But, with education and an appreciation of and for the benefits of data networks on health (and other topics), we can make these big data sets become service providers.

We can and should teach children of all ages and at all stages what the future of data is and let them use their talents and imagination to build a robust data future.  We need our children to see and visualize and appreciate this new vision for data as they will be, for sure, its users over the course of their lifetime. And for the rest of us, we have an obligation to insure that our children and our children’s children inherit a world that enables wise and thoughtful decision-making about the issues we confront daily in our lives.  What could be more important than learning so we can improve our collective well-being?

Adapting The Learning Health System (LHS) for All

While not a panacea for the spiraling costs, poor outcomes, unevaluable silo pipelines of education, outdated public health infrastructures, alarming patient and public safety issues, and significant waste on treatments and courses and information systems of non-personal learning that simply do not work, the Learning Health System ( LHS) model is grounded in a recognition that many of the problems that cost lives and dollars can be seen in part as symptoms of an economy that delivers opportunity for some but not all. The LHS is more than a vision though; it is realizable. It is about enabling “big data” to empower all stakeholders in health care to “learn large;” it empowers advanced data to be a service provider, educating us about and for life. 

The design and operation of the national-scale LHS derive from its core values:

  • Person-Focused;
  • Protective of Privacy;
  • Inclusive;
  • Transparent;
  • Accessible;
  • Adaptable;
  • Trusted governance;
  • Cooperative and Participatory Leadership;
  • Scientific Integrity;
  • Value.

Enabling data to be a service provider through the LHS is not a pipedream; it is not off in the distant future; it is not a mirage.  What we need to do is have the courage to embrace data – hug the previously unhuggable so to speak and enable data to work for us and with us. 

If we delay in making these systems available on an anywhere, anytime basis for real people, we delay making meaningful use of available technology and artificial intelligence too. We curb the capacity to foster wellness and perpetuate remarkable lack of coordination across healthcare sectors. 

And, if now isn’t a good time, when will there be a good time?  Yes, there is some “risk taking” and “faith-believing” and “thinking reorientation” needed to change how we approach our health and the data surrounding it. 

But, if we don’t act, we all stand to lose.  Better to try a win-win.  Now.