About this blog
Welcome to the SAS Health and Life Sciences blog. We explore how the healthcare ecosystem – providers, payers, pharmaceutical firms, regulators and consumers – can collaboratively use information and analytics to transform health quality, cost and outcomes. I’m Jason Burke, Director of Health and Life Sciences here at SAS, and you can read more about me here.Subscribe to this blog
Archives
Consumer-friendly predictive analytics are the key to making CDHPs work
The new year, filled with New Year's resolutions and rolling your accumulated out of pocket health care expenses back to zero. In fourth quarter last year, most employees renewed their current employer provided health plan coverage options without question. But for some, they were learning about new coverage plans – perhaps an increase in their out-of-pocket share- be it premiums or copays – beginning on January 1, 2012. Others may have learned about the only employer offering: the Consumer-Driven (or Directed) Health Plan (CDHP).
Under this broad industry category, offerings vary by benefit design, health savings account (HSA) and employer intent. Yet one theme that unites all CDHPs is the premise that when individuals pay close attention to their medical costs, they make more informed decisions about the services they seek and the providers they visit.
Sound familiar?
The CDHP concept reminds me of what happens when my teenagers tell me they “need” some new trendy item. If they think I’m paying, their “need” is urgent. But when they learn they’re paying, they suddenly find time to shop for the best price and wait for coupons, sales or perhaps forego the purchase completely. That’s the idea behind CDHPs. But they’re not right for everyone – at least, not yet. Nevertheless, the rising cost of health care and the struggling economy coupled with changes to the US health-care landscape almost guarantee we’ll continue hearing more about them.
For healthy consumers, CDHPs offer an economical choice. CDHP members usually pay lower monthly premiums than traditional plans charge. Members who remain healthy and require little more than free preventative-care services will enjoy lower annual costs. Conversely, consumers with new or existing illness may face hardships up front. In addition to their premium costs, they may pay nearly $2,000 in out-of-pocket expenses – hence, the creation of HSAs. And when those consumers reach their out-of-pocket maximums, many go for all the elective services they can handle, which reverses employer intent.
Pros and cons
The number of employers offering CDHPs grew for 2012 coverage, according to Mercer, a benefits consulting firm. More than half of large employers and less than one-fifth of small employers offered at least one CDHP. And several large employers for the first time, – including GE, Wells Fargo and JPMorgan Chase among others – only offered CDHP. (According to Reuters, “Wells will still offer traditional plans in California and other states where switching would force too many employees to change their doctors.”)
Why the move to CDHPs? Cost, for one. Mercer estimated employer costs and employee premium costs average 15 percent less under CDHPs compared to conventional PPOs and HMOs. Some contend CDHPs are a way for employers to shift more costs onto their workers while offering them less coverage. To a point, that’s true. Costs aside, CDHPs attract employers looking to give their workers more say in their coverage and care options. CDHPs works well for employees interested in managing their own health care, maintaining certain providers, or having the freedom to shop for less expensive alternatives without navigating the prior approvals or referrals to specialists gauntlet. According to the Kaiser Family Foundation, 17 percent of covered workers elected to enroll in a high-deductible health plan in 2011, up from just 4 percent five years earlier.
What’s missing?
Given the uptake by employers and employees alike, why do I suggest the CDHP is not ready for everyone?
Effectiveness and success depend on too many factors – among them, consumer attitudes and behaviors, access to services and providers, and utilization of preventative services to ensure no underlying conditions exist. We lack sufficient insight into these factors and thus have no basis for ensuring success across a large, broad patient population. Opponents contend CDHPs, in effect, discourage the cash-strapped, demotivate the thrifty and further disincline those who already avoid trips to the doctor from seeking medical care when they need it – instead relying on self-diagnosis and resorting to self-medication. By foregoing proper care now, many will later suffer otherwise avoidable, preventable conditions that will prove costlier to their health, wallets and job productivity. Advocates, on the other hand, point to the cost visibility and personal accountability that the actual health-care consumer gains under CDHP – components lacking under the current US system.
Individual decision support
What will it take to make the CDHP a suitable offering for the population at large?
If the goals of a CDHP are to save money and empower personal choice, health plans must equip consumers with the tools they need to make wise decisions. With so much at stake, workers cannot rely on their knowledge of the system. Nor can they rely on their usual “back of the envelope” calculations to decide how to fund their future HSAs. Instead, plans should offer consumers a robust predictive health-care cost solution that consolidates and models a variety of health-related data to estimate possible spending corridors. Predictive analytics would quantify for employees how much they’re likely to spend on medical expenses next year and thus how much they should invest in their HSAs to soften the blow of steep out-of-pocket requirements.
Employee-specific data as well as CDHP plan history would populate these predictive models. For the employee data, the model would drive the personal health-risk assessment (HRA) along with that individual’s historical health-care spending. Other factors could include behaviors and health-care spending from the CDHP population at large. Predictive analytics would drive deeper insights, such as the impact of lifestyle modifications on overall costs. By putting such rich knowledge in each member’s hands, plans would allow employees to see in real dollars how their health-care decisions impact their future.
As employees utilize their coverage in 2012, many continue to feel jittery about the economy. Concerns about household income and future health-care costs prompt some workers to change their household spending, delay medical care or feel anxious about future medical costs. In a less-than-certain labor market, individuals will continue to rely on – and expect – better tools for managing personal expenses. Health plans have most of the data, the interest, the power and hopefully soon, the predictive health analytics to help them.
Advanced analytics in commercial life science, the rewards offset the risks
Over the past ten years managed markets organizations have developed their own analytical capabilities to evaluate membership, portfolio and health coverage offerings. The results from these analytical exercises drive annual benefit option refinements. For life science organizations, product coverage at play during managed markets negotiations include tier placement, copay thresholds, and utilization management techniques. Today, due to numerous market factors and legislation, the managed market environment is even more complicated, illuminating a critical business issue for life science organizations: "How to optimize the mix of contract rebate dollars and other financial considerations to managed markets across product portfolios, market segments and geographies to drive maximum revenue, market share or profitability?" Are life science organizations equally armed with analytical firepower to participate and successfully negotiate optimal product coverage contracts? How is optimal defined in the organization? Do the brand and the managed markets teams agree on the metrics for success?
F
or life science organizations to become more competitive in these expensive negotiations, to make the most of their data assets, both internal and third party, and to create value regardless of economic conditions, they must consider the increasingly important role advanced analytics can play. In the right organization, insights gained from advanced analytics will drive future direction, predict the changing healthcare landscape, and identify market opportunities that set the organization apart from their competitors. However, for these initiatives to succeed, organizations need more than just technology, they need a strategy, change management initiatives and an implementation plan to coordinate and align key dimensions of their managed market organizations.
What is at risk with business analytics as usual? Profitability! The market is becoming more complicated and more transparent due to market restructuring. Organizations failing to address the new market run the risk of:
* Creating partnerships with non-performing payers
* Paying for utilization without transparent patients benefits
* Lack of visibility to the purchasing patterns and health status for the non-contracted
* Increasing exposure to government pricing thresholds
* Arbitrary rebate thresholds institutionalized without substantial basis
Yet, a large percentage of life science organizations are still struggling with several aspects of using advanced analytics, including how and where to start, how to take the next step and how to change their internal culture to one that understands the value of analytics as an integral component in the decisions that matter most. These decisions could be automated decisions that enable others to take best actions, one-time decisions that inform executive strategy or a range of decision types in between.
With stock holders watching, patients paying a larger amount out of pocket for products, and health plans scrutinizing products while growing the utilization restrictions the risks are too great to continue analytics and business as usual.
Photo by: http://www.flickr.com/photos/gustavthree/3681251333/sizes/s/in/photostream/
Nurse practitioners and physicians assistants - untapped opportunity for health care and life science companies
National health care reform will create coverage for an estimated 30 million additional consumers in 2014. These consumers will inevitably be seeing a primary care physician on a more frequent basis than they are today. The question is how our healthcare system will engage with this many new patients as the wait for an appointment with a primary care physician exceeding 20 days in certain metropolitan areas.
As the lead time to supply new physicians to address 30 million additional covered patients, is longer than 2 years, our healthcare system needs to leverage more of its existing healthcare providers, specifically nurse practitioners(NP) and physicians assistants(PA). Today, there are approximately 85,000 NPs and 26,000 PAs practice in primary care today. The ability of these two professions to diagnose, treat, and prescribe varies widely from state to state.
Hopefully, the individual states can proactively assess the potential shortage of prescribers within their state and address the changes for additional autonomy by these professions.
If this happens, the life sciences industry will have a new audience for providing patient support materials, co-pay cards, and potentially samples. On average NP’s are writing ~70 prescriptions per week while PA’s are writing ~85 prescriptions. (Physicians are averaging 95 prescriptions per week). With the significant increase of covered lives in 2014, we can expect the NPs and PAs to be seeing more patients and subsequently writing more prescriptions.
Life sciences companies have not typically targeted these prescribers for sales rep details as their ability to link prescriptions written to an individual NP or PA has been difficult and inaccurate due to the fact that many NPs/PAs write on other physicians pads or even their own pad but with a physician's name/number at the very top with other physicians in the office which is captured with the script at the pharmacy.
In the future, this issue may very well be addressed through ePrescribing by which the NPs/PAs name is identified as the prescriber when transfer to the pharmacy.
In the interim, the multitude of prescribing data, AMA address data, office/practice affiliation data enables mapping scripts to NPs and PAs to provide insight into volume of scripts at the individual prescriber level. Leveraging advanced analytics to mine this data will provide a wealth of opportunity to target an audience that has largely been untapped. As life science industry continues to focus on ROI for sales and marketing initiatives, the return on pursuing this audience of professionals promises a solid return.
Disease management programs – Receive help from 'big data' and analytics
Previous disease management programs have not delivered the results that our healthcare system anticipated or needed. Big Data, advanced analytics and a new healthcare ecosystem collectively have the potential to make this time around different for disease management programs for a host of reasons. First, there are multiple stakeholders focused on improving patient outcomes with financial incentives to make this a reality. Second, we will have multiple large scale claims databases that will enable more detailed analysis of patient populations for aligning and refining program investment for optimal impact. That said, there are many complexities in developing a program with analytic rigor around both directing the investment in support services and enabling the financial rigor with the end goal of implementing a sustainable financially viable disease management program. Additionally, the various stakeholders will be required to work collaboratively in order to achieve optimal success.
The healthcare reform act has established financial incentives based upon health outcomes. This places the patient and the patient's care at the center of focus for our healthcare ecosystem. Both payers and providers will be incentivized to attain improved outcomes while managing costs. This is will be a change for provides who have historically been compensated based upon activity(visits, tests, and treatments). This will also be a change for providers which will now share in the risk of managing costs down. Finally, pharmaceutical companies, which are already paying into healthcare reform through a “drug tax”, will be limited in their ability to take price increases as there will be increasing scrutiny on value-based outcomes and with a keen focus on costs. Hence, the pharma industry would be well served if the patient adherence challenge was addressed as part of more comprehensive disease management programs.
If all three parties work collaboratively and shared data/insights, disease management programs will be designed are were cost-effective and deliver improved patient outcomes at a scale that will drive changes to our overall healthcare cost structure.
Each party brings unique insight, data, and experiences to assist in the design and execution of a disease management program that could make a sustainable difference.
- Providers bring the deep insight to a patient's health, longitudinal view to the patient’s disease progression, and hopefully some historical insight to a patient’s past behavior in managing their health.
- Payers bring a comprehensive view to patient medical claims across providers, labs, pharmacies, etc. Additionally, they may have collected one or more health histories to proactively manage at-risk members. Finally, payers bring experience in managing disease management programs on behalf of their employer clients.
- Although pharma companies do not bring individual patient data, they do bring a deep understanding of patient populations from both primary and secondary market research studies and countless adherence programs launched over the last 15 years. The multitude of marginally successful adherence programs yields light on what worked and what did not when attempting to address the ~200 influences on patients' non-adherence.
Together, these parties have the ability to approach disease management with the business acumen and analytic rigor to ensure that resources are aligned at the individual patient level for improved disease management.
One of the first steps is to identify which patients’ require assistance in managing their condition such as the ability to follow healthcare professionals’ directions in modifying lifestyle behaviors, completing follow-up appointments/exams, and maintaining adherence to prescribed treatments. Investing in patients that do not need support wastes resources and jeopardizes the financial viability of a disease management program.
The patient selection process should continue to the next level to also discern patients from those that have the ability to “change their behavior” versus those that do not. Not every patient will have the ability, even with a comprehensive disease management program, to change their behavior relative to managing their condition. Programs need to invest in patients that want to and are capable of changing their behavior.
While disease programs have historically sought to identify those patients that "need" addition support and then invest in those patients, few have also segmented/invested in patients based upon the patient's ability or propensity to "change" their behavior.
Finally, we need to invest at different levels based upon patient need and disease severity. Again, the investment needs to be focused on delivering results with a view to financial sustainability of lower costs while improving outcomes. Is this easy? No. Is it achievable? Yes.
What if all three parties worked collaboratively to deliver the following:
- Patients with a chronic condition are identified in the office for “potential” inclusion in a comprehensive disease management program.
- The treating healthcare professional is presented with a few quick questions with their HER tablet to ask the patient.(The questions were developed based upon analysis of Big Data to identify by disease state, patient medical history, and geo-demo graphic data with a view to which patients are at risk for the inability to manage their disease and the propensity to change their behavior.)
- The answers to these questions combined with the previous analytic models provide a recommendation to the treating healthcare professional as to which disease management program level is appropriate for this patient.
- When the patient leaves the office, depending upon which disease management program they were enrolled, they receive follow-up in-home visits, phone calls, letters, email, text messages, and other patient support materials on a continuing basis.
- Data from sources, payer, provider and third-party, are utilized to initiate follow-ups if the patient does not complete lab tests, attend follow-up appointments, refill their prescriptions, and/or in-home technology that indicates the patient is not being adherent.
There are many program complexities to be addressed but with the three parties working together to leverage Big Data, the opportunity exists to make a disease management programs that delivers that are financially sustainable.
Disease management programs - A time to revisit
The Pharma industry landscape has changed dramatically over the last 5+ years. Consider the patent cliff underway; a weak R&D pipeline for blockbuster drugs; spiraling costs on multiple fronts including R&D and managed market rebates; increasing regulatory scrutiny across sales and marketing, and last but not least healthcare reform.
Healthcare reform is the item on this list which potential provides the greatest concern for the industry's current model of promoting and selling products. Our healthcare reform act provides for the development of national healthcare data repositories which will capture diagnosis data, medical procedure data, prescription claims data, and lab values. Additionally, this data will be available for healthcare entities to mine and analyze. If a healthcare provider is compensated based upon patient outcomes and has a large comprehensive dataset for analyze at its disposal, would the provider prefer to use analysis of the data to development treatment best practices for various patient segments or would the provider prefer to leave the treatment best practices to the decision of the healthcare professional and the influences of sales and marketing? At least in some instances already, the healthcare providers are already using their own data to drive shared best practices across their network. Additioally, these two healthcare providers of large networks that have eliminated access by pharma sales reps and all multi-channel marketing to their employee healthcare professionals.
If this approach was adopted widely, traditional sales and marketing to healthcare professionals by pharma would be marginalized. This is not likely to happen tomorrow but certainly possible if not likely in the future.
Pharma companies have a number of options to make themselves more relevant and an important part of the healthcare ecosystem.
One option would be to develop disease management programs, when implemented with its Branded drug, which would lead to such improved patient outcomes that even after the brand went generic healthcare professionals and payers would continue to prescribe and reimbursement the Brand.
At first pass, this sounds like a rather lofty goal. Pharma has developed countless disease management programs over the last decade. However, these programs have not delivered sufficient value as to make them sustainable nor drive brand sales post Brand loss of exclusivity. There are a multitude of articles reviewing the limited to non-existent financial success of disease management programs. Just one example of a review is linked below which initiated a host of activity in the disease management industry to substantiate the value of programs.(http://www.cbo.gov/ftpdocs/59xx/doc5909/10-13-DiseaseMngmnt.pdf)
To be clear, this "uber-disease management program", which I am suggesting, would not be a masked "adherence program" but a comprehensive provider and patient support program. For the next blog update, we will explore what the "uber-adherence" program might consist of and how various parties within the healthcare ecosystem would need to work together and the benefits for all parties. Topics to be covered include:
- patient assessment for program inclusion driven by analytics to identify which patients need support, which patient's behavior can be impacted, and tiered program support aligned with patient need
- required business partnerships between payer, providers, and Pharma as well as the potential value to each party
- data sharing requirements to enable real-time interaction with patients, providers, and plans
- market readiness/requirements to successfully execute, and
- "why" this time could be different for all parties..
Are your Account Managers equipped for the new healthcare?
I love watching college football! There is the ceremony and the rituals- both on and off the field. The mascots, a beautiful Fall day, and the pre-game jitters all lead to the big event. There is so much more to the game than just the 60 minutes on the field. The teams’ preparation, the pageantry, and the energy- it’s all contagious and gets your adrenaline pumping.
How similar is a Saturday game to an account executive’s contract negotiation with a payer? After all, don’t organizations often call their account executives the “quarterback”? In negotiations aren’t there also lengthy preparations, team coordination, customer research, and contingency plan development? Don’t preparations include coaching and strategy development utilizing information about the payer just as a football team watches the opponent’s previous game tapes, reviews scouting reports and adjusts the strategy accordingly?
What payer insights are you providing to your account managers? Previous negotiation history followed by a payer analysis, the competitive market and projected sales? How deep do the insights take them? Do they understand the copay ranges within each tier based upon claims data? Do they know the true impact of pharmacy utilization management controls? They know the contracting guidelines provided by the home office, but do they know the optimal considerations to maximize sales while minimizing considerations?
Picture yourself at the National Championship Bowl game. What if you knew your favorite college football team was entering the stadium with little knowledge of the opponent, without reviewing game tapes, devoid of scout insights while the opponent utilized detailed competitive intelligence, had practiced in the stadium, and was familiar with the referees? Would you leave the stadium fearful of the game outcome, or would you stay and cheer your team hoping talent and experience could triumph? More importantly, would you bet your salary on a favorable outcome?
How Can You Influence the Balance of Power in the new healthcare?
In the Cutting Edge Information (CEI) report, titled “Securing Market Access: Reimbursement, Payer Relationships & Healthcare Reform”, CEI claims “perfecting these 3 R’s is more critical than ever to meet the payers growing needs and to accelerate the reimbursement decision making process." I agree these three R’s are the cornerstone for negotiating products to the appropriate formulary position. And in order to lay those cornerstones, a company must understand and study the payer and the evolving market with new, more advanced insights. Healthcare reform and the economy have already caused large “lives shifts” for payers, with many more shifts to come. How will these shifts impact the payers? How will your brands be affected? What are you doing to help payers understand and project new patient populations and patient preferences? What insights are you offering payers to foster a positive relationship? Maybe it’s time to enhance the game playbook with renewed emphasis on the 3 R’s supported with more advanced analytics and insights; otherwise the cost of entry will continue to rise, the sales will remain unremarkable and the relationships will not evolve.
Innovation can get you noticed
Innovation can grab your attention- the length of attention varies by the audience, the level of interest, and the novelty. Innovation can occur on purpose, by accident or as a clever disguise. Life Science enterprises built their business on drug innovation- now it is time to use that innovative spirit in other parts of the organization- like Sales & Marketing and Disease Management. IF, life science organizations aspire to have ANY voice in the health care of the future, they need to apply innovation NOW!
Recently I was impressed twice by Life Science innovation, they are transforming a bit!
The first enterprise understands Customer Expectation and Experience in purchasing dynamics, at least as far as health care professionals and their external website are concerned. The second demonstrates Leadership, Values and Culture through creating a thought leader health care collaboration to conquer chronic diseases. I recognize my two examples are purely web offerings, so my impressions for this blog do not scratch the corporate culture surface- these organizations may only have innovative web designers, yet both of these “outward facing initiatives” would require internal efforts, negotiations and ultimate senior executive agreement. I also admit I do not have the ROI on these initiatives- I cannot speak to the success- but I can acknowledge their efforts and innovation.
#1-Servicing health care providers.
This company offers health care providers two portals to purchases vaccines directly with a call center and web ordering capabilities, not unusual, YET they take it one step further- they also offer other manufacturer's vaccines and a variety of physician office supplies. They understand what I call “the American grocery shopping concept”: get in, get all you need at a reasonable price, and get out”. Other manufacturers require providers to purchase their products from their own website only, following my “French shopping experience: pick up one item here, another at a competitor, and a third on the way home”. Most of us, especially those running a physician facility, want convenience- not specialty shopping from multiple sites.
#2-Forming a collaboration to improve health outcomes. This enterprise announced a new healthcare community to address the cost of chronic diseases. “The new online community exclusively for managed care executives and other professionals will be responsible for population-level health outcomes through a variety of online tools that enable participants to collaborate and share health intervention ideas in a closed community.” While this exists elsewhere, it is an attempt to be part of the discussion and resources, while providing the forum to help shape the solution.
In both examples I admire the efforts to take Life Sciences out of the center of the solution, and work to improve health care and delivery. I am not saying I believe they are doing this without thinking of the potential ROI on both- I wonder 18 months in the future how the market will classify these two examples? Will they be deemed successes or marginal attempts lost in the market noise?
Top Photo by:http://www.flickr.com/photos/centralasian/5577865674/ cc- Attribution
How big is 'big data' in healthcare?
In my last post I spoke to the problem of medical data overload as it relates to our innate human inability to store and process more than 7 variables (plus or minus 2) at any one time. This week, I’ll explore what’s driving the data overload and, in particular, how big is Big Data in Healthcare.
So, why all the recent hype about Big Data in Healthcare? Have hospitals suddenly started to treat significantly more patients or double the number of lab tests and EKGs? Have Health Plans recently dramatically increased the number of patients that they provide insurance coverage to? Has there been a sudden surge in the number of Pharmaceutical trials that are underway. Well, the simple answer is no, no and no. Then why do we keep hearing that Healthcare is emerging as the poster child for Big Data? The answer lies in the fact that Big Data isn’t simply about the volume, velocity and variety of the data in storage, it is also about the potential value of those data that already exist but are poorly coordinated and stored in widely disparate formats across industries that haven't typically shared data openly.
I’ll write more about the opportunities and challenges of Big Data next time.
Let’s start by addressing the first part of the question. How big is Big Data in Healthcare? This is deceptively simple question but not one that I could easily find any existing answers to. McKinsey's Global Institute group recently published a Big Data report in which they estimate that, in ten years or so, the total value of Big Data in Healthcare could be as much as $300B. However, I was still unable to find any estimates of the actual global size of healthcare data.
So, not to be deterred, I decided to take a crack at it myself and set out to gather some data that would provide me with a foundation for an estimate of the total size of all digital healthcare data stored in the world today. An ex-colleague of mine from a leading PACS vendor provided me with some very helpful input as did a CIO from a highly prestigious integrated delivery system in the Boston area.
With the disclaimer that all the following data should be taken with a liberal pinch of salt, here are my estimates. In these examples, data sizes are expressed as Exabytes, where one Exabyte = 1x1018 Bytes. If my assumptions sound way off base, my math is wrong and/or any of you have seen any other published estimates with associated sources and assumptions I would be interested to hear from you.
Based on data provided by the PACS vendor, I estimate the total amount of PACS data to be about 78 Exabytes (one Exabyte = 1x1018 Bytes) by using the following assumptions – the vendor apparently has copies of all its (global) customers PACS images in a central archive. That archive contains approximately 550 Billion images. At roughly 20MB per image, and with a claimed 15% global share of the PACS market, that would be a total of 550 Billion * 20MB / .15 = 73 Exabytes. I then added a near random 5 Exabytes to take account of redundant storage in local PACS stores and some regional and national image exchanges. My assumption is also that these PACS numbers contain not only traditional radiological data (Digital X-Ray, CT, MR and USS) but also other diagnostic image and some stored video based data and, if not, the estimate could probably be increased by an additional 30%. The estimate does not include data that are of a transient nature, such as full video from interventional procedures or real time bedside medical device data.
Now, this is where things get really tricky. My gut feel (which may be way off but also sounded reasonable to the two colleagues I mentioned above) is that PACS images currently account for at around 50% of the total medical data stored per patient, the rest includes all other stored data such as demographic data, laboratory results, nurse charted data, other diagnostic reports, physicians notes, scanned documents, claims data, trials data, current genomic stores etc. So my total estimate for the currently stored global healthcare data would be something in the region of 150 Exabytes.
Next, my growth rate estimates are based on the Boston CIO’s data. He sees a growth rate of approximately 100MB per patient per year for hospital generated data. Assuming that Boston is reasonably representative of the average medical data stored per citizen in the US, that would lead to an estimated growth rate for hospital data across the US of 30 Petabytes (100MB * 300 Million people). As this only covers hospital based data, I then multiplied by 4, to estimate the total data growth rate - taking into account of all US healthcare data stored outside of the hospital setting (physician offices, freestanding imaging centers, nursing home, claims data & pharmaceutical trials data) which I felt might be four times the amount of data held within a hospital setting.... and then multiplied that by a factor of 10, which I took to represent the broader world of new healthcare data being generated each year outside of the US. My sense is that the multiple of 10 is probably too small, and it might be quite reasonable to multiply by closer to 20. That leaves us with a growth rate in global healthcare data of between 1.2 and 2.4 Exabytes per year.
In summary, my estimates are that the global size of “Big Data” in Healthcare stands at roughly 150 Exabytes in 2011, increasing at a rate between 1.2 and 2.4 Exabytes per year.
What do you think? Can you come up with a more accurate number?
Healthcare has a problem: big data & the law of seven
In one of the most quoted papers on psychology ever written, George Armitage Miller claimed that he had been persecuted by an integer. That integer was the number 7. Over a period of seven years, he had found through his own research and through an extensive literature review that human beings were capable of processing seven (plus or minus two) variables at any one time. For those of you more technically inclined, if human beings were computers, then seven would define the number of variables that we could store in our working memory. In fact, more recent research has shown that the number might actually be closer to four, rather than seven.
Most of us experience this challenge every day. Studies show that, in this increasingly wired and information intense world, we are interrupted every three to five minutes during the course of a work day, multitasking between email notifications, text messages, cellphone calls, chat requests and reviewing web sites. We frequently end up getting distracted and temporarily forget what we were working on a few minutes before. We are all information junkies and digital connectivity has radically increased the flow of data to feed our habit - yet our brain processing capacity hasn't evolved significantly since the Stone Age. No surprise that many folks feel overwhelmed and less productive and effective than they feel they could be.
So, what does this have to do with healthcare? Actually, plenty! Physicians and other caregivers are pattern matchers by training. They rely on their extensive knowledge base, built painstakingly over many years of training, practice and ongoing education, to identify key signals in a patient’s presentation that will point to the most likely diagnosis, treatment options and ongoing day to day management.
Now consider this. Assuming, for the time being, that your doctor is an unaided human functioning in a way that is not chemically or digitally enhanced, then depending on the amount of sleep and associated daily stress levels, they are capable of weighing approximately seven variables at any one time when dealing with something pretty important – your health! Of course, that might not be so bad if the number of variables that your doctor had to try to juggle at any one time were typically less than seven. In truth, quite frequently it doesn’t even take seven variables to be able to safely determine the most appropriate next steps in weighing evaluation or treatment options. However, on occasion, and as will become increasingly frequent, there are many, many more variables to consider in managing the care of an individual patient as well as the care of large panels of patients.
Add some other emerging dynamics to this basic human limitation and we can see that we have a problem on our hands.
First, the doubling time of medical knowledge is currently estimated to be less than ten years and continues to decrease. Second, the volume of medical literature that a busy doctor needs to read to stay current keeps increasing and all too frequently different research reports present apparently contradictory evidence – frequently, no single treatment approach is suitable for all cases. Finally, as in all other aspects of our life, the amount of useful interpretable health data being collected has started to grow at an explosive rate.
When you consider these factors, combined with the pressure to maintain quality while seeing increasing volumes of patients, it’s easy to understand why many doctors today feel that there has to be a different and more satisfying way to deliver the high quality care that they believe is possible.
The truth is that doctors, like all humans, are excellent pattern matchers but highly fallible complex information processors. They are also not “perfectible” – we all make occasional mistakes, some more serious than others. We tend to overly simplify complex issues and our recent personal experience weighs more heavily than it should when good evidence exists to the contrary.
The bottom line is that healthcare has to change to accommodate these dynamics. What does the future look like? Over the next series of blog posts, I’ll examine how big “big data” in healthcare actually is and how fast it’s growing, its implications to the medical profession and to healthcare more broadly and some predictions about how health analytics are poised to help physicians deal with this dilemma in the same way that smart GPS systems delight even the most experienced motorist.
No cash for clunkers in healthcare
I hate car shopping but
every few years I go through the process of buying one. Regardless of my dislike of buying a car I do not spend time and research. I evaluate my choices based on many dimensions – time, value, price, outcome, and customer service experience. I can easily narrow down my choices using the internet and or calling car dealerships to help with my decision. I determine the value based on my personal expectations along with market variables. There is a baseline expectation and I can add things that I value and am willing to pay for.
I seek health care services more often than I buy cars. I do as much research as I can (some not statistically significant as it is word of mouth). In absence of a price transparency and other quality metrics there is not a whole lot of information for someone as analytically inclined as me to easily make these decisions. Each and every one of us would like some metric that will allow us seek health care that has the best value.
So how do we go about defining and then measuring “value”? Historically quality of health care has been measured by decrease in readmissions, complications, number of days in the hospital, or clinical outcomes. These should help set the baseline expectations. I would not buy a car without working breaks, so why would I go to a facility for a procedure where steps are not taken to reduce readmissions or complications?
Most of us have three basic expectations of health care – access, cost, and quality. The first two are relatively simple to define, measure and compare (if there is available information and transparency) – how far one drives to get care and how much it costs.
Defining healthcare quality on the other hand is not easy, let alone measuring it. We as consumers of health care tend to equate quality with affordable access and often believe more is better than less in terms of care. We value our individual outcome, service we get from medical professionals, autonomy in choosing provider, and other bells and whistles like the waiting room, food, or décor. At a population level health outcomes are somewhat more easily defined – life expectancy rates, disease eradication, impact of immunization, or mortality rate.
In the near future there is going to be considerable focus on defining and measuring health care quality. Once defined the medical expense payment and utilization management will have to be aligned as we all know “you get the behavior that you pay for”.
Our individual inputs to measure quality may be different but we need a common definition of measured outcome. As we all bring many individual variables into our care (medical compliance, genetics, mental, physical, social, economic, and behavioral characteristics to name a few) measuring the value of heath care will need to utilize advanced analytics. The health care roadmap we personally undertake and the care we seek and receive together will help define our health outcome.
I truly hope that we have no lemons in health care, as they would be measured in caskets, and I sure hope we don’t have a health care version of “Cash for Clunkers,” because I don’t need my kids trading me in!
Top photo by http://www.flickr.com/photos/msvg/4007509222/ CC-Attribution



