Blog: Lou Agosta Subscribe to this blog's RSS feed!

Lou Agosta

Greetings and welcome to my blog focusing on reengineering healthcare using information technology. The commitment is to provide an engaging mixture of brainstorming, blue sky speculation and business intelligence vision with real world experiences – including those reported by you, the reader-participant – about what works and what doesn't in using healthcare information technology (HIT) to optimize consumer, provider and payer processes in healthcare. Keeping in mind that sometimes a scalpel, not a hammer, is the tool of choice, the approach is to be a stand for new possibilities in the face of entrenched mediocrity, to do so without tilting windmills and to follow the line of least resistance to getting the job done – a healthcare system that works for us all. So let me invite you to HIT me with your best shot at LAgosta@acm.org.

About the author >

Lou Agosta is an independent industry analyst, specializing in data warehousing, data mining and data quality. A former industry analyst at Giga Information Group, Agosta has published extensively on industry trends in data warehousing, business and information technology. He is currently focusing on the challenge of transforming America’s healthcare system using information technology (HIT). He can be reached at LAgosta@acm.org.

Editor's Note: More articles, resources, and events are available in Lou's BeyeNETWORK Expert Channel. Be sure to visit today!

Recently in comparative effectiveness research Category

The answer is clinical data warehousing, decision support, and analytics. What's the question? Wellpoint (one of the leading Blue Cross branded health insurance companies) is reportedly contracting to use IBM's computing grand challenge system nicknamed "Watson" (after IBM's founder) to address a list of clinical issues in medical diagnosis, treatment, and (potentially) cost. In the spirit of Jeopardy!, the question is will it advance in the direction of enabling comparative effectiveness research (CER) and pay for performance (P4P) while enhancing the quality of medical outcomes? Healthcare consumers tend to get a tad nervous when they suspect that insurance companies are going to deploy a new computer system as part of the physician payment approval process, nor (let us be clear) has anyone actually said that will happen in this case.

The diagnosis of a disease is part science, part intuition and artistry. The medical model trains doctors and healthcare specialists using an apprentice system (in addition, of course, to long schooling and lab work). The hierarchical nature of disease diagnosis has long invited automation using computers and databases. Early expert medical systems such as MYCIN at Stanford or CADUCEUS at Carnegie-Mellon University were initially modest sized arrays of if-then rules or semantic networks that grew explosively in resource consumption, time-to-manage, and cost and complexity of usability. They were compared in terms of accuracy and speed with the results generated by real world physicians. The matter of accountability and error was left to be worked out later. Early results were such that automated diagnoses was as much work, slower, and not significantly better - though the automation would occasionally be surprisingly "out of the box" with something no one else had imagined. One lessons learned? Any computer system is better managed and deployed like an automated co-pilot rather than a primary locus of decision making or responsibility.

Work has been ongoing at universities and research labs over the decades and new results are starting to emerge based on orders of magnitude improvements in computing power, reduced storage costs, ease of administration, and usability enhancements. The case in point is IBM's Watson, which has been programmed to handle significant aspects of natural language processing, play jeopardy (it beat the humans), and, as they say in the corporate world, other duties as assigned.

Watson generates and prunes back hypotheses in a way that simulates what human beings do in formulating a differential diagnoses. However, the computer system does so in an explicit, verbose, and even clunky way using massive parallel processing whereas the human expert distills the result out of experience, years of training, and unconscious pattern matching. Watson requires about eight refrigerator size cabinets for its hardware. The human brain still occupies a space about the size of a shoe box.

Still, the accomplishment is substantial. An initial application being considered is having Watson scan the vast medical literature on treatments and procedures to match evidence-based outcomes to individual persons or cohorts with the disease in question. This is where Waton's strengths in natural language processing, formulating hypotheses, and pruning them back based on confidence level calculations - the same strengths that enabled it to win at Jeopardy - come into play. In addition, oncology is a key initial target area because of the complexity of the underlying disorder as well as the sheer number of individual variables. Be ready for some surprises as Watson percolates up innovative approaches to treatment that are expensive and do not necessarily satisfy anyone's cost containment algorithm. Meanwhile, there are literally a million new medical articles published each year, though only a tiny fraction of them are relevant to any particular case. M.D.s are human beings and have been unable to "know everything" there is to know about a specialty for at least thirty years. In short,  Watson just could be the optimal technology for finding that elusive needle in a haystack - and doing so cost effectively.

A medical differential diagnosis in medicine is a set of hypotheses that subsequently have to be first exploded, pruned, and finally combined based on confidence and prior probability to yield an answer. This corresponds to the so-called Deep Question and Answering Architecture implemented in Watson. Within five years, similar technologies will have been licensed and migrated to clinical decision support systems from standard EMR/EHR vendors.

While your clinical data warehouse may not be running 3,000 Power 750 cores and terabytes of self-contained data in a physical footprint about the size of eight refrigerators, some key lessons learned are available even for a modest implementation of clinical data warehousing decision support:

  • Position the clinical data warehouse as a physician's assistant (think: co-pilot) to answer questions, provide a "sanity check," and fill in the gaps created by explosively growing treatments.
  • Plan on significant data preparation (and attention to data quality) to get data down to the level of granularity required to make a differential diagnoses. ICD-10 (currently mandated for 10/2013 but likely to slip), will help a lot, but may still have gaps.
  • Plan on significant data preparation (and more attention to data quality) to get data down to the level of granularity required to make a meaningful financial decision about the effectiveness of a given treatment or procedure. Pricing and cost data is dynamic, changing over time. New treatments start out expensive and become less costly. Time series pricing data will be critical path. ICD-10 (currently mandated for 10/2013 but likely to slip) will help but will need to be augmented significantly into new pricing data structures and even then but may still have gaps.
  • Often there is no one right answer in medicine - it is called a "differential diagnosis" - prefer systems that show the differential (few of them today do, though reportedly Watson can be so configured) and trace the logic at a high level for medical review.
  • Continue to lobby for tort and liability reform as computers are made part of the health care team, even in an assistant role. Legal issues may delay, but will not stop implementation in the service of better quality care.
  • Look to natural language interfaces to make the computing system a part of the health care team, but be prepared to work with a print out to a screen till then.
  • Advanced clinical decision support, rare in the market at this time, is like a resident in psychiatry, in that it learns from its right and wrong answers using machine learning technologies as well as "hard coded" answers from a database of semantic network.
  • This will take "before Google (BG)" and "after Google (AG)" in medical training to a new level. Watson-like systems will be available on a smart phone or tablet to residents and attendings at the bedside.

Finally, for the curious, the cost of the hardware and customized software for some 3,000 Power 750 cores (commercially available "off the shelf"), terabytes of data and including time and effort of a development team of some 25 people with Ph.D.s working for four years (the later being the real expense), my back of the envelope pricing (after all this is a blog post!) weighs in at least in the ball park of $100 million. This is probably low, but I am embarrassed to price it higher. This does not include the cost of preparing the videos and marketing. One final thought. The four year development time of this project is about the length of time to train a psychiatrist in a standard residency program.

Bibliography

  1. "Wellpoint's New Hire. What is Watson?" The Wall Street Journal. September 13, 2011. http://online.wsj.com/article_email/SB10001424053111903532804576564600781798420-lMyQjAxMTAxMDEwMzExNDMyWj.html?mod=wsj_share_email

  2. IBM: "The Science Behind and Answer": http://www-03.ibm.com/innovation/us/watson/

 


Posted September 14, 2011 4:06 PM
Permalink | No Comments |

There are so many challenges that it is hard to know where to begin. For those providers (hospitals and large physician practices) that have already attained a basic degree of automation there is an obvious next step - performance improvement. For example, if an enterprise is operating eClinic Works (ECW) or similar run-your-provider EHR system, then it makes sense to take the next step and get one's hand on the actual levers and dials
that drive revenues and costs.

Hospitals (and physician practices) often do not understand their actual costs, so they are struggling to control and reduce the costs of providing care. They are unable to say with assurance what services are the most profitable, so they are unable to concentrate on increasing market share in those services. Often times when the billing system drives provider performance management, the data, which is adequate for collecting payments, is totally unsatisfactory for improving the cost-effective delivery of clinical services. If the billing system codes the admitting doctor as responsible for the revenue, and it is the attending physician or some other doctor who performs the surgery, then accurately tracking costs will be a hopeless data mess. The amount of revenue collected by the hospital may indeed be accurate overall; but the medical, clinical side of the house will have no idea how to manage the process or improve the actual delivery of medical procedures.

Thumbnail image for Thumbnail image for riverlogicjpg.JPG

Into this dynamic, enters River Logic's Integrated Delivery System (IDS) Planner (www.riverlogic.com). The really innovative thing about the offering is that it models the causal relationship between activities,
resources, costs, revenues, and profits in the healthcare context. It takes what-if analyses to new levels, using its custom algorithms in the theory of constraints, delivering forecasts and analyses that show how to improve performance (i.e., revenue, as well as other key outcomes such as quality) based on the trade-offs between relevant system constraints. For example, at one hospital, the operating room was showing up as a constraint, limiting procedures and related revenues; however, careful examination of the data showed that the operating room was not being utilized between 1 PM and 3 PM. The  way to bust through this constraint was to charge less for the facility, thereby incenting physicians to use it at what was for them not an optimal time in comparison with golf or late lunches or siesta time. Of course, this is just an over-simplified tip of the iceberg.

 

IDS Planner enables physician-centric coordination, where costs, resources, and activities are tracked and assessed in terms of the workflow of the entire, integrated system. This creates a context of physician decision-making and its relationship to costs and revenues. Doctors appreciate the requirement to control costs, consistent with sustaining and improving quality, and they are eager to do so when shown the facts. When properly configured and implemented, IDS Planner delivers the facts. According to River Logic, this enabled the Institute for Musculosketal Health and Wellness at the Greenville Hospital System to improve profit  by more than $10M a year by identifying operational discrepancies, increase physician-generated revenue over $1,700 a month, and reduce accounts receivable by 62 down to 44 days (and still falling), which represents the top 1% of the industry.  Full disclosure: this success was made possible through a template approach with some upfront services that integrated the software with the upstream EHR system, solved rampant data quality issues, and obtained physician "buy in" by showing this constituency that the effort was win-win.

The underlying technology for IDS Planner is based on the Microsoft SQL Server (2008) database and Share Point for web-enabled information delivery.

In my opinion, there is no tool on the market today that does exactly what IDS Planner does in the areas of optimizing provider performance.River Logic's IDS Planner has marched ahead of the competition, including successfully getting the word out about its capabilities. The obvious question is for how long? The evidence is that this is a growth area based on the real and urgent needs of hospitals and large provider practices. There is no market unless there is competition; and an overview of the market indicates offerings
such as Mediware's InSight (www.mediware.com/InSight), Dimensional Insight (www.dimins.com) with a suite of the same name, Vantage Point HIS  (www.vantagepointinc.com) once again with a product of the same name. It is easy to predict that sleeping giants such as Cognos (IBM) and Business Objects (SAP) and Hyperion (Oracle) are about to reposition the existing performance management capabilities of these products in the direction of healthcare providers. Microsoft is participating, though mostly from a data integration perspective (but that is another story), with its Amalga Life Science offering with a ProClarity frontend. It is a buyer talking point whether and how these offerings are able to furnish useable software algorithms that implement a robust approach to identifying and busting through performance constraints. In every case, all the usual disclaimers apply. Software is a proven method of improving productivity, but only if properly deployed and integrated into the enterprise so that professionals can work smarter. Finally, given market dynamics in this anemic economic recovery, for those end-user enterprises with budget, it is a buyer's market. Drive a hard bargain. Many sellers are hungry for it and are willing to go the extra mile in terms of extra training, services, or payment terms.



Posted April 5, 2010 11:33 AM
Permalink | No Comments |

In the healthcare IT (HIT) market, 'meaningful use' is the term of art used by the HIT Policy Committee (of the federal government) to qualify doctors and hospitals for reimbursement under the HITECH portion of the American Recovery and Reinvestment Act (ARRA). While the definition of 'meaningful use' is a work in progress, the broad outlines are starting to emerge. While a few grants have been 'let', so far dollars have been as scarce as cats in the swimming pool. Things are expected to pick up as the definition is clarified and actually improving the efficiencies of the healthcare system become an even more urgent priority. It is relatively safe to say:

  • Meaningful use is a data integration challenge. Clinical data such as hypertension, diabetes, smoking cessation, recommended tests (mammography, coloectal screening, and so on) have to be cross-referenced with demographics, eligibility for insurance, electronic healthcare records (EHR) in order to compare the effectiveness of treatments and procedures.
  • Comparative effectiveness research (CER) is a data integration challenge. This takes the 'meaningful use' challenge up a level. In order to assess the effectiveness of procedures, treatments, tests, the program has to access both the outcome of the procedure (did it work?) as well as financial data about its cost(s). Cost drivers include the time and effort of healthcare providers, the price of powerful drug therapies (an ongoing area of innovation), and what the payers agree to reimburse. This in turn results in the proposal to provide financial incentives to healthcare providers for improving quality ('performance').
  • Pay for performance is a data integration challenge. Like CER, this takes 'meaningful' use to the next level - providing a structure and incentives in terms of payments to healthcare providers (hospitals and doctors) for 'hitting their numbers'. The definition and production of those numbers is and promises to continue to be obvious in some cases and controversial in others, especially new and emerging treatments and technologies. However, in almost every case, clinical outcomes have to be lined up at a low level of granularity with what the cost is determined to be.

Of course, the healthcare is not a closed system or a completely rationalized one. Note that I say 'rationalized', not 'nationalized' (the latter is a story for another post). Medicare and Medicare payments continue to be the 2-ton elephant; and if Medicare does not pay, then how can a treatment be assessed as 'effective' or impacting quality? Obviously, there is a defined process for including a procedure or drug on the list of payment eligibility, including an act of Congress (I am not making this up), so there are many issues. For example, coordination of care is neglected and under-reimbursed (if paid at all) - where doctors are reimbursed to work together to care for complex illnesses of aging or life-style (not the same thing) such as diabetes, congestive heart failure, and kidney failure. Most of these disease entities require data integration of a life-time of healthcare treatments and transactions - like a 360% view of the client in customer relationship management (CRM).

Thus, as in most areas of the economy and across multiple vertical markets, data integration vendors who are engaging healthcare clients and applications are trying to hit a moving target. IT systems and infrastructure continue to develop in good times and in less good times. The standard relational databases are clean and effective data sources for the storage and manipulation of business and financial data in payment and run-your-healthcare-operation. But on the clinical side, heterogeneous data is ever more heterogeneous and even more inaccessible in proprietary systems such as Cerner, Eclipsys, GE Centricity, McKesson, and a whole host of other software providers. Even MedSphere which boasts about being 'open source' operates with the Mumps data store, not the target for development of new features and functions across vertical industries. I am not saying that Mumps is not 'open', but it does put the definition in context. Naturally, all data is accessible by definition in some form if you need it badly enough; but it might be a relatively inefficient dump to a batch file and clumsy handoff between heterogeneous systems, absent additional automation..

In data integration, connectors and adapters (plug-and-play type components to enable grabbing and transforming data sources into target patterns) are on the critical path to success. As in many markets, significant consolidation has occurred among data integration vendors as they have marched towards building platforms that combine data profiling, data quality, with data transformation and integration. Informatica is still touting its cherished independence as the proper database-neutral role to integrate all comers after Ascential with its famous DataStage technology joined forces with IBM in 2005 to provide the foundation of what is now IBM's InfoSphere data warehousing platform. Oracle has its own suite of tools, which continue to be a good choice for Oracle customers, including those considering Exadata; but Oracle has been slow to break out of the Oracle-to-Oracle niche (albeit a very large 'niche').

Pervasive software is a perhaps lesser known firm with offerings in data integration, service oriented architecture (SOA), and application development. Pervasive Software has contributed steadily to the development of innovative data integration technology for some twelve years, much of that as a publicly traded and scrutinized entity. Pervasive plays across multiple vertical markets from finance to retail, from manufacturing to insurance, from telecommunications to healthcare. The latter (healthcare) has been the target of this discussion. In 2003, Pervasive Software took a lesson from the play book of such Big Guys as HP, IBM, and Oracle - namely, innovation can sometimes be bought in the market easier than it can be developed in-house - and it bought some. Pervasive acquired Data Junction Corporation, makers of the Data Junction ETL (extract, transform, load) technology. This suite of data and application integration tools was rebranded and brought forward with enhancements and now known as Pervasive Data Integrator. No doubt "geographic determinism" played a role in the acquisition - both firms were located in Austin, TX. Pervasive continues to develop and market its high performance, flagship Btrieve database, PSQL This is more than passing interest from a technology perspective, since another B-tree database, Mumps, is quite common in the healthcare IT applications and implementations. Whether this will give Pervasive additional access is an open question, but it will surely give them additional insight into the technical dynamics and challenges of data integration and downstream applications such as business intelligence, pay-for-performance, and comparative effectiveness research, all of which are critical path in healthcare reform. I examined the technology at the time, and Data Junction brought to the market adapters for many relatively obscure data sources in small niches at a reasonable price point as well as all the standard relational databases and corporate data sources. Fast forward some six years, and Pervasive has built on the franchise, earning a spot on the short list of enterprises confronting information reconciliation, consolidation, and rationalization challenges. [This just in...update (12/16/2009): The Pervasive database team noted that 'Btrieve' is a registered trademark for Pervasive and actually based on B-tree technology. So 'Btrieve' would only refer to the Pervasive product whereas MUMPS is a 'B-tree' (rather than 'Btrieve') implemented database. Good catch! ]


Posted December 7, 2009 10:54 AM
Permalink | No Comments |