Some see the primary problems with health care in the United States as issues of affordability or universal access; others see the problems more in terms of the system's ability to reliably deliver an effective result, to rapidly and effectively use new therapies or technologies, or to contain future growth in costs. Under each of these desirable outcomes, for true reforms to be be acceptable for all stakeholders and sustainable over the long term, they must ultimately confront health economics, since all questions ultimately come down to determinations of who is going to pay for care, what care will be covered, and how much will be reimbursed.

As a benchmark, the US already is spending more per capita on health care than nearly all other countries, with the exception of Norway and Iceland. Yet there is a sense of urgency for reforms, since in the absence of properly focused and cost-effective changes, health care spending is expected to grow even more, to 19.6 percent of our gross domestic product (GDP) by 2016.  While many different approaches to reforms have been initiated across the US, by both private and public concerns, the results of these efforts have often been counter-intuitive, as systems theory dictates. Our risk exposure to the unintended effects of reforms may thus be unacceptably high; attempted improvements for one outcome could easily further exacerbate problems with other outcomes, rather than making things better overall. When this fact is combined with the reality that well intentioned reform efforts often fail to deliver on promises, and the net result is understandable concern about revolutionary, rather than evolutionary change. The overwhelming majority of the population would simply be unwilling to trade improved access for others if it meant lower quality or higher costs for themselves.

While there is broad (though unfortunately, not deep) public debate over the best strategy to contain medical costs, most successful cost-containment efforts in other business settings ultimately must address quality issues. Such issues, and the wasteful and expensive practices that produce them, represent the greatest source of inefficiencies in most systems. And in our current health care system, there seems to be a broad consensus about one thing - the quality issues with our current system are a mess:

• In a disturbing 1991 study, Michael Millenson reports that 110 nurses of varying experience levels took a written test of their ability to calculate medication doses. Eight out of 10 made calculation mistakes at least 10% of the time, while four out of 10 made mistakes 30% of the time (Demanding Medical Excellence: Doctors and Accountability in the Information Age)
• The Wall Street Journal has reported that there are 1,000,000 "serious medication errors per year" ... "illegible handwriting, misplaced decimal points, and missed drug interactions and allergies."
• In 1998 alone, the CDC reported 90,000 killed and 2,000,000 injured from hospital-caused drug errors & infections.
• The Independent reported in a study in 2006 that in 53 autopsy studies, there was an overall 24% misdiagnosis rate.
• In another article last year ("What Doctors Hate About Hospitals", Time), the cover article states that "It remains almost a stroke of luck to enter a U.S. hospital and receive precisely the right treatment." and "No day passed-not one-without a medication error. The errors were not rare; they were the norm."
• In a cover article in BusinessWeek last year, Dr David Eddy, of the Kaiser Permanente Care Management Institute, said "The problem is we do not know what we are doing." He goes on to state that only 15% of what doctors do is "backed by hard evidence"
• According to a recent report from the National Committee for Quality Assurance,  a thousand Americans die each week across the US because the care they get is not consistent with the care that medical science tells us they should get.

Despite these problems, health care has a progressive and dramatic record of evolutionary improvements in the effectiveness of individual treatments for specific diseases, each based upon disciplined applications of the scientific method and medical research. Recently, a strategy of using evidence-based medicine techniques has been widely adopted, and this approach has further driven efforts for standardization of measures in support of data analytics. This strategy is now beginning to be mirrored  in other industries (see Evidence-based Management).

Additional pressures are also now being brought to bear on health care providers by patients who are far better informed than they have ever been in the past, through use of internet-based aggregators of bioinformatics databases, and sites such as Curehunter and WebMD. The empowerments of health care consumers, through such information sharing, will only increase pressure on providers to further improve transparency and deliver higher-quality solutions. But providing such solutions will require confronting, and correctly deciding, answers to many difficult questions:

• Is the health care system underused or overused by its customers?
• Is our United States system really worse than other countries, or is it in fact better (and by which measures)?
• Is the system under-regulated or over-regulated?
• Is a more distributed system the answer, with strategic use of market forces, or is centralized management indeed the real way to controlling costs?
• Can (or should) one size 'fit all'?
• Who should lead improvements, and at what pace?
• Should solutions be scaled towards what we can afford, in order to provide equal access for all, or is cost an independent variable?
• Are we already smart enough to 'fix' the system, and is it wise to attempt to do it in one 'big fix'?
• How should runaway costs and quality be managed?
• How much freedom should patients be granted? 
• What priorities are most important in rolling out changes (fix quality problems, or provide universal access)?
• How should costs and benefits be balanced across stakeholders, and what's the 'menu' of choices that would entail?
• How should incentives and competition be utilized?
• How will new technologies and therapies be leveraged in a balanced way that will deliver the most affordable and effective value?

If we were to start from scratch, we would not design the system we have today. Yet we cannot start from a clean sheet of paper, since everyone must continue to use the system as it is improved. Determining the proper choices for any one of these questions, in the complex system that we all interact with, is thus very  difficult. Further, proposed solutions are often driven as much by the political power of stakeholders as by rigorous science and interventions designed to  optimize a particular set of outcomes.

Systems theory dictates that the underlying drivers that can enable  answers to questions such as the above set are often highly interrelated. Further, since such systems are dynamic rather than static in operation, the challenges for implementing useful reforms are even more pronounced. Candidate solutions may only be appropriate for a snapshot in time, or may be sub-optimal when considered in the context of the entire system. The risks of introducing harm may thus be just as great as the opportunities to do good. Further, there is an additional dilemma due to the pace of technology changes, which easily exceed the rate of absorbing these changes into wide use, especially given the complexity of the system itself.

Models and simulations of these system dynamics have been attempted to seek insights about how these interactions may play out over time, but the fidelity of such models with respect to what really is happening over time, or would happen under reforms, cannot be validated. Insights gained from such models, while interesting, have to be carefully considered before using them as a basis for interventions. Conventional wisdom thus dictates that the right thing to do is to conduct cautious, limited interventions and experiments, and exploit learning and leverage insights that are gained along the way. Of course, this all takes time.