Enabling Healthy Decisions

Health Care Reform Now! A Prescription For Change is the latest book by George Halvorson (CEO of Kaiser Permanente). I have been talking about it and using quotes from it for a few months. I finished the book a few weeks ago and figured that I better carve out the time to capture my thoughts now.

First, if you are looking for a great book on why healthcare is a big issue in this election, you don’t have to look any further. As someone running one of the biggest healthcare entities in the US, George clearly knows what he is talking about and speaks from a position of authority. I know that he has talked with all of the candidates about their policies.

If you are in healthcare and trying to be a catalyst for change, you have to read the book. It is pointed, opinionated, and supported with lots of facts and examples. If it doesn’t make you want to change what we have, I would be shocked. Some of the examples of mis-alignment are scary.

Some of the facts he shares:

• Family health insurance rates in CA already exceed the per capita income of 147 countries.
• General Motors now spends more money on healthcare then on steel.
• Nearly 50% of the time, patients in the US are receiving less than adequate, inconsistent, and too often, unsafe care.
• Healthcare costs are unevenly distributed in America.
  • 1% of the population uses 35% of the healthcare dollars
  • 5% uses 60%
• Care linkage deficiencies abound – and can impair or cripple care delivery.
• Economic incentives significantly influence healthcare.
• Systems thinking isn’t usually on the healthcare radar screen.
• Most of our costs are for chronic diseases – primarily diabetes, congestive heart failure, coronary artery disease, asthma, and depression.
• Prevention is a lot less expensive than addressing these chronic diseases at their late stages.
• The US ranks 35^th in the world in infant mortality.
• We could cut the complications of diabetes by 90% with best care and involved patients.
  • We could cut second heart attacks by 40%.
  • We could cut school and work days lost because of asthma by 90%.
• Incentives work…yet while we have 9,000 billing codes for procedures and services not one of them is for curing someone or improving someone’s health.
• There is up to a 60% difference in the 5-year mortality rate for breast cancer patients, depending on which hospital’s surgery team did the surgery.
• 1 in 10 doctors use electronic medical records (EMR) and only 5% of hospitals use computerized physician order entry (CPOE). This means our history exists mostly in paper files with no standards.
• Almost 50 developing nations have higher immunization rates for preventable childhood diseases than the US.
• The Institute of Medicine showed that it takes “seventeen years before a proven new technique becomes the standard of care in a given medical specialty.”
• There were 2,000 published clinical trials in 1985 and 30,000 published in 2005. (Can your provider really keep up without an electronic system?)
• Diabetes is the number one cause of new blindness (90% preventable) and foot and leg amputations (85% preventable). It is the number one co-morbidity associated with death from heart failure.
• Asthma causes – 2M emergency room visits, 500,000 hospital stays, 5,000 deaths, and 14M lost school and work days per year.
• The vast majority of asthma attacks can be prevented.
• If Americans were 5-10% thinner and walked just 30 minutes per day, the incidence of Type 2 diabetes could be cut by more than half. (Culture and incentives matter)
• We spend $250,000 every minute on heart disease.
• More than 15M Americans have depression…and on average, people with depression have 3 other chronic diseases.
• A 10% reduction in spending for the top 0.5% of patients would create enough savings to fund universal coverage for the uninsured.
• The most expensive acute conditions are cancer, maternity, and trauma care. (Acute conditions account for 30% of the health care spend.)
• The median life expectancy across the 117 cystic fibrosis centers is 33, but it is 47 at the highest performing center. (This seems embarrassing that there could be such a difference here.)
• US employers pay an average of $6,600 Per Employee Per Year compared to $600 in Canada.
• 4% of people believe they have insurance…but they don’t. (Who are these people?)
• Government pays 44% of the healthcare bill today; employers 26%; and individuals 30%.

Key Point – I think everyone wishes that we could address the uninsured and underinsured issue here in the US. It is ridiculous. But, I think most people feel it would further complicate the economy and be a downward drag. George presents a good case that today’s model simply cost shifts so that we are paying for care but paying at the high cost of emergency care not preventative care for those people. In the book, they say that this cost represents $922 per employee today in what is paid. Someone has to pay the providers for these real costs that they incur and can’t recoup. We could cover the costs of the uninsured without any real increases in costs.

Some of my favorite quotes:

• “We don’t really have a health care delivery system in this country. We have an expensive plethora of uncoordinated, unlinked, economically segregated, operationally limited Microsystems, each performing in ways that too often create suboptimal performance both for the overall health care infrastructure and for individual patients.” (introduction)
• “Performance reporting that actually exists about either processes or outcomes is almost always regarded in the current culture of American health care as an onerous, externally imposed burden, extraneous and irrelevant to the actual business and profession of care delivery.” (pg. 23)
• “I do not want ‘rules-based’ medicine. I do want accountable care.” (pg. 29)
• “Process reengineering will not happen on any scale in health care until there is a financial reward for doing just that.” (pg. 33)
• From the book Escape Fire: Designs for the Future of Health Care by Don Berwick – “A patient with anything but the simplest needs is traversing a very complicated system across many handoffs and locations and players. And as the machine gets more complicated, there are more ways it can break.” (pg 86)
• “We need highly credible doctors, nurses, and health educators talking to patients in targeted and effective ways to help people make the lifestyle changes necessary to avoid diabetes.” (pg 117)
• “Health care can be improved. The challenge is to do it consistently and systematically, not incidentally and haphazardly.” (pg 122)
• “Improving care by 50 percent for diabetics is wonderful, but not as wonderful as reducing the number of diabetics by 50 percent by preventing the disease.” (pg 206)

Comments:

• He talks about studying the international models and that none of them are the same. They have all been individually developed to fit the culture and needs of the country.
• He talks about creating a “patient-centered American health care marketplace”.
• He is careful about not just pushing the Kaiser model of vertical integration. He focuses on virtual integration which is more achievable.
• More care is not better care.
• He gives several examples of how following best practices for evidence based medicine improved outcomes but reduced revenues for the providers which is a hard model to sell.
• He compares HEDIS scores (which measure how often health plans offer care that complies with best practices) with Six Sigma:
  • Average performance for screening for colorectal cancer is 49% (or 1.5 sigma).
  • Recommended treatment of acute depression is 61.6% (average) and 70.8% (90^th percentile) which are 1.8 and 2.1 sigma performance.
  • Note: 2-sigma performance means 308,000 cases of non-compliance per million patients…6-sigma means only 3.4 cases per million.
• He talks about the fact that 5% of patients experience an adverse drug event. I think the PBM industry has consolidated a lot of data to minimize this, but I am surprised more people don’t talk about samples here. Although they are supposed to track samples, I bet most physicians don’t record them in the chart and they certainly aren’t electronically managed to look for potential drug-drug interactions. (In my opinion, there is still opportunity for improvement, but it is at the pharmacy level not the provider level.)
• He proactively addresses one major excuse about controlling patient behavior. Yes…we can’t control the patients, but we can make sure that the right events happen to align them for success.
• I like his suggestion that a personal health record could be a more logical first-step than a full blown EMR solution due to costs and ability to execute.
  • “That personal health record data set for each patient should show all care received by that patient, all prescriptions paid for, all tests given, all diagnosis made, and all providers who delivered care to each person as a patient. The information should be in an easy-to-use format and available to each patient on demand, either electronically or on paper.”
• He provides a good, quick comparison of PHR and EMR:
  • EMR has the exact Rx dosage and level. PHR may just have the name of the drug.
  • EMR will have the x-rays and scans. PHR will just say the date the test was done.
  • EMR will have notes from physician visit. PHR will just know the patient visited.
• Preventing a CHF (congestive heart failure) crisis might only generate $200 in billable revenue while treating a crisis creates $10,000 – $20,000 in revenue. (And, we really wonder why people aren’t acting preventatively.)
• Preventative care makes me think of two examples:
  • People have to want to be healthy and manage their risk. I know numerous people who are told to be on bed rest when they’re pregnant that don’t listen to their physicians.
  • People have to know there is not a risk of discrimination. I know a friend with MS who didn’t go see a doctor for several years until she had found a job with good health insurance.
• He talks a little about it, but I think the issue of helping patients evaluate trade-offs is a big one. Enabling them with information is important, but how do we help them compare two treatments based on both outcomes and the experience (i.e., pain, functionality). Is it always better to simply live longer even if you have limited functionality and are always in pain?
• He talks about plan design with some very good insight:
  • Deductibles only work if the unit of care being purchased is less than the deductible.
  • Deductibles tend to discourage chronic patients from getting preventative and maintenance care.
  • Percentage copays only work on big dollar differences. Otherwise, paying 10% more of a drug or office visit that costs $20 more is only $2.
• In talking about plan design, he talks about something that in pharmacy is referred to as Therapeutic MAC. (MAC = maximum allowable cost) This allows patients access to any drug, but the plan only pays for the lowest cost drug which produces equal outcomes. Therefore, a patient might get the first $70 of any office visit covered, and they pay the difference. Then they care about where and when they go to the doctor.
• For all the talk about price transparency and driving decisions, he makes a great point that this is thrown out the window at times. For example, when you are having a heart attack, you don’t have time to research your options and make tradeoffs.
• Kaiser saw first-hand what happens after seniors pass a cap on prescription coverage (pg 137):
  • 18% started skipping doses of medication
  • 9% increase in ER visits
  • 13% increase in hospital admissions
  • 22% increase in mortality
• He talks about 8 developments that have made health care reform possible:
  • Common provider number
  • Computerized databases
  • Electronic claims data portability
  • Government transparency about payment data
  • Universal awareness of the quality issues
  • Buyers are ready for change
  • Internet functionality used for care
  • Lawmakers are ready for reform
• He talks about blending virtual care and live care with a technology infrastructure which I think makes a lot of sense. I wonder how we change physicians to be more comfortable with the “DIY” (Do It Yourself) patient that comes in with lots of information and suggestions from other caregivers or even getting “second-guessed” by the rules engine of the EMR.
• He talks about health care needing a Target, Best Buy, or Wal-mart to manages the buy and sell side of health care.
• (I am going to massively over-simplify this) He talks a lot about having the buyers issue an RFP requiring certain things and creating a new type of entity – the Infrastructure Vendor (IV). “The IV should facilitate and operate electronic connectivity support tools for the patients and caregivers and should demonstrate their effectiveness to the buyers.”
  • He doesn’t see the government playing this role which limits who could do this nationwide.
  • Conceptually, I agree that a technology backbone that connects everyone would be key.
  • It sounds a little too build it, and they will come to me. This is a radically and risky change that would need everyone on board.
  • Some mandated change at a government level has to be required.
  • Could you do this at a state level first?? For example, I know a coalition that got all the employers to agree to a RFP and moved all their business to Humana for one area after they won the RFP.
• At many points in the book, I kept thinking about the need for SLAs (service level agreements) on outcomes. (I haven’t studied the capitation modes tried in the US years ago, but there seems to be something there about paying a provider a fixed amount per year. Their job is then to act preventatively.)
• I am a fan of using incentives and penalties in the system with one caveat. I think you need to tie this to genomics. So, someone who has high cholesterol based on their family history and tries to treat it shouldn’t be treated the same way as someone who eats junk food all the time with no family history.
• I think making people buy-up to different providers or drugs works great for events that can be planned, but not for emergency. It would be possible to tell which one was which with a fully integrated system. Of course, you have to manage people not gaming the system, but that is where there should be incentives for being preventative. Trading off metrics in your design to balance behavior will be key.
• Another sad fact that he relays toward the end of the book is some of the data pointing to the racial and ethnic disparities in coverage and care in the US.
  • The death rate from asthma for African American children is 4x the death rate for white children.
  • Minority Americans make up ~ 1/3^rd of our population but over ½ of the uninsured.
• One thing I didn’t see or get was whether any of the international models that he studied had a focus on outcomes.
  • I thought one interesting point he made that in a government system where votes are at stake there is a strong focus on primary care which is used by the masses (i.e., more votes) versus specialists which are used by the minority of patients. Another example of how incentives skew solution design.
• I am always shocked when I see the Federal Poverty Guidelines. How does someone survive on $9,800 or $20,000 for a family of 4? If you ever wonder how all the tasks get done around you and still feel like addressing the uninsured and underinsured is an issue, you should try to live on that income.

My summary after reading the book was:

• Wow! We have a lot of work to do.
• We can make a difference pretty easily.
• There are three things that matter – infrastructure, incentives, and culture.
  
• Employers have to be willing to push incentives or penalties to their employees. The strategy of lowering costs without “disrupting” people doesn’t work.

Go read the book. Help make a change.