Health Insurance Company Response to High Risk Patients



Today’s Managing Health Care Costs Indicator is 25%


Click image to enlarge. Source above

Today’s NY Times has an article focused on high risk patients – who in the under-65 population represent about 1% of patients, and 25% of all medical costs – over $100,000 per year. Full Report Here (Registration Required)

The concentration of the cost of care is an important observation – the bottom 50% of Americans represent about 3% of total medical costs – so interventions to lower the cost of care for these least expensive health plan members are bound to be unsuccessful. This is also why programs to convince Americans to go see the doctor raise costs.   (Good Op-Ed on this topic from Dartmouth University today, too).

The robust and publicly-available Medicare claims database has been studied extensively – but many of the conclusions from the over 65 population are not applicable to the younger cohort insured through employer-based plans

-       Medicare beneficiaries are highly likely to be readmitted to the hospital (almost 20% in 30 days). Readmission rates are much lower  in the non-Medicare population (5-8%) , and many of these readmissions are planned, like followup inpatient chemotherapy.
-       Medicare spends a quarter of its dollars on patients in the last 6 months of life.   Commercial plans actually spend a very small portion of their claims on end of life care, since death is much less common in this population
-       Medicare beneficiaries stay with that insurance plan for the remainder of their lives, while commercial health plans “churn” membership at a rate of 15% or more a year

Care management interventions should not assume that the commercial population is akin to the Medicare populations.

Health plans have hired platoons of nurses to do outbound calls to high-risk beneficiaries– and they charge employers large fees to perform this work.    However, most of the evidence of efficacy of this intervention is highly anecdotal. From Reed Abelson’s article:

When Wendy Meath, a 59-year-old with diabetes, was hospitalized a year ago, she was identified by HealthPartners as someone who needed help to control her disease. She had been admitted for kidney stones, one of many possible complications of diabetes. Although she had insurance through her husband, she was unemployed.

Since leaving the hospital, where she was admitted for 12 days for a series of complications from the surgery to remove the stones, Ms. Meath has been in regular contact with one of HealthPartners’ nurses, who serves as a case manager. The nurse calls at least once a month and checks in after she goes to the doctor for any developments. The health plan also assigned a social worker to help her with the cost of medications and other obstacles that were preventing her from taking better care of herself. “It makes me feel like I’m not alone,” Ms. Meath said.
“They’re trying to prevent the big things from happening, which is great,” she said.

But the iron-clad evidence of the effectiveness of this intervention is still lacking.

Health plans tend to cite compelling anecdotes – but here’s what we should look at to assess efficacy of these programs

Structure 
a. What intervention is in place?    Are the elements evidence-based?  Is there a measurement plan?
Process

a.     How many of the targeted high risk members actually participate?  Many health plan programs have half or more targeted members refuse to participate.   That sharply limits potential effectiveness
 Outcome

a.     Are there changes in clinical course due to the intervention?
b.     Are there beneficial financial outcomes as a result?

There are two major problems with the outcome evaluations I’ve seen – most of which have focused on only program participants.  

The first is regression to the mean.  The likelihood that Wendy Meath will be hospitalized again over the next year is low.  This doesn’t mean that the intervention has been successful – it’s what you’d expect in the year following such an admission.

The second problem is selection bias. The half of patients who eagerly participate in the intervention are fundamentally different than those who refuse.  Even efforts to do “propensity matching,” to try to adjust for known differences between participants and nonparticipants, aren’t effective for adjusting for these differences.

We should only attribute cost savings to these programs if there are cost savings over the entire population.  Further, we should only find claims of success believable if enough members were engaged to credibly lead to the claimed cost savings.

I believe that high risk programs performed in the provider realm are likely to get more patient engagement – and could lead to more success in preventing bad outcomes and lowering costs.    Health plan leverage in the future might have more to do with payment reform than with hiring legions of remote nurses.

Of course, we will have to measure that too!

Perceptions of the Affordable Care Act


Today’s Managing Health Care Costs Indicator is 44%


Click on image to enlarge.   Source 


The Kaiser Family Foundation does periodic tracking surveys, asking Americans for their opinion of the Affordable Care Act.  The January survey showed that 44% of the public regard the bill unfavorably, while 37% regard the bill favorably. 

KFF explored the percent of nonelderly who would benefit from the ACA by PUMA (Public Use Microdata Areas – a bit larger than zip codes).    Interestingly, Republican congressional districts got substantially more benefit than Democratic congressional districts. 



On average, an estimated 17% of the non-elderly population nationwide would benefit from the Medicaid expansion and tax credits. In parts of Florida, New Mexico, Texas, Louisiana, and California, 36-40% of population could benefit.  In areas of Massachusetts, Hawaii, New York, and Connecticut – states that generally have high levels of employer-provided health insurance or have already implemented reforms to make insurance more accessible and affordable – 2-4% of the non-elderly could benefit from the coverage expansions in the ACA.



Click image to enlarge.  Source

Harold Pollack had a thoughtful essay in The Incidental Economist, debunking some of the right wing press which misstates conclusions from Jonathan Gruber who evaluated impact of the ACA in Wisconsin, Minnesota and Colorado.   He concluded with a question:

Millions of people will join health insurance exchanges. Most will be relatively healthy. They will see certain relatively small and immediate things, such as contraceptive coverage and clinical preventive services. They will not (yet) experience chronic illness and thus the resulting interactions with their insurerWill they perceive and value the improved actuarial value of this insurance? Two years ahead of time, it is impossible to answer this basic question.

I don't feel sanguine about this.  The fact that Gruber and others show that on the average households will be better off doesn’t mean that the Affordable Care Act will be popular.  Distribution of benefits from the ACA and our understanding of behavioral economics gives us some reasons to worry

1.     The Affordable Care Act includes defines essential coverage, which will prevent insurers from selling policies that would not cover preventive care.   Many people won’t realize that the preventive care wasn’t covered before (especially if they had HMO policies which traditionally cover such services.) Most employer policies cover preventive care already.  
2.     Medicare cuts in future hospital reimbursement increases are necessary – and there is no credible health care reform plan that does not include such cuts.  However, these cuts will be blamed every time a hospital has a layoff  - and might be highlighted whenever a patient has access difficulty.  This is a framing problem – the ACA will be compared with what we had before 2010 – not what we would have had instead.
3.     The ACA’s prohibition against underwriting for kids and against excluding preexisting illnesses will provide much-needed new benefits to a very small number of people.   Most of those with employer-based insurance already have this.  Many who had previously purchased defective coverage – which would not have protected them if they had a major illness – value their current low premium, which will increase under the ACA.   They will in the future have an insurance plan that is a  “Chevy” rather than a Yugo with a defective engine block – but they didn't realize their old policy was defective if they were in good health.  They liked the low premium for the Yugo though!
4.     The ACA only allows a three-fold difference between the price of coverage for the oldest vs. the youngest health plan beneficiaries.  This means that there will be substantial subsidization of the old by the young.  That might seem politically good (middle age people vote in higher numbers than the young).  However, people hate losses much more than they like gains.  Therefore, the “losers” in the new system will be much more unhappy than the “winners” will be gleeful.
5.     We tend to like what we already have (“Endowment Effect,”) and many people will lose their current policies. They might get better policies – but still change is unpopular.
6.     We also tend to like choices.   While there will be more choices in health insurance exchanges (assuming that each state has a functional exchange) than the current choices available to individuals and small groups – there are mandates now which limit individual choice.
7.     2012 had relatively low rates of health care premium inflation. If that inflation rate is higher in 2013 or 2014, many will blame the ACA – even if actuaries know that the incremental requirements are responsible for only a tiny portion of overall health care premium inflation.


There are a few concepts from behavioral economics to suggest how the Obama administration should approach the political conundrum of gaining more public support for the ACA

1.     Discounting: Be sure to show the benefits as soon as possible.  We discount too much in our minds -  so a benefit in the distant future is like no benefit at all.  2014 seems like an eternity away right now, and the uncertainty around the Supreme Court challenge makes this problem worse.  (The KFF survey showed that the majority of Americans think the Supreme Court will find the individual mandate unconstitutional).  Coverage of young adults on their parents’ policies was an excellent idea.
2.     “Availability”.  We really like stories – much more than facts or statistics.  Therefore, we need more stories about families with hemophiliac kids no longer facing ruin because of lifetime limits, and families covered for dramatically less than they had to spend prior to health care reform.
3.     We overestimate the chances of small-likelihood events – which is why we gamble and play the lottery. It’s also why we worry about being killed in  a plane crash. (Cars kill over 40,000 per year; in the average year the deaths from civilian air crashes with scheduled carriers is less than a dozen).  The Affordable Care Act would protect Americans from terrible outcomes from rare events.   We need to hear more about this.

The Affordable Care Act will be good for the average family – and over time should be quite good for our economy.   But many will continue to doubt the overall benefits of the ACA – and we need an effective political campaign to show its benefits.

Physicians Would Order Screening Inappropriately for Ovarian Cancer


Today’s Managing Health Care Costs Indicator is 28%

Likelihood a Physician Would Order Ovarian Cancer Screening Tests
 Nonadherent to Evidence-Based Medicine

Click Image to Enlarge  Source
We’re often overly enthusiastic about screening efforts – because it makes such good intuitive sense that finding a cancer early would be better (and maybe even less expensive) than finding it late.    But screening for especially rare diseases produces many false positives, and screening for diseases where the treatment is of uncertain efficacy leads to many people living longer with a disease only because it was found sooner.  Screening can sometimes increase medical utilization and cost without helping us gain quality adjusted life years. 

The February 7 Annals of Internal Medicine illustrates how difficult it is for physicians to practice evidence based care.  Over a quarter of physicians reported that they would order ovarian cancer screening on low risk women; while almost 2/3 of physicians (65.4%) reported that they would order ovarian cancer screening for women at medium risk of ovarian cancer.  Both the American College of Obstetrics and Gynecology and the US Preventive Health Services  Task Force have recommended against such testing.

Researchers asked a nationally representative sample of  physicians to review one of three clinical scenarios of women of varying ages who were at low, medium, and high risk of ovarian cancer.  The physicians were then asked how likely it would be that they would order CA125 (a blood test) or transvaginal ultrasounds to screen for ovarian cancer.   One in three surveyed physicians believed incorrectly that there was evidence to support using these screening tests in low or medium risk women.  

Factors that made physicians more likely to order this test
-        The patient asked for a screening test
-        The physician believed incorrectly that evidence supported this screening test in this type of situation
-        The physician overestimated patient risk
-        The physician was older
-        The physician was an obstetrician-gynecologist (I might have expected specialists in this field to perform better, not worse)
-        Physician did no clinical teaching
-        Physician had had cancer him or herself

Patients hate uncertainty – and often believe that a noninvasive test couldn’t hurt them.  We doctors should know better, but we don't.  We study Bayes’ Theorem, which demonstrates that doing moderate specificity  on low risk populations produces an overwhelming number of false positives.  But good decision-making requires more explanation.  Further, the patient who has a false alarm is often grateful for our efforts, while we all know of women diagnosed with ovarian (or breast) cancer when they were very young, and many of them die horrible deaths. 

The authors suggest that costs of this excess screening could be as much as $360 million annually across the US.  I suspect that the cost is much higher, because the authors are not counting the downstream tests initiated by the initial false positive from the inappropriate screening test. 

We clearly need more effective professional education around the efficacy of ovarian cancer screening.   The authors note also that patient education has lowered the demand for inappropriate BRCA 1 and BRCA 2 screening women worried about breast cancer, suggesting that patient education could play an important role.

By the way - good news today that colon cancer screening really DOES save lives.  The study reviewed the experience of 2602 patients who had colon polyps removed with rate of death among the general population; patients were tracked as long as 20 years, and the study showed a 53% decrease in rate of death from colon cancer among those who had the polyp removal.

We need physicians to expend more energy to promote colonoscopies, and less energy to promote screening that has not been showed to lengthen or improve life. 

Accretive vs. Disruptive Innovation in Health Care


Today’s Managing Health Care Costs Indicator is 4 (quadrants)


I've pointed out that the "action" in health care policy is not around increasing use of services that increase quality and decrease cost. Everyone agrees we should increase use of such services, but historically there have been few such services except for childhood vaccinations.  There is also not much "action" in decreasing services that increase cost while they decrease quality.  No one clinically thinks we should use antiinflammatory medications that are wildly expensive and increase the rate of heart attacks (Vioxx).

However, the "action" in health care policy should be around embracing some services that sacrifice an inconsequential amount of quality for a big price cut, or deciding at what price point we will reject a trivial improvement in quality because the cost is just too high.

Here's a post from late 2010 on the issue, and here is the graphic I showed then when I was thinking about a $93,000 prostate cancer vaccine that extends life by an average of 4 months. (By the way - I'm not saying that this is necessarily too high a price to pay for 4 months of life - but there is some price point at which we'll have to pass on small incremental benefits)

I've been thinking more about disruptive vs. accretive innovation - and this fits well in a similar related schematic.  Disruptive innovations are initially less functional than the incumbent good or service (think of tax return software compared to a certified public accountant, or an early personal computer compared to a mainframe.)  Yet, they improve more quickly than the incumbent over time, and often displace the incumbent.  Disruptive innovations are a key to increasing value delivered to consumers.  By now, few would buy a mainframe computer because a laptop (or even a cell phone or tablet) is "good enough" to meet most of our needs.  Disruptive innovations are in quadrant 2 - the top left.

Accretive innovations, which are much more common in American health care, offer at least a little more quality - but they increase cost. Accretive innovations usually don't displace existing goods or services.  That's good for providers who don't like to lose existing revenue, but it means that the value to the consumer/patient often goes up by only a small amount.  Accretive innovations can even decrease value to consumers/patients if the cost increases dramatically for a modest improvement in outcome.   Accretive innovations are in quadrant 4 - the bottom right. 

I'm working on a paper with a colleague describing how the confluence of increasing price sensitivity (and patient cost sharing) and the Affordable Care Act should lead to a real increase in demand for quadrant 2 services - aka disruptive innovation. That, I believe, will be an important key to increasing value in health care, and tamping down the rate of health care cost increases.

Here's the way I'm thinking about this schematic now:
Click image to enlarge
Comments or suggestions on this construct are welcome!

Asthma Program A Good Idea, but Cost Saving Claims Not Credible


Today’s Managing Health Care Costs Indicator is $1.46

Click image to enlarge.  Source 

Today’s Boston Globe front page rings out with a “top of the fold” headline “Children’s Hospital Reports Asthma Progress.”

Hospitalizations for asthma have been dramatically cut by a program that helps families reduce the conditions that trigger attacks, saving $1.46 in hospital care for every $1 spent on prevention, according to a Children’s Hospital Boston study being released today.

The actual article was e-published by Pediatrics today.  The link to the article is hereHarvard Link

The truth, as always, is murkier than the headline.

First of all – this study is laudable for many reasons.   It aimed to lower the asthma morbidity in underprivileged communities – which is where asthma has the most devastating impact.   Asthma hospitalization rates are five times higher among Blacks and Hispanics compared to whites – and the study group was largely Black and Hispanic. The researchers sifted through the literature about what works, and developed a multidisciplinary intervention based on this literature.  They measured carefully.  They counted only “hard” savings like decreased hospitalizations and ED visits, and did not attribute monetary value to

Researchers made notable efforts to be culturally sensitive, and offered home community health worker and nursing visits, supplies to help decrease allergens, and even special vacuum cleaners to decrease airborne particulate matter.  When necessary, the program did extermination to help prevent exposure to allergens that could trigger asthma attacks.  

Now, the concerns.

This is a study with 283 patients, who were handpicked by researchers because they appeared to be at highest risk for recurrent emergency department visit or hospitalization based on their recent history.  The study was offered to 562 families; so the take-up rate was about 50%.    The cost savings are comparing these study patients with patients from different zip codes – who were not subject to the same selection process.

Note that the study was designed in 2003 and carried out from 2005-2008. This gives you the sense of how complicated these studies are to complete.

Here are two reasons why this study likely overstates the benefit from this intervention

  1.  Regression to the mean.  Those asthmatics who are chosen because they appear very ill today will as a group always have far fewer hospitalizations and ED visits going forward than they had in the recent past.   Here’s a link to a 2004 study where the control group (no intervention) had more than a 30% decrease in cost.   Here’s a link to my letter to the editor, where I pointed out that even this understated the true amount of regression to the mean.
  2. Selection bias.   The families that were willing to participate and were able to persist in the intervention likely had more means than those who refused, were unable to be reached, or dropped out during the course of the trial.  The stated control group did not have such selection.


To the Globe’s credit, the last two paragraphs quote a researcher who points out that this was not a randomized controlled trial. 

It’s heartening that it appears that there was some decrease in asthma morbidity that coincided with this intervention.   I would be very cautious about the cost-saving claims from this article.

Two major public health interventions to lower asthma morbidity – decreasing air pollution and decreasing parental and teenage smoking rates, are not mentioned in the article.  We tend to overemphasize interventions in the medical model while we underinvest in effective public health interventions. 

End of Life Discussions: Too Late and Still Too Infrequent


Today’s Managing Health Care Costs Indicator is 73%


Site of End of Life Discussions
 Click image to enlarge. Source

We know end of life care is responsible for an outsize portion of the total medical budget. Many people express a desire to avoid extensive intervention at the end of their lives; nonetheless, a majority of Americans die in hospitals, and there is unbearable variation in how Americans are cared for at the end of life.

Medicare spends about a quarter of all of its dollars on patients in the last six months of their lives.   For most deaths, though, we don’t know in advance exactly when the clock starts ticking on that last six months.    However, those who have Stage 4 metastatic lung and colon cancer have median survivals of 4-8 months and 12-24 months. 

Even in those with diseases known to be associated with high likelihood of early death, there is huge opportunity to improve physician discussion with patients about their preferences for end of life (EOL) care.

This week’s Annals Of Internal Medicine has a painstaking study of end-of-life care discussions with over 2100 patients with end stage lung or colon cancer. The results are not as disheartening as some past studies – and almost three quarters of all patients had either a conversation reported by the patient (or family member) or a conversation documented in the medical record.   The research did not grade the meaningfulness of this conversation, and interviews focused on resuscitation and hospice care only, while record review also included venue for dying and palliative care

The results aren’t pretty

·       64% of patients (or surrogates) reported a conversation with a physician on end of life care preferences; 58% of medical records reported such a conversation.  The concordance rate was 65% - and the authors state that most of the discordant cases involved lack of medical record documentation of the end of life care conversation. 
·       Most of the end of life conversations (64%) happened in the hospital.  Even general physicians had these end-of –life discussions in the hospital 73% of the time.  This means these conversations were likely when the patient had  an acute deterioration.  It also means that few of these discussions were between t he long-time primary care physician and the patient, as many are now cared for by hospitalists.
·       Most of these discussions happened in the final weeks of life.  The first discussion took place a median of 33 days before death among those with documented EOL discussion who died during the study.  Even among those whose cancers were diagnosed over 12 months before their deaths, 29% of patients who had a discussion had this within 30 days of their deaths.

There are real social and cultural reasons why we delay talking to patients about their EOL preferences. As physicians we want  to promote hope, and as human beings we are optimistic.  But the cost of our reticence is that many get care which they want to avoid, at a very high cost.   The debate in 2010 around  “death panels” doesn’t make this issue any easier.  Lead author Jennifer Mack and her colleagues have given us good evidence that we have to improve our capacity to talk frankly about end of life with our patients.

This issue of the Annals also has an interesting survey showing that physicians are highly likely to order inappropriate screening tests for ovarian cancer.   More on that in a future post.

Hospitals and Labs Pay for the Doc Fix


Today’s Managing Health Care Costs Indicator is 19%

It’s good news is that Congress is actually working – and has fast-tracked reauthorizing the Social Security tax break,   unemployment insurance, and the “doc fix” to prevent 27.4% decreases in Medicare physician payment.  Providing relief to those who are unemployed in this terrible economy, avoiding a middle income tax hike, and avoiding a catastrophic decrease in physician fees are all important.  Doing these together makes it harder to defeat the effort.

Doctors are the health care winners in this, at least for 2012.  Let’s look at who are the losers.  Kaiser Health News has links to a number of articles on this topic.  

·          Hospitals are the biggest losers.   They’ve already given back substantial future increases as part of the Affordable Care Act, and the current legislation will remove federal payment for bad debt and decrease payment for disproportionate share hospitals – those which take care of a large share of the most impoverished Americans.   Safety net hospitals in Massachusetts have fared poorly under health care reform, and this legislation makes it likely that this will be repeated across the country.  Rural hospitals maintained their enhanced fees.
·         There will be a $5 billion clawback from the public health funding promised in the Affordable Care Act.   It’s public health efforts that have the highest return on investment – and it’s disheartening to see us lower this investment.  However, many Republicans opposed this public health funding – and it’s easier to stop paying for education to prevent future disease than to make it difficult for Grandma to find a physician
·         Laboratories are big losers as well. Their fees will be shaved another 2%; The Affordable Care Act had already lowered laboratory fees by as much as 19%.
·         Louisiana is a surprise loser.  Senator Mary Landreiu had obtained an additional $2.5 billion in Medicaid payments for her state.  This is being rescinded.

The “doc fix” only cost $20 billion – since it is only for the rest of 2012. This means we’ll be back to this issue again at the end of the year, and by that time the cut to be averted will be even larger than 27.4%.