The thing that worries me most about the long term health system outlook

Economic growth in the second quarter was up over expectations, health care inflation remained historically low, and the ACA was rolled out in a reasonably successful manner (remember in December when people were saying no one would sign up, or the uninsured rate would go up?, or whatever doom & gloom was the flavor of the day). I am a charter member of “The ACA was a good step but we need many more to have a sustainable health care system” and there are tons of uncertainties in forecasting long run health care cost growth, overlaid with difficulties in forecasting econcomic growth. But, the passage and implementation of the ACA has put the entire health system in play and I am optimistic that we will muddle through reasonably, helped by a political deal at some point relatively soon.

However, the thing that really worries me about the future of our health care system and country, is long term care.

Even if we manage to continue on the path of lower than average per capita cost growth in the Medicare program, the transition of the baby boomers into retirement still means more workers (adult children) caring for their parents (Medicare beneficiaries) with fewer siblings to help. This is esssentially the demographic definition of the baby boom. Long Term Care (LTC)-which you could define as help and support to enable someone to live with a disability-is easily the most important public policy issue about which we are not discussing.

The experience of my family in caring for my mother-in-law who suffers from a rare form of dementia has been exceedingly difficult.

• Several years of very close supervision across town, hours spent navigating physicians and health care use, dealing with her financial affairs for the past decade.
• Her moving into our house (me + my wife and  3 teenagers) for around 6 months, and finally moving to an assisted living facility after 3 hospitalizations related to falls and wandering (hallmarks of dementia) in less than 6 weeks.
• Navigating the issues related to assisted living placement and monitoring (she is in her second facility, after being thrown out of the first)

There are many types of burdens:

• Physical in the form of picking my mother in law off the floor when she fell;
• Missed work and lack of being able to focus at work due to worry;
• Emotional strain in many forms, especially for my wife who struggled with anger and depression as we tried to do what was best.
• The hardest thing overall was the uncertainty about what to do to care for her. I am supposed to be an expert in LTC and palliative care, and we were constantly unsure if we were doing the correct thing. Our marriage was actually strengthened by our escalating caregiving responsibilities over the past few years, but it is easy to imagine it having the opposite effect on a marriage.
• Financial. We have been spared much of this burden because my mother in law has an expansive private LTC insurance policy. Great for her and us, but not a population wide solution.

The ACA of course contained the CLASS provisions that were an attempt to set up a private LTC insurance market to cover low level LTC costs, and these provisions were repealed from the ACA as part of the fiscal cliff deal. I think CLASS as passed was unworkable, and think that Sec Sebelius was correct to say that it should not move ahead in October, 2011. However, one of the worst outcomes of the acrimonious debate about CLASS was to turn the discussion of LTC into nothing more that a deficit accounting debate. Let me assure that it is much more than that, and that our country needs to talk about how we should deal with LTC in a much more forthright manner.

This link has many posts on LTC if you want to read more.

Seniors Overestimate Their Likelihood of Going to a Nursing Home

Yesterday, I was helping someone think through options to care for their parent who has dementia and recently suffered a serious fall. This person’s parent does not have private long term care (LTC) insurance (most don’t) and my friend said that “this must be due to the fact that elders underestimate their likelihood of moving to a nursing home”, the most expensive LTC setting.

In fact, it is the opposite based on work I have done using the HRS. From a 2006 paper in Health Services Research:

The simple comparison of the stated probability that a person who was age 65+ would move to a nursing home within 5 years was higher than the actual probability (for full sample, mean perception 0.14 v observed 0.08; see first column of p values).

Private LTC insurance is rare in spite of the elderly over-estimating their likelihood of needing such care. Why?

From a Health Affairs paper on the topic, here are the 6 primary reasons people do not buy such insurance.

• No clear risk signal for need of LTC when young
• myopia about the need for and cost of LTC; so people don’t understand and/or face up to the risk
• 3 in 10 of those achieving age 65 will not use LTC, so if everyone bought private policies, a substantial minority would never claim benefits from their policy
• Medicaid coverage of NH care likely crowds out the purchase of private coverage
• A substantial proportion of the population has insufficient income to pay LTC insurance premiums, and/or insufficient wealth to “protect” from the cost of LTC
• The structure of the policies themselves lead to rational non-purchase: (benefits denominated in dollars per day and not care, which is risky especially when insuring against something that is probabilistically a long way off; history of premium increases when claims experience is higher than predicted; denial of applications).

Keep in mind the paper comparing stated v actual probabilities was done in people age 65 and over. So, if young people are myopic and wait too late to purchase, it is too late. However, there are many rational reasons that people don’t purchase LTC insurance.

Papers cited:

Donald H. Taylor, Jr., Jan Ostermann, S. Will Acuff, Truls Ostbye. Do Seniors Understand Their Risk of Moving to a Nursing Home? Health Services Research 2005;40(3):811-828.

Donald H. Taylor, Jr. Robert Cook-Deegan, Susan Hiraki, Scott Roberts, Dan G. Blazer, Robert C. Green. How Genetic Testing for Risk of Alzheimer’s Disease Could Affect Long Term Care Insurance. Health Affairs 2010;29(1):102-108.

Pre-fund VA health benefits?

Michael Cannon and Christopher Preble have an interesting suggestion in the New York Times–pre fund Veterans benefits via private insurance

We propose a system of veterans’ benefits that would be funded by Congress in advance. It would allow veterans to purchase life, disability and health insurance from private insurers. Those policies would cover losses related to their term of service, and would pay benefits when they left active duty through the remainder of their lives.

To cover the cost, military personnel would receive additional pay sufficient to purchase a statutorily defined package of benefits at actuarially fair rates. The precise amount would be determined with reference to premiums quoted by competing insurers, and would vary with the risks posed by particular military jobs.

This is an interesting idea, and I am not conceptually opposed to the use of private insurance to meet health care commitments to veterans. I suspect this idea is ultimately found to be unworkable, and the best place to get a sense of why are a quick look at the problems in the private long term care insurance market (general post on insuring LTC). The following problems are likely to be particularly tricky with Cannon & Preble’s proposal:

• when you set up an insurance contract with a potentially long lag time between initiation and collection of benefits, there is a very real question of whether the company will be around when needed. Private insurers are exiting the long term care insurance market in droves currently. With major medical insurance, for example, the contract is only for the coming year. For an 18 year old enlistee, their point of need may be 5 or 6 decades away.
• The longer the lead time, the harder it is to get the premium straight and/or the benefits appropriate. Both the insurer and the covered person are at risk here. LTC insurance is denominated in dollars per day (instead of care), and if you have say a 5% rider, and the true cost rises 5.1% for 30 years, that turns out to be a big miss.
• The reason private insurers are leaving the LTC market are because of the difficulty of underwriting LTC risks. It might be simpler with the military in the sense of you don’t have (presumably) 19 year olds with pre-existing reasons to make them more likely to have disability/complicated long term health claims, while private LTC insurance is rife with selection (bad risks want to buy). The exposure to military service is what is causing the need. I am unsure of how well one could underwrite need based on military job?
• One problem with private LTC insurance policies is that while premiums cannot be raised on individuals, they can be raised for an entire class of policies if losses overshoot what underwriters expected. Regulators are left with the un-enviable choice of the insurer saying they will go bankrupt or premiums are raised on everyone. This often leads persons to drop premiums, leaving them uncovered when they have LTC needs. You can say that folks in the military/vets shouldn’t make bad decisions, but the plain reality is that politically we won’t let such actions harm people. So, the insurance will not ever really be private. This is probably the biggest problem with the proposal.
• I am open to the use of private insurance to meet the needs of vets, especially for the non-specialized services that many need. However, it seems much more likely to be a pay-as-you-go as compared to a pre-fund set up, though I give credit to Michael C for proposing a solution (I often hector him about him only being against stuff).

This discussion would be useful to have while we are talking about starting a war, as it would help to make the true costs more evident.It is quite possible that is both the main point, and the most useful aspect of this proposal.

CBO on Long Term Care

CBO released a great report on Long Term Care for the elderly (what is now most typically being called long term services and supports). Several key figures/graphs. The first shows that most people receiving LTC do not live in Nursing Homes, but in the community.

A little over half of LTC care (in terms of its economic value) is provided by families and others on an informal, unpaid basis.

So, when you see this break down of payments for LTC by type of payer, remember that over HALF of the total economic value of all LTC is not accounted for in the graph below, but is paid for implicitly by family members.

Two quick policy take home points:

• When evaluating a long term care proposal of any type, the correct cost to compare it with  is not zero; the default costs are huge, it is just that many of them accrue silently via informal care.
• The lack of a coherent LTC system is causing problems in all sorts of other policy areas. Just two examples are in hospice policy, where long stays are likely driven at least in part by persons seeking care in a way that has turned hospice in some areas into a back door long term care benefit. Similarly, hospital re-admissions are at least partly driven by LTC break downs that are beyond the scope of any hospital-based program to address.

The hope for coherent policy making in this area is negligible in the near future.

Genetic Nondiscrimination Act of 2008 and LTC Insurance

NPR has a story on the legal (h/t Brad Flansbaum) use of genetic information to discriminate (underwrite, set premiums) in the private Long Term Care (LTC) insurance market. I am the first author of the paper that Bob Green (the P.I. of the underlying study) is discussing in the NPR piece and I have blogged about its findings here and here.

The basic story of the paper in Health Affairs is that persons with at least one copy of the e4 variant of the APOE-4 genotype were found to be both more likely to move to a nursing home in a community cohort of elderly persons living in 5 N.C. counties, as well as being found in the REVEAL II study to alter their behavior upon finding out they were at increased risk of AD, including making changes such as buying Long Term Care Insurance. In short, this is adverse selection whereby consumers have information that companies do not, a story that likely keeps insurance executives up at night.

The biggest news of the segment was the report that Genworth, the largest seller of LTC Insurance in the U.S. acknowledged that the The Genetic Nondiscrimination Act (GINA) of 2008 does not ban the use of genetic markers to underwrite their product, and they stated they wished to keep this option available to do so (some states have moved to do so). I blogged at some length on this back on October 2011 (that it was allowed for LTC Insurance), but I have never heard anyone from the LTC Insurance industry acknowledge they wanted to keep this option.

I think it is a mistake to take a reflexive “see the insurance companies are bad” response to this story, and that in Long Term Care especially, we need some renewed thinking about what ‘fairness’ means.

The general idea behind GINA 2008 is that you cannot pick your genes, so to discriminate against someone on that basis is unfair. This makes general sense to me, and likely to most people. However, in a private insurance market with very low penetration (~less than 7-10% of persons in age ranges to consider such products buy them; this post highlights the many reasons) what constitutes fairness is not as clear. An average risk person who wants a LTC policy because they have no children, for example, yet who ends up paying a higher than warranted premium due to adverse selection in the market (those with higher risks signing up) could also be viewed as being harmed in an unfair manner.The chart below lays out how different scenarios of who knows genetic information in LTC could influence uptake and premiums. The table is meant as food for thought.

The bottom line answer to planning for LTC is risk pooling, and if ever there were a risk distribution that called out for social insurance it is LTC. That seems impossible politically at this time. However, the LTC system in this country is a patch work mess that is in need of some sustained policy efforts. This post has a weeks worth of links on planning for LTC, from both an individual and population basis if you want more. Long Term Care is always the forgotten topic until someone in your family needs it.

Long Term Care problems won’t go away

Busy week and little time to blog, but a quick note on this Kaiser story reporting on a SCAN Foundation poll on long term care needs/perceptions/preparations in California, sent my way by Brad Flansbaum. The article nicely summarizes the surprise many families receive when it comes time for a loved one to need LTC:

Long-term care costs can surprise many families who expect Medicare to cover their needs. After a hospital stay, Medicare will pay for 100 days of nursing home care, but after that, families are on their own or are forced to spend down their assets to become poor enough to qualify for Medicaid.

Only 35% of Californians correctly understood that Medicare does not pay for extended nursing home care and only 1 in 5 understood the Medicare home health benefit (they think it is more generous than it is). So, Medicaid is the default nursing home payer in the U.S. and the program pays around half of the nation’s nursing home bill. Block granting Medicaid in a way that reduces fairly costs for the program will pit LTC services for the elderly and disabled against the need for acute care insurance for children and pregnant women, and say to states “tag; you’re it!”.

We desperately need a more coherent LTC policy in this country. Private LTC insurance offers no hope of a population based answer. Families are left to stand in the breach, doing the best they can. As I have said before, one of the worst outcomes of the ACA debate was the demise of CLASS in a way that treated those provisions only through the lens of deficit reduction. Deficit accounting is important, but cannot answer all important policy questions.

CLASS wasn’t a deficit gimmick, it was an attempt to set up a self sustaining LTC insurance benefit that could have helped people age in their homes and delay NH admission. Like so much of health reform, critics know far more about what they are against than what they are for.

The question remains: how will our country insure against the need for LTC?

States develop regulations to limit use of genetic info in LTC insurance

My colleague Bob Cook-Deegan sent this article on states developing LTC insurance regulations restricting the use of genetic markers to underwrite private LTC insurance policies. The background is the hole in the Genetic Information Non Discrimination Act (GINA) of 2008, which did not ban the use of genetic information to underwrite or risk adjust long term care insurance policies/premiums. It has always been unclear to me if this was an oversight, or purposeful. Below is a partial snapshot of a table showing state regulations regarding the use of genetic information in the sale of LTC, disability and life insurance (link to full table, compiled by the Cancer Legal Resource Center):

Some states have moved to regulate the use of genetic information in LTC insurance, while others have not (North Carolina has not).

I did some work with Bob and colleagues from Boston University showing evidence of people using genetic information to behave in a manner consistent with adverse selection. Generally, the use of genetic information is thought to harm the individual whose genetic information may be used and that is the big idea behind GINA 2008, but I think that in voluntary insurance markets with rare uptake like private LTC insurance, there is an argument to be made that the entire risk pool is harmed with one-sided risk information. The key is which notion of “fairness” makes the most sense to you in a given market or context. The “you can’t pick your genes” notion, or the “actuarially fair premium is the only way to increase market share of a rare type of insurance notion” idea. The first is the most common, but I think the second may apply in private LTC insurance markets (which are falling apart in any event).

One of the worse by-products of the debate over the ACA was the labelling of the CLASS long term care provisions as simply a deficit reducing gimmick. Long Term Care is the forgotten policy, that will reliably rear its ugly head any time you want to do something like move toward premium support in Medicare or undertake any Medicaid reform. In short, Republicans really seared the ground for government efforts to jump start private insurance markets in the way they attacked CLASS, and they desperately need something besides the status quo in LTC (Medicaid pays for half the nation’s nursing home costs) to bring about their preferred big ideas in health reform.

Politics aside, the question remains: how will we insure against LTC? The baby boomers are coming….

update: clarified post by noting the LTC insurance market is voluntary; though forced risk pooling would have fixed CLASS and made it workable

Private LTC Insurance is a not a population solution

The WSJ has a good point/counter-point on the question “should you purchase private Long Term Care Insurance?” (h/t Brad Flansbaum). Mark Meiners argues for purchase by saying “you shouldn’t hope for the best” while Prescott Cole says “LTC insurance is too expensive, you should invest what you would spend on premiums.”

Essentially, they are both arguing that you should prepare for LTC, one via purchasing insurance, the other by building flexible assets that could be use for LTC, or bequeathed to your favorite charity upon your death if die without needing it.

The Journal piece outlines an important conversation that misses the main public policy point: private LTC insurance will never be the solution for the LTC needs of the general population due to income and wealth levels.

Perhaps 10 percent of the population has enough income to pay premiums, and/or enough in assets they may wish to partially protect from a potentially catastrophic LTC cost. For example, past work I have done showed that around 4 in 10 elderly persons had income and assets at age 65 low enough to qualify for Medicaid before paying for any LTC. Those people are removed from the complicated decision framed in the WSJ piece, but they are certainly at risk of needing LTC.

If ever there were a risk profile that cried out for social insurance, it is LTC. The reason that seems so laughable, is our countries failure to grasp the most important thing in all public policy debates: the counterfactual, or the costs and benefits of what happens by default, in this case for LTC.

NPR on Informal Long Term Care

NPR has an interesting story asking “who needs LTC insurance?” because other arrangements can be made to provide for such care (h/t Brad Flansbaum).

This puts a personal face and story on a public policy problem that hits close to home, as my mother-in-law is moving in with my family this Summer. The last 8-9 months my family has been consumed with myriad details to work this out (estate planning, new wills, buying a house, trying to sell two others, getting siblings on the same page, etc). I wrote last week on some of the public policy aspects of LTC; I mostly want to point folks to the NPR story, as it is the best, short representation of the many issues related to thinking through inter-generational living that I have heard.

The cost of caregiving and the demise of CLASS

AARP estimates that around 25 Million persons are providing unpaid caregiving to a loved one with a disability, and that those who do so while juggling market work lose around $325,000 in lifetime income after accounting for foregone wages, income from Social Security, and private pensions. The worst part of the demise of the CLASS provisions in the Affordable Care Act has been the fact that the majority of the focus was on the impact of CLASS on the 10 year deficit score assigned to the bill be the CBO (it accounted for around half of the ACA’s deficit reduction), diverting attention from one of the most profound public policy questions facing our nation:

• How will we insure Long Term Care?

CLASS wasn’t an accounting gimmick, but an attempt to set up a self sustaining, voluntary LTC insurance program that would provide modest benefits (not enough for a nursing home; best thought of as aiding aging in place) to persons with disabilities. If CLASS had worked perfectly and been a self sustaining program, it would have decreased the deficit in the short term, while increasing the deficit in the out years. As I wrote in December, 2009 about CLASS:

…Proponents claim that CLASS is self-financing over the long term, and opponents say it will increase the deficit in later years. Both are correct.

It is hard to project what will happen with CLASS, mostly because of uncertainties regarding disability rates 30-plus years from now, but there are numerous provisions designed to ensure that the program is self-financing. Regarding the long-term deficit, when a program runs a surplus, it buys federal securities that pay interest. When interest is later used to pay for care, it is counted as a transfer instead of revenue. Therefore, CLASS will inevitably increase the deficit in years 30 to 75 even if it pays for itself totally through premiums and interest earned on premiums, given current budget accounting rules. [update: it would add to the deficit before the 30th year]

One of the key beneficiaries of a self sustaining LTC insurance program would be the adult children of the disabled–who provide unpaid caregiving, with all the related harms (there are also benefits). CLASS was flawed, but could have been fixed had we focused on the policy goal–how will we insure LTC as the baby boomers move into retirement? I think the federal budget deficit is very important to address, but you cannot answer all important policy questions using deficit accounting. The CLASS episode whereby LTC policy was lost in a sea of deficit accounting scores shows this.