The latest edition of the Cavalcade of Risk is up at Insurance Writer.
You are currently browsing the monthly archive for August 2010.
Tags: CoR
- Cost of Medical Errors: $19.8 billion.
- Betting on your grades.
- One weight loss method.
- Back pain: whether to use hot or cold packs.
- Healthcare.gov (and parody)
Tags: Links
Almost 7 out of every 10 of individuals living to age 65 will require some long-term care (LTC) assistance. Of these, over one-third will spend some time in a nursing home. In general, however, the elderly strongly prefer home based LTC if possible. “Mattimore and colleagues (1997) found that 30% of elderly survey respondents would rather die than enter a nursing home and an additional 26% indicated they were very unwilling to move to an institutional setting.”
A recent paper Goda, Golberstein and Grabowski (GGG hereafter) examines how permanent income shocks affect LTC utilization.
The estimation equation used is the following:
- Uhi = βIh + δXh + ε
where U refers to a LTC utilization measure (e.g., nursing home, paid home health care, unpaid care), I is annual household Social Security Income, and X is a set of exogenous controls.
Identifying income and LTC choices independently is difficult. Individuals with a high probability of needing LTC may more years or longer hours per year to accumulate additional assets to fund LTC expenses. Ideally, identification requires an income shock that is exogenous to the individual’s labor market choices. For that purpose, GGG rely on natural experiment known as the “Social Security Benefits Notch” to serve as an instrument for the income variable. The authors define the Notch as follows:
“Prior to 1972, neither lifetime earnings nor post-retirement payments were indexed for inflation, but rather periodically adjusted by the Congress. In 1972, Congress amended the Social Security Act to provide automatic indexation of credited earnings for those workers who had not yet retired, which created an unanticipated windfall for workers from certain birth cohorts because of an error that led the prior earnings of these workers to be doubly indexed for inflation. The high rate of inflation over the following years led to a large increase in benefits for the affected cohorts. In 1977, Congress passed another law to eliminate the double indexation for future cohorts of retirees. This law change created a large reduction in Social Security payments for those cohorts born in 1917 or later relative to the preceding cohorts. Importantly however, cohorts born prior to 1917 (near retirement in 1977) retained doubly indexed benefits under a grandfather provision. Taken together, these law changes and the high rate of inflation over the mid 1970s created a large and permanent difference in Social Security payments across birth cohorts, which came to be called the Social Security Benefits Notch.”
Econometrically, GGG use a dummy variable for being born in the notch years (i.e., 1915-1917). To apply their model, the authors use data from the 1993 and 1995 waves of the Assets and Health Dynamics Among the Oldest Old (AHEAD). The authors find that this instrument is weak for richer households who have at least a high school education, but much stronger for households whose heads have at least a high school diploma. Due to this result, GGG limit the sample only households without a high school education. Using this specification, the authors find the following results:
“…positive income shocks had a negative effect on nursing home entry, but a positive effect on the use of paid home care. Specifically, a $1,000 (or 10.2 percent) increase in annual Social Security income for those in this low-education group would decrease the likelihood of any nursing home use by 22%-30% (relative to mean) and increase the likelihood of receiving any paid home care use by 24%-34%. Social Security income was not systematically related to the receipt of any informal (unpaid) care across the different specifications.”
At first glance, one might perceive that increased income leads to less nursing home care due to substitution for home health care. Alternatively, higher income could improve health directly and thus lessen the need for institutionalized care.
One obvious mechanism by which increased income would decrease nursing home use is through disqualification of Medicaid eligibility. The authors claim that assets rather than income are the key driver of Medicaid eligibility for nursing home care, but do not investigate how asset accumulation and the Social Security notch are related.
- Gopi Shah Goda, Ezra Golberstein, David C. Grabowski, Income and the Utilization of Long-Term Care Services: Evidence from the Social Security Benefit Notch, NBER Working Paper No. 16076, June 2010.
Tags: Income, Long Term Care, LTC
According to Kathleen Lohr, the most pressing issues for comparative effectiveness research (CER) include: 1) how to conduct CER for heterogeneous patient populations and 2) ways to implement longitudinal investigations to capture long-term health outcomes. Today I will focus on measuring patient heterogeneity. Although it makes sense to take into account differences across patients, measuring this in practice can be difficult.
A paper by Kaplan et al. (2010) discusses a number of ways to measure patient heterogeneity in practice. The article discusses six categories of patient characteristics. Below I present each and–where appropriate–discuss how the authors attempt to measure differences in patient characteristics.
- Immutable characteristics: These are intrinsic factors that the patient cannot change such as demographic characteristics or genetic factors. The authors use age, gender and race as well as education in their study. Although one could of course get more education, few elderly receive advanced degrees. A more appropriate measure might be highest education reached by age 30 (which would not change over time after age 30), but because this measure would likely be very similar to education overall, current education can be used as a substitute.
- Health Profile. This category represents the patient’s current health condition. This can be represented by factors such as a patient’s disease burden, mental/physicial functioning and other measures. The authors use the patient’s Total Illness Burden Index (TIBI). A paper by Willson et al (2000) uses a Comprehensive Severity Index (CSI) to measure patient illness severity over time. To determine physician functioning, the authors use a 10 item physical function scale (PFI-10) on the Short Form 36 of the TIBI. Other studies have used the Functional Independence Measure (FIM) to estimate patient physical functioning. To measure the patient’s mental health and depression, the authors used a modified version of the Center for Epidemiological Studies Depression Scale (CES-D).
- Personality Profile Measures. A patient’s personality can affect health outcomes as well. In Kaplanet al. (2010) study, the authors measure whether the patient has a passive orientation to health and health care according to the Provider Dependent Health Care Orientation (PDHCO) measure. The scale measures whether patients take a passive approach to disease management and the author claims that this measure “has been linked with poor transitions in physical functioning over time.”
- Behavioral Profile. This includes disease management skills and health habits (e.g., smoking, diet, exercise).
- Medical/Treatment Context. The category measures differences across patients in the relationship with their physician, the setting where care is given, and continuity of care.
- Life Context. The goal of this final category is to measure whether the patient experience a stressful life event or whether they have a significant social support system.
In the paper, the authors examine whether diabetes treatment improves glycemic control (i.e., HbA1c levels). Only the first three categories of patient information are included in the regressions. Unsurprisingly, the strongest predictor of differential benefits from treatment is generally the health profile. The TIBI index has the strongest effect on how treatment affects patient outcomes, but the PF-10 physical functioning disability measure is also influential in a statistically significant manner. Patient adherence to drug regimens, unsurprisingly, also improves the affect of treatment.
Although the finding that individual patient characteristics affects treatment benefits is not surprising, Kaplan and co-authors present researchers with some tools to summarize these differences for a number of different patient characteristic categories.
- Kaplan, Sherrie H. PhD, MPH; Billimek, John PhD; Sorkin, Dara H. PhD; Ngo-Metzger, Quyen MD, MPH; Greenfield, Sheldon MD “Who Can Respond to Treatment?: Identifying Patient Characteristics Related to Heterogeneity of Treatment Effects,” Medical Care: June 2010 – Volume 48 – Issue 6 – pp S9-S16.
Tags: CER, Comparative Effectiveness Research, Patient Heterogeneity
PricewaterhouseCoopers recently conducted a survey of about 700 companies to determine the latest trends in employer-provided benefits. The survey, conducted in early 2010, assessed the level of health insurance, retirement, and other benefits provided by firms from over 30 industries.
Today, I will focus on the results with respect to the health insurance. Broadly, PPOs are still the most popular plan, but high-deductible plans are gaining ground. Further, there is a general trend towards increased cost sharing both for in-network and out of network care. An increasing number of firms also utilize wellness (76%) and disease management (68%) programs for their employees.
The table provides a more detailed summary of the trends in employer-provided benefits between 2008 and 2010.
The latest edition of the health wonk review is up at the Disease Management Care Blog. It even has an air travel theme for the jet setters among us.
Tags: HWR
The Healthcare Economist has previously reviewed different forms of Accountable Care Organizations (ACOs). Implementing ACOs in practice, however, may prove more difficult. How could the government or private insurers incentivize providers to provide integrated care? How can they incentivize providers do perform fewer services and, thus, make less money?
An article by Shortell and Casalino (2010) takes a tiered approach. ACOs would be classified into one of 3 tiers.
- Level I: ACOs bear no financial risk, but simply receive a share of savings and bonuses for hitting quality targets. This payment structure is similar to the Medicare Group Practice Demonstration. The organizations would have to establish a legal practice entity that would be able to provide performance measures. A minimum number of PCPs would also be required.
- Level II: ACOs in this level would be eligible for a larger share of savings gains, but would also be liable to penalties if costs rise above predetermined targets. Level II ACO’s would also be required to meet a wider scope of performance metrics. Further, these ACOs would receive more bundled payments and episode of care payments as well.
- Level III: The most “advanced” ACOs could be paid through full or partial capitation. These ACOs would also be required to have public reporting of a comprehensive set of performance measures and electronic health records.
The Level I and Level II ACOs likely will only have a marginal impact on physician behavior. For instance, assume that the cost per patient is $100 per year and ACOs in level II must pay 25% of any additional costs above $100. If an ACO has $200 of charges per patient, even after the $25 penalty, they will still make $175, which is greater than what they would have made by reducing volume. [This example does not take into account that performing more services also increases the cost for the physician.]
There is little doubt that paying physicians in Level III ACOs via capitation will decrease cost. Physicians will not have an incentive to provide excessive services since they will not receive additional revenue. The question is will quality remain the same? Policymakers may claim that it will because of all the necessary quality metrics. Nevertheless, most healthcare outcomes are difficult to measure and measuring some outcomes may cause physicians to transfer their efforts from improving more important, unmeasured outcomes to less important but measured outcomes.
Currently, Medicare is a fee-for-service system with the exception of the Medicare Advantage system. Instituting a capitation-based system may meet much resistance from physician and hospital organizations such as the AMA and AHA. It might be easier to simply transfer all Medicare beneficiaries to managed care rather than create ACOs. Regardless, implementing ACOs in the U.S. will not be as easy as it seems.
- Shortell SM and Casalino LP (2010) “Implementing Qualifications Criteria and Technical Assistance for Accountable Care Organizations” JAMA. 2010;303(17):1747-1748.
- A more detailed version is available here.
Tags: Accountable Care Organization, ACO
What is power? Merriam Webster defines power as the “possession of control, authority, or influence over others.” The power I will talk about today, however, is statistical power. Statistical power measures the ability of a statistical test to determine whether the null hypothesis is false. For instance, in the U.S. judicial system, the null hypothesis is that the defendant is innocent. Trials that can more accurately determine when the defendant is in fact guilty have more power.
In statistics, there are two types of errors: Type I and Type II. The probability of a Type II error, a false negative, is represented by the symbol β. Thus, the probability of correctly rejecting the null (i.e., the power) is 1-β.
The larger the magnitude of the hypothesized effect, the higher the power. It is much easier to detect a large effect than a small effect. Also, as the size of the sample increases, so does a test’s statistical power.
The more variation that exists in the data, however, the lower the power. If there is a lot of variation in the data, it is difficult to determine if null hypothesis is false or if observing a phenomenon that contradicts the null is simply due to the excessive amount of variability in the data. On the hand, if the variability (i.e., standard deviation) is low, then one can generally conclude that that the null hypothesis is false, since the low variability indicates that the anomaly is not caused by normal variation in the data.
Tags: Econometrics, Statistics
In Virginia, there are over one million people age 60 and older and over 90,000 Virginians age 85 and older. These figures will only grow in the upcoming decades. Thus will put increasing strain on public programs and will require service providers to reorient medical care toward providing continued, high-quality long term care services. Long term care is of growing importance to health care sector. Although the aged and disabled populations make up 30% of Virginia’s Medicaid population, these individuals account for 70% of the state’s $4 billion Medicaid budget.
Yet providing long term care to those in need is a confusing a bureaucratic process. For instance, in Virgina, there are 6 agencies that provide long term care services:
Tags: Long Term Care, LTC, No Wrong Door
Recent comments