1. FINDINGS FROM ATTORNEY GENERAL’S EXAMINATIONS OF HEALTH CARE COST TRENDS AND COST DRIVERS PURSUANT TO G.L. c. 118G, § 6½(b)
OFFICE OF ATTORNEY GENERAL MARTHA COAKLEY
ONE ASHBURTON PLACE • BOSTON, MA
January 5, 2012

2. EXAMINATION APPROACH
We issued dozens of subpoenas for data, documents, and testimony to major health plans and many different types of providers.
We conducted dozens of interviews and meetings with providers, insurers, health care experts, consumer advocates, employers, and other key stakeholders.
We engaged experts with extensive experience in the Massachusetts health care market.
We greatly appreciate the courtesy and cooperation of payers and providers who provided information for these examinations.
-By way of brief background, c. 118G provides General Coakley with the power to intervene in the DHCFP annual hearings re: cost trends and cost drivers. Pursuant to that authority, General Coakley has completed two examinations in 2010 and 201 of health care cost trends and cost drivers.
Her approach was to gather information from insurers and providers in order to hold up a mirror to the Massachusetts health care market.
We issued subpoenas for testimony and documents, and obtained confidential data from insurers and providers in MA to help inform the health care system.
Today I’m only going to cover a few highlights from our findings. Encourage you to read both our reports to see more details about our approach and findings.

3. MEASURING HEALTH CARE COSTS
CONFIDENTIAL DRAFT - FOR DISCUSSION PURPOSES ONLY
9/15/2009
MEASURING HEALTH CARE COSTS
TOTAL MEDICAL EXPENSES (TME): The total cost of all the care that a patient receives, including the payments by the health plan for the care of the patient, and any copayment or deductible for which the patient is responsible. TME reflects both price of services and volume of services.
PRICE: The contractually negotiated amount that an insurance company pays a health care provider for providing health care services; we reviewed relative price information, which shows the prices paid by health plans to providers for all services in aggregate as compared to other providers in the health plan network.
To set the stage for explaining our analysis, I want to review a few key metrics and definitions that played an important roll in our examination so that we’re all working from a common baseline.
We focused on two important measures of health care costs.
“TME” is the total cost of medical care associated with an individual patient, regardless of where the patient incurred those costs. It includes both the price of the service, and the volume, or number of times the service is utilized. This includes not only payment for medical services, but also other payments to providers such as quality incentive payments and payments to support infrastructure.
TME can be adjusted by health status to control for differences in health and demographics between two patient populations.
We believe that TME is the single best measure of provider efficiency, since it reflects both the price of the service and the volume of services provided.
We also reviewed price information. “Price” is the negotiated amount the insurance company pays a health care provider for the health care services the provider delivers.
Price is important because it tells us the amount that one provider gets paid compared to another for providing the same service. Rather than negotiation prices on a service by service basis, insurers and providers negotiate a price across all services, and that is the information that we reviewed. 3

4. 2010 and 2011 EXAMINATION HIGHLIGHTS
Prices paid by health insurers to hospitals and physician groups vary significantly.
Variations in prices are not adequately explained by value-based differences in the services provided.
Variations in prices are correlated to provider and insurer market leverage.
Global budgets vary significantly and globally paid providers do not have consistently lower TME.
Variations in prices impact the increase in overall health care costs.
Here are a few highlights from our examinations that I will go over briefly today:
1) Prices paid by commercial insurers to providers vary significantly.
2) Variations in prices are not adequately explained by value-based differences.
3) Variations in prices are correlated to relative leverage of providers and insurers.
4) Global budgets, or risk budgets, vary significantly and globally paid providers do not have consistently lower TME.
5) Variations in prices impact the increase in overall health care costs.
I’ll review each of these highlighted findings with a some additional detail.

5. PRICES PAID TO PROVIDERS VARY SIGNIFICANTLY
To understand better how the health care market in MA functions, we first explored the very basic question of how insurers reimburse providers. We reviewed relative payments and prices paid by each of the 3 largest commercial health insurers in Massachusetts to physicians and hospitals within their networks in 2008 and That analysis shows that in there was significant variation in the prices paid by health insurers to physicians and hospitals in MA.
For example, this slide shows relative payments made to hospitals in one health insurer’s network. There is a 300% difference (4x) in the prices paid by the health insurer to the lowest and the highest reimbursed physician groups in the insurer’s network.
Next, we asked what could cause these large differences in prices.

6. DIFFERENCES IN PRICES ARE NOT ADEQUATELY EXPLAINED BY
VALUE-BASED FACTORS
We looked at five factors that many in the market have cited as reasons for variations in prices:
Quality
complexity of care or sickness of patients treated
teaching status
proportion of government insured patients, and
underlying costs.
We found that none of these factors adequately explains the differences in prices.
For example, this slide shows just one measure of quality that we reviewed: hospital performance across four clinical areas (heart attack, heart failure, pneumonia, surgery). In general, MA hospitals perform within a tight range of each other, and 52 of the 61 MA hospitals (or 85%) exceed the national average performance. We saw similar results when we examined other measures of hospital and physician quality.
So we didn’t find that any of those 5 factors could adequately explain variation in prices paid to providers. However, we did find that the relative leverage of providers and insurers in contract negotiations is related to the prices paid to providers. 6

7. HIGHER PRICES ARE EXPLAINED BY MARKET LEVERAGE
For providers, leverage typically results from variables such as: size, geographic location, “brand name,” and/or niche or specialty service lines offered. For insurers, leverage tends to result from size and penetration into a geographic area.
This slide shows an example of one way in which we reviewed leverage and its relation to price. This shows 6 adult AMCs in Boston, and compares their size – expressed through both mm and through system revenue – to their relative price. We see here that the larger the hospital, the higher its relative price.
The DHCFP, the Price Reform Commission, and others have echoed the AG’s finding that significant variation in commercial prices paid to providers is not value-based.

8. VARIATIONS IN PRICES PAID TO PROVIDERS EXIST IN GLOBAL RISK BUDGETS AS WELL AS IN FEE-FOR-SERVICE ARRANGEMENTS
We found wide variations in the health status adjusted global payments made by health plans to at-risk providers.
For example, in one health plan’s network in 2009, one globally paid provider had a health status adjusted budget of approximately $428 per member, per month, while another had a health status adjusted budget of only $276 per member per month.
The issue of price variation isn’t a FFS issue; this issue exists throughout all types of payment methodologies, because all commercial contracts involve negotiations between insurers and providers with varying amounts of leverage. Therefore, variations in prices paid by insurers to providers exist regardless of the payment methodology. We saw significant variation in FFS payments, in supplemental payments, in global payments, etc.
To focus on the global payments for a moment - as you all know, some providers are paid through global risk payments on their HMO patients, instead of through FFS payments. In these contracts, providers are given a pmpm budget target. For example, $300 pmpm budget for the total care of each HMO patient. Our examination found wide variations in these types of payments structures as well.
In one health plan’s network in 2009, one globally paid provider had a health status adjusted budget of approximately $428 per member, per month, while another had a budget of only $276 per member per month to care for patients of comparable age and health.
These wide differences in payment, whether fee for service or global payments, means higher paid providers have more to spend on many things: on building new facilities, on expanding operations into more communities; on technology for patient care; on recruiting and salaries; on advertising and amenities.
It also means that higher paid providers don’t have the same incentives to perform efficiently and to control costs that providers with lower budgets have. [At $300 actual, $128 surplus vs. $24 deficit for the same performance]
As a result, although global risk budgets are meant to reward providers for efficiency, rather than for volume, as you’ll see on this next slide, we do not see that global risk providers have consistently lower TME than FFS providers.

9. GLOBALLY PAID PROVIDERS DO NOT HAVE CONSISTENTLY LOWER
TOTAL MEDICAL EXPENSES
This slide shows the TME of providers in one health insurer’s network. Global risk providers are shown in red, and FFS providers are shown in blue.
As you can see, paying these providers on a global risk basis has not resulted in lower TME as compared to FFS providers. That includes providers who have been in global risk contracts for a long period of time. There are some providers in MA who have been at risk for decades; here, we indicate those providers who have been at risk for 5 or more years with the yellow bubble.
We reviewed all available TME information for the three largest insurers from and did not find that globally paid providers have consistently lower TME.
Again, one potential reason for this is that there are significant variations in prices paid by insurers to providers in both FFS and global payment arrangements.
So we know that there are wide variations in the prices paid by insurance companies to hospitals and physicians in MA, regardless of the payment methodology, that are not adequately explained by differences in provider value. But why do we care?

10. PRICE INCREASES CAUSED THE MAJORITY OF THE INCREASES IN HEALTH CARE COSTS IN THE LAST SIX YEARS
We care for a number of reasons. First, because increases in prices paid to providers caused most of the increases in health care costs during the past few years.
Each year, the cost of health care continues to grow faster than inflation, wages, and other economic indicators. Two main reasons why our costs could be growing are price and utilization. We could be consuming ever more services, or we could be paying hospitals ever higher prices.
This slide shows the increase in health care costs for one insurer for each year since 2005; you can see that for this major insurer, more than half of the increase in health care costs is due to paying providers more (shown in blue), versus about 25% from consuming more services (shown in green). This suggests to us that if we want to moderate the trend in escalating health care costs, we must address provider price increases.
-Price disparities also contribute to differences in total medical spending. Recall that total medical spending is the sum of all of the cost of care for a patient over a given year, or month. This metric can be health status adjusted to control for factors like age and health status, so we are comparing the level of medical spending for “comparable” patients. So for patients with higher TME, the additional amount that’s being spent on their care is not due to age or health.
Instead, three factors explain differences in health status adjusted TME. The patient with higher TME may be getting more health care services (utilization), the patient may be using higher-priced providers more often than the patient with lower TME, and the patient may be using more expensive treatments (service mix). We examined whether there are differences in health status adjusted TME across MA, and there are. These results are shown on the next slide.

11. TOTAL MEDICAL SPENDING IS HIGHER FOR THE CARE OF COMMERCIAL PATIENTS FROM HIGHER-INCOME COMMUNITIES
- For each zip code in MA, we compared the TME of patients to the avg income.
We found that differences in TME correlate with income. Looking at the bar furthest to the left, you see that for the 135 zip codes with low total medical expenses, more than half are also low-income zip codes. Among the zip codes that have the highest medical spending, shown in the bar furthest to the right, almost 60% are also the zip codes with the highest levels of income. And remember, that TME isn’t higher because they are sicker or older.
If the increased cost of caring for patients with high TME is spread throughout a larger risk pool, as may happen in the small group market, or in a single large employer group, those with lower TME may be subsidizing the care of those with higher TME in the same risk pool.
Now AAG Johnson is going to highlight some of the AG’s recommendations based on these findings.

12. TIERED AND LIMITED NETWORK PRODUCTS HAVE INCREASED CONSUMER ENGAGEMENT IN VALUE-BASED PURCHASING
Health insurance products that do not differentiate among providers based on value do not give consumers an incentive to seek out more efficient providers, because consumers are not rewarded with the cost savings associated with that choice.
As a result: (1) consumers are de-sensitized from value-based purchasing decisions and (2) providers are not rewarded for competing on value.
There have been recent developments in tiered and limited network products; these types of innovative products should be encouraged.
Another way that increased transparency may help increase healthy market functioning is through insurance product design.
We reviewed various health care insurance products available to consumers in the MA market and found that, typically, insurance products do not reward consumers with cost savings when they use high quality, low cost providers. Similarly, health care providers aren’t rewarded with more patient volume when they offer high quality, low cost care.
Limited and tiered network products encourage value-based purchasing by rewarding consumers who choose more efficient providers, and by shifting patient volume to high quality, low cost providers.
The market has already started to move towards these type of products, and we should continue to encourage innovative product design that rewards value-based purchasing decisions.
We reviewed:
Hospital Choice Cost-Share, which differentiates copayments by as much as $1,000, goes further than other tiered products in the market in the level of point-of-service savings available to consumers for choosing a high-value hospital. Even so, it should be apparent to employers and consumers from Section A.1 above that the differences in payments received by high-cost versus high-value providers for quality care exceeds $1,000.
Information that shows savings in medical costs nearly twice as large as the premiums savings passed on to consumers (for Fallon, its Direct Care product has TME 22.1% lower than its Select Care product in Central MA; in Boston, the TME differential is 31%, and in Eastern MA, the differential is 34%).

13. DELIVERING VALUE-BASED HEALTH CARE
1) How can we best improve market function?
2) How can we improve care coordination?
Market Function
The health care market, like any competitive market, must be responsive to the purchaser – employers and consumers – who must have an incentive and the information necessary to make more efficient and effective use of health care. We must change how health coverage is sold to maintain pooling of the risk associated with health conditions and accidents, but eliminate the current pooling of costs that result from inefficient use of health care.
We have found significant dysfunction in the market place:
Unwarranted price disparities – even in global payment/risk arrangements
Have and have not providers
Lack of incentives for consumers to choose value
Disparate medical spending by income/Underwriting inefficiency
Lack of transparent information – made great strides with c. 288 on transparency of price and quality
Market is not self- correcting – we need to find ways to improve functioning of market for consumers
Care Coordination
There are significant opportunities for providers and health insurers to improve care coordination. Care coordination can be best achieved when the patient, the provider, and the health insurer agree on how care coordination will function. Product design should reward patients with lower rates when they enroll in plans that allow for care coordination.
Coordination is not automatic with provider consolidation or global payment – there are significant challenges:
Capacity and Information
Cost of infrastructure for population management - expertise
HMO plans – designed around PCPs

14. MOVING FORWARD ON COST CONTAINMENT
Promote tiered and limited network products to increase value- based purchasing decisions.
Reduce health care price distortions through temporary statutory restrictions until tiered and limited network products and commercial market transparency can improve market function.
Encourage consumers to select a primary care provider who can assist consumers in coordinating care based on each consumer’s needs and best interests.
Promote coordination of patient care through primary care providers by recognizing the need to improve funding of care coordination, including the infrastructure necessary to coordinate care, and by giving providers timely access to relevant patient data regardless of their size or payment methodology.

15. MOVING FORWARD ON COST CONTAINMENT
5. Consider steps to improve the use of the all payer claims database (“APCD”) by: (i) developing reports for providers and the public to guide development of patient care coordination improvements and system accountability, and (ii) increasing the standardization of claim level submissions by reducing differences in how payers report payment level information.
6. Develop appropriate regulations, solvency standards, and oversight for providers who contract to manage the risk of insured and self-insured populations.

16. FINDINGS FROM ATTORNEY GENERAL’S EXAMINATIONS OF HEALTH CARE COST TRENDS AND COST DRIVERS PURSUANT TO G.L. c. 118G, § 6½(b)
OFFICE OF ATTORNEY GENERAL MARTHA COAKLEY
ONE ASHBURTON PLACE • BOSTON, MA
January 5, 2012