According to the latest census, nearly 85% of Americans have some form of health insurance. We know that the implementation of Obamacare will result in an increase in this number, but we have no guarantee that the additional costs associated with better access to care will be counteracted by improvements in overall health.

Of course, we are hopeful that health care reform will help (and, at least not hurt), but we believe a simpler approach to cost containment would be to bring the most American of pastimes to health care: shopping. Today with the intricacies of, and variancies in, medical billing, patients simply have no insight into the actual costs of care. And the information that consumers can access about the quality of available physicians and facilities is woefully inadequate.

In every other market of the U.S. economy, consumers have the ability to compare prices—but not in health care. Need a colonoscopy? Well, it turns out the cost of that single procedure can vary by a factor of four or more right in your own neighborhood. The athenahealth point of view is that the ability to shop begins with information, and that pricing transparency is a great place to start. (Side note: The idea of transparency has recently been getting increased exposure in the national press).

Health care prices—referred to as “contracted rates” or “allowed amounts”—are largely determined by contractual agreements between insurance companies and health care providers. Let’s put a fine point on this: the charges established by health care providers may have little bearing on what the insurance companies will actually pay. In many cases, providers receive contracted rates under their fee schedule in exchange for participating in the company’s plan. So what are the “real” rates?

Actual commercial contracted rates are some of the best kept secrets in health care, and while we guard individual provider data with our lives, our massive database allows us to compare and benchmark contracted rates across 40,000 providers, in all 50 states. We can access medical billing rates for commercial payers, Medicare and Medicaid, and have done so: the result can be explored in a new app we’ve developed called CodeView.

CodeView displays the maximum, minimum and average dollar amounts that insurers pay providers. Why does this matter? Because having access to accurate price information is essential to decrease cost. Furthermore, as any free marketer knows, having visibility to the differences in prices is key (we won’t get into the issue of private vs. single-payer systems). Those differences, the gaps, are what CodeView really attempts to highlight.

CodeView Data – The Big Picture

On average for ambulatory / physician services, across all levels of procedure complexity, we generally see an expected pattern for pay: Medicaid rates are less than Medicare rates, and Medicare rates are less than Commercial Payer rates. As seen in this first chart, the median cost for a procedure increases as the procedure’s complexity increases:

Note: Complexity is measured numerically here by Medicare’s total RVU, or Relative Value Unit, which typically serves as a basis for how much a provider should be reimbursed for a procedure or service, and incorporates complexity factors.

It is generally understood that commercial payers have a much wider variation in contracted medical billing rates than Medicare for a given procedure. What surprised us a bit with this data is that some providers are actually paid less by commercial insurers than by Medicare, for the same procedures. You can see both of these phenomena in CodeView for procedures ranging from office visits to complex cardiology imaging procedures. Here’s how the 25th percentile to 75th percentile ranges compare for Medicare and commercial payer rates, for a handful of procedures:

It’s hard to imagine the price of anything outside of health care varying so broadly. If you rename a ham sandwich a Croque Monsieur, maybe you can go from $7 to $15… but the variability on display here says one thing: opportunities to decrease health care expenditures could benefit from a little price transparency. Say what you will about governments as payers, but they do drive a better bargain.

Using CodeView

To dig further into CodeView data, let’s compare rates for a cardiology imaging procedure known as a Myocardial Perfusion SPECT. (Note: If you’d like to follow along using CodeView, you will have to change the specialty you’re viewing to ‘Cardiology,’ or use this direct link. According to our data, Medicare says a doctor is typically owed between $489 and $531 for this procedure, depending on the region of the country where it was performed. Medicaid rates vary by state, but we typically see that they are 50-90% of Medicare rates; this particular procedure falls on the lower end of that range, as Medicaid payers typically allow between $206 and $294 for this procedure.

The commercial market is where we really start seeing a huge variation. While the 25th percentile seen here (at $407) is below that of Medicare, the median rate from commercial payers for this procedure was $520, toward the upper end for Medicare. And at the 75th percentile for commercial payers’ reimbursement, caregivers received $686 for this procedure, illustrating that many private contracts are paying far more than Medicare.

CodeView: Bringing More Transparency to Healthcare Prices

Here’s the simple, obvious summary: Health care spending across the country is rising. Patients’ responsibility is growing and doctors are on the hook to collect that rising share of revenue. Wouldn’t it be great if patients, doctors—anyone—knew how much a service actually cost? Surfacing this information is just another step in demonstrating the power of athenahealth’s data in bringing transparency and openness to health care. Explore CodeView and you’ll find that we’ve highlighted the most common procedures for several specialties, across all regions of the country.

You can access the CodeView app here. Take a look. What do you see that surprises you? What data do you want to discuss with us?

Two weeks ago, our Cloud Analytics team presented some flu season trends right here in the blog, and discovered that others had great interest in our findings and figures. Now that the flu is past its peak, let’s dig into some more data to take a broader look at flu vaccinations: We’ll compare flu rates for patients who have been vaccinated and those who haven’t.

We’ll limit our discussion to pediatrics, given that most children receive their vaccinations in the doctor’s office (see Figure 1). By doing so, this means that our data on vaccination rates best reflect national rates for children. We can see this in the fact that our vaccination rate for children via athenaNet data—37%—is similar to the CDC’s estimate of 40% (Figure 2). We can also see, not surprisingly, that we underestimate rates for adults since they tend to get their shots from a variety of alternative locations.

Source for Figure 1 and Figure 2: CDC

So, how did vaccinated children fare this season, compared to unvaccinated children? 2.5% of children who were unvaccinated were later diagnosed with the flu, while only 0.9% of vaccinated children were diagnosed with the flu (Figure 3).

This means that patients who get the flu vaccine are 63% less likely to get the flu (0.9% is 63% less than 2.5%). This 63% figure is what is commonly referred to as the “effectiveness” of the vaccine. Now, 2.5% and 0.9% may seem like low percentages, but in a population of millions, those translate into large numbers. And those large numbers are why the flu is a public health issue.

We’ll continue to dig into the data on our cloud-based network to determine trends, analyze results and report back to you. Want to know more about the pool of information we refer to? Check out our first Cloud Analytics blog post from November 2012 for details.

It’s that time of year again: flu season. We’ve been reading about this year’s flu in the news, and (as a group of data nerds) tracking it online through CDC reporting and Google’s Flu Trends.

We thought we’d see what we could add to the conversation. So we dug into our athenaNet data set of 40 million patient records to see how this flu season is developing.

Here is what we found so far:

This flu season is severe, but may have already peaked. By our metric, you’d have to go back to the 2009-2010 flu season (the second wave of H1N1 “swine flu”) to find a season that reached the intensity levels we’re seeing right now. However, our data is showing a glimmer of hope: Two weeks ago, flu incidence decreased for the first time since the start of the 2012-2013 flu season, and last week it stayed level. Our data is right in line with other sources right up until this past week (CDC numbers come out on Friday), so we’re confident these patterns are an excellent real-time “flu view” (sorry, CDC)!

Last week’s flu uptick is driven by regional variation. The South, Midwest and Northeast regions of the country seem to have hit their peak levels already, but the flu has continued to ramp up in the West over the past few weeks. Based on flu patterns from the past several years, our guess is that the West reached its peak levels last week, though there have been flu seasons with a second strong peak (2001-2002, for example).

The flu has hit the West. At the end of December, most of the South and Northeast were starting to experience intense flu rates. Over the two weeks that followed, this year’s flu had taken over almost the entire country: 

The news coverage and vaccination campaigns seem to be working: more people are getting vaccinated. In October, we saw the usual peak in vaccinations this flu season. However, following the January 4 CDC report that recommended vaccination in light of the flu continuing to spread across the country, we saw an uptick in vaccination rates last week, as seen in the accompanying chart. Hopefully, these preventative steps will get us to the end of this flu season more quickly.

We’ll continue monitoring this situation over the coming weeks, and will post updated findings as they come in. Hopefully, our next post will continue to show the flu abating across the country. Until then, join the growing number of people who are getting vaccinated (find a vaccine near you) if you haven’t already, wash your hands frequently, and stay healthy!

Welcome to our second installment of Cloud Analytics! Today, we’re delving into patient responsibility with regard to deductibles, co-pay and co-insurance – for patients with private insurance.

Across the country, total patient responsibility is on the rise, increasing financial stress not only for patients, but also for practices that now have to collect more money from patients.

How much has the patient payment burden been increasing? Calculated as a percentage of allowables (the contracted amount that insurers agree to pay providers, rather than the amount the providers charge), patient responsibility has increased in each of the last two years, rising from 18.6% in 2009 to 19.9% in 2011:

This points to patient responsibility increasing at an even higher rate than annual increases in total allowables.
What’s driving this increase? Deductibles, a fact that becomes apparent when we break out the data by types of patient responsibility. Between 2009 and 2011, deductibles have increased considerably, while co-pays have actually decreased and co-insurance has stayed flat:

Some believe that increasing deductibles are a good thing, acting as an incentive for consumers to take greater personal responsibility. Some think it’s a negative, penalizing those who are chronically ill and/or have lower incomes. Either way, rising deductibles affect both patients and practices.
Over the last three years, unpaid deductibles (deductibles not paid within a year), as a percentage of allowables, have increased. In contrast, unpaid co-pays have stayed constant:

In other words, practices are having more difficulty collecting what they are owed by the patient. This is hardly news; the more money patients owe, the harder it is to collect. A struggling economy does not help either. What is new, however, are the trends toward greater participation in high deductible plans, whether by experimentation or employer fiat.

As for co-pays, why are they decreasing? We’re not policy or plan design experts (if you are, we’d love to hear from you!), but we think the Affordable Care Act (ACA) may be an influencing factor. A key part of the ACA is a provision that brings the co-pay for preventative care visits to $0 for new plans, starting in September of 2010. And with the ACA extending subsidies to many new patients and allowing them to get lower-end individual plans from private insurance, many of these patients will end up in high-deductible plans over the next few years. Measuring the impact of the ACA is new territory for us so we’ll explore related issues in future posts.

In the meantime, let’s look at the most recent data we have: Is 2012 offering a reprieve from the rise in patient responsibility? It doesn’t appear so. Looking at the first three quarters, for 2009 through 2012, we see similar patterns (increasing deductibles, decreasing co-pays):

Note that percentages for deductibles in Figure 4 (Q1-Q3) are higher than in Figure 2 (full year), because payments towards deductibles are higher towards the beginning of the year. This is due to deductibles becoming reset every year, usually at the beginning of the calendar year, meaning practices have an uneven revenue stream beyond other seasonal variations:

If we break these numbers down by region and specialty, we also see rising deductibles. If you’re interested in drilling down further, you can see these numbers in our recent infographic, also published in Medscape. It appears this trend in rising patient responsibility is real, across regions and specialties.

The bottom line is that this is an extra burden for medical practices. Can patient collection be added to what’s needed to meet Meaningful Use requirements, as well as the additional responsibilities of being a Patient-Centered Medical Home (PCMH) or part of an Accountable Care Organization (ACO)? How will things change by region or specialty?

While we think about these questions, let us know what you think!

In our next installment, we’ll explore reimbursement differences between specialties. Thanks for reading!

Welcome to the inaugural Cloud Analytics blog post, brought to you by the athenaNet Intelligence team. We are a crew of data geeks who firmly believe athenahealth’s greatest asset—other than our clients and coworkers—is our data. While our colleagues are busily building the country’s first health information backbone, our team digs through and mines our massive and growing cloud-based database with an eye toward improving physician and practice performance and nudging the health care industry toward a better future.

Just how big a data set are we talking?

• 40 million patients
• 27,000 physicians on the network
• 11,000 non-physician practitioners
• More than 12 years of data
• Input from clients in 48 states

With that kind of scale, we can slice and dice and then share out knowledge with a thoroughness and intelligence that only a cloud-based service can provide.

In future posts, we’ll discuss important or simply interesting themes and trends, each offering a data-driven look at a facet of health care in the United States. This is a tremendous opportunity for us to share our point of view, in real time, on the industry with the greatest impact on our economy and society.

So, to kick it off, here’s our first data insight of the series, a demographic profile of our physician and patient populations.

What does our physician population look like? Very similar to national averages!

Compared to the national population, athenahealth physicians are slightly older (82% over 40 for athenahealth vs. 74% over 40 for U.S.), slightly more male, and very slightly less specialized. Here, we exclude residents and pinch-hitting docs (locum tenens) given year-to-year fluctuations of these two groups. In future posts, depending on the research topic, data for all physicians and non-physician practitioners may be included.

Geographically, we are dispersed across all the lower 48 states, with physicians representing 77 specialties. This map shows athenahealth’s primary care physician (PCP) percentage (the “Specialty” row in the first chart) for each state. Here, you can see that states in the Midwest and West have a higher concentration of PCPs than in the East. In absolute terms, we are well represented in states with large metropolitan populations such as California, Texas, Pennsylvania and Florida.

What does our patient population look like? Again, similar to national averages.

Compared to the national population, athenahealth patients skew older and more female. They are also more likely to be insured, based on patients with claims reimbursed between August 2010 and July 2012. This is consistent with the observation that females, older patients, and patients with insurance are more likely to seek care.

A geographical view of the percentage of patients in each state who are insured shows variation across different areas of the country. As of November 2012, our data draws from approximately 29 million patients with encounters in the last 24 months, or ~9% of the US population. This large footprint forms the core of our research, enabling us to provide a robust and representative view of population health and care delivery patterns.

So that’s a baseline view of our data. What do you think? We already have the data lined up for subsequent posts, on a wide array of topics, but we’d really like to hear from you. What kind of data and trends would you like us to examine more closely? Drop a comment to let us know.

As a tee-up to our next post, here’s a preview of out-of-pocket costs data that we’ll be exploring next. More specifically, we’ll look into deductibles as a percent of what insurers agree to pay providers, or the “contracted rate.”

What does it look like for 2010-2011? Do these trends portend the impact of health care reform, even before its enactment in 2014? Stay tuned!