Got a Cough? Open Data may be one Cure

Open Health Data in New Hampshire

We’ve been plugging away with our open data project.  Open data is data that is free to the public and is generally released by federal and state governments.  NH has probably the most progressive policy in the nation when it comes to providing consumers access to health care claims data; it was literally a one-page application.  The consumers of the state should frankly use this to their advantage, whether its on our site or a different one; we truly believe that a lack of education and engagement translates into consumers leaving millions, if not billions of dollars on the table.  After all, wage growth has been anemic and health costs continue their inexorable increase.  How often do we see laws passed that truly potentially save consumers money??

Along with the federal data we are ‘refining’ and publishing, we also have some wonderful cost data from NH.  This post is intended to illustrate how consumers can leverage our site as an educational resource.

Deep Dive:  Chest X-Ray

So let’s take a quick look at what will hopefully be the first of many deep dives into the data.  We were assessing one particular service recently:  a chest X-ray, from 8 random hospitals in the state.  First, we zeroed in on one hospital’s claims data from 2013.  We examined 3,725 claims.  Average total costs (out-of-pocket for the consumer and amount paid by insurance company) was $144.55 while the average cost paid out-of-pocket (deductibles, co-pays, co-insurance) was $63.95.  The first chart below plots all 3,725 claims.  Now the average reader probably does not get too excited by the scatter plot that looks like something my two year-old did as an afterthought.  But to those who have toiled in the depths of the wide-open sea of open data, the first observation is clear:  The darker horizontal lines likely represent varying negotiated rates across insurance plans – so in examining the two darkest horizontal lines, it appears that one plan negotiated a much lower rate than the other (average total of $200 vs. $100).  In total, it appears there are around 8 plans that comprise the majority of the claims.

Who cares?  Well, it goes to show that insurance plans negotiate varying rates, and to the extent that competition is healthy, aggregate costs should be kept at bay.  But what if there is a large merger in a small state like we’ve seen in NH with the impending merger between Anthem and Cigna (a cool $54 billion)?  Well, as consumers, we sure better be monitoring the aggregate price levels in the state.  Studies have been issued citing NH as the state that will potentially face the most adverse impact from the merger.  Clearly this may not come to fruition and prices may fall, if you believe managed care companies are altruistic…


Chest x-Ray – one NH hospital (2013)


So we were pretty excited about our finger paint discovery.  But we then asked “Why are people getting chest x-rays?”  We looked at the most common diagnoses for the chest x-ray and we must be honest, we had a VERY good laugh when we discovered most chest x-rays were performed due to a ‘cough’.  The average out-of-pocket cost was $62.52 for this particular diagnoses and the average total paid was $130.61, meaning consumers on average paid about 48% of the cost of the visit.  In total, consumers paid $35,451 in out-of-pocket costs at this one hospital in 2013.  This diagnoses was made for 567 patients.

$74,059 was spent between consumers and payers for having a cough, at one hospital, in one year.  In fact, when we look at the vast amount of money spent in healthcare, this actually seems almost ‘cute’.  But if we scaled these numbers to all hospitals in the state and then to all hospitals in the country, we eventually move the needle.  We actually did just that – consumers paid $449,750 out of pocket, total cost was $1,088,768 for the state (again, only looking at institutional providers like hospitals).  So gross that up on a population basis and for the country that would amount to $115 million on out-of-pocket and $278 million total.  This assumes a direct linear relationship between NH population diagnoses and procedural recommendation, of course.


The plot below looks at out-of-pocket and Total costs for those with a cough that received the chest x-Ray.



There are all sorts of things we could consider from here.   Like age, for example.  Before we let ourselves get on the soapbox too much and rant about healthcare costs, we asked ‘maybe these are young kids?  We have children, so of course we would want the doctor to take all precautions’ (though we would be equally foolish for thinking like this for a cough, but we do tend to go overboard for our kids).  So we took a look at the age distribution of those getting the chest x-ray.  Alas, we could stop pulling on the heart strings after all; it’s wasn’t like the average age skewed towards infants and toddlers:



Turns out the average age for our scenario was 39.2 years.  Also, the male/female breakdown was about 49/51 (not plotted).

So that is across plans, for one hospital.  But what about some more relevant analysis that is more actionable?  Well, let’s take a sampling of 8 hospitals throughout the state and isolate the same procedure (chest x-ray) that was linked to the same primary diagnoses:  a cough.  We also isolated just HMO claims, so its more relevant for consumers.

Here we see that the Elliot Hospital by far performed the most of these cases, well over 1,000.  Exeter Hospital, Concord Hospital, Mary Hitchcock, and Wentworth accounted for the other volume.



We can then examine the distribution of average total costs across all 8 hospitals.  Total costs include how much the insurance company paid as well as the consumer in out-of-pocket.


So now we can see, by way of a visual analysis, how the total costs in 2013 varied across our sampling of 8 hospitals.  The chart above is referred to as a box-plot chart (or box and whisker chart).  The black horizontal line in each box represents the median observation, while the upper end of the white boxes marks the 75th percentile while the bottom captures the 25th percentile of total amount paid.

What can we make of this?  Well, for virtually every hospital, the interquartile ranges (25% to the 75% of total paid) are relatively wide.  In other words, there isn’t just “one price”.  The reason for this may have to do with modifiers, or adjustments to the bills submitted by the facilities that could make the procedure more or less expensive.  Huggins & Monadnock have relatively tight ranges.  This is likely the result of each hospital having a lower count of chest x-rays, thus a lower probability of experiencing cases where modifiers were necessary.  But a more subtle observation is that these hospitals also tend to be the more expensive hospitals.  One would expect to pay more at those facilities that have more experience performing and analyzing x-rays, but as we all know, healthcare is a bit challenging when it comes to defining “value”.

How about cash cost to the consumer?  Below we’ve plotted out-of-pocket costs as defined by deductible, coinsurance, and co-pays (all available data points with our open data).  Valley Regional and Wentworth both have low medians amounts.  This is most likely because the consumers who went there have lower deductibles, etc., based on the particular plan designs in that geography.  (One day we hope to get more detail about the actual plan designs; open data is progressing, but it still has room for improvement.)  It appears that Monadnock patients experienced the highest median out-of-pocket cost in this example.  Hitchcock, despite it being a teaching hospital, which are generally pricier, looks quite competitive.  Concord, Elliot, and Exeter patients on average spend around the same amount in out-of-pocket costs.


OK, that’s enough numbers and charts.  What does it all mean?  Well, high deductible plans are on the rise.  Consumers in NH already have a leg up on the other states – by a MILE.  The open data can be brought to life as in this example.  Our next post will examine the cost differences between different facilities.  The hospital in this post actually looks pretty good compared to others in the state when it comes to cost.

Please let us know if there are other services of interest.

In the meantime, maybe we ought to get into the x-ray business if our open data project does not materialize.

Comments are closed.