Emergency Department Visits Down For First Time Since Health Reform, Survey Finds

One of the key goals of health reform here in Massachusetts was to improve people’s health through prevention and primary care to the point that costly trips to the emergency room would decline. Until now, that hasn’t happened, according to several earlier reports.

But a just-released analysis, by the Blue Cross Blue Shield of Massachusetts Foundation found, for the first time, that ER visits (among non-elderly adults) have started to drop. Here, according to the report, which is also published in the current issue of the journal Health Affairs are the numbers:

Between fall 2006 and fall 2010, there were reductions in emergency department use overall (down 3.8 percentage points), frequent emergency department visits (down 1.9 percentage points), and the use of the emergency department for non-emergency conditions (down 3.8 percentage points). This is the first reduction in emergency department use among nonelderly adults in Massachusetts observed in the MHRS.

The reduced reliance on the emergency department among nonelderly adults may reflect many factors, including the increases in use of other types of health care (e.g., increases in preventive care visits, multiple doctor visits, specialist visits, and dental care) or increases in cost sharing under their health plans.

I asked John McDonough, Professor of Public Health Practice & Director of the Center for Public Health Leadership at the Harvard School of Public Health for his thoughts on the apparent drop in emergency department visits. Here’s his email response:

First, on the face of it, it’s good news to see ED use going down after so many years of no changes, and especially because that was an outcome many expected out of MA health reform. We expected this to happen, and when it did not, many were puzzled, and many used the non-drop as evidence of health reform’s failure.

Second, given the time lag, it’s not clear this most recent drop is because of MA health reform or because of other factors. I don’t know of any specific policy changes that might account for this drop at this point in time — so one would have to view a connection with some skepticism.

Third, there is a belief that cost sharing generally is going up, and that is leading folks to use fewer medical services; and the health plans in Massachusetts have confirmed that drop in usage. So it’s not that surprising to see it begin to show up in this indicator.

Finally, it seems we need more data and analysis to give us a better picture of what is really going on. Are people just not getting necessary care, or are they getting care in other places? Are people not getting medically necessary care, or are they more likely to not obtain unnecessary services? What groups in the population are dropping their use the most?

All these matter in reaching some more definitive conclusions.

I also asked health policy expert Nancy Turnbull, Associate Dean for Educational Programs at Harvard’s School of Public Health, for a quick reaction. She was even more upbeat than her colleague. She emails:

This is very good news and exactly one of the results that we hoped for from expanding health insurance: a reduction in the use of the ED for conditions that don’t need to be treated in the ED. Unnecessary use of the ED is expensive, both for individuals and the health care system, and it results in longer waits and a drain on resources that can be better used taking care of people who do need emergency care. The quality of care provided in the ED is also episodic so it’s not high quality care for people who have chronic conditions. Health insurance is a powerful means to help people get the care they need in the most appropriate setting, and this report confirms that’s happening as a result of the coverage expansions here. Great news here and a sign of what’s ahead nationally as the ACA is fully implemented.

Please follow our community rules when engaging in comment discussion on this site.
  • Peter Smulowitz

    Three huge problems with this study: 1) based on a survey of 3,000 individuals only.  Survey data has huge limitations. 2) The 2010 data is one point in time compared to several years of consistent trends.  To make any meaningful conclusions (and not just poltical sound-bites), we need longer term studies. And 3) Any decrease in ED use (that was statistically significant) was seen in the higher income groups, which are presumably those individuals that did not benefit whatsoever from health reform.  So, this study really only says something about some factor that is impacting those who already had health insurance, again perhaps higher co-pays, maybe better access to primary care appointments, etc.  We simply cannot conclude that health reform had any impact on ED visits for the population that actually benefited from health reform.  Further studies are absolutely needed for this purpose. 

    Finally, Nancy Turnbull’s comments represent wishful thinking.  There are major problems with measuring “non-emergent” ED use, in particular by asking patients what they think since about 1/3 of patients over-estimate how sick they are and 1/3 under-estimate the severity of their illness.  Furthermore, there are not substantial cost savings associated with diverting low-urgency patients away from the ED.  The real cost savings are associated with reducing admissions and the complex testing that comes with hospital admissions.  This is where the health system should focus it’s efforts.  While it may be politically popular, policy-makers and the media really should stop emphasizing the concept of “unnecesary” ED visits.  What does that mean?  Is the chest pain visit that turns out to be heartburn an “unnecessary visit.”  What about the visit for a sore throat that turns out to be a neck abscess?  The point is, there is a lot more behind this than most policy experts take into consideration.

  • http://byrondennis.typepad.com/masshealthstats/ dennis byron

    @ rachel@8337cc08ef400be6b8ef9b242d7bf8b4:disqus 

    My comment was not directed at you but was literally a “comment” directed to your readers in general and to the comment by another reader about some hospital’s ER usage being up 10%.  It is very possible that ER usage by non-elderly adults was down but ER usage for everyone was up 10%. Or his or her hospital could be an outlier. Or…. 

    This Blue Cross of Massachusetts report draws multiple conclusions based on projections from 3000 phone calls after applying ”adjustment” techniques that are not documented. Typically a statistican would say that his or her data is “weighted” so we don’t know if that is what is meant by adjustment (but, if so, why not say it and provide the weighting factors?) or if some other “adjustment” was done (and, if so, what and why?). Or were there multiple adjustments? 

    So, for example, the Blue Cross report’s claim that ESI coverage is up after the unexplained adjustments (see Exhibit III.1) does not agree with the unadjusted data on the same subject (see Appendix Exhibit III.1). Strangely, the “adjustments” only seem to lower the earlier years (2006 and 2007) of the analysis timeframe. The current years (2009 and 2010) adjusted and unadjusted totals are almost identical. That would not be the typical result of weighting. It is the adjustments that allow the analyst to reach the “ESI is up” conclusion. If I were to dig into the ER data, I suspect I would find similar issues.

    That “ESI is up” conclusion is the opposite of findings based on survey work by the U.S. Census(surveys are more reliable than phone calls and the Census ‘n’ was larger)  and on actual ‘hand counting’ of Massachusetts residents by the DHCFP using almost every one of the 3,000,000-4,000,000 healthcare insurance policies in force in Massachusetts in those years. 

    Here is why this is important and the little sound bites are highly misleading. It is possible that ESI for all residents under 65 is down 6% (the ‘hard’ state data) while the ESI for non-elderly adults is up 6% (the Blue Cross projection or “adjusted” estimate or whatever it is). But that would mean that the total for everyone else covered by ESI (which is almost totally residents under 18, but does include some seniors still employed) is down by a couple of hundred thousand people.  That would be important information for legislators to know in discussing the ongoing reform of Massachusetts healthcare insurance and delivery system. It could mean, for example, that the crowd-out predicted for RomneyCare and PPACA is happening (although the Blue Cross analyst says it isn’t) but that it is happening primarily among children. That is, potentially, workers are buying (or getting) their own insurance through employers but putting their children on public or — highly unlikely — individually purchased programs in very high numbers.

    On the other hand, the state and U.S. Census methodology might be faulty.  Or the Blue Cross conclusions might be wrong 

  • Rachel Zimmerman

    Hi Dennis –

    In the post, I clearly say the statistic about ED use is by non-elderly adults only. I will try to find out the overall number. Thanks for the nudge. RZ

    • http://byrondennis.typepad.com/masshealthstats/ dennis byron

      – rachel

      I think Blue Cross will tell you that that the data on the entire population does not exist in its survey work. That was the impression I received in an email from Professor Long, one of the authors of the Blue Cross report (who also told me “there is not a public use file available for the survey data that I used in that paper”).  

      Analogous data for the entire Massachusetts population is available for public use — in different versions collected and normalized using different methodologies but pretty exhaustively — on the DHCFP web site. But Professor Long says, among other observations on the DHCFP ESI-related data, that the DHCFP effort is not audited and that the results from some smaller carriers were eliminated. Interestingly, she bases her observations on very clear footnotes in the DHCFP reports (some of which I think she prepared). It would be good to see similar detail in the Blue Cross report, as I described in another comment on this thread. 

      Unfortunately, after millions of dollars spent on data collection and analysis by the state of Massachusetts and “reform” beneficiaries such as Blue Cross (much of it with out of state providers, which really irritates me) we do not have a mutually exclusive, collectively exhaustive, taxonomically consistent view of the Massachusetts health care insurance and delivery market. 

      All we have is the statistic du jour, trotted out to support one politician’s or another’s pet project.

  • http://byrondennis.typepad.com/masshealthstats/ dennis byron

    Keep reading the fine print.  The statistic is about ER use by non-elderly adults only, not everyone in Massachusetts.  Is that a big problem in general?  I dunno (but as with the ESI data — see my other comment — I think it disagrees with state of Massachusetts data).  The problem is that only the elite can see the data

  • Guest

    Our ER visits are up 10% in 1 year!

    • Rachel Zimmerman

      What hospital?

  • http://byrondennis.typepad.com/masshealthstats/ dennis byron

    As usual, the elites won’t release the raw data.  Trust us, is their attitude. You paid for this research with your Blue Cross premium dollars and your Massachusetts tax dollars that went to pay for other people’s Blue Cross policies but you can’t see the research your money paid for. 

    The one small bit of data from this Blue Cross funded research that I could find – with a lot of effort - relates to employer sponsored insurance (ESI).  Somehow this survey projection based on 3000 phone calls finds that ESI is up among non-elderly adults over the last five years even though U.S. Census-based data based on over 10,000 surveys and Massachusetts state data — based on actually counting the insured via looking at their policies — says it’s down for all Massachusetts residents except those on Medicare. 

    Both statements could be true IF a couple of hundred thousand kids are no longer covered by employer sponosred insurance.  So the crowd out starts in the kindergarten I guess. This would be very useful information to the debate but it is being hidden for some reason?