In reporting on effects of hospital closures, granular tales tell the real story

Sometimes the best ideas are forced upon you.

Here at the Pittsburgh Post-Gazette, like many papers now with smaller staffs, nearly all of the reporters have to regularly work a weekend shift once a month or so.

So it was on Aug. 21, 2011 when my turn came up and I had to work a Saturday shift. I was told that one possible story was the opening of a free medical clinic in Braddock, Pa., a struggling former steel town in the Monongahela Valley that had lost its hospital earlier in the year.

I spent most of the morning and early afternoon at the clinic, talking to the volunteer doctor and nurse, and, most importantly, patients. There weren’t a lot of them that first day, just four in the morning. But I was struck by their stories and how the closure of one small, community hospital impacted the lives of so many.

Some predicted that with three hospitals within a 30-minute ride of Braddock, and a federally qualified health center (FQHC) available in Braddock, the free clinic would not be needed. But as word got out over the next year, the patient volume grew to a steady 20 to 30 people every Saturday and Sunday.

Over the next two years, while covering other stories, I began to wonder not just about Braddock, but the other parts of our region that lost hospitals, many of them also former steel and manufacturing towns: Aliquippa, New Kensington, Jeannette and, of course, Pittsburgh.

I began gathering clips on the hospitals I knew had closed and learning why.

In early 2013, I put together a spreadsheet of all the hospitals in our eight-county region and the amount of “uncompensated care” – that’s a combination of charity care and bad debt – that they reported year-by-year, going back to 1997.

The data is kept by a unique state agency in Pennsylvania, the Pennsylvania Health Care Cost Containment Council (PHC4, as it’s known), which is mandated to collect this and other hospital-specific data every year. The data I used was available in annual reports on PHC4’s website, though because there were at least four different types of spreadsheets used by the state since 1997, and various formats of presentation that made combining data sets by computer impossible, I had to type in each hospital’s annual figures for each by hand – a tedious task that paid off.

What I saw when the spreadsheet was completed astounded me: 11 of the 39 hospitals in our region had closed between 2000 and 2010. That simple statistic shocked even our most devoted readers, who tended to recall each closure story but hadn’t thought of them in the aggregate.

I also noticed was that all but one of those 11 hospitals that closed typically provided more than average amounts of uncompensated care. In short, the more care a hospital provided to the poor in the area, the more likely it was likely to close. This really grabbed our readers’ attention.

And it led me to wonder: Where did all the poor who used to go to the now-closed 11 hospitals go to get their care now? What challenges do they face, particularly if they were in more far-flung, less urban areas like Braddock, where the nearest hospital is at least a 30-minute car drive away? And, if hospitals had to close in poor neighborhoods and towns that were losing money, why were the same health systems that closed them also expanding or building new hospitals or “medical malls” in wealthier towns and neighborhoods at the same time?

For four months I attempted to answer those questions by sitting in lobbies of the local free clinics and health centers that serve the poor and working class, while also asking the hospitals themselves why the system seemed to be working against the interests of these patients.

One of my best decisions was to visit other cities and see how they provided care to the poor. I visited Indianapolis, Cleveland, Philadelphia and Baltimore – cities that were all chosen because their medical communities were similar to Pittsburgh in that they all had at least one prominent academic medical center with strong competitors.

The visits to those four cities was particularly illuminating because I ventured out only after most of the on-the-ground research here in Pittsburgh was completed. By then I had a very clear picture of the challenges in Pittsburgh and I could accurately question leaders in other cities about how they met the same challenges.

It was while choosing which clinics to visit in Pittsburgh that I had another local realization: There were a lot more free clinics and FQHCs now than there were in 1997. But what I found from months of talking with patients, doctors, nurses and clinic managers was that despite the rise in the number of free clinics and FQHCs, our poor were still too often going without health care and, in particular, specialty care.

While I had done extensive research – talking to health care leaders, reading journal articles, visiting health organizations websites – nothing was as informative as spending time with patients. The granular stories they offered me were so much more complicated and illuminating than anything the smartest health care executive could tell me.

One challenge I encountered was that some of the same financial reasons that brought poor patients to a free clinic or FQHC – having lost a job, etc. – made staying in touch a challenge. Many rely on temporary, pre-paid cell phones that quickly become disconnected. Their home addresses they gave you in February may not be their home address in April. Emails may go weeks or months without an answer because the recipient doesn’t have a computer.

To deal with this, I would ask when they would return to the clinic and agree to meet them there the next time. Or I’d agree to meet them, say, at a coffee shop or restaurant near them to talk again in the near future.

The more of those stories that describe the same problem, the better: One story is an anecdote. But if you can say you heard it a dozen times, from different patients in different settings, then it becomes a trend worth discussing more broadly. That was the case with my story on the difficulty of accessing specialty care.

That doesn’t mean the search for meaningful data isn’t important – and challenging.

Trying to get a handle on data that would demonstrate how bad the health care situation is for the poor at a local level also proved to be a daunting task. Nearly all of the federal and state data that is available only goes down to the county level. Too few counties – including ours – do regular health surveys in their communities that would tell you how an individual town or neighborhood is doing.

And no one out there has a count on, for example, how many poor people are unable to get specialty care because they can’t afford it, or were told to bring cash to an appointment and they simply did not have the money.

We did have some success combining data sets to show not only where the poor and disabled in our region are, but where the doctors are not – and how that was often the same place. We created an interactive map with federal data that went down to census tracks that combined physician shortage areas with percentages of poverty. The results were not surprising: Many of the same areas that lost hospitals had high concentrations of poverty and low numbers of doctors.

To get this data, one of my colleagues at the paper, Lillian Thomas, who was on her own fellowship and working with the Milwaukee Journal-Sentinel and their staff, obtained readily available data on poverty, median income and disability from the five-year American Community Survey from the U.S. Census bureau.

Data on the location of primary care physician shortage areas – that’s an area with less than one primary care physician per 3,500 people – used data from the U.S. Health Resources and Services Administration.

I also attempted – and failed – to gather federal data on FQHCs to show how, over time, the number of poor these clinics serve has increased as hospitals in our region closed.

The federal government makes the data on the most recent year readily available online, and will send additional data going back a year pretty quickly. But anything older was going to take a significant amount of time for Health and Human Services to get to me, and I simply ran out of time before it could. It always pays to begin your data requests as early as possible.

Despite the challenges and dead-ends on data, our series revealed a simple truth: The poor continue to have difficulty accessing health care in one of the most medically advanced cities in the world. How to solve that problem is something I will continue to pursue with my current and future work.

Photo by Michael Fleshman via Flickr.