How a veteran Post reporter solved the data puzzle for story on young patients stuck waiting for psychiatric beds
(Photo by Matt McClain/The Washington Post)
If you’re in a mental crisis and get taken to the ER, you could end up waiting days in the hospital for a psychiatric bed. It’s even worse if you’re a teen; you could wait for weeks or even months.
While covering mental health for my newspaper, The Washington Post, I kept hearing about this phenomenon, called psychiatric boarding. In interviews, families likened their agonizing waits in the emergency room to imprisonment. One teenager described it this way: “Imagine already wanting to die, and then someone locks you up in the middle of hell.”
The story, however, sat on my to do list for more than a year. It felt urgent, important, but I kept putting it off because I wasn’t sure I could pull it off in a way that would really make an impact. Like many problems in the world of mental health, it was widespread, occurring at hospitals across the country. And responsibility for the situation was diffuse, with no clear entity to hold responsible. In some ways, society at large was most to blame for not caring — by not giving as much in insurance reimbursement to mental ailments as physical ones, like heart attacks or cancer. For not creating the spectrum of community resources that has been needed for decades, a vacuum that has resulted in ever-longer backups and delays all the way up to the emergency room door.
There was also a lack of national data on the problem. No one knows how many children sit waiting specifically for mental help in ERs across the country and for how long.
I started looking for ways to narrow the problem down to a single state and a single patient. Focusing on a single state would help me find data and explore the roots of the problem and question those in charge who had let it deteriorate to this point. Focusing on a single patient would give me a narrative to convey the important stakes involved.
Among states with generally long ER wait times, Maryland — which our newspaper cares about deeply — ranked nearly the worst in the country. I found a report in the meeting minutes of the Maryland Health Care Commission indicating there were state records that could be examined to show exactly how long and how many children and teenagers were waiting in the ERs. And I saw the kernel of the story there — a story that combined data and accountability reporting with the power of narrative. But I knew I’d need help pulling it off.
I pride myself on being a Swiss Army knife reporter — someone editors can throw into any situation. For years, I had worked as The Washington Post’s roving national correspondent. I kept a packed bag at my desk, ready to drop everything at a moment’s notice and parachute into disaster zones and mass shootings. Before that, as a foreign correspondent, I had written investigative narratives about those suffering human rights abuses at the hands of their government. But in pursuit of that versatility, I’d never taken the time needed to develop the data and computer-assisted reporting skills required for such a story. That’s why I applied to the 2021 Center for Health Journalism Data Fellowship.
Here are some of the key lessons I learned along the way:
In the face of roadblocks (and there are always going to be roadblocks), find other ways of getting the data.
I was correct that a state agency had the records I needed — a massive database documenting every patient case admitted to every ER in Maryland and the length of stay for each.
But despite negotiations, it became clear obtaining even one year’s worth of that data would cost thousands of dollars. I needed 10 years, a small fortune’s worth.
Through relationship-building and digging, I learned that the state data was forwarded every year to an obscure federal agency. Obtaining data from that federal agency would not only be faster but require a fraction of the cost.
It’s okay if the task is far beyond your abilities, as long as you find (and ask) the right people for help.
As a former English major, math and science are not my forte. And I had only basic experience with Excel. To filter through hundreds of thousands of case records, I would need to learn how to code in the R programming language.
I had the outrageous fortune of being assigned Liz Lucas as my mentor for the fellowship. In the months that followed, she became my coding coach, data editor, therapist, spirit guide and (most of all) friend. Liz and I devised a plan to learn enough coding in R during my fellowship to carry out the rudimentary functions needed. The rest of the analysis, we decided, I could do in Excel.
Because I was so busy learning R during the fellowship, I designed a mini-boot camp for myself after the class portion of fellowship to get up to speed on how to do remedial data analysis. I relied on online courses other fellowship instructors had assembled over the years — MaryJo Webster, Christian McDonald and Andrew Tran. (For complete novices like myself, I would especially recommend Christian’s videos in which he uses Starburst candies to explain basic concepts like sorting, filtering, etc. And MaryJo’s Excel tutorials in her Data Journalism Academy.)
Outside experts can be invaluable and it’s worth source building with them.
Figuring out how to analyze the data in a correct and meaningful way took way longer and much more effort than I hoped. The details are too long and painfully boring to spell out here (sleepless nights trying to figure out, for example, “crosswalks” and mental health codes for ICD-9 to ICD-10 conversion). When certain tasks outpaced my skill, my mentor Liz stepped in to help. We also consulted with federal and state agency experts I had built a relationship with. We were able to use past analyzes they’d conducted as a roadmap to make sure we were doing ours correctly.
Be patient and find the right characters for your story.
It took months of searching and in-depth interviews with dozens of families who experienced mental crises to find the right ones to focus our story on. It took even longer to find a single hospital willing to let us into its emergency room to see what the bed shortage crisis looked like up close (and to take the photos needed for visual power). Psychiatric boarding was a problem all hospitals hated and were frustrated with, but few institutions wanted their failings and inner problems depicted in photos and print. It helped to explain what we were trying to do and cultivate advocates within hospitals.
Perhaps the biggest lesson for me was the mental and emotional endurance that ambitious stories demand.
There were so many challenges at every stage: obtaining the data, learning enough skills in R and Excel to decode it, convincing families and state officials to talk, and reducing those gigabytes and piles of interview notes into a compelling story. At times, it felt like the universe was working against this story I was struggling to bring into existence.
I learned the importance of building and leaning on your support network. I talked constantly with my editor at the Post, Lynda Robinson, who is quite likely the best narrative editor now working in journalism. I checked in almost weekly with my fellowship mentor, Liz. I talked to friends and colleagues about my doubts about the story and my abilities to pull it off, which often buoyed my confidence.
The article resulting from all my late nights of work and worry finally published on Oct. 20, 2022, with the headline, “An autistic teen needed mental health help. He spent weeks in an ER instead: Zach Chafos languished for a total of 76 days in a Maryland ER waiting for a psychiatric bed — part of a growing mental health treatment crisis for teens across the country.”
This is how the lede began: “By his fourth week waiting for help in the emergency room, Zachary Chafos’s skin had turned pale white from lack of sun. His mother, Cheryl Chafos, bathed her autistic teenage son daily in the ER’s shower, trying to scrub the sickly pallor off him. His father, Tim Chafos, held the 18-year-old’s hand, trying to soothe his son’s pain and confusion over what was happening.”
It was one of our newspaper’s most-read stories for days, garnering hundreds of thousands of readers. Dozens of readers and mental health activists wrote in. It sparked debate among state medical leaders and new efforts by the Maryland Hospital Association to solve the problem.
We used our exclusive data analysis to show exactly how Maryland officials kept promising but delaying for the past decade attempts to work on the problem, and how over that same period, wait times for mental health patients in ERs soared to dramatic lengths, especially among children and teens. We documented how Maryland health officials were now violating a law that required them to have a statewide online bed registry already up and running — a tool to match vacant psychiatric beds with ER patients. Instead, every hospital in Maryland was still manually calling and faxing psychiatric wards almost hourly across the state in search of beds.
The day after I interviewed the Maryland health secretary about these failings, he sent a letter to every hospital in the state, alerting them to the creation of a makeshift first attempt at a psychiatric bed registry. And he began convening monthly meetings with stakeholders to fast-track the creation of a better, fully functioning system.
It was gratifying to see. And it made all those sleepless nights worth it.