E101: Access to Autopsy Reports

Rebecca Snyder sits down with Ryan Little, data editor at The Baltimore Banner, to explore the power of data in investigative journalism. They discuss the role of data analysis in uncovering major trends, particularly in Baltimore’s opioid crisis. Ryan shares insights into the challenges of accessing public records, the legal battle over autopsy data and how their reporting revealed surprising overdose hotspots. They also touch on the implications of a new Maryland bill that could restrict access to vital public health data. Tune in for an eye-opening conversation on transparency and the fight for public information.

Chapters

• 0:00 Meet Ryan Little of The Baltimore Banner
• 2:20 The role of data in journalism
• 7:30 Investigating Baltimore's opioid crisis
• 15:45 Legal battles over public data
• 30:10 The future of data access & journalism

Transcript

0:01

Hi, I am delighted today to have Ryan Little, who's the data editor at the Baltimore Banner, join us today. And we're going to be talking about data sets and autopsies and a bill that is in Maryland's legislature right now that would limit access to data. So welcome to the program, Ryan. Thank you for having me. So before we get deep into the grim things around autopsies and what's publicly available, tell me a little bit about your work. The data editor is a newish kind of role in journalism these days. So tell us kind of how you arrived at that desk and what you do. Sure. So the data team at the Baltimore Banner is responsible for analyzing massive data sets to identify newsworthy trends, newsworthy people, stories that couldn't be possible if we didn't have some folks on the team who are journalists who are tech savvy enough to analyze large data sets. And when I say large, I mean mega. Like the largest data set we've ever analyzed is 1.2 billion with a B rose. So So this is something we're doing a lot, right? We're using coding languages, primarily R, which is a programming language created by statisticians, but we also use some Python. And we have used these skills to uncover some pretty powerful stories. I came to this while I was in graduate school. I once was a city hall reporter in Florida back in the early part of my career. I left for a little bit to get my wife through law school, supporting her so she didn't have to work. And then I went to the graduate school at Maryland to become just a reporter. I took intro to data journalism just the same way I took a radio class, I took a video class, and I just loved the puzzle. I felt like I was so powerful that I was unlocking information that would be impossible to have without these technical skills. And I threw myself at it. And from there, we've been able to... I've been able to report on some things I'm incredibly proud of, and some of our most impactful work at the banner has originated with a data set and us analyzing it, and sometimes even stories where it's not really possible, or it's not obvious that it looks like it's a data story. Well, and certainly like when sort of traditional, like when you think of reporting, it's gum shoe, you're out talking to people in the street and that gives you those single stories, kind of those little snapshots. And there's nothing more powerful than a data set to connect all those dots to say, you know, it's not just Cindy Jones over here who's having this problem. It's, you know, Freddie over here and so and so. And so I want to center our conversation on some of really interesting, and it's all interesting work, don't get me wrong. But the work that is kind of ended up as a focus of Senate Bill 115 slash House Bill 290, which is a bill out of the medical examiner's office that is trying to change the definition of final autopsy diagnosis. And for those... of you following along. That sounds so super boring. Like who cares? Oh, it's a definition. Big deal. let's and data is all about all those little points. So let's kind of go big picture and talk about the, uh, the autopsy reports that you were seeking to report on the opioid crisis in Baltimore. And we'll go on the way back machine. Cause I think we're talking about reporting that happened in 2022. I getting that right? Okay. this process back at the beginning of the launch of the Baltimore banner, which was May 2022. And we knew that opioids and overdoses were a big problem in Baltimore. But nobody had really done much of a look at it, at least a deep dive. The sun had done something maybe 10 years ago, I think my memory serves. And we wanted to take another close look at it. Mayor Brandon Scott was sort of well known to say in press conferences, you know, that, you know, when talking about a homicide counts to always mention, well, there are actually more overdoses. So we really wanted to quantify how many overdoses we wanted to put that count into perspective, both nationally, right? Looking at a big macro trend, but then also micro. We're like, let's look to see which neighborhoods, which specific groups of people who live in different types of neighborhoods are experiencing this. And so to do that, what we need is data. So we went to two sources. First, we went to the CDC. The CDC collects mortality data for every county in the United States. And so we got some academics-only access to that data using former data reporter Nick Teems' academic credentials. He is an adjunct at the Columbia University, where he teaches data journalism. We also went to the Office of the Chief Medical Examiner. We ideally would have gotten death certificates, but death certificates are not a public record in Maryland. So we went to the next best option, which was autopsy reports. it, sure. is do you think every death that's an overdose would have a linked autopsy report or I mean could some people just have died and you would never know? The, I do not know the specific rules off the top of my head, but so I cannot tell you exactly why every autopsy is counted or every overdose is counted in autopsy, but they are. okay. Interesting. We're learning things all the time. Great. Okay, so then you know your data set is complete when you're looking at the opioid crisis. Every opioid death or every overdose death has a linked autopsy report. So you went to the medical examiner's office and was this sort of new, like in the course of your career and writing and so forth? Is it unusual? Like was it? Unusual to go to the medical examiner's office? It's not unusual to go to medical examiner's office. The rules and the laws around access to autopsy reports are different in different states. Nick, who was the data reporter and really leading this process for us at the time, had come from Georgia, where this was something that they had regular access to and something he was used to being able to get pretty easily. I can also add that we ended up having a national partnership with nine other local newsrooms plus the New York Times. working with some of the same data and in some states people were able to get a lot of data like we can in some states it was a little bit harder for them, right? And let's contextualize a little bit because you were working within Baltimore. so I'm just thinking of all the links when you have an overdose, all the records that are generated. So oftentimes police are called EMTs. Contextualize, because I don't mean to let the cat out of the bag, but you had some trouble getting this information from the medical examiner's office. So I'm just trying to contextualize, comparatively speaking. You know, is that information or similar information available through the police or through other channels? Yeah, there's some stuff we were able to get through the police. know, we've seen some Narcan distribution data from, you know, EMTs from the police, specifically in Baltimore. The most valuable data set we were able to get access to is a public data set for 911 calls. So, you know, we can tell you which neighborhood in the city has the most 911 calls for overdoses. But that's not a real count of overdose deaths. Right. And, you know, important thing to remember that everyone, even who works on this project, sometimes forgets is that there's a difference between an overdose and an overdose death. In fact, Narcan has expanded so much that there are lots of overdoses that do not result in deaths. And what we are interested in is primarily those that do result in deaths. The other primary difference here is that generally speaking, a deceased person doesn't have privacy rights to their medical records the same way that a living person would. So, you know, if John Doe overdosed in Baltimore, Narcan was administered, he went to the hospital. That's not something you would typically see in a public database because that is a private medical record. Whereas the death of someone who has died of an overdose usually is. And it's pretty typical to be able to get some information about that autopsy. And in Maryland, it's been the standard that we've been able to get, you know, the autopsy report, what was initially defined to be a multi-page piece of paper that talked about the existence of the death, the dead person's body, some information about the damage or harm that had come to that body and then what the medical examiner has then said the cause of death is. That had been the norm, but that page of that piece of paper did not include all the information we needed in it. And we also wanted to get a data set of every autopsy instead of just pieces of paper for every autopsy. So we had sort of two points of contention with the Office of the Chief Medical Examiner. One is that we couldn't get the cover page and that we could not get the toxicology report, which is sometimes stapled together and ultimately was decided by a judge is in fact part of an autopsy report, but the Office of the Chief Medical Examiner was saying is not part of an autopsy report. This information shows us, you know, for an overdose death, how much of a given drug was in a person's system. know, xylazine is an important part of the overdose conversation. It will tell us if xylazine was involved. It will tell us if fentanyl was involved. The cover page gives us demographic information. How old was this person? what gender was this person, what race was this person, helping us understand trends along different age groups, among different racial and ethnicity groups, and our data points that ultimately became fundamental to the reporting that we did. Through the course of litigation with the Office of the Chief Medical Examiner, we ultimately sued to get access to these reports. It became clear that A database also existed of this. This database included even more important information. Not only was it a database of all the autopsies performed, we ultimately got it for 10 years from 2014 to the end of 2023. It has information about where the person lived, where the incident occurred. If there was an institution involved, what institution was that? That's typically the hospital, right? The types of information that makes it much more easy for us to understand trends in geospatial, right? Like where are things happening? And this information, once we ultimately prevailed at court, was fundamental to our finding that in Baltimore, ground zero for overdoses is not the street corner, right? It is subsidized senior housing where older black men are dying of the 10 properties in the city. where the most people are dying of overdose, I think half of them are senior subsidized housing apartments. So earlier you said, and I kind of want to get into some of the data fields and the sensitivity around them, but I'm transfixed right now because you had the residential address. You know where people are living. You know where the incident or the overdose death is taking place. In that senior housing, are they one in the same or are people just dying at the senior housing and living somewhere else? There tends to be a correlation with where the residential address and the incident address. fascinating. And so again, I just want to better understand the results of this reporting. That seems like that would be such an actionable piece of information from the city just in terms of public health. Has anything happened to your knowledge? So again, context, this report came out in 2023, right? We first published in 2023. Yeah. actually, maybe we, maybe it was 2024. I think it might've been 2024. Let me just check. Maybe like January. But there's been a minute that this has been out in the world. Has anyone sort of taken this up to say, oh my gosh, this is a real issue? Yeah, no, the stories came out in 2024, not 2023. Sorry about that. You know, to my knowledge, there has not been a wide scale movement from any group to address overdoses in senior homes. We have, you know, done our reporting has covered specific buildings where there have been a number of deaths and work by, you know, the people who run that building who own that building to get resources and get help for the people living in that building. But in terms of what the city is doing, the response to our reporting has mostly been silence. There have been calls from the city council to hold hearings about overdose rate. We did also ultimately show that Baltimore is currently experiencing the worst overdose epidemic of any major American city in American history. We went through a great deal of work creating a one-of-a-kind data set to be able to prove that. And in response, the mayor simply had those hearings canceled. The mayor has used ongoing litigation with opioids manufacturers to shield any public conversation about what the city should be doing or is doing about this. and has also suggested that the banner and the reporters who have worked on this were part of a conspiracy from Big Pharma to publish this story, which of course we're not. That's in... how deadly opioids are. how can that be a con... I feel like we're getting off track here and we could talk for so long about when you have actionable information and authorities don't sort of take it up to make the world a better place. So maybe Ryan, we'll talk more about that another day. But I feel like that's a huge, that is a huge issue. But let's go back to, because we wanted to talk about the... the data set and public information and sort of how things are proposed to be changed. So you said earlier that in the autopsy report cover page and in the data set, which there was just to summarize the medical examiner's office rebuffed your attempts or your requests under the Public Information Act to provide this data. And so there is an option under the Public Information Act for people to litigate, to file suit. Usually everyone tries to go to the ombudsman, Lisa Kirchner, to have her sort of mediate those disputes. But if that dispute is not corrected or remedied to the satisfaction of anyone in the dispute, you can go straight to court. which is what ultimately the banner did. And so after a tumultuous process, Judge Nugent ordered the medical examiner to release the records and pretty unequivocably shared his rationale. And so I know you weren't the attorney on it, but tell us a little bit about what came out of that order and Honestly, the fact that attorney's fees were awarded to the banner, to me, is always telling because courts don't like to award attorney's fees. So what's your take on it? You know, I'm not a lawyer. I am not super familiar with exactly what the the judgment says, but I know what we got from the judgment. And we got all of the things we were asking for. Autopsy reports that include the cover letter and the toxicology report, as well as this database of every autopsy the Office of the Chief Medical Examiner has performed from the beginning of 2014 until the end of 2023. In a subsequent request, we have also been able to get some of 2024. I believe it ends through the end of August right now. And this is the data that we've used to do all of this important reporting. And for the record, also have a lot of other plans for this data. This is incredibly powerful and interesting data that really helps us understand macro and micro trends, sort of like the ones we've been talking about across time and space and across groups of people who are dying. And we intend to continue to use this data as a powerful tool to hold people account and to look for problems that could be solved. So when you, I want to get to some of the privacy concerns because you said earlier that the death certificates are not open to public inspection and autopsy records are. And when you think about, okay, this is somebody's residential address, the address of the incident leading to the death, the name, age, gender, that's a lot of sensitive information. Where is, knowing that the Public Information Act is open to everyone, know, journalists or professional users of the MPIA, how do you contend with the privacy issue and sort of the idea of there's purient, nefarious people wanting to paw through autopsy reports? A, does that happen? Has there ever been? an instance in your understanding where these records are at risk and B, why do you need all that stuff? Aren't you nosy neighbor? Why? Why? You know, I don't know of any nefarious acts that have ever taken place. But I also don't know of any other time that the Office of Chief Medical Examiner has actually released this data to anyone except the Baltimore banner. know, autopsy reports have been being released as public records for a long time, to my knowledge. And as far as I know, no one has used them to be to do nefarious acts. have not gone out of my way to check that. So I can't say for certain, you know. I think I and the banner certainly understand and recognize that there is some pretty sensitive information in here. This includes the residential address of every person who's had an autopsy for the last 10 years. Typically the Baltimore Banner data team likes to release the data that we're analyzing as part of our reporting so that people can check us to question our methodology, to understand every subjective decision we have to make when we analyze data, but we haven't done it in this case because We don't want to be the upstream source of this data and nefarious acts from someone else. So I think we understand that, yes, this is very sensitive data that maybe not everyone should have. But there should be an opportunity for us to acquire data that gives us some information that is detailed enough to be able to still do the types of you know, small and large data analyses that we want to do, but isn't so detailed that we know the exact address. A great example of data that is aggregated, so taken from its specificity and moved up a notch to some sort of broader data point is the Baltimore Police's crime data. When a crime occurs in Baltimore, you know, if it occurred at, you know, 558th Street, it wouldn't say 558th Street. in the crime database that is released publicly, would say 500. Everything is aggregated to the block level. So we know broadly where it happened, but we can't say with certainty that it happened in a specific place. If we were committed, we could do the shoe leather reporting to know for certain. But when we're doing large data analyses across multiple years, there's no way that we could do that sort of reporting. But we can still. understand what happened at a broader scale block by block. That's a reasonable compromise that I think everyone should be willing to consider. But that will also take an understanding that data can be modified and that there are people who can use computer programming languages to make these sorts of changes in databases so that they can be released publicly. In our experience, working with a lot of different governmental agencies here in Maryland, you know, over the last two and half years, that's been difficult, you know. Some lied, you know, some government agencies don't employ the people who have the skill sets to do these things, or they have consultants who do that for them. And it can often be very difficult to get the agencies to sort of understand how it can be done and to work with us in a way that lets us get access to really important public information, but do so in a way that follows the law and also helps privacy. Of course, the MPIA says that if the records are kept in a specific electronic format, they must be released in that electronic format. Absolutely. And just for those of you who may not live and breathe the PIA, the Public Information Act gives really a broad comprehensive right of access to records that are created by government or are in the possession of government. Now, with that being said, there's, I think, around 25 exemptions to those that do cover everything from a lot a lot of times around juvenile records, health and safety, trade secret information. The Public Information Act is both very vague and ultra-specific, all at the same time. And I've always thought that it's deliberately vague because it seeks to foster a culture of transparency, that the drafting language can never suppose what's going to happen in the future and you don't want to open it up every year because you now have like for instance, you can submit PIA requests for communications about certain that government officials have. It's not only phone conversations. It's not only emails. Recently, it was litigation was done to broaden it to, you know, text conversations and communication platforms. So we always want to be, we don't want to foreclose the opportunity that things change. And so that is, that's really helpful. And I was just scanning through Judge Nugent's order here. And it specifically says when we talked about a mandatory exception for medical or psychological information, just like you talked about earlier. The exception says that a custodian, the person holding the records, is required to deny inspection for medical and psychological information about an individual other than an autopsy report of a medical examiner. So certainly the Maryland law has been very clear in the past about that autopsy reports are different. than medical records. And so I think your point also earlier when you were talking about the skill sets that are available in Maryland's agencies, Maryland has had a diminishment of public servants. There's just not as many people in government as there used to be. I know that it's a super stressful situation for custodians to feel like. I have to provide this, there's timelines, there's requirements and deadlines, but I don't have the skills to figure out how to manipulate it. And so there's always a tension there. And that's why that culture of transparency is so important, that government has a responsibility to the public to share information. So I'm gonna adroitly step off my soapbox for a minute and... go back to sort of the whys of these data sets. you finally got those records, and these are unique records. You really can't get anywhere else, but there are corollaries to the Baltimore City databases. Other states, you said you were with nine other newsrooms in here. Tell us a little bit about what the experience was like in other states as far as you know. Yeah, we worked with other partners who got data aggregated to different levels. Some better than others. I think Pennsylvania are very near neighbor. The Philadelphia Inquirer was able to get information about autopsies to the zip code level. So they could say, this zip code had this many and this other had that many zip codes are better than nothing, but often too large to really be able to make the sorts of, to get the understanding of demographics and income changes to outcomes. They're just too big and they tend to include so many neighborhoods that are so different from each other that it's just not a great source. Whereas when you can get tighter, you can get too closer to the neighborhood level. These would typically be US census tracts, regions of cities and counties that are split up into about the size of medium-sized neighborhoods, where you get very targeted groups where you can really understand the demographics and the income and the education of these areas, the people who live in these specific areas. And so being able to understand what's happening to different groups is super helpful. Other states get line by line sort of the way we do, don't get as much specificity as we did, where we were able to get the exact address of the incident, the exact address of the residence. They get something aggregated to another level as well, which is really what I think probably where Maryland should fall, which is. Hey, know, this person lived broadly in this area, right? This person died broadly in this area. And so we can't be so specific, but have a really good idea, you know. Well, I think it's important and I think very telling Baltimore has lived through an opioid crisis that is decades standing. This is not a new thing that's come like a house of fire. And I find it just kind of shocking that no one's pulled this data before to really look at that. If there is such a strong, and you all were able to show through analysis, I mean, it's down to 10. Like there's a locus here, like it's not random and nothing has happened on that. And so I think it is through the news gathering efforts of local media that this is really, this is the mission. I mean, you're shining light where there was none previously and creating, showing the data to say, okay, here's where it goes. Um, and so we're going to fast forward. You've, you've done the reporting kind of sinks like a stone. It sounds like, but we're here cheering you on. It, nothing, you know, kind of happened with that. Um, but you published it, data sets are out there and we come to, uh, this legislative session or in the 2025 session. And we find that the medical examiner is putting a bill in to define autopsy final. autopsy diagnosis. And so I think there's a pretty clear line that Judge Nugent reaffirmed that there's significant precedent for autopsy reports to be clear under the PIA and that the medical information exemption is to be narrowly construed. And so Now it seems like the medical examiner saying, well, let's change the definition. Let's create this final autopsy diagnosis. And so right now, of course, that's a very loose definition and advocates are working to make that a less spongy definition. tell us, like, what happens if this just pops on through? Like, what is the sort of through line if government agencies can be ruled against and then they just decide to change the rules. For us, it certainly means we won't get access to this data anymore. We won't be able to know what's happening in those senior homes that have been ground zero for the overdose crisis in Baltimore. We won't be able to look at thousands of other issues that we could analyze. It means that we will be stuck in time. You can't take the data you have away from us, but we won't be able to look forward. It also means that maybe media organizations won't be, you know, we'll maybe take another thought about, well, is it worth suing? Right? Maybe I can win, but even if I win, maybe I'll lose the next time the General Assembly convenes. You know, I think everyone, you know. Obviously, you know, we sued because we believe that to be the right thing and that those were public documents and we wanted to use these public documents to do some really important reporting and we want to continue to use those documents to do that important reporting so that we can shine light where it hasn't been shined before and and and hold officials accountable, but also, you know, understand terrible trends that are So, and thank you for that. think that's, that is compelling and important to bring out that this, you know, this changes the landscape, not just for these particular, for that data set, but it creates this, it turns that culture of transparency into one of, into a more opaque culture where people don't have access to information. So. just to kind of sum up, hearing that there hasn't been an issue before except that perhaps people haven't asked for this data. There's no, in your reporting, you you didn't, you haven't published the data set anywhere. You've published the results of the reporting. The privacy of those individuals, like what happens to that information as you've reported, you have access, you know that, you know, Rebecca Snyder, aged 42, died, you know, in this space. What did you do at a sort of granular level with any of that information? And where is that information now? you know, we have the closest we've come to publishing to publishing the data is making a map to show where overdose deaths are happening in Baltimore. You know, we give some information, gender, age and approximate location. You know, we've taken, you know, the latitude and longitude points and added some random, you know, jitter so that whenever a point falls on a very specific part of the map, it's not. actually that place so that you couldn't use our map to say, this happened in this home. It didn't happen in this home. But we wanted people to see that this isn't a black butterfly thing. This isn't a white L thing. The two popular terms for racial segregation in Baltimore and the way it looks when you visualize majority black and majority white neighborhoods in Baltimore on a map. But that it's everywhere. You look at that map, it shows up all across the city. You know, we have used it as a reporting tool to identify people who have died to tell their stories. know, a really sad note of the fentanyl crisis is that it has become more deadly for young children because fentanyl is so strong. All they have to do is touch a smallest bit of fentanyl and then take, you know, touch their mouth with their hand. And that's probably enough for a young child to overdose. And so we were able to find stories of young children who died of overdose and tell those stories because we had access to this data. Now, we also haven't published names of anyone who's died of an overdose who we haven't done the reporting to. So we know of some significant Marylanders, or at least Marylanders who are significant in the communities in which they live, who have died of overdose, where it is not public knowledge that they died of overdose. Mm-hmm. some of those stories we would like to tell, but we have not told those stories because we have not been able to talk to families, talk to friends, get the type of information and do the type of reporting we would want to do to share that information. What that line is is a little hard to say for certain, but that line is there and we have gone out of our way to make sure that we haven't violated families' privacy. Absolutely. And so we've talked a lot about your specific like this specific report. And so I just wanted to Google search. And so if people listening along, they can go to the Banner website, which is the Baltimore Banner. I always forget the dot com forward slash specials forward slash overdose dash. crisis and you can see this reporting which came out June 3rd in 2024, specifically about the black men are dying from overdoses in Baltimore. So that is that you can sort of read what that reporting engendered. But I also wanted to check in about what data sets you're working with now. Like what's next? Is the data editor? You're always working with something. So what's on the horizon? God, I wish I could tell you what we're working on right now because boy, I'm so excited about it. And it's possible that by the time this airs, we will have published it. So really excited about that. All I can say is that we have taken a long, hard look at some databases that have to do with education and have gone about some means basically capturing some public data feeds. that just anyone can get access to, capturing them for months, and then analyzing millions and millions of rows of this public data feed to understand something that's going on with education in Baltimore. And it's something I'm incredibly excited about. that sounds really exciting. Well, we'll have to have you back to talk about that. And I really appreciate your time and your energy in speaking with us about this particular situation and this particular bill. We'll have to see where that goes. Advocates are very concerned about the diminishment of autopsy records availability across the state. It doesn't match up with what we see in other states. it's... from your experience and from what you've shared, it really is a significant issue for news gathering and the type of reporting that can really move cities and jurisdictions in the state forward. So thank you for your time and in joining us. Thank you for having me.