War on a Thumb Drive: How SigAct Data Could Transform Care for Veterans
I deployed to Fallujah, Iraq in May of 2004. It was my second deployment to the country, and I was looking forward to my tour ending as I was separating from the Marines shortly thereafter. Toward the middle of the month, I was sent to 1st Marine Division’s headquarters to work on a vehicle.
We were unable to fix the vehicle on location and waited for a couple of days for an escort to Al Taqqadum (where engine mechanics were stationed). In those few days, I briefly came to know a staff sergeant assigned to Charlie Company of 1st Light Armored Reconnaissance Battalion. He took care of me and the other mechanics and even chatted with me about going back home. He talked fondly of his three sons and struck me as a decent man and a great marine.
A couple of days after returning to Fallujah, I learned that his vehicle was attacked and that he was killed in our area. Furthermore, I was tasked with pressure-washing the vehicle to remove the blood so that the vehicle could be scrapped. After finishing the work, I returned to my trailer covered in specks of blood from the back spray, emotionally stirred.
If a veteran diagnosed with a mental health condition other than post-traumatic stress disorder submits a claim for disability benefits or care to the Department of Veterans Affairs, they are often required to provide enough information for the agency to validate that the event that caused their condition likely occurred. This impacts tens of thousands of veterans per year. I know this because I have worked for the Department of Veterans Affairs for over 10 years.
Stressors, as they are called, historically took on average 60 days for the Department of Veterans Affairs to verify. Veterans Affairs personnel submit requests to Department of Defense entities for research into the validity of the stressor based on source documents. For many, this prolonged research has delayed not only timely payment of benefits but also free mental healthcare from the Department of Veterans Affairs.
As a veteran, I have always been drawn to the verification process. We are often required to verify things that were commonplace in theater, such as mortar attacks. Early on in my career, I was also intrigued by the stories of the Iraq War files that were given to Wikileaks, which, per reports, had detailed combat event data. However, being that I was a federal employee and could not use or view the website, I quickly forgot about the existence of such data.
It was not until many years later, as the supervisor of someone assigned to verify these stressors, that I remembered that the Department of Defense kept what’s called SigActs (or significant activities) databases in Iraq and Afghanistan. Googling late at night, I found the work of Vincent Bauer, a PhD student at Stanford. On his website, he posted a dataset that listed hundreds of thousands SigActs that were also geolocated, meaning each individual event could be mapped in space and time. Along with colleagues at the Department of Veterans Affairs, we took this initial dataset and created a proof of concept for something that would later streamline stressor verification for many veterans.
After developing a proof of concept, I searched and searched for more datasets that we could use. I stumbled across the work of Andrew Shaver, founder of the Political Violence Lab at Dartmouth College and a post-doctoral researcher at Stanford University. Additionally, I noticed that he was a member of the Empirical Studies of Conflict group, which analyzes conflict data in fascinating ways. On his site he referenced having extensive SigActs datasets for both Iraq and Afghanistan. After exchanging some emails and some calls, Shaver shared these datasets with the Department of Veterans Affairs at no cost. My colleagues and I leveraged these powerful datasets by mapping them in Tableau, and the result was the Official Military Activities Report, or OMAR.
The Power of Combat Mapping
OMAR is a shift in the way that the Department of Veterans Affairs concedes stressors. The Tableau workbook pulls together more than 1 million events and casualties. The events are mapped in place and time, so you can filter the maps by a veteran’s service dates and location overseas and see just how much combat was in the area. Rather than waiting 60 days or more, research for Iraq and Afghanistan veterans takes users minutes.
We are not stopping there. Based upon datasets readily available from the National Archives and Records Administration, we are putting together a combat dataset for the Vietnam War. This will further ensure that we can quickly and accurately validate stressors and help veterans receive access to disability compensation and free mental healthcare.
Fighting for Data
A major issue that is exists is that OMAR is largely operating on a snapshot. While expansive, it is still limited in its scope. For instance, Afghanistan data stops at 2015. Additionally, it took Shaver more than 2 years of Freedom of Information Act requests to obtain what we have now. Instead of giving him spreadsheets with structured data housed within, he was given PDFs with messier, more unstructured data. He had to write code to extract the data from the scores of documents.
But OMAR is not the only use for this data. While I am mainly interested in obtaining access to combat datasets to help veterans with their claims, The Political Violence Lab and the Empirical Studies of Conflict group have much broader scopes that answer important questions. Questions like, “Is it better to deploy troops in seven-month increments (like the Marines) or in 14-month increments (like the Army)?” The answers that they provide inform commanders in theater and arm them with vital information on how to operate in a combat zone.
Imagine, however, having to answer such questions with a multi-year (sometimes even four-year) lag. When you do get a response, the first one is often incomplete, and you must go back to the relevant official over and over until you receive what you need. The responses are not provided in a format that is helpful, and people have had to write special code to extract information. Lesser people would give up, but the Political Violence Lab and Empirical Studies of Conflict group presses on.
I have no lasting effects from that fateful day in Iraq. However, I sometimes wonder if I could prove my event to be true in front of the Department of Veterans Affairs. My event is difficult to corroborate, as the marine was not in my unit. Furthermore, I wasn’t present at the attack. Complicating matters further, cleaning the vehicle was probably something not documented in my unit’s command chronology, which is essentially an official diary on what happened to the unit. My event, for lack of a better term, would not be verifiable through normal means. I would have to ask others with whom I served for supporting statements to assist with the verification of the event in question.
This perspective informs and drives my thinking in terms of improving stressor verification for all veterans. A combat zone is a place where fate intervenes daily and where incredulous horror can be commonplace. The least I could do is share this perspective and work to make it easier for veterans raising their hands in need of help. However, to be truly successful, my colleagues and I need help as well. We need open and free access to declassified combat and operational datasets maintained by the Department of Defense. Otherwise, we must play the Freedom of Information Act game and wait years. In the interim, veterans are waiting for care and benefits.
Chris Aragao is a coach and civil servant at the Providence Veterans Affairs Regional Office in the Veterans Benefits Administration. He enjoys working on projects to streamline the claims process for veterans and their families. He lives with his family in Rhode Island.
Image: U.S. Army, 3/2 Stryker Brigade Combat Team