New York Times, May 17, 2012
FORT CAMPBELL, Ky. — To those unfamiliar with a battlefield’s bleak routine, Col. Michael D. Wirt’s database could be read like a catalog of horrors. In it, more than 500 American soldiers are subjected to characteristic forms of violence of the Afghan war.
Faces are smacked with shrapnel, legs are blasted away near knees, bullets pass through young men’s abdomens. Vehicles roll over, crushing bones. Eardrums rupture. Digits are severed.
Dozens of soldiers die. Hundreds more begin journeys home, sometimes to treatment that will last the rest of their lives.
Each was listed in a small but meticulous computer entry by Colonel Wirt, a doctor intent on documenting how soldiers were wounded or sickened, how they were treated and how they fared. For those seeking to understand war and how best to survive it, the doctor on his own initiative created an evidence-based tool and a possible model.
His database is one part of a vast store of information recorded about the experiences of American combatants. But there are concerns that the potential lessons from such data could be lost, because no one has yet brought the information together and made it fully cohere.
Colonel Wirt was a brigade surgeon from the 101st Airborne Division during the American-led effort in 2010 and 2011 to dislodge the Taliban from their rural stronghold along the Arghandab River. His database was part official record, part personal research project.
His commander required him to keep tabs on ailing and wounded soldiers, and to inform him of their prognosis and whereabouts in the medical system.
To this, Colonel Wirt added layers of information. He documented which weapons caused which wounds. He tried to record increased or decreased risk factors — whether the victim was wearing larger or smaller body armor, whether a bomb-sniffing dog was present, when a tourniquet was applied.
He recorded which accidents and diseases took which soldiers off duty, and for how long. He mapped where on a human body bullets most often struck.
A year after he returned to the states, Colonel Wirt and his database point to the promise and obstacles related to studying more than a decade of American war.
The amassed information on combatants over 10 years amounts to the most detailed data ever assembled on battlefield trauma and its care, American military officers say. But the records are scattered.
The Defense Department’s trauma registry has information on roughly 66,000 patients who were admitted to modern military hospitals, including American and coalition troops, Afghans and Iraqis, contractors, and the odd journalist, diplomat or aid worker. It is a record, largely, of those who survived.
The Armed Forces Medical Examiner’s Office maintains separate sets of records, including full-body CT scans, for service members killed in action. And each patient’s medical records, often with narrative details of wounds and a thorough chronicle of treatments, are available in electronic form, but only to those involved in each person’s care.
Supporting documents for Purple Heart awards can also include medical and tactical data.
In certain contexts, some of the data is merged, as at quarterly meetings of a special committee that has been seeking ways to improve prehospital care.
So far these disparate storehouses of information have not been joined in a permanent place, much less made widely available for cross-disciplinary study.
Col. Jeffrey A. Bailey, a surgeon who directs the Joint Trauma System at the Institute of Surgical Research at Fort Sam Houston, confirmed what several military doctors noted: There as yet is no standardized medical database that enables researchers to look back comprehensively on the experiences of Afghanistan and Iraq.
Colonel Bailey said his institute, a research arm to promote improvement in medical care, eventually hopes to combine a “K.I.A. module” from the medical examiner with the registry of patients treated in hospitals.
He added, however, that discussion of merging the data is in its early phases, and that while “I think we will get there, I can’t tell you when it will happen.”
Against this background, Colonel Wirt, a neuroradiologist who volunteered for duty in an infantry brigade, set out in 2010 to make his own record of one brief but bloody chapter of the Afghan war. “This was a way to take something away,” he said, “so that all of the casualties mean something.”
His commander wanted a high level of detail, he said. Curiosity drove him further.
“If you don’t take data and analyze it and try to find ways to improve, then what are you doing?” Colonel Wirt asked in an interview at Fort Campbell, where he is a deputy commander at Blanchfield Army Community Hospital. “In my humble opinion, a consolidated database with standardized input consisting of mechanism of injury and resulting wounds, classified by battle and nonbattle injuries, would be something you could actually use.”
Other officers agreed. Maj. Kirk W. Webb, formerly responsible for compiling casualty data for the 101st Airborne Division, said each unit tracked its casualties, although not to the detail that Colonel Wirt pursued. Most of the information has probably vanished, he added.
“It’s kind of sad, actually, because there is a lot information out there that gets lost,” Major Webb said.
Dr. Dave Edmond Lounsbury is a retired colonel and medical doctor who was co-author of “War Surgery in Afghanistan and Iraq,” a textbook from the Office of the Army Surgeon General that, over considerable internal dissent, published case studies of combat wounds.
He said that data like that compiled by Colonel Wirt would also be valuable for those who study workplace safety, for historians and for officers who hope to rise above collecting anecdotes to examine how insurgent and counterinsurgency forces fought each other and evolved.
But he and other officers noted a potential obstacle: Many people in the military have opposed sharing detailed medical data. The reasons, Dr. Lounsbury said, include concerns about patient privacy and a desire to present an airbrushed picture of war for public consumption.
One military official also said restricting access to the data could prevent potential enemies from studying it.
(The Office of the Surgeon General forbade Dr. Wirt to share with The New York Times his data on how American soldiers were wounded, even though the newspaper asked for the data in the format in which the data is released and updated monthly by the Office of the Secretary of Defense for the entire Iraq and Afghan wars.)
The data can be politically charged. Records from the Arghandab offensive, for example, show that 530 Americans from Dr. Wirt’s former unit were wounded in a roughly one-year period, compared with 150 Afghan soldiers and police officers. The contrast belies the official insistence that Afghan forces led the campaign, or even participated equally in it.
Dr. Lounsbury suggested that whatever the political content or concerns, compiling data and circulating it broadly was important for the practice of wartime medicine, and for the American military and public to understand better a long period of war.
“I can’t think of a higher lesson learned than to put all of that data together and find out what weapons were used and who got killed and who lived and with what therapy and treatment,” he said. Dr. Wirt, he added, “should be applauded for what he has done.”
“Big Data” has received a ton of buzz in recent years, and with good reason. Big Data is becoming more the norm in many industries, and it is important to be able to turn that data into actionable insight and knowledge. However, many datasets in business contain loads of information themselves while missing the moving target of being “Big Data.” We should not forget that these datasets might lead to just as much insight as their bigger siblings.
Before Big Data became the latest buzz word, many people were doing more traditional analytics on their desktops. In the advent of the Big Data age, I feel as though people assume that small data is a solved problem, but that is far from the case. Experts in analytics have been able to provide answers to small data problems for years, but something will always be missing so long as the domain experts are separated from the process. The key to better solutions for small data problems is to provide a an analytics platform powerful enough to provide real answers but simple enough that domain experts can use it to ask the right questions.