Experts: Football Death Reports Aren’t Valid Epidemiology

The University of North Carolina’s self-published football casualty statistics are deemed ‘unverified’ and ‘unreliable’ while its claim of zero deaths in 1990 is debunked by two case finds

By Matt Chaney

Posted Monday, January 12, 2015

Copyright ©2015 by Matthew L. Chaney

When sportswriters report 17 football players died in America during 2013, and medical writers and researchers cite the same figure, everyone’s common information source is the National Center for Catastrophic Sports Injury Research (NCCSIR) at the University of North Carolina in Chapel Hill.

The annual UNC numbers on football fatalities have been quoted worldwide since the 1970s and include statistical framework back to 1931, the year football organizations assumed control of annual recording. Reports are posted at the NCCSIR website and everyone presumes the data meet epidemiological standards for charting mortality risk in American football.

But the numbers do not qualify as epidemiological research, not remotely, say two experts who have monitored faults and recurring issues in UNC postings for three years.

“There is no independent verification of the data,” said Don Comrie, the CEO of PanMedix, a New York company that designs statistical analysis protocols for medical and pharmaceutical research. “When we look at the UNC data, there is no reliability.”

Epidemiologist Charles E. Yesalis says NCCSIR reports fail to pass muster for his discipline. Yesalis, author of epidemiological studies, journal articles and books on sports, identifies an historical misstep by UNC workers who lack medical and scientific credentials—their reliance on limited news content for casualty assessment of millions of football players, the vast majority being juveniles.

“When you’re dealing with (information) as what might get reported in the news, versus trying to identify accurately what’s reported in emergency rooms, or hospital records, that’s problematic,” said Yesalis, professor emeritus of public health for Penn State University, with his doctorate of science from Johns Hopkins University in 1975.

“It’s tough, it really is,” Yesalis said, empathizing with the football academics at UNC. “You’re trying to glue this all together, which is what it appears they’re doing. And they’re not the first people to do this on a variety of disease states, ER conditions, injuries and all that.”

Yesalis and Comrie have followed my email updates and blog critiques of UNC football publications since 2011, when I began collecting news reports of catastrophic player casualties found online.

For fatalities, I’ve located more than 1,400 deaths among active football players from 1960 to 2014, cases both confirmed and still suspect for links to the sport. And when comparing to the approximate 1,050 deaths UNC logs in the timeframe, methodological holes emerge for the NCCSIR.

Center officials refuse to answer my inquiries but an overriding fact is clear: more people die from football than the game-funded “studies” convey to the public.

Faults beset six decades of UNC football data, with inconsistent classifications and case omissions exposed by contemporary electronic search. Missing are deaths caused by football collision, defined as “direct” fatalities of the sport, along with likely hundreds of “indirect” or game-related fatalities—particularly for cardiac arrest, its various mechanisms that can include delayed attack from previous chest blows, according to a recent study.

Casualties omitted from 2011 UNC statistics, for example, include 13-year-old Kansas school player Alec Mounkes, who died of blood clots following an ankle injury. Kishon Cooper, 8, collapsed and died in Florida while training at home for “youth league” football. Two young men succumbed, Marcellis Williamson and Andy Collins, of blood clots and cardiac arrest, respectively, as free-agent hopefuls for professional football.

During 2012, Pennsylvania semipro player Willie Mims collapsed at football practice and later died, as did prep player Temoc Castellanos, 15, stricken during off-season conditioning for his school team.

An aspiring player in Texas, 15-year-old Jacob Gatlin, collided heads with another on a football field at school. The boys were participating in a “7-on-7” passing drill without helmets during “athletics class” directed by coaches, school officials told media, and Gatlin suffered a skull fracture and fatal brain hemorrhaging.

None of these deaths is recorded by UNC.

Several high-school players died of cardiac arrest in 2012, during exercise and restive state like sleep, including: Anthony Vaeao, California; Austin Lempera, Illinois; Cody Stephens, Texas, David Widzinski, Michigan; and Tyler Miller, New York. Such cases require specialized postmortem applications for diagnosis and determining a possible link to sport, according to a host of experts worldwide.

None of these deaths is included in UNC “research.” Many more of possible football ties since 1960, found online, require proper scientific evaluation.

Errors likewise dog UNC’s recorded cases, such as the 2010 death logged in the wrong year. Youth player Quadaar White, 15, died of a broken neck in Philadelphia on Aug. 31, 2010, but the NCCSIR recorded his case for 2011.

I repeatedly emailed then-center director Frederick Mueller. Instead of correcting his mistake, Mueller—a lifelong football man with a PhD in education who goes by “Dr. Mueller”—demanded that I cease contacting him. Mueller has since retired as an exercise professor, after co-authoring football surveys at UNC from the 1970s to 2012.

The error on Quadaar White remains standing in NCCSIR reports at the website.

1990 Collision Deaths, Retrieved Online, Nullify Old UNC Claim

Now a substantial mistake has emerged involving Mueller, his inaccurate declaration publicized for three decades from Chapel Hill, such as it reads here, typos intact, on Page 1 of the UNC report for 2013 football deaths:

The 1990 report was historic in that it was the first year since the beginning of the research, 1931, that there was not a direct fatality in football at any level of play.(Mueller & Schindler 1991)  This clearly illustrates that data collection and analysis is important and plays a major role in prevention.

Wrong, at least on the first point.

Minimally two direct football deaths occurred during 1990, both of violent “sandlot” incidents, according to reports I recently retrieved from NewsBank database.

One case occurred merely 80 miles from Chapel Hill: Jamarl Gentry, 17, died on Nov. 7, 1990, of a broken neck suffered in a pickup tackle game at Winston-Salem, reported The Greensboro News & Record.

The second 1990 football death retrieved from NewsBank is Christopher Mock, 19, a college student from Bluffton, Ind. Mock died on Dec. 1 of a brain injury suffered in sandlot tackle football, reported The Fort Wayne News-Sentinel.

During December I forwarded these cases to Mueller, NCCSIR medical director Dr. Robert Cantu in Boston, and other officials like Kristen Krucera, PhD, the athletic trainer who’s replaced Mueller as center director. I requested their comments, repeatedly.

Only one person replied, Mueller, by email: “I told you to take me off of your mailing list,” he griped.

Modern Emergency Care Dramatically Cuts Football Deaths Since 1960s

NCCSIR officials argue that their dubious football statistics and recommendations like Heads Up “safe contact,” the latest version of stale “head up” theory, have reduced player deaths by as much as two-thirds since the 1960s.

Historical news shows no such evidence. Rather, the major reason for fewer reported football fatalities was faster, better medical care that kept seriously injured players alive, with an assist from modern helmets.

Most football fatalities in the 1960s were connected to inadequate medical care, based on news reports. Players died of brain bleeding, spinal fracture, chest impact, ruptured spleen, lacerated kidney, blood clotting, heat stroke, cardiac arrest and more maladies that became better managed in America by end of the Vietnam War.

“By the early 1970s, many influential members of medical society (in the United States) believed that lessons learned on the battlefields in Korea and Vietnam in terms of triage, rapid transport of trauma patients to definitive care centers, and standardization of pre-hospital and in-hospital care could be applied effectively to civilian patients,” recounts a Canadian medical review.

The 1970s advancement of emergency medicine in America—led by widespread establishment of EMTs, modular ambulances, life flights, emergency rooms and trauma surgery—saved countless athletes who would have died previously without it.

The steel head-and-neck brace or “halo cast,” breakthrough technology available nationwide by 1973, stabilized vertebral fractures that previously killed people. Numerous tragedies were averted in football and all walks of life, with immobilization techniques for spinal casualties and treatment like the halo brace.

Even Mueller admits emergency response has cut football deaths from brain injuries, compared to a half-century ago. “The line is going down with fatalities. I think that’s related to kids getting better medical care on the field,” Mueller told HealthDay.com, after reviewing news reports of two-dozen players who survived catastrophic brain bleeds in 2011. “They’re not dying, but they’re having permanent brain damage.”

Left unsaid? I had forwarded Mueller those cases, or he probably would’ve missed most.

UNC Changes in Death Definitions Trim Football Numbers Since 1960s

Factors beyond emergency care and improved helmets also have reduced football deaths in UNC records since 1960. Based on available information, numbers have been shaved as much as one-fifth since 1960 just by altering definitions to qualify game fatalities.

Background begins in 1931, when the American Football Coaches Association hired Floyd R. Eastwood, a college professor with a PhD in education, to record yearly football casualties—formerly the task of media entities like The Associated Press. News accounts describe Eastwood’s method for collecting cases and defining types of football death.

“Dr. Eastwood” analyzed news reports of deaths among football players for 35 years, working for both the coaches association and the NCAA, groups which continue to fund the NCCSIR today. Relying heavily on the NCAA’s “clipping service” of major newspaper and wire-service articles, Eastwood gathered stories of casualties and looked for football causes or possible links, sometimes basing a case decision on news content alone.

Eastwood had to track national football casualties while grounded on campuses where he taught PE pedagogy and gym classes. Limited in information access and funding, he tried to assimilate medical protocol despite a personal résumé far short in education and training for the mission.

Eastwood followed up many football incidents he learned of, making phone calls and mailing information forms to witnesses and authorities. But their responses likely varied in substance and it was difficult and costly to obtain medical files and death certificates from across the country. Moreover, Eastwood surely understood that official information was frequently tainted by simple incompetence and/or football allegiance among local authorities.

“Keep in mind…,” intoned Comrie, who has compiled football casualty data, “many doctors in many parts of the country don’t want to blame football. So on those death certificates, is that information reliable? I don’t know. We don’t have a clue.”

Most significantly, Eastwood defined and qualified several types of football death differently than his present-day successors at UNC.

For example, Eastwood believed that football exertion and impacts could trigger congenital brain bleeds in players, caused by “AVM” and Chiari arterial malformations since birth. Medical opinion was divided, but many doctors determined that football spurred these cerebral vessel ruptures of natural origin, and Eastwood embraced the stance.

UNC researchers dismiss these incidents today, meanwhile, like the 2010 death of college player Ben Bundy, killed by his genetic brain-artery malformation that launched a blood clot during a team workout. Bundy wasn’t counted in the annual NCCSIR report.

Eastwood included meningitis fatalities among players and field deaths of referees for his 1960s football statistics. Neither type of incident figures into current UNC data.

Eastwood counted players dead of blood clots originating from leg injuries, classifying them as indirect fatalities of the game. In recent decades, however, UNC publications include only an occasional death via non-cerebral blood clots; many additional cases are omitted without explanation, like Mounkes, the aforementioned schoolboy, and Ben Jordan, 16, a South Carolina prep player dead of a pulmonary embolism in 2012 after he was hospitalized for blood clots during successive football seasons.

Altogether, the types of deaths counted by Eastwood but not counted by UNC comprise as much as 20 percent of the 350 football fatalities from the 1960s that I’ve collected.

And the “sandlot” classification has become the biggest area of NCCSIR shell games that produce smaller numbers, over time, and inspire the rhetorical mirage of “safer” football today.

UNC Qualifies Few ‘Sandlot’ Fatalities for Football Statistics Since 1980s

Erstwhile PE professor Floyd Eastwood held a broad view of what constituted a “sandlot” death, qualifying any person who died from injuries suffered while playing any type of football: organized or informal; tackle or “touch”; games in vacant lots, flag leagues and PE classes; and even passing and catching during school recess. In 1962, for example, Eastwood counted a young father who died of striking a telephone pole during a backyard touch game with family as one of the professor’s 19 direct football fatalities that year.

Eastwood also logged indirect or game-related sandlot deaths, numerous cases in his 1960s data, for causes such as cardiac arrest and heat stroke.

Thus far, NCCSIR officials decline to provide me with the names and locations for football deaths in their multi-decade collections. They decline to address inconsistencies in data classifications back to Eastwood’s tenure. They do not offer, or possess, a single peer-validated document incorporating detailed cases, formal literature review and a complete research method that identifies limitations.

UNC transparency isn’t needed, however, to deduce that Mueller et al. have basically counted only “youth league” players for sandlot fatalities since 1986, based on NCCSIR postings, cases found online, and public statements of Mueller and colleagues.

UNC no longer counts fatalities of flag football, “touch” games, PE classes, “athletics classes,” recess periods, and intramural competition at colleges; deaths from tackle sandlot games are no longer included—yet all types still load those numbers from the 1960s and ’70s.

It would seem “sandlot” deaths were disappearing by the 1980s, according to football-funded researchers. Indeed, UNC statistics from 1986 to 1998 do not list any direct fatalities in the category.

On the contrary, numerous deaths of impacts occurred in the period that Eastwood would’ve counted for sandlot classification, including the following cases I’ve located:

1986: Ervin Kolk, 27, died of “being kicked in the head during a touch football game” at Tukwila, Wash., reported The Seattle Times.

1987: Joshua Arruda, 12, died of “injuries he received when he fell and hit his head on a rock during a tag football game at school,” reported The Daily News of Los Angeles.

1988: An adult male, unidentified, died of “a severed aorta after taking a blow to the chest” in touch football at North Conway, N.H., reported The Sporting News.

1989: Walter Jackson, 27, died “as a result of the head injury he suffered” in touch football, reported The Buffalo News.

1990: Jamarl Gentry, 17 (aforementioned case), died of a broken neck sustained in pickup tackle football at Winston-Salem, N.C., reported The Greensboro News & Record.

1990: Christopher Mock, 19, (aforementioned case), died of head injuries sustained in pickup tackle football in Indiana, reported The Fort Wayne News-Sentinel.

1991: Timmy Hysinger, 29, died “of a head injury he suffered… playing touch football in the street” at Mauldin, S.C., reported The State newspaper.

1994: Chris Hart, 18, dead “from head injuries suffered in a flag football game” at Texas A&M University, reported The Houston Chronicle.

1996: Terry Crayton, 16, died after “being knocked unconscious in a gym-class collision… playing a game called ‘speedball,’ a combination of soccer and football” at a Milwaukee school, reported The La Crosse Tribune.

1996: Jason Boone, 19, died of “receiving a severe head injury… in a touch football game” at Maryville, Tenn., reported The Knoxville News-Sentinel.

1996: Derek McMillen-Morgan, 16, died of “massive spinal injuries (sustained) when tackled” during a pickup game at Canton, Ohio, reported The Akron Beacon Journal.

The following deaths also are omitted from UNC statistics that log zero “sandlot” collision fatalities in said years:

2000: Maurice Doty, 16, “died of cardiac arrhythmia due to blunt force impact of the chest” in a pickup tackle game at Dayton, Ohio, reported The Daily News.

2005: Kenny Luong, 19, “died from (head) injuries received during a UC Irvine fraternity football game,” reported The Orange County Register.

2005: Steve Lynes, 19, “died of (head) injuries suffered during a pickup football game” at Brigham Young University, reported The Associated Press.

2005: Robert Meza, 24, died of a brain injury sustained in flag football at Taylor, Mich., reported The Detroit News.

2006: Logan Honsinger, 10, “died after his diaphragm was ruptured, an injury authorities… suspect he received during practice” for his youth-league team at Hemlock, Mich., reported The Associated Press.

2006: Andre Thibault, 12, “died from injuries suffered… when he tripped and fell into a pole while playing football” at Halstead, Kan., reported The Kansas City Star.

2008: John Buzzard, 15, “died of heart and brain-related conditions” sustained “during a touch-tackle football game” at Brooklyn, N.Y., reported The Staten Island Advance.

2008: Coty Bluford, 14, died of injury sustained when he knocked “heads with another boy” in football play during PE class at school in Lenoir City, Tenn., reported The Associated Press.

2008: Dominique Edwards, 19, died of a ruptured kidney sustained “when he dove for a football… and struck his left side” in a pickup game at Macon, Ga., reported The Telegraph.

2008: A boy, unidentified, 11, died “after being struck in the throat during a recess game” at school in Lake Oswego, Ore., reported The Oregonian.

2012: Jacob Gatlin, 15 (aforementioned case), died of skull fracture and brain hemorrhaging after a collision during a 7-on-7 passing session in school “athletics class” at Hawkins, Texas, reported The Longview News-Herald.

2012: Alex Lott, 17, died of a neck fracture “received playing touch football” at Richton, Miss., reported The Jackson Clarion-Ledger.

Again, these cases require examination by accredited authorities for qualification in valid football research, but this news batch demonstrates the stark, unannounced change between 1960s record-keeping and modern death data from UNC.

“We hear, ‘Oh, football’s become safer,’ ” Comrie said. “I don’t know if it’s become safer. I’m hearing this but I don’t know. Because no one’s willing to sift through the (UNC) data, we don’t know how many kids died or what they died of. We don’t know how many football deaths were purely preventable.”

Comrie, who has consulted for the U.S. Air Force and NFLPA regarding brain injuries and assessment, sees professional incompetence for NCCSIR publications that claim to reduce football mortality rates.

“It’s certainly uninformed,” he said. “Science is about asking questions. If you have incomplete or inaccurate data, you’re probably asking the wrong questions. The lack of information is bad for everybody, including the NFL, the NCAA and the national high schools.

“Reliable data is the key to making strong recommendations about what to do, but decisions are being made in football based on nothing,” Comrie said. “Everyone should know the data is just crap, period.”

Like Eastwood in the past, UNC researchers erroneously announce football deaths in absolute numbers, never mind that their primary means of incident details–news content–remains inadequate. Any accuracy would be a lucky guess and unverifiable anyway.

The total for a given year “could be stone-cold accurate or it could be off by a million miles,” said Yesalis, the epidemiologist. “It’s all based on how you count the population at risk, football players. It’s how you count the death events, how you acquire information, with bias for relying on news reports. Even of a death, there could be an injury where the kid is taken from the field and dies maybe three weeks later. Well, how confident can you be that will be reported (in news)?

“Basically going by news media alone? No epidemiologist would say that’s ideal.”

Comrie believes enough information exists empirically to resolve deficiencies through collaborations among investigators like me, the UNC academics, and appropriate authorities from medicine and science. Amassing death certificates would be a scientific start.

Vital data on football mortality risk “probably exist somewhere, in some form,” Comrie said, but politics stymie progress. “They (game officials) have just made it as difficult to get and to analyze as they possibly can because they want no change in the status quo. I have my beef in all this, because I can’t make reasoned decisions because the data’s unreliable. And the scientific community just goes along.

“We’ve been operating in a world where no one’s ever checked to see if the (football) data are real or not,” Comrie said. “They publicize it, of course, under UNC, but what was their research method? How did they do it? But since the method, regardless, produced exactly what the media wanted to hear and the parents wanted to hear, no one’s ever questioned it.”

Modern Myth of Safer Football, Research Heroes and Saving Lives

In 1998, the United States Sports Academy gave Fred Mueller an award for “lasting contributions to the growth and development of sports medicine through practice or scholarship.”

A UNC-Chapel Hill press release heralded the university’s “Life-Saving Dr. Mueller,” stating:

Statistics he compiled, first with Dr. Carl Blyth, also of UNC-CH, on football injuries and deaths helped lead to rule changes and improved coaching credited with saving dozens of lives a year in the United States.

Comrie scoffs in New York, pondering what’s really happened around the so-called National Center for Catastrophic Sports Injury Research—which has no street address in Chapel Hill, no facility on campus, nothing of validated research.

It’s all a façade fronted by Mueller types and feel-good statistics, Comrie alleges, designed to lead naïve kids and parents along the Yellow Brick Road to Safer Football.

“It’s the mythology being confused for reality,” he surmised. “And the way to promote mythology is to make sure there are no hard facts or evidence.

“So we go through the curtains and we find out the great, mysterious Oz is not who he appears to be.”

Matt Chaney is a writer, editor, teacher and restaurant cook in Missouri, USA. Chaney’s 2001 MA thesis at the University of Central Missouri involved electronic search for thousands of news reports on performance-enhancing drugs in American football, a project inspired by his experience of injecting testosterone as a college player in 1982 (Southeast Missouri State). Email him at mattchaney@fourwallspublishing.com. For more information, including about Chaney’s 2009 book Spiral of Denial: Muscle Doping in American Football, visit the homepage at www.fourwallspublishing.com.