Eric D. Zemper, Ph.D.

Exercise Research Associates of Oregon

Football injury data were collected by the Athletic Injury Monitoring System (AIMS) from a national sample of 56 high schools and 46 colleges for the 1997 season. AIMS is a national sports injury data collection system capable of doing injury surveillance on a variety of sports, and which meets the major criteria for reliable studies of sports injury rates outlined in 1987 by the American Orthopaedic Society for Sports Medicine [Am J Sp Med 15:117-124 (1987)]. This paper presents some of the results dealing specifically with head injuries in football. The high school sample was distributed by four geographic regions, and the college data was distributed by four geographic regions and by size of program (NCAA and NAIA divisions). Statistical analyses of the distributions in these samples indicated that the college sample was representative of the entire country by region and size of program, and therefore the results presented here for the collegiate level can be reasonably projected to the country as a whole. The nature of the high school sample precludes a rigorously representative sampling. The data collection procedures require a medically trained person on site (specifically, the certified athletic trainer), and only a relatively small proportion of the more than 1700 high schools that sponsor football teams in the country have athletic trainers on staff (compared with essentially all colleges). Very few of the athletic trainers who are working at the high school level are in smaller schools. In this case the most reasonable approach is to sample a large number of high schools (albeit a small proportion of the total) and at least keep the sample representative by geographic region. The 56 high schools in this sample were statistically representative by geographic region, and with over 4000 players it is considered adequate for the purposes of this project.

Utilizing simple, easy-to-use forms, data on exposure to the possibility of injury in practices and games, and on any injury that kept a player out for one day or more (including any head injury whether on not time loss was involved), were returned on a weekly basis throughout the season by certified athletic trainers at each participating school. As data arrived, it was logged in and screened for completeness and consistency before being entered into computer storage files. One problem with surveillance systems of this sort is incomplete data submission from the field creating "gaps" in the data. To prevent this problem in this project, the AIMS staff maintains frequent mail and telephone contact with the athletic trainers, obtaining corrections of any inconsistencies in the reported data and reminding them to submit missing data. Through this process of "preventive maintenance" the response rate for data submission was extremely high. For the college sample 99.8% of the weekly reports were submitted, and for the high school sample the response rate was 100%.

In addition to the weekly exposure and individual injury forms, at the beginning of the season the athletic trainers completed a form indicating what football helmet models were being used by their teams, and how many players had a history of cerebral concussion anytime during the previous five years. The data on exposures and on injuries were used to calculate injury rates based on the following equation:

((# injuries reported) ÷ (# exposures)) x 1000 = injury rate per 1000 Athlete-Exposures

(An Athlete-Exposure is one player participating in one game or one practice where he is exposed to the possibility of being injured.) The data collected for the 1997 season included 8782 player-seasons (4311 HS and 4471 college) and 612,527 athlete-exposures in games and practices (297,390 HS and 325,137 college). The total injury rate for the high school players was 5.06/1000 A-E and for college players it was 5.91/1000 A-E.

During this season there were 360 reported concussions (157 HS and 203 college). The cerebral concussion rates at the high school and college levels were quite similar, with a rate of 0.55 concussions per 1000 Athlete-Exposures (A-E) for high school players and 0.60 per 1000 A-E for college players. The combined rate is 0.58 concussions per 1000 A-E. Looked at another way, this is equivalent to one concussion in every 1725 times a player participates in a game or practice. For a team of 100 players (generating 100 A-E in every daily practice and every game, assuming they all played in the game), this would be one concussion in every 17.25 days of activity. A less accurate and less useful "rate" that is still commonly seen in sports medicine literature is the number of injuries per 100 players. For concussions in this sample, there were 3.78 concussions per 100 high school players and 4.54 concussions per 100 college players. While this appears to indicate that the rate for college players is about 20 percent higher, because college players typically have more exposures (practices and games) during a season, the actual rates for high school and college players are really quite similar. This is one illustration of how trying to make comparisons using injuries per 100 players as a rate can be misleading.

Because this project collects data on all types of injuries, it is possible to look at the occurrence of cerebral concussions in relation to other types of injuries. The head is the third most frequently injured body part at the high school level, exceeded only by the ankle and the knee. At the college level, the head is the fourth most frequently injured body part, following the knee, ankle and shoulder in that order. With regard to type of injury, cerebral concussions are the fourth most frequent injury at both levels, after ligament sprains, muscle strains and contusions. What cannot be seen in these raw data summaries is how cerebral concussions rank in relation to other specific body part/injury type combinations (for instance, not all knee injuries are ligament strains; they include injuries such as contusions and meniscus injuries as well). When broken down further, cerebral concussions turn out to be the second most frequent injury at the high school level (ankle sprains being the most common), and the third most frequent injury at the college level, following ankle sprains and knee ligament sprains and tears.

The rate of concussions in practice for high school players was 0.30 per 1000 A-E, while in games it was 2.33 per 1000 A-E. For college players the concussion rate in practices was 0.39 per 1000 A-E, and in games it was 3.03 per 1000 A-E. Both high school and college players were 7.8 times as likely to sustain a cerebral concussion in a game as they were in practice. This is not surprising, since the rate for all injuries in games has previously been found to be about eight times higher than in practices. This most likely is due to a much higher intensity level that is sustained throughout a game as compared to most practices.

At the high school level, 52.1% of the concussions occurred in offensive players and 39.9% in defensive players, with the remaining 8.0% in kickers and special team players. At the college level the situation was reversed, with 38.4% of the concussions occurring in offensive players and 55.7% in defensive players. The remaining 5.9% occurred in kickers and special team players. This difference between high school and college also is reflected in the mechanism of injury for concussions. At the high school level, being tackled accounted for 27.6% of the concussions and blocking accounted for 19.6%, while tackling was the mechanism for 21.5% and being blocked was the cause of 14.7% of the concussions. Tackling was the causative mechanism for 28.6% of the college concussions and being blocked for 22.7%. On the offensive side of the ball, being tackled caused 20.7% and blocking caused 14.8% of the college concussions. At both the high school and college levels, impact with the playing surface caused about 4% of the concussions.

The athletic trainers noted whether or not an injury was directly caused by impact from another player’s helmet. Combining the high school and college data, direct impact from another player’s helmet appears to be the causative mechanism of a minimum of 16.6% of all football injuries, and possibly as much as 26.0% of all football injuries. (This is estimated by using the percentage of injuries definitely attributed to helmet impact as the minimum value and adding the percentage of injuries where the athletic trainer indicated that it was uncertain whether or not a helmet was involved, to give a maximum estimate.) When only cerebral concussions are considered, impact from another helmet was the direct cause of 55.5% of all cerebral concussions in football, and possibly as much as 76.3% of all concussions in this sport.

This project uses a grading system for severity of cerebral concussion developed by the American Academy of Neurology [Neurology 48:581-585 (1997)]. Grade 1 on this scale indicates transient confusion that resolves in less than 15 minutes (the "bell ringer"); Grade 2 indicates symptoms that last longer than 15 minutes, but there is still no loss of consciousness; and Grade 3 is used for any concussion involving loss of consciousness. For the purposes of this project, in order to provide greater sensitivity at the "low" end of the scale, we use an additional grade, Grade 0, to pick up situations where there is no immediate indication of concussive injury but the player later complains of headache and difficulty concentrating. From the data collected during this first season, the average time loss for a Grade 1 concussion was 4.4 days, the average time loss for a Grade 2 concussion was 16.2 days and for Grade 3 it was 10.3 days. The average time loss for those injuries classified as Grade 0 was 5.5 days. Because the average number of days lost can be skewed to the high side by a few instances where players were held out for the rest of the season, more informative statistics in this case would be the median and the mode. (The median being the midpoint in a listing of the number of days lost per injury in ascending order, and the mode being the most frequently occurring number of days lost.) The median number of days lost for Grade 0 concussions was 3.5 days, for Grade 1 concussions it was 2 days, for Grade 2 it was 6.5 days, and for Grade 3 it was 8 days. The mode for Grade 0 was 1 day, for Grade 1 it was 1 day, for Grade 2 it was 2 days and for Grade 3 it was 8 days.

On the forms the athletic trainers used to submit data on the helmets being used by their teams, they also indicated who was primarily responsible for fitting the football helmets and whether or not these individuals had received specific training on helmet fitting. At the college level just over 60% of the responsible individuals were equipment managers, of whom all but 14% had specific training in fitting helmets. About a quarter of the individuals were athletic trainers, all of whom had training. Just over 20% of the responsible individuals were coaches, and 60% of them did not have any training. All of the individuals at the college level who did not have training were in the smaller (Division II and Division III) schools. There were a total of 19% of the individuals responsible for fitting helmets at the college level who did not have any specific training. At the high school level the situation was different, because high schools usually do not have professional equipment managers and the coach is most often the one responsible for fitting helmets. Coaches comprised nearly 70% of the responsible individuals, and 18% of the coaches did not have any training in fitting helmets. Most of the rest of the responsible individuals were the athletic trainers, and nearly all of them had training. Overall, in this sample about 15% of the individuals responsible for fitting helmets at the high school level had no training. Given that most high schools do not have either equipment managers or athletic trainers, it can be estimated that at about one in five high schools the person responsible for fitting football helmets has no training to do so.

An unknown factor for many years has been the relative risk for sustaining a cerebral concussion among football players who have previously received such an injury compared with players with no history of concussion. As part of the data collected from the athletic trainers regarding helmets used by their teams, they also indicated how many of their players had any history of concussion during the previous five years. From this information, plus data from the individual injury forms indicating whether or not a concussion being reported occurred in a player with a previous concussion, it is possible to calculate a relative risk for those players with a previous history of cerebral concussion.

At the high school level the sample for the calculation of relative risk consisted of 4311 players (data from two teams had to be deleted when the athletic trainers were not able to provide complete data regarding previous history). Of those, 187 players (4.3%) had a history of concussion sometime during the previous five years. There were a total of 158 concussions recorded in this sample, 30 among those players with a previous history, and 128 in players with no previous history. Using this information, the relative risk is calculated as follows:

Among those high school players with a previous history of cerebral concussion, 16% sustained a new concussion during the season, while among the players with no previous history of concussion only 3% sustained a concussion. At the high school level the relative risk of sustaining a cerebral concussion during the season among those players with a previous history of concussion is just over five times greater than for those with no previous history of concussion.

At the college level the sample for this calculation consisted of 3696 players (data from seven teams had to be deleted because of incomplete information on concussion history). A total of 347 players from this sample (10.4%) had sustained a concussion within the previous five years. Out of the total of 154 concussions recorded during the season, 46 occurred in players with a previous history and 108 occurred in players with no previous history. The relative risk is therefore calculated as follows:

Just over 13% of college players with a previous history of cerebral concussion sustained a new concussion during the season, while just over 3% of those players with no previous history sustained concussions. At the college level the relative risk of sustaining a cerebral concussion during the season among those players with a history of concussion during the previous five years is slightly more than four times greater than for those with no previous history of concussion.

Combining the high school and college data results in the following calculation of risk:

Therefore, in this large national sample of high school and college football players, the risk of sustaining a cerebral concussion during the season was 4.5 times greater for those players with a history of concussion during the previous five years than for those players with no previous history of concussion.

It has been known for many years that players with a previous concussion appear to be more likely to incur a new concussion than players with no history of previous concussions, but there has been little data on exactly how much more likely. The data presented here indicates that players with a history of concussion within the previous five years are 4.5 times as likely to suffer a new concussion during the season than those with no history of concussion. Among other areas, this information will have implications for questions regarding when it is appropriate to allow a concussed player to return to activity, and for research on the "second impact" syndrome.

Several of the results presented here, when taken together, appear to imply the need for a more vigorous and sustained education program for athletes and coaches regarding the head and the helmet in this sport. The fact that cerebral concussion is the second or third most frequent injury by itself should raise some warning flags, especially when considering research showing that measurable cognitive deficits in memory and information processing for up to thirty days occur following a closed head injury, even when there is no loss of consciousness [e.g., Lancet 2:605-609 (1974); Lancet 2:995-997 (1975)]. It should be kept in mind that these players have to perform in the classroom as well as on the field. In addition, anywhere from one-half to three-quarters of all football concussions involve direct impact from another player’s helmet, which implies that the head is still being used as an initial contact point far too often. Given the nature of this sport, there undoubtedly are many instances where such contact is accidental or unavoidable. But when the helmet is the causative agent in three to four times the percentage of concussions as it is for injuries in general, this is a strong indication that players still tend to drop their head and use it as a battering. (This also has implications for the risk of neck injuries.) This is supported by the observation that tackling and being tackled are generally the most frequent mechanisms for concussion. Finally, the observation that at about one in five high schools the person responsible for fitting helmets has no training to do so (and nearly that proportion of the smaller colleges) has implications for ensuring the optimal protective capabilities of the helmet as well. To a certain degree all of these problems may be inter-related, and can be addressed by a comprehensive education program aimed at players (and their parents at the high school level), coaches and medical personnel. It needs to be more coordinated and more comprehensive than the sporadic attempts of the past. And as noted at another session on concussion earlier at this conference, it should include educating these groups about the importance of not returning to play while concussion symptoms are still present. It also would be appropriate to encourage governing bodies (state and national high school associations, NCAA, NAIA, etc.) to put much more emphasis on enforcement by game officials of the existing rules regarding use of the head as an initial contact point.

Concussions occur more frequently in football than many people seem to realize, the majority of them the result of head-to-head contact. Since a previous concussion makes a player 4.5 times more likely to suffer a new concussion, and since there is growing evidence of a "second impact syndrome" and of measurable cognitive deficits for up to a month after even mild head injury, the common practice of sending a player back into action as soon as he can "see straight" needs to be re-evaluated.