Search tips
Search criteria 


Logo of jathtrainLink to Publisher's site
J Athl Train. 2007 Jan-Mar; 42(1): 11–19.
PMCID: PMC1896066

The Influence of Various Factors on High School Football Helmet Face Mask Removal: A Retrospective, Cross-Sectional Analysis

Erik E Swartz, PhD, ATC,* Laura C Decoster, ATC, Susan A Norkus, PhD, ATC, and Thomas A Cappaert, PhD, ATC, CSCS§


Context: Most research on face mask removal has been performed on unused equipment.

Objective: To identify and compare factors that influence the condition of helmet components and their relationship to face mask removal.

Design: A cross-sectional, retrospective study.

Setting: Five athletic equipment reconditioning/recertification facilities.

Participants: 2584 helmets from 46 high school football teams representing 5 geographic regions.

Intervention(s): Helmet characteristics (brand, model, hardware components) were recorded. Helmets were mounted and face mask removal was attempted using a cordless screwdriver. The 2004 season profiles and weather histories were obtained for each high school.

Main Outcome Measure(s): Success and failure (including reason) for removal of 4 screws from the face mask were noted. Failure rates among regions, teams, reconditioning year, and screw color (type) were compared. Weather histories were compared. We conducted a discriminant analysis to determine if weather variables, region, helmet brand and model, reconditioning year, and screw color could predict successful face mask removal. Metallurgic analysis of screw samples was performed.

Results: All screws were successfully removed from 2165 (84%) helmets. At least 1 screw could not be removed from 419 (16%) helmets. Significant differences were found for mean screw failure per helmet among the 5 regions, with the Midwest having the lowest failure rate (0.08 ± 0.38) and the Southern (0.33 ± 0.72), the highest. Differences were found in screw failure rates among the 46 teams (F1,45 = 9.4, P < .01). Helmets with the longest interval since last reconditioning (3 years) had the highest failure rate, 0.47 ± 0.93. Differences in success rates were found among 4 screw types (χ21,4 = 647, P < .01), with silver screws having the lowest percentage of failures (3.4%). A discriminant analysis (Λ = .932, χ214,n=2584 = 175.34, P < .001) revealed screw type to be the strongest predictor of successful removal.

Conclusions: Helmets with stainless steel or nickel-plated carbon steel screws reconditioned in the previous year had the most favorable combination of factors for successful screw removal. T-nut spinning at the side screw locations was the most common reason and location for failure.

Keywords: cervical spine, emergency management, airway access, football injury

Key Points

  • Stainless steel or nickel-plated screws best predicted successful removal of the face mask.
  • Longer interval lengths between helmet reconditioning were associated with higher screw removal failure rates.
  • A total of 84% of face masks were successfully removed with the cordless screwdriver.

Football helmets must be designed and constructed in such a way as to meet specific certification standards imposed by the National Operating Committee on Standards for Athletic Equipment (NOCSAE).1 These specifications were devised to protect the wearer during head impact. Used helmets must continue to meet those standards and, toward that end, helmets are inspected and recertified according to NOCSAE standards.2 This recertification takes place when helmets are periodically sent for reconditioning to NOCSAE-approved agents. Although the NOCSAE standards address impact management, they do not consider emergency face mask (FM) removal requirements. Such requirements may be important in light of recent research establishing that the design of football equipment has an effect on the amount of movement, time, and degree of difficulty associated with removal of the FM.3

The importance of minimizing the time required and the movement created during football helmet FM removal is reflected in the recommendations of the Inter-Association Task Force for the Appropriate Care of the Spine-Injured Athlete, which recommended removing the FM “as quickly as possible and with as little movement of the head and neck as possible.”4 The Task Force further recommended that “all loop-straps of the face mask be cut ….”4 Since publication of those guidelines, however, researchers have shown this task to be relatively difficult3 and sometimes unacceptably time intensive.3,5 Despite the availability of cutting implements manufactured expressly for this purpose, cutting loop straps has been reported to require significantly more time3,6 and to create more movement3 or place more torque on the helmet6 than removing them with a cordless screwdriver. Furthermore, failure to successfully remove the FM in a recent study with the cutting tool was more common than was failure with the cordless screwdriver.3 Based on current literature, therefore, the screwdriver minimizes the time required and the movement created during FM removal and may be more reliable than originally theorized.

However, an important deficiency exists in the current body of literature on this topic. Most authors investigating FM removal have used new equipment,3,5–8 and concerns persist about the screwdriver's effectiveness on hardware that has been exposed to the elements. Past research, therefore, is limited in clinical applicability to equipment that has been used throughout a season with regard to the potential effects of the environment and daily wear and tear.

Recently, a small cross-sectional FM removal study was performed on high school football helmets at the end of the 2003 football season.9 Researchers performed FM removal on 222 helmets at 3 local high schools in northern New England. During the removal trials, helmets were secured to a board and screws were removed with a cordless screwdriver. The overall FM removal success rate was 82.4% (183/222). However, 1 of the high schools had a success rate of only 52% (24/46). The authors suggested that a possible cause for a low success rate may have been that certain hardware materials were subject to rust and corrosion. They also suggested that differences in helmet brands may be a cause for different failure rates.9 However, because of the relatively small number of helmets, limited geographic diversity, and unknown metallurgic characteristics of the screws, the authors cautioned that the conclusions were speculative and that further research was warranted in order to establish potential relationships.9

Therefore, we thought a larger, more representative study considering such factors for helmeted (football) athletes was required. We theorized that differences in weather characteristics among the regions of the country could affect the ability to remove the screws from FMs and that the metal composition of the helmet hardware (screws and T-nuts) also could influence FM removal. We also expected that longer intervals between helmet reconditioning would increase rates of screw failure. The purpose of our study was twofold. The first aim was to assess the ability to remove FMs from high school football helmets using a cordless screwdriver after at least 1 season of play and to compare success rates among 4 associated factors (region, team, helmet characteristics, and reconditioning/recertification history). Therefore, we developed 4 specific research hypotheses to guide our methods and design: (1) No differences would be seen in FM removal success rates among football helmets from 5 regions of the country. (2) No difference would be seen in FM removal success rates among football helmets from different teams. (3) No differences would be seen in FM removal success rates among football helmets containing screws with different metallurgic characteristics. (4) No differences would be seen in FM removal success rates among football helmets reconditioned in different years. Our secondary purpose was to determine the level of association between FM removal success and various factors (eg, helmet brand and model, helmet components, region, environment) that influence the condition of the football helmet and associated equipment in order to identify the most important predictive variables.


Project Design

A cross-sectional retrospective research design was created for this study. The independent variables included geographic region based upon The Library of Congress Web site10 (Northeast, Midwest, Pacific, Rocky Mountain, and Southern), regional weather characteristics (averages for precipitation, high temperature, low temperature, daily temperature, percentage humidity, barometric pressure, and elevation of school's geographic location), helmet brand (eg, Riddell, Schutt), helmet model (eg, VSR4, Air Advantage, Revolution), hardware components (loop strap type, screw type, presence of T-nut walls, presence of washers, type of washer), year helmet was last reconditioned, and reconditioning/recertification plant from which data were collected (5 NOCSAE-certified athletic reconditioning/recertification facilities). Our dependent variable was FM removal success or failure.

A pilot study conducted at a local reconditioning/recertification plant on 61 helmets allowed us to estimate the number of helmets, teams, and data collection hours that would be needed to meet a minimum goal for the study. Based on pilot data and previous research,9 we determined that a minimum of 5 high school football teams' helmets from each of the 5 regions of the country, providing at least 300 helmets per region (1500 helmets overall), would serve as the minimum sample size necessary for our comparisons.

Sample Selection

Between May and December 2004, we contacted all National Athletic Equipment Reconditioners Association (NAERA) plants by fax, e-mail, and phone to explain the purpose of our study and to request their participation. Of the plants that responded, 5 were selected based on geographic location, operating schedule, availability, and customer profiles. Before our arrival, we sent a list to a representative from each plant with the number of teams and approximate number of helmets from each specific region needed during the visit. All schools whose helmets fit our criteria were identified, and their helmets were set aside by the plant representative. We attempted to include schools from different regions at each plant to reduce the influence of each individual plant's reconditioning/recertification process on the total sample.


Face mask removal was performed using 7 battery-operated, rechargeable, cordless screwdrivers (3.6-V pivot driver; Black & Decker, Towson, MD). Screwdrivers were outfitted with No. 2 Phillips head screwdriver bits. Two custom-made helmet mounts were designed and constructed specifically for this study (Figure). The mounts were fixed to the top of a standard-height table using clamps to prevent them from moving during the data collection process. The helmet mounts allowed for secure fixation of the helmet without the need for an additional investigator to provide stabilization. The metallurgic analyses were performed using a scanning electron microscope (model SM-300; Topcon, Tokyo, Japan), a metallograph (model Neophot 21; LECO Corp, St Joseph, MI), and an Olympus binocular microscope (model SZ-61; Tokyo, Japan) with a PaxCam digital camera (model PX-CM; MIS, Inc, Franklin Park, IL).

figure i1062-6050-42-1-11-f01
Helmet strapped to mounting fixture before attempted removal of face mask


A total of 5 investigators (certified athletic trainers with a minimum of 8 years' experience) contributed to the FM removal process throughout the study, with either 2 or 3 of the 5 being present at each location. Before data collection, all investigators involved in FM removal practiced the techniques using the screwdriver. In order to establish intertester reliability, we performed pairwise comparisons among all the investigators' success rates after the study and found a difference in failures between only 2 of the testers. However, this difference may not be attributable to skill, because teams and regions differed in the number of helmet failures that occurred. In other words, these 2 testers' rates may have differed simply because by chance, one came across more failures than the other did. In order to further explore intertester reliability, we also performed an additional discriminant analysis using investigator as an independent variable. The result was that investigator had the lowest prediction, or least effect, on success rate in removing the FM. Based on these results, we concluded that there was no effect from the different investigators on the results.

We traveled to each of the 5 reconditioning/recertification plants to collect data on 5 separate trips between February and April 2005. Upon arrival, the plant representative designated an area within the facility specifically for the data collection process. We then were provided with an inventory of available helmets from schools that met our geographic criteria. From each list, we selected schools for the data collection session based on school location, the number of helmets available, and our current need in each region. Helmets were not visible during the selection process. Once the school's helmets were retrieved, the school name and address were recorded so that we could later obtain the school demographics and environmental characteristics of the area. Each of the helmets that the high school had sent in for reconditioning/recertification then was removed from storage, one at a time, and was assigned an identification number. The following helmet information was then recorded: helmet brand and model, the most recent reconditioning year (identified by either the NAERA plant sticker or NOCSAE sticker or both), loop strap brand and model at each of the 4 screw locations (left ear, left top, right top, right ear), screw color (screw samples to be analyzed later to determine metal content) at each of the 4 locations, the presence of T-nut walls molded into the inside of the shell as an aid to screw removal, and the presence of washers within the helmet.

After the helmet demographics were recorded, we strapped each helmet to a mount. Face mask removal then was attempted using a cordless screwdriver to remove the screws from each of the 4 screw locations. The 7 screwdrivers and 10 fully charged batteries were used throughout the study and were rotated every 45 helmets to ensure consistent torque output and to minimize wear on the screwdriver bits. During data collection, we rotated the responsibility for removing screws every 15 helmets to eliminate fatigue. Upon completion of the FM removal attempt, the investigator documented success or failure to remove each screw at each screw location. If the screw could not be removed, it was classified further by its reason for failure. The 4 failure categories created for the study were (1) screw failed (slots stripped), (2) T-nut spinning, (3) screw and T-nut fused, and (4) other (ie, a foreign substance in screw head).

During the FM removal process, we randomly selected 2 to 5 representative screws from each team and placed them in individual storage bags labeled with the school's identification number. We also randomly selected and used alternate means to remove screws that failed to be removed by the screwdriver and stored them in the same manner. After data collection, metallurgic analysis was performed by a private laboratory (New Hampshire Materials Laboratory, Inc, Somersworth, NH). This analysis included examination under a low-power binocular microscope and subsequent analysis using energy-dispersive spectroscopy. Energy-dispersive spectroscopy provides a semiquantitative elemental analysis of the analyzed surface and can be used to detect the presence of elements with atomic numbers of 6 (carbon) or greater. The purpose of these analyses was to determine alloy composition for the 4 screw color classifications we recorded through gross observation and also to attempt to explain the causes for screw corrosion and failure. In the initial classification, screws were identified by color: silver, black, gold, or “undetermined” (degraded, corroded, or rusted to the extent that we were unable to determine the original metal color).

After the FM removal data for all helmets from each plant location had been collected, we obtained demographic information for each school, including school enrollment, school type (public versus private), length of season, and primary playing surface (grass or artificial turf). Retrospective weather data were obtained for the geographic location for each school for the duration of the specific season, including averages for the following items: high and low temperature, daily temperature, precipitation, humidity, and barometric pressure. The elevation for the school's geographic location also was obtained.

Statistical Analysis

We tracked frequencies of all the helmet demographics (eg, screws, walls, loop straps), means (±SDs) of the weather data, frequency of failures, and the mean number of screws failed per helmet. The total number of screw failures per helmet was recorded, and these totals were then aggregated for each region and team. Separate analyses of variance with Tamhane post hoc comparisons were conducted to compare failure rates among the 5 regions, 46 teams, and 6 reconditioning year categories (new helmet, 2004, 2003, 2002, 2001, no label). Failure rates for each screw classification were examined. The number of screws successfully removed by type was aggregated and a χ2 test for independence was used to compare the differences in success rates among the 4 types of screws. A multivariate analysis of variance with Tamhane post hoc comparisons was conducted to examine differences among each of the regions for the following variables: precipitation amount, average high temperature, average low temperature, average daily temperature, average humidity level, average barometric pressure, and elevation. We performed a discriminant analysis to determine whether 15 variables (the 7 weather measures, region, helmet brand and model, year of reconditioning/recertification, and screw type at each of the 4 locations) could predict the successful removal of the FM. All tests were performed on a confidence level of .05 with SPSS (version 13.0; SPSS Inc, Chicago, IL).


Sample Descriptive Data

Helmets from 42 public and 4 private high schools were included in the sample. The mean enrollment at the schools was 734.9 ± 541.3 (range = 42–2450). The teams participated in an average of 10.6 ± 1.7 games and 13.4 ± 2.3 weeks of practice, with the earliest starting August 2, 2004, and the latest ending December 18, 2004. Only 2 of the schools played and practiced primarily on an artificial surface. A total of 2584 helmets were sampled and had FM removal attempted. The schools were categorized based upon the 5 regions (Table 1). The helmet brands found in the sample included Riddell (Elyria, OH: 1027 helmets), Schutt Sports (Litchfield, IL: 1469 helmets), and Adams USA (Cookeville, TN: 88 helmets). The helmets were classified further based upon model (Table 2). The helmets were examined for the presence of T-nut walls; 75.8% (1957) had no walls, 19% (492) had 2 T-nut walls, and the remaining 5.2% (135) had 4 T-nut walls. Washers were installed under the T-nuts in 70% (3591/5168) of the ear locations of these helmets. When washers were present, most of these washers were metal (2532, 49%). The remaining washers were plastic (1059, 20.5%); 1577 locations (30.5%) had no washer present. The helmets also were examined for evidence of previous reconditioning, and the majority of helmets had been reconditioned within the previous 2 years (Table 3). The color of screw found at each of the 4 locations on each helmet also was examined for classification. Out of 10 336 screw locations, 6849 (66.3%) had silver screws, 2010 (19.4%) were black, 1374 (13.3%) were gold, 51 (0.5%) were of an unknown or undetermined color, and 52 (0.5%) of the locations had a missing screw. In some instances, a single helmet had more than one type or color of screw installed. The most common loop strap found on the sample helmets was the standard ArmourGuard (Schutt Sports)–style loop strap (4862/47%).

Table 1
Distribution of Sample Schools and Helmets Based Upon Geographic Region (n = 2584)
Table 2
Distribution of Sample Helmets Based Upon Brand and Model (n = 2584)
Table 3
Distribution of Sample Helmets Based Upon Previous Reconditioning (n = 2584)

Face Mask Removal and Screw Failure

Of the 2584 helmets assessed, 2165 (84%) successfully had all 4 screws and the FM removed. This resulted in a failure rate of 16% (419/2584), in which at least 1 of the screws could not be removed with the screwdriver (Table 4). Of those helmets, 66% (277/419) had 1 screw that could not be removed, 25% (107/419) had 2 screws that could not be removed, 5% (21/419) had 3 screws that could not be removed, and 14 helmets (3%) had all 4 screws that could not be removed. Differences were found among the 5 regions for mean screw failure per helmet (F4,2579 = 16.62, P < .01). The Midwest region contained the lowest failure rate (0.08 ± 0.38), and helmets from the Southern region had the highest failure rate (0.33 ± 0.72) (Table 5). Significant differences were found among the 46 teams for mean screw failure per helmet (F1,45 = 9.4, P < .01). Specific post hoc group differences among all of the teams are beyond the primary focus of our objective and are not provided. However, the percentage of failure among individual teams' helmets ranged from a low of 0% to a high of 57%. Significant differences were found among the 6 reconditioning year categories for mean screw failure (F1,5 = 16.6, P < .01). Helmets with the longest interval since last reconditioning (3 years) had the highest failure rate of 0.47 ± 0.93, whereas those that were unused or new before the previous season had the lowest failure rate (0.08 ± 0.29) (Table 6). From the 2584 helmets, 9673 of 10 284 (94%) screws were removed. We identified significant differences in success rate among all 4 types of screws (χ21,4 = 647, P < .01). The best success rate (96.6%) was in the silver group, and the worst (25.5%) was for screws that were classified within the unknown group (Table 7).

Table 4
Failure Rates and Failure Reasons at Each Screw Location, All Helmets (n = 2584)
Table 5
Regional Screw Failure Rates (n = 2584)
Table 6
Reconditioning Year Screw Failure Rates (n = 2584)
Table 7
Removal Success by Screw Type (n = 10 284)

Regional Weather Comparison

Significant differences were found among the regions on all 7 weather variables (P < .01), confirming the effectiveness of our geographic breakdown and justifying comparisons based on that division. The Southern region had the highest average precipitation (35.30 ± 13.76 cm), highest average high temperature (26.17°C ± −14.11°C), and highest average daily temperature (20.67°C ± −14.33°C), whereas the Northeast region had the highest average percentage humidity (77.8 ± 4.4) (Table 8).

Table 8
Environmental Data by Region

Successful Screw Removal

A discriminant analysis was conducted to determine whether 15 variables (the 7 weather measures, school region, helmet brand and model, year of reconditioning/recertification, and screw type at each of the 4 locations) could predict the successful removal of all 4 screws and the FM. Because it failed the tolerance test, average daily temperature was excluded from the analysis. One function was generated and was significant (Λ = .932, χ214,n=2584 = 175.34, P < .001), indicating that the function of predictors significantly differentiated between successful and unsuccessful screw and FM removal. Removal success was found to account for 26% of function variance. Standardized function coefficients and correlation coefficients revealed that the screw type variable at each of the 4 locations was associated most with the function (Table 9). Original classification results revealed that 99.8% of the successful removals were classified correctly and 95.3% of the unsuccessful removals were classified correctly. For the overall sample, 84.3% were correctly classified. Cross-validation derived 84.3% accuracy for the total sample. The means of the discriminant functions are consistent with these results. Successful removals had a function mean of −0.119; unsuccessful removals had a mean of 0.613.

Table 9
Correlation Coefficients and Standardized Function Coefficients

Metallurgic Analysis

According to a qualitative analysis of the screw samples, the 2 primary alloys from which all screws were fabricated were stainless steel and carbon steel. The silver-colored screws in our study actually comprised 2 types of screws, 1 made from a 300-series nonmagnetic stainless steel and the other made from magnetic carbon steel with a covering of nickel plating. The black-colored and gold-colored screws both were constructed from carbon steel with a covering of zinc plating. However, although the black and gold carbon steel screws were both zinc plated, the gold screws also contained an additional chromate conversion coating, accounting for their different color.

Stainless steel screws rely on the intrinsic corrosion resistance provided by a thin, passive surface film. This surface film is not completely stable and can be broken down under some conditions.11 As for the silver-colored carbon steel screws, the nickel-plated coating is more noble (less corroding) than the base metal. In the case of the gold carbon steel screws, the chromate coating provides better, although relatively short-lived, corrosion resistance, whereas the zinc plating acts as a layer of sacrificial anode.12 Our semiquantitative analysis of screws revealed that most screws experienced a uniform corrosion. The corrosion was primarily a combination of moisture and “atmospheric” corrosion, with moisture providing the driving force for the general corrosion. Evidence of localized chloride corrosion was detected but was not the primary factor in the corrosion process.


The purpose of our study was to assess the ability to remove FMs with a cordless screwdriver after at least 1 season of play in high school football helmets and to identify factors that influence FM removal. All 4 hypotheses developed for the study were rejected, suggesting that FM removal success rates differed among teams, regions, helmets containing different screw types, and year of last reconditioning/recertification. In addition, several factors thought to be related to FM removal were identified as strong predictors for success in that task. The screw type used to hold the FM loop straps in place was the strongest indicator for successful FM removal.

Failure Rates

Overall, 16% (419/2584) of the FMs could not be removed from the helmets because 1 or more screws failed during the process. This percentage is similar to the findings of Decoster et al,9 in which 17.5% (39/222) of the FMs could not be removed. Decoster et al9 also reported differences among teams in that 1 of the 3 high schools investigated had a failure rate of nearly 48%, compared with the other 2 schools, whose rates were 11% and 9%. Similarly, our investigation revealed differences in failure rates among teams, with a wider range of rates: 3 schools had a 0% failure rate and another 3 schools had a more than 50% failure rate (Table 10). Anecdotally, it appeared that schools with very high failure rates had poorer equipment maintenance (ie, the helmets appeared extremely worn, loop straps were broken, FMs were taped to helmets, and so on) and less frequent helmet reconditioning/recertification than the schools with the best results. The higher failure rates among many of the teams' helmets are disturbing and obviously unacceptable, but what is encouraging from our results is that it is still possible to obtain 100% success in FM removal using a cordless screwdriver in helmets used for at least 1 season of participation.

Table 10
The 5 Lowest and 5 Highest Failure Rates and Failure Percentages Among the 46 Teams

The sample from the Decoster et al9 study was taken from a focused location in the New England region. Our sample represented multiple regions of the country, and we found differences in FM removal failure rates among those regions. Although the regions also differed in some of the weather characteristics analyzed, those variables did not have as strong an effect on FM removal as other factors (eg, screw type), according to the results of the discriminant analysis. The differences in failure rates among regions is likely multifactorial and may have more to do with the use of hardware that resists degradation or corrosion and the practice of more regular equipment maintenance. Obviously the weather is not something we can control; however, with weather having less to do with FM removal success than other factors that are more controllable, we can strive to make the situation more favorable for FM removal with a screwdriver.

Overall, the most common reason for failed FM removal during our study was the T-nut spinning with the screw during the removal attempt. This also has been reported by authors investigating both used9 and unused3,6,8 football equipment. Decoster et al9 suggested that a potential factor influencing whether or not the T-nut spins may be the presence of a T-nut wall within the shell of the helmet. The T-nut wall is a helmet feature that is intended to capture the base of the T-nut and prevent it from spinning with the screw as the screw initially turns. Riddell was the first helmet company to introduce this feature in their VSR-4 model helmet. Schutt Sports has since added a similar feature in their new DNA helmets. Twenty-four percent of the helmets in our sample contained a T-nut wall at the 2 ear hole screw locations, whereas approximately 76% had no T-nut walls. Unexpectedly, several of the teams from our sample installed washers in helmets that also had a T-nut wall, such as the Riddell VSR-4, negating the intended function of the T-nut wall. The presence of washers installed between the T-nut and inner portion of the shell, a standard installation feature in traditional style helmets, also was tracked. To our knowledge, no standards have been set forth by NOCSAE regarding the correct construction of T-nut walls. Schutt Sports recommends against placing washers in their DNA helmets, which do have T-nut walls. In addition to improper installation, another possible complication in helmets with T-nut walls is the potential for the plastic walls to wear down, again making the walls nonfunctional. Whether a T-nut wall or washer by itself or in combination affects FM removal is unknown at this time. Keeping these limitations in mind, our results revealed helmets with T-nut walls averaged a success rate of 91% (572/627), whereas those without had a success rate of 82% (1598/1957).

An additional cause of removal failure came in the form of screws stripping at the head. Most of the screws were of the round-head type with a combination slotted and Phillips head screw face. This combination allows for use with either a flat head or Phillips head screwdriver. The Phillips head screwdriver is the ideal choice, because it allows for a more secure fixation into the screw face; it is the type we used in our study. Whether a different type of screw face (eg, hex, square) would lead to fewer failures of this type is unknown but may be worth exploring. However, the type of screw slot might not matter when a screw can not be removed due to plastic or other foreign substances being embedded in the screw face, which was one cause of failure in this study. Finally, in some cases, the screw and T-nut apparently were so corroded that they were fused, or literally were “stuck” in place.

Decoster et al9 reported that rates of failure were higher for the 2 screw locations on the sides of the helmet than for the 2 screw locations on the top. In our study, 610 screws could not be removed from the helmet with the cordless screwdriver. Of those failures, more than double occurred at the side locations (412) than at the top screw locations (198). Decoster et al9 speculated that this difference in failure rates of the side screw locations might have to do with the underside of the screw and T-nut on the inside of the helmet shell being more exposed to sweat and other elements than the top screw locations, which are covered with thick forehead padding. Although we did not investigate this possibility directly, it does seem a logical assumption. Results from both studies should serve as a warning to athletic trainers and equipment managers that a screw failure is more likely at the side locations than at the top. Athletic trainers and equipment managers would be justified in devoting particular attention to the side screw assemblies for evidence of malfunction and corrosion during regular maintenance checks throughout the season.

The fact that 419 face masks, or 610 screws, could not be removed with a cordless screwdriver—regardless of the region, team, or cause—should concern athletic trainers, coaches, and equipment managers alike. Although current multi-organizational guidelines4 recommend the use of a cutting device rather than a screwdriver due to the potential for screw failure, researchers have reported that the cordless screwdriver is the easiest to use,3 creates the least head movement3,6,8 and torque at the head,6 and finishes the retraction6 or removal task3 in the least amount of time. Authors3 of a recent study analyzing the effect of different styles of football helmet equipment on FM removal recommended using a cordless screwdriver as the initial tool for FM removal, not only due to its superiority in time, difficulty, and movement but also because 29 of 40 failed removal attempts occurred during trials using a cutting tool (primarily due to a 4-minute maximum trial definition of failure). Because of the potential for screw failure, although less common in their findings, these researchers recommended having an appropriate backup cutting tool ready.3 Based on the results of this study, we concur with that recommendation.

Metallurgy and Screw Corrosion

The longer or more frequently a material is exposed to atmospheric elements such as humidity, temperature extremes, and precipitation, the more likely the material is to break down or corrode.13 When steel is exposed to moisture for extended periods of time, oxidization of the steel surface forms iron oxide, or rust.14 Although corrosive-resistant materials such as stainless steel and nickel plating often are used to delay this process,12 steel hardware is still prone to corrosive effects. Some metals are more vulnerable to corrosion than others; for example, carbon steel is more vulnerable to corrosion than is stainless steel. Furthermore, the rate of “attack” is affected by the fluid concentration, temperature, fluid velocity, and stress in the metal parts.11,12,14 As a general rule, the rate of corrosion may double with an 18°F (10°C) rise in temperature of the metal part.14 This combination of factors occurs regularly in football. The cumulative effect of moisture exposure to the screws from weather and profuse sweating, temperature changes of the helmet hardware due to ambient extremes, and stress or vibration to the helmet's metal secondary to blocking and tackling through the course of a season all could combine to accelerate the rate of corrosion.

The more weather-resistant stainless steel screws and nickel-plated carbon steel screws from our sample (both originally classified as silver) had the highest success rate, and screw type was the strongest predictor of FM removal success. The highest rate of failure among screw types was for those whose type could not be determined, because the screws were so thoroughly corroded and rusted that their original color was beyond recognition. Evidence from other analyses indirectly supports the possibility that exposure to the weather or the environment in itself could affect the ability to remove screws from a helmet's FM. For example, differences in failure rates between the number of years helmets had gone without reconditioning/recertification (ie, longer exposure) were detected, and this could certainly result in higher failure rates secondary to the longer exposure of the hardware to the elements.


Generally, the longer the helmets in our study had gone without being reconditioned, the higher the screw failure rate per helmet. When we looked at those helmets that were new before the 2004 season (ie, only used through one season), they had significantly fewer screw failures (0.08) per helmet than those that had not been reconditioned since 2003 (ie, 2 years since last reconditioning/recertification), which had 0.37 screw failures per helmet (Table 7). The discriminant function analysis demonstrated a low to moderate level of prediction for FM removal success for the reconditioning year variable. Other researchers dealing with FM removal using new, unused equipment, have reported failure rates of 8%8 and 7%6 in removing screws with a screwdriver. The results of these studies, along with ours and those of Decoster et al,9 suggest that as helmets are exposed to wear and tear from regular use, FM removal failure rates increase.

Although helmets undergo an initial and periodic certification process,1,2 neither NOCSAE nor organizations such as the National Collegiate Athletic Association or the National Federation of State High School Associations, which adopt and endorse these standards, have a specific requirement for when or how many times a helmet must undergo recertification.15–17 For football helmets, NOCSAE recommends that the consumer periodically have helmets recertified, but because of the differences in the amount and intensity of usage for each helmet, the consumer is urged to use discretion regarding the frequency with which helmets are recertified.16 Helmet manufacturers differ in their reconditioning/recertification requirements, with most encouraging reconditioning/recertification every 2 years to maintain their warranty. The National Federation of State High School Associations leaves the decision regarding regular maintenance to the individual state associations or high schools (Bob Colgate, oral communication, 2006).

A closer look at the recertification standards from NOCSAE reveals that the reconditioning/recertification plant is required to demonstrate that a percentage of helmets being recertified meet the impact-testing guidelines.2 The standards also state that “helmet configuration must be the same as originally certified, all components must function as originally certified, and helmet shells must be free from cracks.”2 Literal interpretation of these standards allows for the conclusion that as long as the hardware and components are functioning as originally certified, they do not necessarily need to be replaced. This could be interpreted to include used screws and T-nuts that hold the FM in place. At the reconditioning/recertification plants that we visited, the entire helmet was disassembled and all loop strap, screw, washer, and T-nut components were replaced with new ones. However, the percentage of helmets that secured the FMs with carbon steel screws versus stainless steel screws varied among the reconditioning/recertification plants. We feel that a global standard in which a helmet is required to be recertified on a regular basis (annually), with mandatory replacement of all hardware components (corrosion-resistant screws, T-nuts), would greatly reduce the risk of FM removal failure using a screwdriver.


Although we obtained the length of season for all teams, we made no attempt to determine the precise exposure of individual helmets to participation. Besides the difficulty inherent in such a task, we think this would have had little to no effect on our sample due to its size. We also did not make any effort to learn the maintenance habits or procedures that individual teams may or may have not have undertaken with their helmets regarding helmet hardware components. The results from this study may not be generalizable to different levels of play (ie, youth, collegiate, professional) due to variations in exposure, intensity, and maintenance. Finally, despite the differences among all the regions in several of the weather variables measured (Table 9), combined with the possibility that temperatures, precipitation, or any other weather characteristics may have contributed to failure in removing a FM, we cannot establish a direct cause-effect relationship from the current analyses.


A total of 84% of FMs were removed from a large sample of high school football helmets that were used for at least 1 season of play. Factors related to the use and maintenance of football helmets affect the ability to successfully remove the FM using a cordless screwdriver. Stainless steel or nickel-plated carbon steel screws and more frequent reconditioning are favorable factors for successful screw removal. Differences in screw removal failure rates existed among teams, regions, screw types (stainless steel and nickel-plated carbon steel had the lowest failure rate, unidentifiable screws had the highest), and screw location. The strongest predictor of successful removal was the screw type. Other predictors of screw removal were helmet brand, reconditioning/recertification plant, average low temperature, reconditioning year, and region. The lowest failure rate in a region was 0.08 screw failures per helmet, and the lowest failure rate by a team was 0.0 failures per helmet. These may be considered obtainable standards for FM removal.


We recommend that helmets be fitted with stainless steel or carbon steel hardware with an appropriate protective coating that resists degradation when exposed to football environments. We also recommend regular inspection and maintenance of helmets throughout a season and annual helmet reconditioning/recertification in order to decrease the chances a screw will be not be removed with the cordless screwdriver. Based on past research demonstrating minimal time required, level of difficulty, and movement created during FM removal, we concur with the use of a cordless screwdriver as the primary tool for emergency FM removal and encourage adoption of the above recommendations to decrease the chances of screw failure. We also strongly advise that the sports medicine professional carry an equipment-specific backup cutting tool because the potential for failure still exists.


This project was fully funded through the Byron Goldman Research Award of the National Operating Committee on Standards in Athletic Equipment (NOCSAE). We thank Brian Campbell, PhD; Steven Tucker, MS, ATC; Eric Gattie, ATC; and Kimberly Dolak for their assistance during data collection. We thank David Halstead of the Southern Impact Research Center for his technical review of the manuscript. We also thank Kenneth Swartz for his help in designing and constructing the helmet mounts used during data collection and Tim Kenney of New Hampshire Materials Laboratory (Somersworth, NH) for his assistance with the metallurgic analysis.


  • National Operating Committee on Standards for Athletic Equipment (NOCSAE). Standard Performance Specification for Newly Manufactured Football Helmets. 3rd ed. Overland Park, KS: NOCSAE; 2003:1–5. ND002-98m03.
  • National Operating Committee on Standards for Athletic Equipment (NOCSAE). Standard Performance Specification for Recertified Football Helmets. 6th ed. Overland Park, KS: NOCSAE; 2004:1–5. ND004-96m04a.
  • Swartz EE, Norkus SA, Cappaert T, Decoster LC. Football equipment design affects face mask removal efficiency. Am J Sports Med. 2005;33:1210–1219. [PubMed]
  • Kleiner DM, Almquist JL, Bailes J. Prehospital Care of the Spine-Injured Athlete: A Document From the Inter-Association Task Force for Appropriate Care of the Spine-Injured Athlete. et al. Dallas, TX: National Athletic Trainers' Association; 2001.
  • Swartz EE, Norkus SA, Armstrong CW, Kleiner DW. Face-mask removal: movement and time associated with cutting of the loop straps. J Athl Train. 2003;38:120–125. [PMC free article] [PubMed]
  • Jenkins HL, Valovich TC, Arnold BL, Gansneder BM. Removal tools are faster and produce less force and torque on the helmet than cutting tools during face-mask retraction. J Athl Train. 2002;37:246–251. [PMC free article] [PubMed]
  • Swartz EE, Armstrong CW, Rankin JM, Rogers B A. 3-dimensional analysis of face-mask removal tools in inducing helmet movement. J Athl Train. 2002;37:178–184. [PMC free article] [PubMed]
  • Knox KE, Kleiner DM. The efficiency of tools used to retract a football helmet face mask. J Athl Train. 1997;32:211–215. [PMC free article] [PubMed]
  • Decoster L, Shirley CP, Swartz EE. Football face-mask removal with a cordless screwdriver on helmets used for at least one season of play. J Athl Train. 2005;40:169–173. [PMC free article] [PubMed]
  • Regions of the United States. Library of Congress. Available at: Accessed December 2004.
  • Sedriks J. Corrosion of Stainless Steels. New York, NY: John Wiley & Sons; 1979.
  • Talbot D, Talbot J. Corrosion Science and Technology. Boca Raton, FL: CRC Press; 1998.
  • Veleva L, Kane R. Atmospheric corrosion. In: Cramer SD, Covino BS, eds. Corrosion: Fundamentals, Testing, and Protection. Vol 13A. 9th ed. Materials Park, OH: ASM Intl; 2003:196–209.
  • Silverman D. Aqueous corrosion. In: Cramer SD, Covino BS, eds. Corrosion: Fundamentals, Testing, and Protection. Vol 13A. 9th ed. Materials Park, OH: ASM Intl; 2003:190–196.
  • Crisco J. FAQs and FACTS. NOCSAE Newsletter. Spring 2005;4.
  • National Operating Committee on Standards for Athletic Equipment (NOCSAE). Frequently Asked Questions and Answers. Overland Park, KS: NOCSAE; 1998.
  • Halpin T. 2005 NCAA Football Rules and Interpretation Book. ed. Indianapolis, IN: National Collegiate Athletic Association; 2005.

Articles from Journal of Athletic Training are provided here courtesy of National Athletic Trainers Association