HIPRC: Best Practices

Recreational Injury Interventions

Knee Braces and Other Orthotics

Background

Knee injuries are a common occurrence in a variety of sports. Football is no exception; the most common knee injuries are ligament injuries (grades I-III) with the medial collateral ligament (MCL) the most commonly sprained.⁶ The knee is the most frequent injury site in football constituting 18-23% of all football injuries in high school and college.⁷ Use of knee braces is a common prophylactic measure. Studies which evaluated the effectiveness of braces are reviewed below. Unfortunately, none of these studies is a randomized control trial.

Review of knee brace studies:

Author	Zemper, 1990
Study design and target population	Prospective cohort study. Population: 5% proportionally representative sample of NCAA/NAIA collegiate football teams 1986-1987 seasons. Data obtained from National Sports Injury Surveillance System (NSISS).
Intervention	Knee braces of all types
Outcomes	Total injuries, knee injuries, MCL injuries, time lost from practice and games
Results	6,229 players; 445,856 AE (athlete exposures). 1,744 players (28%) wore knee braces. All injuries: 6.3 per 1000 AE (3.9 per 1000 AE in practices and 33.9 per 1000 AE in games) 521 knee injuries, 1.2 per 1000 AE. Those wearing braces had an increased risk of knee injury, RR 1.42 (1.2-1.8) compared to nonbraced players. MCL injuries were not significantly increased in braced players RR=1.2 (0.8- 1.9) Days lost from participation for all knee and MCL injuries were significantly different from those not wearing braces. Results followed the same pattern when game injuries were analyzed separately. Braces had no impact on ankle injuries; incidence of injury the same for contact and non-contact situations. Interaction found between braces and artificial turf resulting in increased MCL injuries when playing on turf and wearing braces.
Study quality and conclusions	Analysis of 2 years of data from the nationwide sports database (NSISS) shows no protective effect of knee braces for football injuries. All data were analyzed using chi-square comparisons. A larger sample size is needed to carry out multivariate analyses. In order to isolate the effect of knee braces it will be necessary to adjust for practice/training methods, practice and playing surfaces (grass vs. astroturf), player position and conditioning and playing time.

Author	Grace et al., 1988
Study design and target population	Prospective cohort study. Population: AAAA high school football players in Albuquerque and Santa Fe NM. N=580 players over 2 seasons
Intervention	Single upright single-hinged and double-hinged braces.
Outcomes	Knee sprains and strains and lower extremity (LE) injuries. Classified using modified National Athletic Injury/Illness Reporting System (NAIRS) based on time lost from practice and competition. Grades 0-IV.
Results	For the analysis braced and nonbraced players were matched by size, weight and playing position. Results of McNemar chi-square for matched data indicate 3.7 times more injuries in players wearing single hinged braces compared to non-braced players (p<0.001). Increased severity and rate of injury in wearers of double braces, but not significant. Lower leg and ankle injuries 3 times more common in braced players, p<0.01. Crude relative risk for knee and LE calculated from Table II: RR for knee injuries=3.4 (1.7-7.2) comparing all braced players to nonbraced players and RR for lower extremity injuries=3.8 (1.6-8.8)
Study quality and conclusions	Single-upright, single-hinged prophylactic knee braces associated with significantly increased rate of injury. Double hinged braces did not decrease the incidence of injuries. Both types of braces were associated with increased injuries of the ipsilateral ankle and foot.

Author	Shaw et al., 1987
Study design and target population	Before/after study design. Population: junior varsity and varsity football players. Texas City High School, Texas City, TX
Intervention	Single sided brace (McDavid) 1983-no braces worn 1984-varsity fit with bilateral braces 1985-braces provided for all players. Wearing brace not mandatory
Outcomes	Knee injuries requiring surgical/physician care or time lost due to injury.
Results	Knee injuries declined from 19 to 6 per season over 4 year period (1983-86). Time lost declined from 143 days to 53 days per season. Number or percentage of players wearing braces not given.
Study quality and conclusions	Data presented was incomplete and no denominators were given (number of players, practices or games) Coaching staff, team physician and athletic trainer unchanged for period of study.

Author	Rovere et al., 1987
Study design and target population	Before/after study design. Population: Football players, Wake Forest University, 1983-85
Intervention	Anderson Knee Stabilizer. Double hinged single sided brace. All players required to wear brace for practice and games, 1983-85. Comparison (no brace period), 1981-82. Two fall and two spring seasons included in each group
Outcomes	Grade I, II, III knee sprains as diagnosed by orthopedic surgeon.
Results	Increased injuries in players wearing braces. 7.5 per 100 players with braces vs 6.1 per 100 players not braced. Total number of players not specified, relative risk calculated from above proportions. RR=1.2 (0.7-2.1) Majority of injuries Grade I MCL sprains 4.8 per 100 players braced vs. 4.0 per 100 players braced. Relative frequency of injury by player position unchanged.
Study quality and conclusions	Non-significant increase in knee injuries when braces used. No change in coaching practices over 2 year period.

Author	Teitz et al., 1987
Study design and target population	Prospective cohort study. Population: NCAA Division I varsity football players, 1984 and 85 seasons.
Intervention	Knee braces Types included were: McDavid, Omni/Anderson, Jon Joy, American Prostheses
Outcomes	Incidence and severity of knee injuries
Results	1984: 6,307 players, 2297 (36%) braced. 1985: 5,390 players, 2389 (44% braced). All knee injuries higher in players wearing braces; 11 per 100 players braced vs 6 per 100 players nonbraced (1984) and 9.4 vs 6.4 per 100 players, braced vs nonbraced for 1985. p<0.001 MCL injuries more common in braced players (for 1984, 7.6 compared to 3.5 percent; for 1985 5.4 compared to 3.6 percent). Subset analyses conducted to evaluate incidence of injury between braced and nonbraced players for the following: player position, playing surface, mechanism of injury (contact/no contact), skill level (varsity string). Severity of injury was based on loss of playing time, specific grading of MCL injuries, and injuries requiring surgery. No difference in injury severity was found between braced and nonbraced players.
Study quality and conclusions	Preventive braces are not preventive and may be harmful. The consistency of the data from year to year and the corroboration of the overall findings by the results of subset analysis strongly support this conclusion.

Author	Hewson et al., 1986
Study design and target population	Before/after study design Population: University of Arizona intercollegiate football players, 1977-1985
Intervention	Knee brace (Anderson knee stabler) 1977-80 no brace used 1981-1985 mandatory braces for all "at risk" players (offensive and defensive linemen, linebackers, tight ends)
Outcomes	Number, type and severity of knee injuries. Exposure defined as attendance at a practice or game.
Results	28,191 braced exposures (224 players) and 29,293 non-braced exposures (226 players). No statistically significant difference between braced and nonbraced players for knee injury rate, injury type or severity.
Study quality and conclusions	Prophylactic knee bracing did not prevent knee injuries. NCAA rule changes and changes in treatment techniques beginning in 1981 designed to reduce knee injuries.

Author	Hansen et al., 1985
Study design and target population	Retrospective cohort study. Population: University of Southern California (USC) football players (1980-1984)
Intervention	Knee braces (Anderson Knee Stabler)
Outcomes	Number and type of knee surgeries. Outcome determined by review of team’s medical records.
Results	477 players; 148 (31%) braced. 11% injury rate in nonbraced group; 5% injury rate in braced group. 4% of braced linebackers were injured vs 25% of nonbraced linebackers. Similar results for offensive and defensive linemen.
Study quality and conclusions	Authors recommend Anderson brace for linebackers and interior linemen at USC. Simplistic analysis. Results based on small numbers of injuries; no confidence intervals given. No information on characteristics of players, or any other confounders. Self-selected brace use.

Summary of knee brace/orthotic studies

The studies reviewed above provide conflicting results on the protective effect of knee bracing. Some smaller studies and studies involving only one team have found braces protective,¹⁰^,¹⁴ while other single school studies found no protective effect or a non-significant increase in injuries.¹¹^,¹³ Two nation wide studies of collegiate athletes found no evidence that braces offered protection from knee injuries.⁸^,¹² There are a wide variety of factors which can contribute to knee injuries. Some of these (i.e., playing surfaces, shoes, player position, mechanism of injury, history of previous injury) can be analyzed directly. Other factors (particularly player conditioning, coaching methods, rules, treatment, and natural fluctuation in the incidence of injury) are more difficult to measure. Large national studies which were able to evaluate injury rates in different sub groups of players provide the best evidence available to date.⁸^,¹²

Recommendations on knee braces and other orthotics

At the present time prophylactic use of knee braces does not appear to protect players from knee injuries. Studies which found an increased risk of injury with knee braces may simply be reflecting the fact that individuals with prior knee injuries—who are most at risk of re-injury—are also the ones who wear braces.

Recommendations for future research

There is a definite need for well-designed, large-scale randomized controlled trials (RCTs) to evaluate the protective effect (if any) of knee bracing. This is the only way that all the other confounders can be controlled for.