ACL Tears in School-Aged Children and Adolescents Over 20 Years
- Nicholas A. Beck, MDa,
- J. Todd R. Lawrence, MD, PhDb,
- James D. Nordin, MD, MPHc,
- Terese A. DeFor, MSc, and
- Marc Tompkins, MDd
- aUniversity of Minnesota, Minneapolis, Minnesota;
- bChildren’s Hospital of Philadelphia, Philadelphia, Pennsylvania;
- cHealthPartners, Minneapolis, Minnesota; and
- dTRIA Orthopaedic Center, Bloomington, Minnesota
Dr Beck conceptualized and helped design the study, drafted the initial manuscript, and edited the final manuscript; Dr Lawrence conceptualized and helped design the study and reviewed and revised the manuscript; Dr Nordin helped design the study and reviewed and revised the manuscript; Mrs DeFor collected the data, carried out the initial analyses, and reviewed and revised the manuscript; Dr Tompkins conceptualized and helped design the study, acted as the senior author, and reviewed and revised the manuscript; and all authors approved the final manuscript as submitted.
BACKGROUND: Anterior cruciate ligament (ACL) tears are thought to occur with increasing frequency in young patients. No study has shown increased incidence over time. We hypothesized the incidence of ACL tears in young patients has increased over the past 20 years.
METHODS: This descriptive epidemiology study is a retrospective review of insurance billing data of all patients aged 6 to 18 years with Current Procedural Terminology, Fourth Revision codes for ACL tear and reconstruction or International Classification of Diseases, Ninth Revision, Clinical Modification codes from 1994 to 2013. Injuries were normalized to persons per year enrolled in the insurance database based on age and sex. Analysis was performed based on sex and age (6–14, 15–16, and 17–18 years).
RESULTS: The rate of ACL tears per 100 000 person-years averaged 121 ± 19 (range 92–151). All trends increased significantly except for the male 6- to 14-year-old and 17- to 18-year-old age groups. Overall there was an annual increase of 2.3%. Females had significantly higher incidence except in the 17- to 18-year-olds. Females peaked at age 16 years and males at age 17 years, with rates of 392 ACL tears and 422 ACL tears per 100 000 person-years, respectively.
CONCLUSIONS: The incidence of ACL tears in pediatric patients increased over the last 20 years. Females were at higher risk except in the 17- to 18-year -old group. Peak incidence is noted during high school years. These data help target the most at-risk patients for ACL prevention programs.
- ACL —
- anterior cruciate ligament
- ICD-9-CM —
- International Classification of Diseases, Ninth Revision, Clinical Modification
What’s Known on This Subject:
Recent reports of anterior cruciate ligament (ACL) injuries in young patients suggest the incidence is increasing. However, true incidence data are scarce. Female patients are at higher risk of ACL injury than males participating in similar activities.
What This Study Adds:
We present incidence data for ACL tears in young patients and show that the incidence increased over the past 20 years. Girls exhibit higher risk until age 17. The ages most at risk are 16 for girls and 17 for boys.
Midsubstance tears of the anterior cruciate ligament (ACL) were classically thought to be rare injuries in young patients, with tibial spine avulsion fractures supposedly representing the equivalent injury in patients with open physes.1 Others have noted that ACL tears seem to be presenting with greater frequency in pediatric patients, including patients as young as 5 years old.1–6 The increase in ACL surgical consultations in young patients has been attributed to multiple factors, including increased participation in high-demand year-round sports at an earlier age, better clinician awareness and recognition of the signs and symptoms of ACL tears, increased female athletic participation, and the expanded role of magnetic resonance imaging in the diagnosis of intra-articular knee pathology.1,7,8 Identifying these injuries may be important to avoid additional intra-articular knee injuries and chondral damage through activity modification or recently developed surgical reconstruction techniques that minimize or avoid physeal damage.9–13
Most studies of ACL injury rates report on adult athletes or military personnel.4,14–21 Large population-based epidemiologic studies often group young patients together and report very low rates of ACL injuries in school-aged patients, and none to our knowledge have looked at the change in incidence over time.22,23 Thus, the goal of this study was to review the billing database of a large metropolitan insurance network to determine whether the number of ACL tears has increased over the past 20 years. We also sought to examine differences between sex, age, and delay in surgery over the study period. We hypothesized that there would be an increase in the incidence of ACL tears in patients ≤18 years of age over the past 20 years, with female patients showing a higher rate of increase than male patients.
A retrospective review of insurance billing data in a large metropolitan network was performed for all children and adolescents aged 6 to 18 years with International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes for ACL tears and Current Procedural Terminology, Fourth Revision (CPT-4) codes for ACL reconstruction from January 1994 to December 2013. Patients were included if they had ≥2 occurrences of ACL diagnosis codes (844.2 or 717.83) on medical claims ≤4 months apart, or ≥1 ACL procedure codes (29888) during the study period. Therefore, all patients with ACL tears were included, not just those with surgical reconstruction. Multiple ACL injuries per patient were counted if >4 months elapsed between diagnoses. Patients were excluded if they were outside the included age ranges, if the first diagnosis code was outside a member’s enrollment period, or if they had only 1 ICD-9-CM code.
Analysis included year-by-year total incidence of ACL injury and a breakdown by sex and ages grouped 6 to 14 years, 15 to 16 years, and 17 to 18 years. Denominators were normalized to patients per year enrolled in the insurance database based on age and sex to calculate rates per 100 000 person-years specific to that cohort. Because we used the insurance network, all patients in the network were captured by the insurance data search. In addition, because we queried it on a yearly basis, we were able to control for patients entering and leaving the insurance network. Delay in surgery was calculated by time between first ICD-9-CM code and Current Procedural Terminology code for each ACL injury. The annual percent of ACL tears surgically reconstructed was calculated.
The trend of annual incidence for each age and sex cohort was tested for significance with Poisson regression analysis. Logistic regression was used to look for significant change in annual surgery rates. Linear regression was used to look for significant change in the delay in surgery. Differences between sexes were tested with Poisson regression analysis for change in annual incidence and for differences in overall and age-specific rates.
Total patients enrolled in the insurance database within our study age range averaged 136 000 ± 15 000 per year, with approximately an even number of male and female patients each year. A total of 3303 ACL tears were identified over the study period, 1584 (48%) ACL tears in male patients and 1719 (52%) in female patients. The overall annual incidence of ACL tears averaged 121 ± 19 (range 92–151) per 100 000 person-years (or 0.121%) (Fig 1 and Table 1). The incidence in male patients averaged 114 ± 22 (range 83–160) per 100 000 person-years (or 0.114%). The incidence in female patients averaged 129 ± 24 (range 98–184) per 100 000 person-years (or 0.129%). The breakdown by age groups over time for overall patients, male patients, and female patients is shown in Figs 2, 3, and 4, respectively.
The incidence of ACL tears by age overall is shown in Fig 5. Poisson regression analysis indicated that differences in incidence between male and female patients by age were statistically significant for each age category (P < .001). Female patients peaked at age 16 years with an incidence of 392 ACL tears per 100 000 person-years (or 0.392%), and male patients peaked at age 17 years with an incidence of 422 ACL tears per 100 000 person-years (or 0.422%). We found no significant difference between the amount of change in ACL tear incidence over time between male and female patients overall (P = .6) and across the age groups (P = .2); however, overall incidence was significantly higher in female patients (129 ACL tears per 100 000 person-years vs 114 ACL tears per 100 000 person-years; P < .001) and in the 6- to 14-year-old (33 ACL tears per 100 000 person-years vs 26 ACL tears per 100 000 person-years; P < .01) and 15- to 16-year-old (371 ACL tears per 100 000 person-years vs 250 ACL tears per 100 000 person-years; P < .001) age groups over our study period. Male patients had significantly higher incidence in the 17- to 18-year-old age group (394 ACL tears per 100 000 person-years vs 337 ACL tears per 100 000 person-years; P < .01).
Results of the calculated changes in incidence over time are summarized in Table 2. All analyzed trends increased significantly over time except for the male 6- to 14-year-olds and male 17- to 18-year-olds. Overall we found that the number of ACL tears per 100 000 person-years increased 2.3% per year. Male patients had an increase of 2.2% per year, and female patients had an increase of 2.5% per year. The subgroup with the highest increase was female 15- to 16-year-olds, whose incidence increased 2.6% per year.
Logistic regression showed that the percentage of ACL tears surgically reconstructed significantly increased (P < .001) (Fig 6 and Table 3), and linear regression analysis identified a significant increase in the delay to surgery over the study period (P = .04) (Fig 7 and Table 4).
Through analysis of insurance data on a captured population of patients, we were able to calculate true incidence numbers for a large metropolitan population of pediatric patients. Over the 20 years analyzed, we found a 2.3% annual increase in the number of ACL tears in patients aged 6 to 18 years. Female patients had significantly higher rates of injury in the younger ages, but in the 17- to 18-year-old age group male patients had a significantly higher incidence. Subgroup analysis showed that female patients in all 3 age groups had a significant increase in incidence of ACL tears over the study period, whereas only the male patients aged 15 to 16 years had an increased incidence. We also found that the percentage of ACL injuries surgically reconstructed increased, and the average time from injury to surgery increased significantly over our study period.
This study is the first of its kind to identify an increase in the incidence of ACL tears in young patients over the last 20 years. These data are novel in that they report incidence of ACL injury within a known, closed cohort. Data were obtained on a yearly basis, so the insurance database allows for correction of patients entering and leaving the insurance plan. In addition, using the insurance database means data are not restricted to a single institution because the ACL tear could be diagnosed in >1 location, or the surgery could be performed in a separate location, and all of this information would be captured within the insurance database. These measures should add to a more complete understanding of what we have suspected in terms of increases in pediatric ACL injury over time.
The average annual incidence of ACL injuries found in our study (0.12%) is similar to those found in large population-based studies. Nordenvall et al23 used data from the Swedish National Patient Register to estimate the sex- and age-specific incidence of ACL tears in Sweden. They found the average annual incidence to be 78 per 100 000 inhabitants (or 0.078%). Very few injuries were found in patients <10 years old, and the incidence was similar in male and female patients in the 11- to 20-year-old age group of 144 per 100 000 inhabitants (or 0.14%). Gianotti et al22 reported data from New Zealand’s Accident Compensation Corporation showing a similar breakdown of incidence by age and sex. They also showed very few ACL injuries in patients <10 years old. Their incidence in 15- to 19-year-olds was between 65 and 100 per 100 000 person-years for female and male patients respectively (or 0.065%–0.10%).
It is important to note that previous studies define rates of injury in different units. Studies looking at risk for ACL injuries in athletes or military personnel present their rates based on exposure-based denominators to the sport or activity by either hours or season. Population-based studies often define rates based on absolute time-based denominators such as incidence per 100 000 person-years. Moses et al24 attempted to compare multiple studies by converting incidence rates in exposure-based studies to annual units. They found studies reporting population incidence with a low annual median incidence of 0.03%. Studies of ACL injury data from military and professional sports showed much higher annual incidences of up to 2.14% and 3.67%, respectively. Dodwell et al25 showed an increase in the rate of ACL reconstructions performed on pediatric (defined as ages 3 to 20 years old) patients in New York State from 1990 to 2009. In their study, population estimates from the US Census Bureau were used as the denominator to obtain rates per 100 000 population. They found that the rate rose from 17.6 ACL reconstructions per 100 000 population in 1990 to 50.9 ACL reconstructions per 100 000 population in 2009, or a 190% increase over their study period. This compares to the more modest 2.3% yearly increase in our population. Their analysis by sex found that male patients had a 15% higher rate of ACL reconstructions, where our study showed higher rates in female patients except for the 17- to 18-year-old patients. Lively and Feathers17 found an increasing incidence of ACL injuries in a collegiate population based on incoming athletes’ injury history at a single institution from 1996 to 2008. An athlete had a 2.7 greater odds of having a previous ACL injury in 2008 than those who arrived in 1996. Although their analysis showed a nearly linear increasing trend by year, there was no significant correlation with sport (P = .41) or sex (P = .71).
In this study, female patients on average had a higher incidence overall and across the 6- to 14-year-old and 15- to 16-year-old age groups than their male counterparts. ACL injury rates have been shown in multiple studies to be 2 to 8 times higher in female than male patients when comparing exposure to sex-comparable sports.4,14,20 Reasons for this observed difference are thought to be related factors such as geometry of the intercondylar notch and smaller size of ACL, as well as biomechanical and neuromuscular factors such as higher quadriceps-to-hamstring ratio and landing from a jump with less hip and knee flexion and more hip adduction, leading to increased dynamic knee valgus with greater knee abduction angles.26 It is interesting that male patients in the 17- to 18-year-old age group of our study surpassed female patients, with significantly higher rates. Proposed explanations for this observation include higher intensity of collision sports in older male athletes, fewer female adolescents participating in sports at this age, and female athletes prone to ACL injury having already sustained injuries and no longer competing.
Through subgroup analysis by age and sex, we attempted to identify the most at-risk patients over our study period. Much effort has recently been put forth in the sports medicine community to develop prevention programs with neuromuscular and proprioceptive training, especially for female athletes.26 Noyes and Barber-Westin27 conducted a systematic review of 8 neuromuscular retraining studies in female athletes. Three studies found a significant reduction in the number of noncontact ACL injuries, with a relative risk reduction ranging from 75% to 100%. The remaining 5 studies found no significant decrease with their programs; however, the overall ACL injury rates in these studies were much lower. There was a large discrepancy in how the programs were performed, based on the frequency, duration, supervision, and compliance. Myer et al28 performed a meta-analysis of neuromuscular training studies, specifically looking at the effectiveness the programs for different age participants. They included 14 clinical trials and found a significant age-related reduction in injury rate, with an odds ratio of 0.28 in the 14- to 18-year-old female athletes, compared with an odds ratio of 0.48 in the 18- to 20-year-olds and 1.01 in the >20-year-olds. Thus the programs seem to be more effective in younger patients. In our cohort, all subgroups of the female patients and the male 15- to 16-year-old subgroup showed a significant increase in ACL injury incidence over time. Because the injury rates in these cohorts seem to be increasing, they may also be the most likely to benefit from injury prevention programs.
Two additional findings of this study are that the percentage of injuries surgically reconstructed increased and the average time to surgery increased over our study period, although the time to surgery increase was small (<1%) and clinically insignificant. A traditional treatment approach for ACL injuries in young patients has been to attempt a trial of conservative treatment until the child is close to the end of growth, based on the assumption that the risk of limb length discrepancy and angular growth disturbances decreases as the patient approaches skeletal maturity.1,3,7,29 Recent studies have shown poor outcomes with nonoperative treatment. Lawrence et al11 showed that delaying surgery of a complete ACL tear for >12 weeks correlated with a significant increase in the number of irreparable medial meniscal tears and lateral compartment chondral injuries. Because of this risk, multiple new surgical techniques have been developed that range from physeal-sparing, to partial transphyseal, to complete transphyseal reconstruction with smaller, more vertical tunnels.1 Studies reporting on these techniques are small retrospective cohorts, and the procedures are technically demanding, but the outcomes are encouraging. Our data suggest that because of the encouraging results in the recent literature, surgeons are more confident treating these injuries with surgical reconstruction at younger ages.
Our study has several limitations. The data were pulled from a specific regional geographic insurance database. Therefore, the injury and surgical treatment rates may reflect regional differences and thus may not be representative of other areas of the United States or in other countries. The people covered under this insurance plan also probably represent people of a certain socioeconomic class. Patients with a different quality of insurance or no insurance may have more or less exposure to at-risk activities and higher barriers to access of care, which may affect the overall incidence rate. The insurance database itself has limitations. We attempted to identify repeat injuries in patients who have a new diagnosis code after 4 months, and this assumption fails to include new injuries that occur in the first 4 months after return to sport. For patients with multiple injuries, we were unable to distinguish between primary tear of the contralateral knee and a recurrent tear of the originally injured knee. In addition, patients with ACL injuries could have been missed by entering or leaving the health insurance plan before getting the 2 ICD-9-CM codes required for inclusion, so the incidence may actually be higher than that which is reported here. Despite these limitations, this is the first study quantifying and showing an increase in the incidence of ACL tears in school-aged children and adolescents.
This study is the first to show that the incidence of ACL tears in pediatric patients (age 6 to 18 years) has significantly steadily increased by about 2.3% annually over the last 20 years. Female patients in all age categories and male patients in the 15- to 16-year-old category showed the greatest increases and accounted for most of the overall increase. Both male and female adolescents have a peak incidence during high school years (age 16 years for female and 17 years for male adolescents). Future studies identifying the mechanism of injury in young patients will be helpful for proposing prevention training.
- Accepted December 20, 2016.
- Address correspondence to Nicholas A. Beck, MD, 2450 Riverside Ave South, Suite R200, Minneapolis, MN 55454. E-mail:
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: Funds provided by the University of Minnesota Clinical Research Committee.
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
- Copyright © 2017 by the American Academy of Pediatrics