Trends in Australian knee injury rates: An epidemiological analysis of 228,344 knee injuries over 20 years

Abstract

Background: Acute knee injuries are a key predisposing risk factor for knee osteoarthritis. Public health interventions require in-depth epidemiological evidence to determine which knee injuries are problematic in critical age and sex demographics.

Methods: Descriptive epidemiological analysis of longitudinal data on knee injuries (July 1998 - June 2018) from the National Hospital Morbidity Database in Australia were studied. The main outcomes where the population-related knee injury frequency, incidence per 100,000 and annual growth rate (%) over the 20-year observation period. Age-group and sex differences were also studied to determine demographic-specific trends.

Findings: 228,344 knee injuries were diagnosed over the 20-year analysis period. Significantly rising annual incidences were observed for total knee injuries, anterior cruciate ligament (ACL) injuries and knee contusions in males and females. Posterior cruciate ligament (PCL) injuries and knee dislocations were also rising in females, but not males. Greater annual growth rates were observed for females compared to males for total knee injuries, knee contusions, PCL injuries and knee dislocations. Demographic analysis revealed that the highest annual growth rate in injury incidence (10.4%) was observed for ACL injuries in females aged 5-14 years old.

Interpretation: Increasing annual incidence of knee injuries was observed over the 20-year period. Males have a higher incidence of knee injury per capita than females, but the gap appears to have narrowed over the 20-year analysis period. Younger Australians show a precipitous rise in the annual number of ACL injuries, particularly for females aged 5-14 years. These trends warrant urgent intervention.

Funding: None.

Figure 1

Knee injuries diagnosed in Australian hospitals from 1998-1999 to 2017-2018. Panel A) population pyramid of the total number of acute knee injuries in the 20-year analysis period diagnosed in males (green) and females (purple) aged 5 years and older; panel B) annual knee injury frequency for males and females from the analysis period and extrapolated until the year 2030-2031, panel C) annual knee injury incidence per 100,000 population for males and females for the analysis period and extrapolated until the year 2030-2031. For panels B and C: circles, annual knee injury number or incidence; solid line and shaded region, negative binomial regression model and 95% confidence interval, dashed and dotted lines, extrapolated negative binomial regression model and 95% confidence interval. Note the x-axis value in panels B and C indicate the index year i.e., 2000 represents from July 2000 to June 2001.

Figure 2

Population pyramid of each specific knee injury diagnosed in Australian hospitals from 1998-1999 to 2017-2018. Each panel shows the total number of injuries diagnosed in the 20-year analysis period in males (green) and females (purple) aged 5 years and older.

Figure 3

Annual injury incidence (per 100,000 population) for specific knee injuries diagnosed in Australian hospitals in individuals aged 5 years and older from 1998-1999 to 2017-2018. Circles, annual knee injury incidence; solid line and shaded region, negative binomial regression model and 95% confidence interval, dashed and dotted lines, extrapolated negative binomial regression model and 95% confidence interval. Note the x-axis value indicate the index year i.e., 2000 represents from July 2000 to June 2001.

Figure 4

Demographic (age by sex) heat map analysis of the annual growth rate (%) of specific knee injuries diagnosed in individuals aged 5 years and older in Australian hospitals from 1998-1999 to 2017-2018. Panel A) males; panel B) females. Growth rates were determined as the exponentiated co-efficient from negative binomial or Poisson regression (depending on the distribution of the data).