Epidemiology of patellar dislocations in the United States from 2001 to 2020: results of a national emergency department database

ABSTRACT Objectives Recent studies have shown an increasing incidence of patellar dislocations among children and adolescents. Updated, population-based studies of all patellar dislocations in the United States (US), however, are lacking. This study investigated recent trends in injury rates and demographics among patients sustaining patellar dislocations in the US from 2001 to 2020. Methods This descriptive epidemiologic study retrospectively analyzed the National Electronic Injury Surveillance System (NEISS) database to identify cases of acute patellar dislocations presenting to US Emergency Departments (EDs) from 2001 to 2020. Annual, overall, and age-adjusted incidence rates (IRs, expressed per 100,000 at-risk person-years) and additional patient/injury characteristics were analyzed. Average annual percent change (AAPC) estimates are presented to indicate the magnitude/direction of trends in annual injury rates. Results An estimated total of 159,529 patellar dislocations occurred over the study period for an overall IR of 2.58 (95% CI = 2.04–3.12). Accounting for population growth, the overall annual incidence increased significantly from 2.61 in 2001 to 3.0 in 2020 (AAPC = 2.8, p < 0.0001). When considering sex and age, statistically significant increases in annual IRs were observed among males aged 10–19 years (AAPC = 3.8, p < 0.0001), females aged 10–19 years (AAPC = 5.3, p < 0.0001), and females aged 20–29 years (AAPC = 3.5, p = 0.0152), while no significant changes were observed in any other age groups. Two-thirds of patellar dislocations involved sports-related injury mechanisms. The annual incidence of both sports-related and non-sports-related injuries increased significantly over the study period (sports-related: AAPC = 2.6, p = 0.0001; non-sports-related: AAPC = 3.4, p = 0.0001). Athletic patellar dislocations occurred most commonly in basketball and dance. Conclusion The number of patients sustaining patellar dislocations is increasing in the US. Similar increasing trends were observed in both males and females aged 10–19 years, whereas injury rates increased in the third decade only among females. A large percentage of injuries occur during athletic activity, but both sports- and non-sports-related patellar dislocations are on the rise.


Introduction
Patellar dislocations are relatively common injuries and have been described as being the second most frequent cause of traumatic hemarthrosis of the knee (after anterior cruciate ligament tears) [1] and the most common cause of osteochondral fractures of the pediatric knee [2].Following an acute, traumatic patellar dislocation, patients may experience pain, decreased knee function, recurrent dislocations, chronic instability, and patellofemoral arthritis [3][4][5].These injuries typically affect young and active individuals [3,5,6], and a large percentage of patellar dislocations occur during athletic participation [3,5,7,8].
There has been a great deal of recent literature focused on the treatment of these injuries, which remains a controversial and evolving topic [9].Other areas which have received increased attention include the evaluation/identification of risk factors for recurrent instability and rehabilitation/returnto-sport protocols [9].Updated epidemiological studies regarding patellar dislocations, however, are less abundant.With the growing popularity of exercise and organized sports and steady increases in athletic participation among both children and adults [10][11][12], it's possible that the numbers of sports injuries, including patellar dislocations, may be increasing.Two population-based studies in the United States (US) which analyzed the incidence of patellar dislocations occurring during the years 2003-2008 and 1990-2010, however, did not demonstrate any changes in the annual incidence of these injuries [13,14].These findings, however, may not reflect changes in injury patterns during the most recent decade.On the other hand, two more recent studies utilizing a large, US pediatric database documented a significant increase in the number of pediatric hospital admissions for patellar dislocations and a significant increase in the number of pediatric patients undergoing procedures for patellar dislocations [15,16].Given the recent increases in patellar dislocations occurring among children and adolescents, an investigation of recent trends in injury rates in other age groups, and in the overall population, is warranted.
The purpose of the current study, therefore, was to identify trends in acute, first-time patellar dislocation injuries in the US over the last two decades.In addition, this study sought to define the demographic characteristics of individuals suffering these injuries.While the demographics of patellar dislocations have been well-described, certain age-, sex-, and ethnicityspecific differences have been reported across the various patient populations under study [8,17], and updated analyses may demonstrate changes in demographic patterns.By analyzing a nationally representative sample of US hospital Emergency Departments (EDs), this descriptive epidemiology study aimed to improve the scientific knowledge regarding the epidemiology of patellar dislocations.We hypothesized that an increasing number of patellar dislocations presenting to US EDs over time would be observed, primarily driven by an increase in the number of athletic/sports-related dislocations, given sustained high levels of athletic participation among the US population in recent years.

Data source and case selection
The National Electronic Injury Surveillance System (NEISS) is operated by the US Consumer Product Safety Commission (CPSC) to collect product-related injury data from ~100 hospital EDs selected as a probability sample of all EDs in the US [18,19].The NEISS database is a nationally representative sample of US EDs from which nationwide estimates of the number of injuries associated with (but not necessarily caused by) specific consumer products can be made [20].Trained coders at each participating facility collect data variables including demographic information, injury diagnoses, and a brief narrative description of the incident [18,20].NEISS data has been widely used to analyze various injuries and those associated with specific activities [21][22][23][24][25].The NEISS dataset is de-identified and publicly available and thus this study was deemed exempt from institutional review board approval.
Cases from the years 2001 through 2020 were considered for selection.To identify all potential patellar dislocations, the NEISS database was queried for all knee (code 35) injuries diagnosed as a dislocation (code 55).The initial query resulted in n = 17,739 unweighted cases of injuries classified as dislocations in the knee region.From these records, a combination of computer-assisted and manual review of the injury narratives was used to select only those cases diagnosed as acute, first-time patellar dislocations.Figure 1 outlines the case selection process.
First, potential cases were identified with a computerassisted search of the narrative text field for the following case-insensitive strings: 'PATEL' (n = 4,948 unique cases), 'KNEECAP' and 'KNEE CAP' (n = 214 unique cases), and the common misspellings 'PATTEL' (n = 22 unique cases), 'PLATEL' (n = 1 unique case), and 'PATER' (n = 1 unique case).From these records, each case narrative was manually reviewed to identify only those cases with a clear diagnosis of an acute, traumatic, first-time patellar dislocation.Cases without a clear diagnosis (including any 'rule out,' 'possible/suspected,' or 'probable' diagnosis) and cases describing an alternate diagnosis were excluded.Alternate diagnoses which were excluded were those describing patellar subluxation (without dislocation), chronic/recurrent patellar dislocations, patellar dislocations in the setting of prior ipsilateral knee surgery (including arthroplasty), knee (i.e.tibiofemoral) dislocations, patella/knee fractures, and knee strains/sprains.After exclusions, n = 4,567 unweighted cases of acute, firsttime patellar dislocations were identified for analysis.Each NEISS database record carries a weighting factor for the conversion of cases (unweighted records, n) to national injury estimates (weighted estimates, N) by accounting for the NEISS stratified probability sampling design.This allows the total number of ED visits nationwide (related to a certain product/activity) to be estimated from the sample of cases which are reported in the NEISS dataset.Thus, results include the number of unweighted cases (n) as well as the national total case numbers (N) which are estimated from the sum of individual case weights.In total, the n = 4,567 unweighted cases identified corresponded to a national estimate of N = 159,529 cases of acute, first-time patellar dislocations (95% CI = 126,330-192,728) occurring in the US over the 20year period from 2001-2020.

Variables
Certain variables were created/regrouped from the NEISS data as follows:

Injury mechanism
Each case narrative was used to assign an injury mechanism.Mechanism of injury was classified as either (1) sports-/exercise-related or (2) not sports-/exercise-related.The latter category was further subdivided into low-energy versus highenergy mechanism.Each case was assigned a single injury mechanism.A high-energy mechanism was defined as an impact from a height of 6 feet or greater or an injury involving a motor/powered vehicle (e.g.motor vehicle crash, 'pedestrian-versus-automobile,' etc.).If this could not be distinguished from the narrative, the injury was considered a lowenergy mechanism.

Ethnicity
The primary Ethnicity categories in the NEISS dataset include 'Not Specified,' 'White,' 'Black/African American,' 'Other,' 'Asian,' 'American Indian/Alaska Native,' and 'Native Hawaiian/Pacific Islander.'Due to the small number of cases coded with the latter three ethnicity categories, these were combined with the 'Other' ethnicity category, which is coded when either the ED record indicates more than one ethnicity, or when the ethnicity stated in the ED record does not fall into any of the other categories.Of note, Hispanic ethnicity is a separate variable in the NEISS, which is coded exclusive of the primary ethnicity.This variable is used to indicated whether the patient is a person of Hispanic, Latino/Latina, or Spanish origin.

Analysis
Statistical analysis was performed using the survey data commands (svyset) in Stata/IC, version 17.0 (StataCorp, College Station, TX, US), accounting for sample weights and the complex survey design.Results are reported as numbers of unweighted cases (n) and/or as weighted national estimates (N) with corresponding 95% confidence intervals (CIs).Each unweighted case (n) in the NEISS represents a single patient ED encounter and has a sample weight which represents the inverse of the probability of selection of the case.Sample weights are summed to provide national estimates (N).The US Census Bureau population estimates for the years 2001 through 2020 were used to calculate at-risk person-years during the study period [26].Thus, the population at risk was defined as the entire US population in order to derive national incidence estimates.Incidence rates (IRs) are expressed as the number of injuries per 100,000 person-years at-risk (PYR) and are calculated as the number of estimated injuries (N) divided by person-years at-risk.Student t-test/analysis of variance and design-adjusted Rao-Scott chi-square analysis was used for direct comparisons of means for continuous variables and proportions for categorical variables, respectively [27,28].Temporal trends in annual incidence rates and/or annual injury estimates over the course of the study period were assessed with Joinpoint regression analyses (Joinpoint Regression Program, Version 4.8.01 -April 2020; Statistical Methodology and Applications Branch, Surveillance Research Program, National Cancer Institute) [29].Average annual percent change (AAPC) estimates are presented to indicate the magnitude and direction of trends in injury rates for each time period as determined by Joinpoint regressions.A p-value of less than 0.05 was the threshold for statistical significance.Of note, the CPSC considers a national estimate unstable and potentially unreliable when the number of cases used is less than 20 or when the estimate is less than 1,200 [30].These numbers, therefore, should be interpreted with caution.

Results
During the 20-year period from 2001-2020, an estimated total of N = 159,529 (95% CI = 126,330-192,728) acute, first-time patellar dislocations occurred in an at-risk population of 6,183,899,410 person-years for an overall incidence rate of 2.58 PYR (95% CI = 2.04-3.12) in the US.Accounting for population growth, the annual incidence increased significantly over the course of the study period, from 2.61 PYR in 2001 to 3.0 PYR in 2020 (AAPC = 2.8, p < 0.0001; Figure 2).
Peak incidence occurred in the second decade (IR = 11.57PYR, 95% CI = 9.05-14.1).Overall, 61.2% of all patellar dislocations occurred in the second decade, and >80% of cases occurred between the ages of 10-29 years.Mean patient age was 21.6 years (95% CI = 20.9-22.2;median age = 17 years), and the mean age of patients sustaining patellar dislocations did not change over the course of the study period (p = 0.4255).When looking at temporal trends in ageadjusted annual incidence rates over the study period, changes did not affect all age groups proportionally (Figure 3).Statistically significant increases in annual incidence rates over the study period were observed among patients aged 10-19 years (AAPC = 4.5, p < 0.0001) and among patients aged 20-29 years (AAPC = 1.8, p = 0.04), while no significant changes were observed in any of the other 10-year age groups (Figure 3).Additionally, the increase in annual incidence observed in patients aged 10-19 years was significantly greater than that observed in patients aged 20-29 years (p = 0.0033).When looking more closely at patients in the second decade (separated into 5-year age groups), annual incidence increased significantly and at similar rates in patients 10-14 years and 15-19 years of age (10-14 years: AAPC = 4.7, p < 0.0001; 15-19 years: AAPC = 4.1, p < 0.0001; comparison: p = 0.488).

Sex
Nearly half of cases occurred in male patients (50.5%) and the other half in female patients (49.5%).There were N = 80,544 cases in males (95% CI = 63,198-97,890) for an overall male incidence rate of 2.65 PYR (95% CI = 2.08-3.22),and N = 78,984 cases in females (95% CI = 62,567-95,402) for an overall female incidence rate of 2.51 PYR (95% CI = 1.99-3.04).The annual incidence of patellar dislocations increased significantly over the study period in both males (AAPC = 2.3, p < 0.0001) and females (AAPC = 3.2, p = 0.0002), and there was no significant difference in the rate of annual increase between the sexes (p = 0.2586).When looking at trends in age-specific injury rates, the annual incidence of patellar dislocations in patients aged 10-19 years increased significantly over the study in both males and females (males 10-19 years: AAPC = 3.8, p < 0.0001; females 10-19 years: AAPC = 5.3, p < 0.0001), with no significant difference in the rate of increase between the sexes (p = 0.1177), while the annual incidence in patients aged 20-29 years increased significantly only in female patients (females 20-29 years: AAPC = 3.5, p = 0.0152) while remaining unchanged in male patients (males 20-29 years: AAPC = 0.6, p = 0.6).
The age distributions of patellar dislocations for both males and females are illustrated in Figure 4.While peak incidence occurred in the second decade in both sexes, female peak incidence occurred during the first half of the second decade (age 10-14 years), while male peak incidence occurred during the second half (age 15-19 years; Figure 4).Overall, the age distribution in females was somewhat wider than that in males, and a larger percentage of injuries in females occurred in older age groups when compared with males.Accordingly, mean age was higher in females (22.6 years) when compared with males (20.6 years; p = 0.0012).Mean age did not change over the course of the study period for males (p = 0.8719) or females (p = 0.108).Injury mechanisms differed somewhat between the sexes.A significantly higher percentage of patellar dislocations in males were the result of sports-/exercise-related mechanisms when compared with those in females (75.6% versus 56.2%, respectively; p < 0.0001).Table 1 lists the most common sports-related mechanisms for both sexes.

Ethnicity
Ethnicity was specified in 74.5% of cases.Of those cases in which ethnicity was specified, White patients accounted for the largest percentage (75.9%),followed by Black/African American patients (15.6%).N = 90,256 patellar dislocations (95% CI = 67,316-113,196) occurred in White patients for an incidence rate of 1.87 PYR (95% CI = 1.39-2.34),and N = 18,570 dislocations (95% CI = 10,625-26,515) occurred in Black/African American patients for an incidence rate of 2.29 PYR (95% CI = 1.31-3.28).Very few cases of patellar dislocations occurred in Asian, American Indian/Alaska Native, and Native Hawaiian/Pacific Islander patients and thus these groups were included in the 'Other' ethnicity category.There were N = 10,086 patellar dislocations (95% CI = 5,171-15,001) in patients in the 'Other' ethnicity category for a combined incidence rate of 1.88 PYR (95% CI = 0.96-2.8) in these patients.The annual incidence of patellar dislocations increased significantly over the study period in White patients (AAPC = 3.5, p = 0.0001), Black patients (AAPC = 4.6, p = 0.0049), and 'Other' patients (AAPC = 3.3, p = 0.0105), with no significant differences in annual increases noted among the three ethnicity categories (p = 0.4916).Data regarding Hispanic ethnicity was available in just 84.8% of cases, which prevented the calculation of reliable estimates of injury rates in this group.

Injury mechanism
Two-thirds (66%) of patellar dislocations were the result of sports-/exercise-related injury mechanisms, while 32.9% were due to non-sports low-energy mechanisms (i.e.activities of daily living) and 1.1% were due to non-sports high-energy mechanisms.Differences between sports-related patellar dislocations and non-sports-related patellar dislocations are highlighted in Table 2.The annual incidence of both sports-related and non-sports-related patellar dislocations increased significantly over the study period (sports-related: AAPC = 2.6, p = 0.0001; non-sports-related: AAPC = 3.4, p = 0.0001; Figure 5), and the annual incidence increased at a similar rate in both groups (p = 0.3312).There was no significant change over the study period in the percentage of patellar  dislocations which were due to sports versus non-sports mechanisms.
The five most common sports overall in which patients sustained patellar dislocations were basketball (19.3%, N = 20,314), dance (14.6%, N = 15,389), football/rugby (9.8%, N = 10,319), soccer (9.3%, N = 9,833), and baseball/softball (6.9%, N = 7,279).Annual numbers of patellar dislocations increased significantly for basketball (AAPC = 2.5, p = 0.042) and dance (AAPC = 4, p = 0.008), but no statistically significant changes were observed in the other three most common individual sports.Additional information regarding the most common injury mechanisms for the athletic/sports-related patellar dislocations, both overall and stratified by age group and sex, is provided in the Supplement.
When grouping sports according to degree of physical contact/collision risk (with other athletes or with inanimate objects, including the ground) [31], contact sports accounted for the largest percentage of all sports-related patellar dislocations (34.2%), followed by noncontact sports (32.7%), limitedcontact sports (18.2%), and collision sports (14.9%).When comparing males to females, a significantly larger percentage of sports-related patellar dislocations in males occurred in collision/contact sports, while a significantly larger percentage in females occurred in noncontact sports (Table 3).Annual numbers of patellar dislocations increased significantly over the study for all sport categories except collision sports (contact sports: AAPC = 2.8, p = 0.0031; noncontact sports: AAPC = 5.1, p < 0.0001; limited-contact sports: AAPC = 2.7, p = 0.0196; collision sports: AAPC = 0.6, p = 0.5798).

Discussion
The current study presents recent nationwide epidemiologic data on acute, first-time patellar dislocations in the US over an observation period of twenty years.This investigation, which included an estimated 159,529 patients with patellar dislocations presenting to US EDs from 2001 to 2020, reinforces many of the findings demonstrated in previous studies regarding the demographic characteristics of individuals sustaining these injuries, while also highlighting a steady rise in overall, sex-, and ethnicity-specific annual injury rates.Accounting for population growth, incidence rates increased significantly in both males and females and among white, black, and 'other' ethnic groups, and no significant differences in the rates of increase between the sexes or between the different ethnic groups were observed.Furthermore, increases in annual incidence rates disproportionately affected younger patient age groups, with significant trends noted only among patients in the second and third decades.Finally, a large percentage of patellar dislocations occurred during athletic participation, but the incidence of both athletic and nonathletic injuries increased significantly over the study period.
Regarding patient demographic characteristics, the findings of the current study are generally in line with those of previously published investigations.A large percentage (61%) of patellar dislocations in the current study occurred in patients aged 10-19 years and the vast majority (>80%) occurred in the second and third decades of life.This is consistent with previously published reports which indicate that incidence rates are highest in adolescents and youth [3,5,13,14,32].The sex distribution in the current study was nearly equal between male and female patients, with each group accounting for 50.5% and 49.5% of injuries, respectively.While the female peak incidence occurred slightly earlier than the male peak incidence (10-14 years versus 15-19 years, respectively), overall incidence rates were similar between the sexes.While female sex has traditionally been proposed to be a risk factor for patellar dislocation, several recent studies have documented similar incidence rates in males and females [5,13,14,32,33], which was also the case in the current investigation.
Individuals of certain ethnic groups have also been shown to sustain patellar dislocations at higher rates than others, although studies on this topic are quite limited.Hsiao et al. [17] analyzed active-duty US military personnel and found the incidence of patellar dislocations in white and black service Other Ethnicity includes more than one ethnicity and/or any ethnicity that is not either White or Black/African American (including but not limited to Asian, American Indian/Alaska Native, Native Hawaiian/Pacific Islander).members to be significantly higher than that of 'other' service members ('other' including all patients not either white or black ethnicity), with no differences between white and black ethnicities.Waterman et al. [13] also noted similar injury rates in white and black patients in the US general population, but these were significantly higher than injury rates in the Hispanic population.In the current investigation, incidence rates were similar in patients of white, black, and 'other' ethnicity categories.This discrepancy may be the result of differing patient populations (i.e.Hsiao et al. studied an activeduty military population, which may not reflect the general population) and/or methods of categorizing ethnic groups (i.e.Waterman et al. analyzed Hispanic, Asian, and Native American patients separately, whereas all individuals of ethnicity other than white or black were combined into one group in the current study, and the pooling of ethnic group categories may explain the higher overall injury rate).Additionally, although Waterman et al. [13] observed a significantly lower incidence among Hispanic patients when compared with White and Black patients, we could not reliably calculate incidence rates among Hispanic patients as information regarding Hispanic ethnicity was missing in >80% of cases in the current study, precluding similar comparisons.Finally, apart from missing Hispanic ethnicity data, 25.5% of cases had incomplete primary ethnicity data, limiting the strength of the ethnicityrelated findings of the current investigation.
In the current study, the overall annual incidence of acute, first-time patellar dislocations in the US was found to increase significantly from 2001 to 2020.Previous research examining trends in the incidence of patellar instability, however, have generated mixed results.Earlier population-based studies in the US demonstrated no significant changes in overall annual incidence [13,14], but these investigations were limited to data only up to the year 2010, which not may reflect the current epidemiological patterns of patellar dislocations in the US.The current study included data through the year 2020, and differences between the time periods which were analyzed may explain the differences in the observed trends.
In a population-based study in Europe, Gravesen et al. [32] analyzed a Danish national registry of >24,000 patients over a 20-year period (1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013) and observed a significantly increased incidence of patellar dislocations in patients aged 10-17 years during the years 1998-2013, but noted no other significant trends (overall or age-/sex-specific) during that timeframe.Similarly, a significantly increased annual incidence was observed in adolescent patients in the current study (10-19 years), but other significant trends were also noted, including in the overall and sex-specific incidence rates and also in patients aged 20-29 years.The reasons for these discrepancies are unclear, but the sociodemographic context of Denmark and factors unique to its population may limit the translation of the findings of the former study to patients from other nations, including the US, which may explain these differences.
Two more recent studies of a US pediatric population (both utilizing a large, administrative pediatric database) provide data to suggest that injury rates are indeed rising in pediatric/adolescent patients in the US.Poorman et al. [15] analyzed the Pediatric Health Information System (PHIS) database and reported a significant increase in annual pediatric hospitalizations for patellar dislocations occurring from 2004-2016.McFarlane et al. [16], also utilizing PHIS data, reported that the number of pediatric patients who underwent surgery for patellar instability from 2004-2014 increased significantly (2.1-fold increase), as did the number of patients who were seen in the ED and underwent a closed reduction of a patellar dislocation (2.7-fold increase).The PHIS dataset, however, is limited to tertiary-care referral pediatric hospitals, and it's unclear if these trends would also be present at community or non-specialty hospitals.Additionally, these studies analyzed pediatric patients (≤18 years of age) as one group and did not stratify patients by age, limiting age-specific comparisons within the pediatric population.In the current study, while injury rates increased substantially in patients aged 10-19 years, there was no significant change in patients aged 0-9 years.Furthermore, injury rates increased proportionally in patients 10-14 years of age compared to those who were 15-19 years of age.The lack of change in injury rates among patients aged 0-9 years is likely due to a very low overall rate of injury in patients of this age group.Taken together, these findings demonstrate an increasing national burden of patellar instability among young people, specifically among patients in the second decade of life.
The only other 10-year age group in which a significant increase in annual patellar dislocation injury rates was observed was in patients 20-29 years of age.Interestingly, within this age group, this trend was observed only in female patients, while the annual incidence of patellar dislocations in males 20-29 years of age did not change over the study period.The reasons for this are unclear.Similar sex-related differences were not observed in patients 10-19 years of age, as annual incidence increased significantly and at similar rates in both males and females of this age.No significant trends in age-or sex-specific incidence rates over the study period were observed in patients ≥30 years old.One possible explanation for increasing injury rates is an overall increase in exercise and athletic participation among the US population, given that a large percentage of these injuries occur during sporting activities.Indeed, recent data indicates that in 2021, the total number of active sports and fitness participants in the US reached its highest level in the previous six years, with an overall estimated participation rate of over 76% of the population, up from <73% in 2016 [34].Additionally, basketball, the most common sports-related mechanism of injury in the current study, has remained the most popular sport in the US with the largest total number of participants overall [34].Therefore, increasing athletic participation and a general preference for sporting activities associated with a high risk of patellar dislocation (e.g.basketball) among the US population may in part explain the observed trends.This is supported by an observed increase in athletic patellar dislocations in the current study.On the other hand, the annual incidence of nonathletic patellar dislocations also increased and at a similar rate, and there was no significant change in the percentage of patellar dislocations which were due to sports versus non-sports mechanisms over the study period.These findings suggest that increased athletic participation within the population alone cannot account for the trends in overall injury rates demonstrated in the study.
There are other possible explanations for the increased injury rates observed.One possible explanation is that the number of individuals with certain injury risk factors is increasing.Obesity, for example, which has been described as an important risk factor for patellar dislocations [7,35], has been steadily rising within the US population [36], and therefore a growing number of individuals may be at a higher-thannormal risk for these injuries.The influence of other important predisposing factors, such as structural and/or mechanical abnormalities (e.g.trochlear dysplasia, genu valgum, etc.), is unclear as relatively less is known about the prevalence of these factors at the population level and whether the prevalence has changed recent years.
It is also possible that the observed trends are partly due to factors other than an actual change in the incidence of patellar dislocations, such as more accurate coding within the database, increased recognition and diagnosis by ED physicians, or increased ED attendance among patients with patellar dislocations.While the CPSC employs rigorous data collection methodologies and although all of the data have been reviewed and edited carefully for consistency and valid codes [18,20], minor inconsistencies in coding may be possible, which have the potential to influence national injury estimates.Onsite hospital coders are trained to transcribe exactly what medical providers have recorded about each ED encounter, and most narrative descriptions are based on physicians' and nurses' notes.Improved recognition and documentation of patellar dislocations by ED providers, therefore, could potentially contribute to the apparent increase in injury rates over time.Given the long-standing standardization of injury data recording and consistent methodology within the NEISS over time, it's unlikely that these factors played a large role, but it's possible.Finally, there is data to suggest that overall ED utilization in the US is increasing [37].It's unknown what percentage of patients with acute, first-time patellar dislocations present to EDs for treatment, but if similar trends are present within this group, then increased ED attendance could potentially play some role in the increased injury rates observed in this study.
Given an apparently increasing rate of acute, first-time patellar dislocations within the US population, and in particular among young athletes, there is a need for more research to define the optimal rehabilitation strategies and return-to-sport protocols for individuals who sustain these injuries.Despite the abundant research on surgical treatment options [9,38], there is a relatively scarce amount of evidence available regarding rehabilitation and return-to-play after patellar dislocations, and there currently exists wide variation in recommendations [38].Future research is needed to establish standardized guidelines for optimizing rehabilitation, minimizing time loss, and allowing for a safe return to sports after these injuries.
There are several limitations to the current study.Because the NEISS database contains information regarding ED visits only, any patients with patellar dislocations who received care in an alternate setting (e.g.outpatient or urgent care clinic) were not captured in the data.For this reason, the injury rates presented are certainly conservative estimates.Additionally, sampling bias may have resulted from an omission of patients receiving care in alternate healthcare settings, who may represent a different demographic.Another significant limitation is the reliance on the case narratives (which are limited to a specific number of characters) to identify patients with patellar dislocations.We sought to identify only patients with acute, first-time patellar dislocations, but information regarding any prior history of patellar instability was inconsistently reported in the narratives.Thus, increasing injury rates may in part reflect an increasing number of recurrent dislocations.Furthermore, the ability to distinguish patellar subluxation from true luxation was limited by the small amount of available injury data, which is a major limitation.Also, the accuracy of analysis is dependent upon the correctness of the narrative, which is inherently prone to reporter bias.The available data also do not allow for an assessment of whether the observed differences over time resulted from an actual change in incidence or for other reasons, such as those previously described.Finally, the system does not include any radiographic data, which limited our ability to identify and describe the presence of concomitant injuries such as articular cartilage injuries and osteochondral fractures, which have important treatment and prognostic implications.Despite these limitations, the NEISS is an established model for epidemiological surveillance of musculoskeletal injuries, and its strengths include its large scale, heterogenous patient population, and substantial external validity.Finally, while this analysis, including an estimated 159,000+ acute, first-time patellar dislocations occurring over a 20-year period, represents the largest epidemiological study of patellar dislocations (to the authors' knowledge), the sociodemographic context of the US (and factors unique to its patient population) may limit the translation of these findings to patients from other nations.

Conclusion/summary
The number of patients presenting to US EDs for acute, firsttime patellar dislocations is increasing, suggesting that the nationwide injury burden in the US is growing.Increased injury rates were observed among males and females aged 10-19 years and among females aged 20-29 years.An increasing number of athletic/sports-related patellar dislocations may reflect increasing levels of athletic participation, an increase in sport specialization, and/or a general preference for certain sports which are associated with a high risk of injury (e.g.basketball) among the US population.Given that the rate of acute, first-time patellar dislocations appears to be increasing within the US population, and in particular among young athletes, future research is needed to define the optimal rehabilitation strategies and return-to-play protocols for individuals who sustain these injuries.
Word count: 4,904 words

Figure 1 .
Figure 1.Flowchart outlining the methods used to select cases of patellar dislocations from the National Electronic Injury Surveillance System (NEISS) database.

Figure 2 .
Figure 2. Trends in the annual incidence of all patellar dislocations in the United States, 2001-2020.

Figure 3 .
Figure 3. Trends in the annual incidence of patellar dislocations in the United States, 2001-2020, by 10-year age group.Average annual percent change (AAPC) estimates are reported for each age group to indicate the magnitude and direction of trends in annual incidence rates from 2001-2020.

Figure 4 .
Figure 4. Age-and sex-specific incidence of all patellar dislocations in the United States, 2001-2020.

Figure 5 .
Figure 5. Trends in the annual incidence of all sports-related and non-sportsrelated patellar dislocations in the United States, 2001-2020.

Table 1 .
Most common sports responsible for all athletic patellar dislocations in the United States, 2001-2020, by sex.
N, estimated national total number of injuries; %, percentage of all sports-related patellar dislocations by sex.*Includes aerobics, jogging, running, etc. (excluding track & field).

Table 2 .
Comparison of all sports-related and non-sports-related patellar dislocations in the United States, 2001-2020.
*Incidence rates are expressed per 100,000 person-years at-risk.ǂ

Table 3 .
Broad categories of sports (according to degree of physical contact/collision risk) responsible for all athletic patellar dislocations in the United States, 2001-2020, by sex.