Clinical Prediction Models in Sports Medicine: A Guide for Clinicians and Researchers

Prevalence of physiological and perceptual markers of low energy availability in male academy football players: a study protocol for a cross-sectional study

Low energy availability (LEA) is a core feature of the female athlete triad and relative energy deficiency in sport (REDs). LEA underpins multiple adverse health and performance outcomes in various athletic populations, including weight category, endurance and aesthetic sports. Recent reports suggest LEA is highly prevalent in female football, volleyball and netball, with little known on male team-sport athletes. Therefore, the study aims to identify the prevalence of LEA among male academy football players (16-23 years), using surrogate markers that align with the International Olympic Committee REDs Clinical Assessment Tool-Version 2. A cross-sectional study design will be used with physiological and perceptual markers of LEA measured. The study will seek to recruit 355 players to complete several online questionnaires believed to be associated with LEA, measured using a 24-hour food and activity diary. Of the 355 players, a subsample (n=110) will complete an additional 3-day food and activity diary, provide a venous blood sample to measure levels of total testosterone and free triiodothyronine, and have resting metabolic rate (RMR) measured to determine RMRratio. The prevalence of LEA will be determined using the low (<30 kcal·kgFFM-1·day-1) domain of energy availability and divided by the total number of participants. Descriptive statistics will be used to summarise the whole group and difference status of energy availability (eg, low, reduced, optimal, high). A univariable and multivariable binary logistic regression analysis will be modelled to assess the association of various surrogate markers with the presence of LEA.

Training-Load Management Ambiguities and Weak Logic: Creating Potential Consequences in Sport Training and Performance

BACKGROUND

The optimization of athlete training load is not a new concept; however in recent years, the concept of "load management" is one of the most widely studied and divisive topics in sports science and medicine.

PURPOSE

Discuss the challenges faced by sports when utilizing training load monitoring and management, with a specific focus on the use of data to inform load management guidelines and policies/mandates, their consequences, and how we move this field forward.

CHALLENGES

While guidelines can theoretically help protect athletes, overzealous and overcautious guidelines may restrict an athlete's preparedness, negatively influence performance, and increase injury risk. Poor methods, wrong interpretation of study findings, and faulty logic do not allow for systematic scientific evaluations to inform guidelines. Practical Solutions: Guidelines and mandates should be developed through a systematic research process with stronger research designs and clear research questions. Collaborating with statistical and epidemiological experts is essential. Implementing open science principles and sharing all sports training load data increase transparency and allow for more rapid and valid advancements in knowledge. Practitioners should incorporate multiple data streams and consider individual athlete responses, rather than applying broad guidelines based on average data.

CONCLUSION

Many current training load guidelines and mandates in sports come from good intentions; however, they are arbitrary without sound knowledge of the underlying scientific principles or methods. Common sense guidelines are helpful when there is sparse literature, but they should be careful to avoid arbitrarily choosing findings from weak research. Without precise scientific inquiries, implementing training load interventions or guidelines can have negative implications.

Factors associated with non-specific low back pain in field hockey: A cross-sectional study of Premier and Division One players

OBJECTIVE

To determine the extent to which various factors are associated with greater or lesser odds of reporting non-specific low back pain (NS-LBP) in field hockey.

METHODS

To meet the objective of the study, a cross-sectional study design was used with a purposive sampling strategy. A total of 194 responses (~18% of those accessible) from Premier and Division One players within the UK were received using a UK-based online survey. Data collected included information on NS-LBP, participant characteristics, injury history, training related factors, and work and personal factors. The overall and category-specific prevalence of NS-LBP was calculated. Univariable and multivariable logistic regression was used in conjunction with clinical value to identify associations.

RESULTS

The overall prevalence of NS-LBP was 44.0%, with this varying from 23.5 to 70.0% for categories with responses of "yes" and "no" to experiencing NS-LBP. A total of ten individual factors associated with a greater odds ratio (OR) of reporting NS-LBP (OR = 1.43-7.39) were identified in Premier and Division One players. Five individual factors were associated with reduced odds (OR = 0.11-0.60) of reporting NS-LBP. Seven factors (age, stature, playing position, playing internationally, performing a drag flick, low back stiffness/tightness and occupational factors) were deemed particularly pertinent to those working in field hockey given the magnitude of association and clinical value to clinicians.

CONCLUSIONS

Clinicians working in field hockey can consider the key risk factors identified in this study that are associated with NS-LBP when assessing injury risk, movement screening approaches, and overall athlete management.

A machine learning framework to classify musculoskeletal injury risk groups in military service members

Background

Musculoskeletal injuries (MSKIs) are endemic in military populations. Thus, it is essential to identify and mitigate MSKI risks. Time-to-event machine learning models utilizing self-reported questionnaires or existing data (e.g., electronic health records) may aid in creating efficient risk screening tools.

Methods

A total of 4,222 U.S. Army Service members completed a self-report MSKI risk screen as part of their unit's standard in-processing. Additionally, participants' MSKI and demographic data were abstracted from electronic health record data. Survival machine learning models (Cox proportional hazard regression (COX), COX with splines, conditional inference trees, and random forest) were deployed to develop a predictive model on the training data (75%; n = 2,963) for MSKI risk over varying time horizons (30, 90, 180, and 365 days) and were evaluated on the testing data (25%; n = 987). Probability of predicted risk (0.00-1.00) from the final model stratified Service members into quartiles based on MSKI risk.

Results

The COX model demonstrated the best model performance over the time horizons. The time-dependent area under the curve ranged from 0.73 to 0.70 at 30 and 180 days. The index prediction accuracy (IPA) was 12% better at 180 days than the IPA of the null model (0 variables). Within the COX model, "other" race, more self-reported pain items during the movement screens, female gender, and prior MSKI demonstrated the largest hazard ratios. When predicted probability was binned into quartiles, at 180 days, the highest risk bin had an MSKI incidence rate of 2,130.82 ± 171.15 per 1,000 person-years and incidence rate ratio of 4.74 (95% confidence interval: 3.44, 6.54) compared to the lowest risk bin.

Conclusion

Self-reported questionnaires and existing data can be used to create a machine learning algorithm to identify Service members' MSKI risk profiles. Further research should develop more granular Service member-specific MSKI screening tools and create MSKI risk mitigation strategies based on these screenings.

Eating disorder symptoms and weight pressure in female rowers: associations between self-concept, psychological well-being and body composition

BACKGROUND

Female rowers may be at risk of eating disorders and high weight pressure.

AIM

The purpose of the study was to investigate the prevalence of disordered eating symptoms and weight-related pressure and the associations with self-concept, psychological well-being, socio-demographic data, experience, performance level and body composition in female fixed-bench rowers.

METHODS

Female rowers (n = 208; age ranged mean ± SD 23.6 ± 6.5 years) completed the SCOFF scale, Weight-Pressures in Sport-Females (WPS-F), Physical Self-Concept Questionnaire and the Ryff scales of psychological well-being and provided information on their experience and level of competition. In a subgroup of 115 athletes, body composition was assessed using bioimpedance.

RESULTS

It was found that 42.3% of the athletes scored ≥ 2 on SCOFF and mean ± SD value of WPS-F score was 3.65 ± 0.82. Stepwise regression analysis revealed that self-concept of strength and pressure from teammates and the uniform were associated with higher ED symptoms, whereas better psychological well-being in terms of autonomy, self-concept of attractiveness, and age were protective factors for ED symptoms. BMI, athletes' physical condition, strength, and experience were associated with more weight-related pressure and better self-concept of attractiveness and physical well-being of autonomy were significantly associated with less pressure. In body composition analysis, higher extra cellular water, self-acceptance, and physical condition were associated with more weight-related pressure in female rowers, being attractiveness and the environmental mastery protective elements.

CONCLUSIONS

The prevalence of ED symptomatology and weight-related pressure are high in female fixed bench rowing. The psychological factors of well-being and self-concept, team environment, body image concerns and body composition analysis should be considered to promote healthy eating behaviours in female rowers.

Can a Psychologic Profile Predict Successful Return to Full Duty After a Musculoskeletal Injury?

BACKGROUND

Psychologic variables have been shown to have a strong relationship with recovery from injury and return to work or sports. The extent to which psychologic variables predict successful return to work in military settings is unknown.

QUESTIONS/PURPOSES

In a population of active duty soldiers, (1) can a psychologic profile determine the risk of injury after return to full duty? (2) Do psychologic profiles differ between soldiers sustaining injuries in the spine (thoracic or lumbar) and those with injuries to the lower extremities?

METHODS

Psychologic variables were assessed in soldiers returning to full, unrestricted duty after a recent musculoskeletal injury. Most of these were noncombat injuries from work-related physical activity. Between February 2016 and September 2017, 480 service members who were cleared to return to duty after musculoskeletal injuries (excluding those with high-velocity collisions, pregnancy, or amputation) were enrolled in a study that tracked subsequent injuries over the following year. Of those, we considered individuals with complete 12-month follow-up data as potentially eligible for analysis. Based on that, approximately 2% (8 of 480) were excluded because they did not complete baseline surveys, approximately 2% (11 of 480) were separated from the military during the follow-up period and had incomplete injury data, 1% (3 of 480) were excluded for not serving in the Army branch of the military, and approximately 2% (8 of 480) were excluded because they were not cleared to return to full duty. This resulted in 450 soldiers analyzed. Individuals were 86% (385 of 450) men; 74% (331 of 450) had lower extremity injuries and 26% (119 of 450) had spinal injuries, including soft tissue aches and pains (for example, strains and sprains), fractures, and disc herniations. Time-loss injury within 1 year was the primary outcome. While creating and validating a new prediction model using only psychological variables, 19 variables were assessed for nonlinearity, further factor selection was performed through elastic net, and models were internally validated through 2000 bootstrap iterations. Performance was deciphered through calibration, discrimination (area under the curve [AUC]), R 2 , and calibration in the large. Calibration assesses predicted versus actual risk by plotting the x and y intersection of these values; the more similar predicted risk values are to actual ones, the closer the slope of the line formed by the intersection points of all subjects is to equaling "1" (optimal calibration). Likewise, perfect discrimination (predicted injured versus actual injured) presents as an AUC of 1. Perfect calibration in the large would equal 0 because it represents the average predicted risk versus the actual outcome rate. Sensitivity analyses stratified groups by prior injury region (thoracic or lumbar spine and lower extremity) as well as the severity of injury by days of limited duty (moderate [7-27 days] and severe [28 + days]).

RESULTS

A model comprising primarily psychologic variables including depression, anxiety, kinesiophobia, fear avoidance beliefs, and mood did not adequately determine the risk of subsequent injury. The derived logistic prediction model had 18 variables: R 2 = 0.03, calibration = 0.63 (95% confidence interval [CI] 0.30 to 0.97), AUC = 0.62 (95% CI 0.52 to 0.72), and calibration in the large = -0.17. Baseline psychologic profiles between body regions differed only for depression severity (mean difference 1 [95% CI 0 to 1]; p = 0.04), with greater mean scores for spine injuries than for lower extremity injuries. Performance was poor for those with prior spine injuries compared with those with lower extremity injuries (AUC 0.50 [95% CI 0.42 to 0.58] and 0.63 [95% CI 0.57 to 0.69], respectively) and moderate versus severe injury during the 1-year follow-up (AUC 0.61 [95% CI 0.51 to 0.71] versus 0.64 [95% CI 0.64 to 0.74], respectively).

CONCLUSION

The psychologically based model poorly predicted subsequent injury. This study does not minimize the value of assessing the psychologic profiles of injured athletes, but rather suggests that models looking to identify injury risk should consider a multifactorial approach that also includes other nonpsychologic factors such as injury history. Future studies should refine the most important psychologic constructs that can add the most value and precision to multifactorial models aimed at identifying the risk of injury.

LEVEL OF EVIDENCE

Level III, prognostic study.

Association of quantitative diffusion tensor imaging measures with time to return to sport and reinjury incidence following acute hamstring strain injury

Hamstring strain injuries (HSI) are a common occurrence in athletics and complicated by limited prognostic indicators and high rates of reinjury. Assessment of injury characteristics at the time of injury (TOI) may be used to manage athlete expectations for time to return to sport (RTS) and mitigate reinjury risk. Magnetic resonance imaging (MRI) is routinely used in soft tissue injury management, but its prognostic value for HSI is widely debated. Recent advancements in musculoskeletal MRI, such as diffusion tensor imaging (DTI), have allowed for quantitative measures of muscle microstructure assessment. The purpose of this study was to determine the association of TOI MRI-based measures, including the British Athletic Muscle Injury Classification (BAMIC) system, edema volume, and DTI metrics, with time to RTS and reinjury incidence. Negative binomial regressions and generalized estimating equations were used to determine relationships between imaging measures and time to RTS and reinjury, respectively. Twenty-six index injuries were observed, with five recorded reinjuries. A significant association was not detected between BAMIC score and edema volume at TOI with days to RTS (p-values ≥ 0.15) or reinjury (p-values ≥ 0.13). Similarly, a significant association between DTI metrics and days to RTS was not detected (p-values ≥ 0.11). Although diffusivity metrics are expected to increase following injury, decreased values were observed in those who reinjured (mean diffusivity, p = 0.016; radial diffusivity, p = 0.02; principal effective diffusivity eigenvalues, p-values = 0.007-0.057). Additional work to further understand the directional relationship observed between DTI metrics and reinjury status and the influence of external factors is warranted.

Beyond yellow flags: The Big-Five personality traits and psychologically informed musculoskeletal rehabilitation

BACKGROUND

Psychosocial variables are known to play an important role in musculoskeletal pain. Recent efforts incorporating psychological theory into rehabilitative medicine, as part of patient-centred care or psychologically informed physical therapy, have gained broader acceptance. The fear-avoidance model is the dominant psychosocial model and has introduced a variety of phenomena which assess psychological distress (i.e., yellow flags). Yellow flags, such as fear, anxiety and catastrophizing, are useful concepts for musculoskeletal providers but reflect a narrow range of psychological responses to pain.

OBJECTIVE

Clinicians lack a more comprehensive framework to understand psychological profiles of each patient and provide individualised care. This narrative review presents the case for applying personality psychology and the Big-Five trait model (extraversion, agreeableness, conscientiousness, neuroticism and openness to experience) to musculoskeletal medicine. These traits have strong associations with various health outcomes and provide a robust framework to understand patient emotion, motivation, cognition and behaviour.

KEY RESULTS

High conscientiousness is associated with positive health outcomes and health promoting behaviours. High neuroticism with low conscientiousness increases the odds of negative health outcomes. Extraversion, agreeableness and openness have less direct effects but have positive correlations with important health behaviours, including active coping, positive affect, rehabilitation compliance, social connection and education level.

CLINICAL APPLICATION

The Big-Five model offers an evidence-based way for MSK providers to better understand the personality of their patients and how it relates to health. These traits offer the potential for additional prognostic factors, tailored treatments and psychological intervention.

The Trade Secret Taboo: Open Science Methods are Required to Improve Prediction Models in Sports Medicine and Performance

Clinical prediction models in sports medicine that utilize regression or machine learning techniques have become more widely published, used, and disseminated. However, these models are typically characterized by poor methodology and incomplete reporting, and an inadequate evaluation of performance, leading to unreliable predictions and weak clinical utility within their intended sport population. Before implementation in practice, models require a thorough evaluation. Strong replicable methods and transparency reporting allow practitioners and researchers to make independent judgments as to the model's validity, performance, clinical usefulness, and confidence it will do no harm. However, this is not reflected in the sports medicine literature. As shown in a recent systematic review of models for predicting sports injury models, most were typically characterized by poor methodology, incomplete reporting, and inadequate performance evaluation. Because of constraints imposed by data from individual teams, the development of accurate, reliable, and useful models is highly reliant on external validation. However, a barrier to collaboration is a desire to maintain a competitive advantage; a team's proprietary information is often perceived as high value, and so these 'trade secrets' are frequently guarded. These 'trade secrets' also apply to commercially available models, as developers are unwilling to share proprietary (and potentially profitable) development and validation information. In this Current Opinion, we: (1) argue that open science is essential for improving sport prediction models and (2) critically examine sport prediction models for open science practices.

What Factors Do Clinicians, Coaches, and Athletes Perceive Are Associated With Recovery From Low Back Pain in Elite Athletes? A Concept Mapping Study

OBJECTIVE: Identify factors that elite sport clinicians, coaches, and athletes perceive are associated with low back pain (LBP) recovery. DESIGN: Concept mapping methodology. METHOD: Participants brainstormed, sorted (thematically), and rated (5-point Likert scales: importance and feasibility) statements in response to the prompt, "What factors are associated with the recovery of an elite athlete from low back pain?" Data cleaning, analysis (multidimensional scaling, hierarchical cluster analysis, and descriptive statistics), and visual representation (cluster map and Go-Zone graph) were conducted following concept mapping guidelines. RESULTS: Participants (brainstorming, n = 56; sorting, n = 34; and rating, n = 33) comprised 75% clinicians, 15% coaches, and 10% athletes and represented 13 countries and 17 sports. Eighty-two unique and relevant statements were brainstormed. Sorting resulted in 6 LBP recovery-related themes: (1) coach and clinician relationships, (2) inter-disciplinary team factors, (3) athlete psychological factors, (4) athlete rehabilitation journey, (5) athlete non-modifiable risk factors, and (6) athlete physical factors. Participants rated important recovery factors as follows: athlete empowerment and psychology, coach-athlete and athlete-clinician relationships, care team communication, return-to-sport planning, and identifying red flags. CONCLUSION: Factors perceived as important to LBP recovery in elite athletes align with the biopsychosocial model of community LBP management. Clinicians should consider that an athlete's psychology, relationships, care team communication, and rehabilitation plan may be as important to their LBP recovery as the formulation of a diagnosis or the medications or exercises prescribed. J Orthop Sports Phys Ther 2023;53(10):1-16. Epub 10 August 2023. doi:10.2519/jospt.2023.11982.

Musculoskeletal injuries in US Air Force Tactical Air Control Party trainees: an 11-year longitudinal retrospective cohort study and presentation of a musculoskeletal injury classification matrix

INTRODUCTION

Little is known of the epidemiology of musculoskeletal injuries (MSKIs) in US Air Force Special Warfare (AFSPECWAR) Tactical Air Control Party trainees. The purpose of this longitudinal retrospective cohort study was to (1) report the incidence and type of MSKI sustained by AFSPECWAR trainees during and up to 1 year following training, (2) identify factors associated with MSKI, and (3) develop and present the MSKI classification matrix used to identify and categorise injuries in this study.

METHODS

Trainees in the Tactical Air Control Party Apprentice Course between fiscal years 2010-2020 were included. Diagnosis codes were classified as MSKI or non-MSKI using a classification matrix. Incidence rates and incidence proportion for injury types and regions were calculated. Measures were compared for differences between those who did and did not sustain an MSKI during training. A Cox proportional hazards model was used to identify factors associated with MSKI.

RESULTS

Of the 3242 trainees, 1588 (49%) sustained an MSKI during training and the cohort sustained MSKIs at a rate of 16 MSKI per 100 person-months. Overuse/non-specific lower extremity injuries predominated. Differences were seen in some baseline measures between those who did and did not sustain an MSKI. Factors retained in the final Cox regression model were age, 1.5-mile run times and prior MSKI.

CONCLUSION

Slower run times and higher age were associated with an increased likelihood of MSKI. Prior MSKI was the strongest predictor of MSKI during training. Trainees sustained MSKIs at a higher rate than graduates in their first year in the career field. The MSKI matrix was effective in identifying and categorising MSKI over a prolonged (12-year) surveillance period and could be useful for future injury surveillance efforts in the military or civilian settings. Findings from this study could inform future injury mitigation efforts in military training environments.

How to Develop and Validate Prediction Models for Orthopedic Outcomes

Prediction models are common in medicine for predicting outcomes such as mortality, complications, or response to treatment. Despite the growing interest in these models in arthroplasty (and orthopaedics in general), few have been adopted in clinical practice. If robustly built and validated, prediction models can be excellent tools to support surgical decision making. In this paper, we provide an overview of the statistical concepts surrounding prediction models and outline practical steps for prediction model development and validation in arthroplasty research. Please visit the followinghttps://www.youtube.com/watch?v=9Yrit23Rkicfor a video that explains the highlights of the paper in practical terms.

Development of an Injury Burden Prediction Model in Professional Baseball Pitchers

Background

Baseball injuries are a significant problem and have increased in incidence over the last decade. Reporting injury incidence only gives context to rate but not in relation to severity or injury time loss.

Hypothesis/Purpose

The purpose of this study was to 1) incorporate both modifiable and non-modifiable factors to develop an arm injury burden prediction model in Minor League Baseball (MiLB) pitchers; and 2) understand how the model performs separately on elbow and shoulder injury burden.

Study Design

Prospective longitudinal study.

Methods

The study was conducted from 2013 to 2019 on MiLB pitchers. Pitchers were evaluated in spring training arm for shoulder range of motion and injuries were followed throughout the season. A model to predict arm injury burden was produced using zero inflated negative binomial regression. Internal validation was performed using ten-fold cross validation. Subgroup analyses were performed for elbow and shoulder separately. Model performance was assessed with root mean square error (RMSE), model fit (R2), and calibration with 95% confidence intervals (95% CI).

Results

Two-hundred, ninety-seven pitchers (94 injuries) were included with an injury incidence of 1.15 arm injuries per 1000 athletic exposures. Median days lost to an arm injury was 58 (11, 106). The final model demonstrated good prediction ability (RMSE: 11.9 days, R2: 0.80) and a calibration slope of 0.98 (95% CI: 0.92, 1.04). A separate elbow model demonstrated weaker predictive performance (RMSE: 21.3; R2: 0.42; calibration: 1.25 [1.16, 1.34]), as did a separate shoulder model (RMSE: 17.9; R2: 0.57; calibration: 1.01 [0.92, 1.10]).

Conclusions

The injury burden prediction model demonstrated excellent performance. Caution should be advised with predictions between one to 14 days lost to arm injury. Separate elbow and shoulder prediction models demonstrated decreased performance. The inclusion of both modifiable and non-modifiable factors into a comprehensive injury burden model provides the most accurate prediction of days lost in professional pitchers.

Level of Evidence

2.

Improving Clinical Utility of Real-World Prediction Models: Updating Through Recalibration

ABSTRACT

Bullock, GS, Shanley, E, Thigpen, CA, Arden, NK, Noonan, TK, Kissenberth, MJ, Wyland, DJ, and Collins, GS. Improving clinical utility of real-world prediction models: updating through recalibration. J Strength Cond Res XX(X): 000-000, 2022-Prediction models can aid clinicians in identifying at-risk athletes. However, sport and clinical practice patterns continue to change, causing predictive drift and potential suboptimal prediction model performance. Thus, there is a need to temporally recalibrate previously developed baseball arm injury models. The purpose of this study was to perform temporal recalibration on a previously developed injury prediction model and assess model performance in professional baseball pitchers. An arm injury prediction model was developed on data from a prospective cohort from 2009 to 2019 on minor league pitchers. Data for the 2015-2019 seasons were used for temporal recalibration and model performance assessment. Temporal recalibration constituted intercept-only and full model redevelopment. Model performance was investigated by assessing Nagelkerke's R-square, calibration in the large, calibration, and discrimination. Decision curves compared the original model, temporal recalibrated model, and current best evidence-based practice. One hundred seventy-eight pitchers participated in the 2015-2019 seasons with 1.63 arm injuries per 1,000 athlete exposures. The temporal recalibrated intercept model demonstrated the best discrimination (0.81 [95% confidence interval [CI]: 0.73, 0.88]) and R-square (0.32) compared with original model (0.74 [95% CI: 0.69, 0.80]; R-square: 0.32) and the redeveloped model (0.80 [95% CI: 0.73, 0.87]; R-square: 0.30). The temporal recalibrated intercept model demonstrated an improved net benefit of 0.34 compared with current best evidence-based practice. The temporal recalibrated intercept model demonstrated the best model performance and clinical utility. Updating prediction models can account for changes in sport training over time and improve professional baseball arm injury outcomes.

Including Modifiable and Nonmodifiable Factors Improves Injury Risk Assessment in Professional Baseball Pitchers

OBJECTIVES

To (1) evaluate an injury risk model that included modifiable and nonmodifiable factors into an arm injury risk prediction model in Minor League Baseball (MiLB) pitchers and (2) compare model performance separately for predicting the incidence of elbow and shoulder injuries.

DESIGN

Prospective cohort.

METHODS

A 10-year MiLB injury risk study was conducted. Pitchers were evaluated during preseason, and pitches and arm injuries were documented prospectively. Nonmodifiable variables included arm injury history, professional experience, arm dominance, year, and humeral torsion. Modifiable variables included BMI, pitch count, total range of motion, and horizontal adduction. We compared modifiable, nonmodifiable, and combined model performance by R2, calibration (best = 1.00), and discrimination (area under the curve [AUC]; higher number is better). Sensitivity analysis included only arm injuries sustained in the first 90 days.

RESULTS

In this study, 407 MiLB pitchers (141 arm injuries) were included. Arm injury incidence was 0.27 injuries per 1000 pitches. The arm injury model (calibration 1.05 [0.81-1.30]; AUC: 0.74 [0.69-0.80]) had improved performance compared to only using modifiable predictors (calibration: 0.91 [0.68-1.14]; AUC: 0.67 [0.62-0.73]) and only shoulder range of motion (calibration: 0.52 [0.29, 0.75]; AUC: 0.52 [0.46, 58]). Elbow injury model demonstrated improved performance (calibration: 1.03 [0.76-1.33]; AUC: 0.76 [0.69-0.83]) compared to the shoulder injury model (calibration: 0.46 [0.22-0.69]; AUC: 0.62 [95% CI: 0.55, 0.69]). The sensitivity analysis demonstrated improved model performance compared to the arm injury model.

CONCLUSION

Arm injury risk is influenced by modifiable and nonmodifiable risk factors. The most accurate way to identify professional pitchers who are at risk for arm injury is to use a model that includes modifiable and nonmodifiable risk factors. J Orthop Sports Phys Ther 2022;52(9):630-640. Epub: 9 July 2022. doi:10.2519/jospt.2022.11072.

Strength and Biomechanical Risk Factors for Noncontact ACL Injury in Elite Female Footballers: A Prospective Study

PURPOSE

This study aimed to determine if a preseason field-based test battery was prospectively associated with noncontact anterior cruciate ligament (ACL) injury in elite female footballers.

METHODS

In total, 322 elite senior and junior female Australian Rules Football and soccer players had their isometric hip adductor and abductor strength, eccentric knee flexor strength, countermovement jump (CMJ) kinetics, and single-leg hop kinematics assessed during the 2019 preseason. Demographic and injury history details were also collected. Footballers were subsequently followed for 18 months for ACL injury.

RESULTS

Fifteen noncontact ACL injuries occurred during the follow-up period. Prior ACL injury (odds ratio [OR], 9.68; 95% confidence interval (95% CI), 2.67-31.46), a lower isometric hip adductor to abductor strength ratio (OR, 1.98; 95% CI, 1.09-3.61), greater CMJ peak take-off force (OR, 1.74; 95% CI, 1.09-3.61), and greater single-leg triple vertical hop average dynamic knee valgus (OR, 1.97; 95% CI, 1.06-3.63) and ipsilateral trunk flexion (OR, 1.60; 95% CI, 1.01-2.55) were independently associated with an increased risk of subsequent ACL injury. A multivariable prediction model consisting of CMJ peak take-off force, dynamic knee valgus, and ACL injury history that was internally validated classified ACL injured from uninjured footballers with 78% total accuracy. Between-leg asymmetry in lower limb strength and CMJ kinetics were not associated with subsequent ACL injury risk.

CONCLUSIONS

Preseason field-based measures of lower limb muscle strength and biomechanics were associated with future noncontact ACL injury in elite female footballers. These risk factors can be used to guide ACL injury screening practices and inform the design of targeted injury prevention training in elite female footballers.

Red flags for spinal pain in patients diagnosed with spinal infection in Nigeria: A 10-year medical records review

BACKGROUND

Spinal infection is a diagnostic challenge, the personal and economic consequences of misdiagnosis can be significant resulting in paralysis and instability of the spine and can ultimately be fatal. To aid identification of those at risk of spinal infection, a better understanding of the red flags for spinal infection is needed.

OBJECTIVE

To better understand which red flags may help to identify spinal infection.

DESIGN

and Methods: A 10-year medical records review of red flags for spinal infection in Nigeria, using a bespoke data extraction tool. Univariable and multivariable logistic regression was used to identify the main independent predictors of spinal pain.

RESULTS

124,913 records were reviewed, 1,645 patients were diagnosed with spinal infection. 79% of patients presented with spinal pain Univariable analysis revealed nine factors (some centres, all age groups above 16 years, co-morbidities, environmental factors, history of TB, radicular pain, pins and needles, numbness and spine tenderness.) were associated with greater odds (OR = 1.77-21.7, p < 0.001), whilst four (some centres, fatigue, fever and myotomal weakness) were associated with lower odds (OR = 0.51-0.59) of spine pain. Six factors were included in the final multivariable model associated with higher odds of spine pain: age groups above 16 years (OR 2.57 to 5.33, p < 0.05), co-morbidity (OR = 1.68, p < 0.05), history of TB (OR = 3.02, p < 0.05), weight loss (OR = 1.75, p < 0.01), radicular pain (OR = 19.88, p < 0.001); spine tenderness (OR = 6.54, p < 0.001). Myotomal weakness (OR = 0.66, p < 0.05) and fatigue (OR = 0.50, p < 0.01) were associated with lower odds of spinal pain in the final model.

CONCLUSION

Using data from ten hospitals in Nigeria within a ten-year period, we have produced a shortlist of red flags that can inform clinical decision making about potential spinal infection.

Comparing "people-like-me" and linear mixed model predictions of functional recovery following knee arthroplasty

OBJECTIVE

Prediction models can be useful tools for monitoring patient status and personalizing treatment in health care. The goal of this study was to compare the relative strengths and weaknesses of 2 different approaches for predicting functional recovery after knee arthroplasty: a neighbors-based "people-like-me" (PLM) approach and a linear mixed model (LMM) approach.

MATERIALS AND METHODS

We used 2 distinct datasets to train and then test PLM and LMM prediction approaches for functional recovery following knee arthroplasty. We used the Timed Up and Go (TUG)-a common test of mobility-to operationalize physical function. Both approaches used patient characteristics and baseline postoperative TUG values to predict TUG recovery from days 1-425 following surgery. We then compared the accuracy and precision of PLM and LMM predictions.

RESULTS

A total of 317 patient records with 1379 TUG observations were used to train PLM and LMM approaches, and 456 patient records with 1244 TUG observations were used to test the predictions. The approaches performed similarly in terms of mean squared error and bias, but the PLM approach provided more accurate and precise estimates of prediction uncertainty.

DISCUSSION AND CONCLUSION

Overall, the PLM approach more accurately and precisely predicted TUG recovery following knee arthroplasty. These results suggest PLM predictions may be more clinically useful for monitoring recovery and personalizing care following knee arthroplasty. However, clinicians and organizations seeking to use predictions in practice should consider additional factors (eg, resource requirements) when selecting a prediction approach.

Just How Confident Can We Be in Predicting Sports Injuries? A Systematic Review of the Methodological Conduct and Performance of Existing Musculoskeletal Injury Prediction Models in Sport

BACKGROUND

An increasing number of musculoskeletal injury prediction models are being developed and implemented in sports medicine. Prediction model quality needs to be evaluated so clinicians can be informed of their potential usefulness.

OBJECTIVE

To evaluate the methodological conduct and completeness of reporting of musculoskeletal injury prediction models in sport.

METHODS

A systematic review was performed from inception to June 2021. Studies were included if they: (1) predicted sport injury; (2) used regression, machine learning, or deep learning models; (3) were written in English; (4) were peer reviewed.

RESULTS

Thirty studies (204 models) were included; 60% of studies utilized only regression methods, 13% only machine learning, and 27% both regression and machine learning approaches. All studies developed a prediction model and no studies externally validated a prediction model. Two percent of models (7% of studies) were low risk of bias and 98% of models (93% of studies) were high or unclear risk of bias. Three studies (10%) performed an a priori sample size calculation; 14 (47%) performed internal validation. Nineteen studies (63%) reported discrimination and two (7%) reported calibration. Four studies (13%) reported model equations for statistical predictions and no machine learning studies reported code or hyperparameters.

CONCLUSION

Existing sport musculoskeletal injury prediction models were poorly developed and have a high risk of bias. No models could be recommended for use in practice. The majority of models were developed with small sample sizes, had inadequate assessment of model performance, and were poorly reported. To create clinically useful sports musculoskeletal injury prediction models, considerable improvements in methodology and reporting are urgently required.