Harnessing information from injury narrative in the 'big data' era: Understanding and applying machine learning for injury surveillance

Objective Vast amounts of injury narratives are collected daily and are available electronically in real time and have great potential for use in injury surveillance and evaluation. Machine learning algorithms have been developed to assist in identifying cases and classifying mechanisms leading to injury in a much timelier manner than is possible when relying on manual coding of narratives. The aim of this paper is to describe the background, growth, value, challenges and future directions of machine learning as applied to injury surveillance. Methods This paper reviews key aspects of machine learning using injury narratives, providing a case study to demonstrate an application to an established human-machine learning approach. Results The range of applications and utility of narrative text has increased greatly with advancements in computing techniques over time. Practical and feasible methods exist for semi-automatic classification of injury narratives which are accurate, efficient and meaningful. The human-machine learning approach described in the case study achieved high sensitivity and positive predictive value and reduced the need for human coding to less than one-third of cases in one large occupational injury database. Conclusion The last 20 years have seen a dramatic change in the potential for technological advancements in injury surveillance. Machine learning of ‘big injury narrative data’ opens up many possibilities for expanded sources of data which can provide more comprehensive, ongoing and timely surveillance to inform future injury prevention policy and practice.

The global burden of injury: Incidence, mortality, disability-adjusted life years and time trends from the Global Burden of Disease study 2013

Background The Global Burden of Diseases (GBD), Injuries, and Risk Factors study used the disability-adjusted life year (DALY) to quantify the burden of diseases, injuries, and risk factors. This paper provides an overview of injury estimates from the 2013 update of GBD, with detailed information on incidence, mortality, DALYs and rates of change from 1990 to 2013 for 26 causes of injury, globally, by region and by country. Methods Injury mortality was estimated using the extensive GBD mortality database, corrections for ill-defined cause of death and the cause of death ensemble modelling tool. Morbidity estimation was based on inpatient and outpatient data sets, 26 cause-of-injury and 47 nature-of-injury categories, and seven follow-up studies with patient-reported long-term outcome measures. Results In 2013, 973 million (uncertainty interval (UI) 942 to 993) people sustained injuries that warranted some type of healthcare and 4.8 million (UI 4.5 to 5.1) people died from injuries. Between 1990 and 2013 the global age-standardised injury DALY rate decreased by 31% (UI 26% to 35%). The rate of decline in DALY rates was significant for 22 cause-of-injury categories, including all the major injuries. Conclusions Injuries continue to be an important cause of morbidity and mortality in the developed and developing world. The decline in rates for almost all injuries is so prominent that it warrants a general statement that the world is becoming a safer place to live in. However, the patterns vary widely by cause, age, sex, region and time and there are still large improvements that need to be made.

The experience and meaning of injury among migrant farmworkers from Starr County, Texas

Occupational injuries suffered by migrant farmworkers in the United States continue to be a significant byproduct of the hazards of the agricultural work sector. Although the fields of study involving occupational health and safety have helped to further an understanding of the importance of injury prevention in the workplace, farmworkers have typically not been the beneficiaries of significant improvement in working conditions; similarly, public health policies have generally not been successful in significantly improving their protection. The primary objective of this research was to examine more closely, through in-depth interviews, the underlying reasons attributed by migrant farmworkers to their having suffered occupational injuries. By re-contacting and re-interviewing migrant farmworkers who had reported injuries in a study entitled A Cohort Study of Injuries in Migrant Farmworker Families in South Texas, the current research study employed qualitative research methods and determined several reasons that may help to explain why this specific group has suffered injuries and may not have reported such injuries, as well as to consider what their experiences may say about injury prevention within the context of public health.^

Validation of an instrument for injury data collection in strength training

Purpose: To provide for the basis for collecting strength training data using a rigorously validated injury report form. Methods: A group of specialist designed a questionnaire of 45 item grouped into 4 dimensions. Six stages were used to assess face, content, and criterion validity of the weight training injury report form. A 13 members panel assessed the form for face validity, and an expert panel assessed it for content and criterion validity. Panel members were consulted until consensus was reached. A yardstick developed by an expert panel using Intraclass correlation technique was used to assess the reability of the form. Test-retest reliability was assessed with the intraclass correlation coefficient (ICC).The strength training injury report form was developed, and the face, content, and criterion validity successfully assessed. A six step protocol to create a yardstick was also developed to assist in the validation process. Both inter-rater and intra rater reliability results indicated a 98% agreement. Inter-rater reliability agreement of 98% for three injuries. Results: The Cronbach?s alpha of the questionnaire was 0.944 (pmenor que0.01) and the ICC of the entire questionnaire was 0.894 (pmenor que0.01). Conclusion: The questionnaire gathers together enough psychometric properties to be considered a valid and reliable tool for register injury data in strength training, and providing researchers with a basis for future studies in this area. Key Words: data collection; validation; injury prevention; strength training

Injury incidence in judokas at the Spanish National University Championship

Background and Study Aim: Understanding injury incidence rates will be a great help with regards to preventing potential future damages. It is for this reason that this study suggests studying a large number of variables. The purpose of research is the relationship of events (empirical variables) that are usually taken into account in developing injury prevention programs during the battles and training in judo tournament. Material and methods: In this research project, 57 male judokas taking part in the Spanish National University Championship in 2009 were asked to complete a retrospective questionnaire. We analysed the following events: the most commonly injured body regions, the medical diagnosis, how and when the injury happened, the type of injury, the side of the body and the type of medical attention received. For the statistical analysis, we used the SPSS statistics programme to apply the Chi-square test in order to determine the significance levels for non-parametric tests from p<.05. Results: Significant differences were found in the most commonly injured body region, the shoulder/clavicle (p<.05), and in the most common diagnosis, the sprain (p<.05). Impact injuries (p<.05) are the most common and training (p<.05) is the most dangerous time. About the type of injury, 78.38% are new injuries (p<.05) and 69.05% affect the right hand side of the body (p<.05). Doctors are the most consulted specialists, but the physiotherapists obtained the best marks. Have been out due to injury for over 21 days 36.36% of the participants, but not for the entire season. Conclusions: The most common diagnosis in university student judokas coincides with those of elite judokas, with the sprain being the most common. University student judokas have a higher rate of shoulder/clavicle injuries, while professional judokas are prone to a higher rate of knee injuries. Training is the most common moment in which injuries occur, both in university student judokas and professional judokas. New injuries are the most common types of injuries in university student judokas and, while doctors are the most consulted specialists, the physiotherapists obtained the best marks.

Burden of disease and injury in Aboriginal and non-Aboriginal populations in the Northern Territory

Objective: To quantify the burden of disease and injury for the Aboriginal and non-Aboriginal populations in the Northern Territory. Design and setting: Analysis of Northern Territory data for 1 January 1994 to 30 December 1998 from multiple sources. Main outcome measures: Disability-adjusted life-years (DALYs), by age, sex, cause and Aboriginality. Results: Cardiovascular disease was the leading contributor (14.9%) to the total burden of disease and injury in the NT, followed by mental disorders (14.5%) and malignant neoplasms (11.2%). There was also a substantial contribution from unintentional injury (10.4%) and intentional injury (4.9%). Overall, the NT Aboriginal population had a rate of burden of disease 2.5 times higher than the non-Aboriginal population; in the 35-54-year age group their DALY rate was 4.1 times higher. The leading causes of disease burden were cardiovascular disease for both Aboriginal men (19.1%) and women (15.7%) and mental disorders for both non-Aboriginal men (16.7%) and women (22.3%). Conclusions: A comprehensive assessment of fatal and non-fatal conditions is important in describing differentials in health status of the NT population. Our study provides comparative data to identify health priorities and facilitate a more equitable distribution of health funding.

Community-based interventions for the prevention of burns and scalds in children

Background Burns and scalds are a significant cause of morbidity and mortality in children. Successful counter-measures to prevent burn and scald-related injury have been identified. However, evidence indicating the successful roll-out of these counter-measures into the wider community is lacking. Community-based interventions in the form of multi-strategy, multi-focused programmes are hypothesised to result in a reduction in population-wide injury rates. This review tests this hypothesis with regards to burn and scald injury in children. Objectives To assess the effects of community-based interventions, defined as coordinated, multi-strategy initiatives, for reducing burns and scalds in children aged 14 years and under. Search strategy We searched the Cochrane Injuries Group's specialised register, CENTRAL, MEDLINE, EMBASE, CINAHL, PsycINFO, National Research Register and the Web of Knowledge. We also handsearched selected journals and checked the reference lists of selected publications. The searches were last updated in May 2007. Selection criteria Included studies were those that reported changes in medically attended burn and scald-related injury rates in a paediatric population (aged 14 years and under), following the implementation of a controlled community-based intervention. Data collection and analysis Two authors independently assess studies for eligibility and extracted data. Due to heterogeneity between the included studies, a pooled analysis was not appropriate. Main results Of 39 identified studies, four met the criteria for inclusion. Two of the included studies reported a significant decrease in paediatric burn and scald injury in the intervention compared with the control communities. The failure of the other two studies to show a positive result may have been due to limited time-frame for the intervention and/or failure to adequately implement the counter-measures in the communities. Authors' conclusions There are a very limited number of research studies allowing conclusions to be drawn about the effectiveness of community-based injury prevention programmes to prevent burns and scalds in children. There is a pressing need to evaluate high-quality community-based intervention programmes based on efficacious counter-measures to reduce burns and scalds in children. It is important that a framework for considering the problem of burns and scalds in children from a prevention perspective be articulated, and that an evidence-based suite of interventions be combined to create programme guidelines suitable for implementation in communities throughout the world.

Quantifying the association between physical activity and injury in primary school-aged children

BACKGROUND. Regular physical activity is strongly advocated in children, with recommendations suggesting up to several hours of daily participation. However, an unintended consequence of physical activity is exposure to the risk of injury. To date, these risks have not been quantified in primary school-aged children despite injury being a leading cause for hospitalization and death in this population. OBJECT. Our goal was to quantify the risk of injury associated with childhood physical activity both in and out of the school setting and calculate injury rates per exposure time for organized and non-organized activity outside of school. METHODS. The Childhood Injury Prevention Study prospectively followed a cohort of randomly selected Australian primary school- and preschool-aged children (4 to 12 years). Over 12 months, each injury that required first aid attention was registered with the study. Exposure to physical activity outside school hours was measured by using a parent-completed 7-day diary. The age and gender distribution of injury rates per 10 000 hours of exposure were calculated for all activity and for organized and non-organized activity occurring outside school hours. In addition, child-based injury rates were calculated for physical activity-related injuries both in and out of the school setting. RESULTS. Complete diary and injury data were available for 744 children. There were 504 injuries recorded over the study period, 396 (88.6%) of which were directly related to physical activity. Thirty-four percent of physical activity-related injuries required professional medical treatment. Analysis of injuries occurring outside of school revealed an overall injury rate of 5.7 injuries per 10 000 hours of exposure to physical activity and a medically treated injury rate of 1.7 per 10 000 hours. CONCLUSION. Injury rates per hours of exposure to physical activity were low in this cohort of primary school-aged children, with < 2 injuries requiring medical treatment occurring for every 10 000 hours of activity participation outside of school.

Socio-economic distribution of environmental risk factors for childhood injury

Objective: Childhood injury remains the single most important cause of mortality in children aged between 1-14 years in many countries. It has been proposed that lower socio-economic status (SES) and poorer housing contribute to potential hazards in the home environment. This study sought to establish whether the prevalence of observed hazards in and around the home was differentially distributed by SES, in order to identify opportunities for injury prevention. Methods: This study was a cross-sectional, random sample survey of primary school children from 32 schools in Brisbane. Interviews and house audits were conducted between July 2000 and April 2003 to collect information on SES (income, employment and education) and previously identified household hazards. Results: There was evidence of a relationship between prevalence of household environmental hazards and household SES; however, the magnitude and direction of this relationship appeared to be hazard-specific. Household income was related to play equipment characteristics, with higher SES groups being more likely to be exposed to risk. All three SES indicators were associated with differences in the home safety characteristics, with the lower SES groups more likely to be exposed to risk. Conclusion:The differential distribution of environmental risk factors by SES of household may help explain the SES differential in the burden of injury and provides opportunities for focusing efforts to address the problem.

Injury risk from popular childhood physical activities: results from an Australian primary school cohort

Background: Children engage in various physical activities that pose different injury risks. However, the lack of adequate data on exposure has meant that these risks have not been quantified or compared in young children aged 5-12 years. Objectives: To measure exposure to popular activities among Australian primary school children and to quantify the associated injury risks. Method: The Childhood Injury Prevention Study prospectively followed up a cohort of randomly selected Australian primary and preschool children aged 5-12 years. Time (min) engaged in various physical activities was measured using a parent-completed 7-day diary. All injuries over 12 months were reported to the study. All data on exposure and injuries were coded using the International classification of external causes of injury. Injury rates per 1000 h of exposure were calculated for the most popular activities. Results: Complete diaries and data on injuries were available for 744 children. Over 12 months, 314 injuries relating to physical activity outside of school were reported. The highest injury risks per exposure time occurred for tackle-style football (2.18/1000 h), wheeled activities (1.72/1000 h) and tennis (1.19/1000 h). Overall, boys were injured more often than girls; however, the differences were non-significant or reversed for some activities including soccer, trampolining and team ball sports. Conclusion: Although the overall injury rate was low in this prospective cohort, the safety of some popular childhood activities can be improved so that the benefits may be enjoyed with fewer negative consequences.

A review of risk factors for hamstring injury in soccer: a biomechanical approach

The aim has been to review the literature about the risk factors of hamstring injury in soccer from a biomechanical point of view. METHODOLOGY. Data bases of bibliography references were Medline, Scopus and SportDiscuss. RESULTS AND DISCUSSION. Many prospective studies have shown that the previous injury is the greatest risk factor of sustaining the injury. However the primary causes of the injury are unclear in soccer. A lack of hamstring flexibility has been one of the main injury risk factors with controversies on the results. Imbalance of isokinetic force is a risk factor but electrical coactivation of all muscles participating during knee flexion and extension are unknown in football. While the importance of lumbopelvic-hamstrings muscles synchronization during running seems to be crucial for understanding the risk of injury, no research has been developed in this topic in football. CONCLUSIONS. More research using new data recording procedures as Dynamic Scanners, Surface EMG, Inverse Dynamic Analysis are needed. The analysis of more specific movements as running, kicking or jumping is clearly required. Managers, coaches, physical trainers, physiotherapists, sport physicians and researchers should work together in order to improve the injury prevention and rehabilitation programs of football players. Key Words: sports biomechanics, soccer, hamstring injury, risk factors

The Burden of Injury in Iowa, December 2008

Injuries are a major public health concern that affects the lives of all Iowans, regardless of age, race, gender, or size of county. Unintentional injuries are the leading cause of death for Iowans between the ages of 1 and 34, while suicides and/or homicides also rank among the top 5 leading causes of death for Iowans between the ages of 1 and 54. Unintentional injury is the 5th leading cause of death for all Iowans, with over 1,500 injury deaths occurring on average each year (from 2002‐ 2006) in Iowa. Injuries also lead to more than 17,000 hospitalizations in Iowa each year and more than 250,000 emergency department (ED) visits. Further, untold numbers of Iowans do not seek medical care for many of the injuries they may incur. Because injuries are preventable, they lead to unnecessary medical costs, economic losses, reduced productivity, and immense physical and emotional strain.

Causes of injuries resulting in hospitalisation in Australia : assessing coder agreement on external causes

Objective: To assess extent of coder agreement for external causes of injury using ICD-10-AM for injury-related hospitalisations in Australian public hospitals. Methods: A random sample of 4850 discharges from 2002 to 2004 was obtained from a stratified random sample of 50 hospitals across four states in Australia. On-site medical record reviews were conducted and external cause codes were assigned blinded to the original coded data. Code agreement levels were grouped into the following agreement categories: block level, 3-character level, 4-character level, 5th-character level, and complete code level. Results: At a broad block level, code agreement was found in over 90% of cases for most mechanisms (eg, transport, fall). Percentage disagreement was 26.0% at the 3-character level; agreement for the complete external cause code was 67.6%. For activity codes, the percentage of disagreement at the 3-character level was 7.3% and agreement for the complete activity code was 68.0%. For place of occurrence codes, the percentage of disagreement at the 4-character level was 22.0%; agreement for the complete place code was 75.4%. Conclusions: With 68% agreement for complete codes and 74% agreement for 3-character codes, as well as variability in agreement levels across different code blocks, place and activity codes, researchers need to be aware of the reliability of their specific data of interest when they wish to undertake trend analyses or case selection for specific causes of interest.

The research pathway to developing a health promotion intervention for the high school curriculum

Introduction: Schools provide the opportunity to reach a large number of adolescents in a systematic way however there are increasing demands on curriculum providing challenges for health promotion activities. This paper will describe the research processes and strategies used to design an injury prevention program.----- Methods: A multi-stage process of data collection included: (1) Surveys on injury-risk behaviours to identify targets of change (examining behaviour and risk/ protective factors among more than 4000 adolescents); (2) Focus groups (n= 30 high-risk adolescents) to understand and determine risk situations; (3) Hospital emergency outpatients survey to understand injury types/ situations; (4) Workshop (n= 50 teachers/ administrators) to understand the target curriculum and experiences with injury-risk behaviours; (5) Additional focus groups (students and teachers) regarding draft material and processes.----- Results: Summaries of findings from each stage are presented particularly demonstrating the design process. The baseline data identified target risk and protective factors. The following qualitative study provided detail about content and context and with the hospital findings assisted in developing ways to ensure relevance and meaning (e.g. identifying high risk situations and providing insights into language, culture and development). School staff identified links to school processes with final data providing feedback on curriculum fit, feasibility and appropriateness of resources. The data were integrated into a program which demonstrated reduced injury.----- Conclusions: A comprehensive research process is required to develop an informed and effective intervention. The next stage of a cluster randomised control trial is a major task and justifies the intensive and comprehensive development.

A 5 year review of information requests to QISU

The Queensland Injury Surveillance Unit (QISU) has been collecting and analysing injury data in Queensland since 1988. QISU data is collected from participating emergency departments (EDs) in urban, rural and remote areas of Queensland. Using this data, QISU produces several injury bulletins per year on selected topics, providing a picture of Queensland injury, and setting this in the context of relevant local, national and international research and policy. These bulletins are used by numerous government and non-government groups to inform injury prevention and practice throughout the state. QISU bulletins are also used by local and state media to inform the general public of injury risk and prevention strategies. In addition to producing the bulletins, QISU regularly responds to requests for information from a variety of sources. These requests often require additional analysis of QISU data to tailor the response to the needs of the end user. This edition of the bulletin reviews 5 years of information requests to QISU.