The literature illustrates that agriculture continues to be the leader in the nation for accidents. The attitudes and behaviors of farmers and farm workers towards farm equipment and its safe use are major concerns. Farm safety education and research are strongly needed to prevent or reduce farm injuries. The primary purpose of this study was to investigate health and safety problems among limited resource farmers in selected counties of North Carolina. The survey method with a mail questionnaire was used to collect the data. The sample for this study consisted of 297 limited resource farmers in North Carolina. Selected conclusions drawn from the findings were: 1. The risks associated with limited resource farmers and their workers, such as use of tractors, machinery, and farm chemicals are very high. 2. Farmers between the ages of 36-50 have higher expected injuries compared to other age groups. 3. Training associated with the use of farm tractors, machinery, and personal protective equipment (PPE) is very essential for limited resource farmers. 4. The research suggested that hazards exposed by using tractors without rollover protection structure (ROPS) could result in higher injuries among limited resource farmers and workers. Selected recommendations: 1. Federal and state governments should give incentives to farmers to encourage them to retrofit ROPS on their old tractors. 2. Mandatory training should be required for small scale or limited resource farmers on utilization of farm equipment operation and personal protective equipment (PPE) utilization. 3. New design of preventive guards and shields that can be easily removed or opened for maintenance and remounted or closed afterward are needed. / Ph. D.

Produce is responsible for 46% of all foodborne illnesses in the USA. Salmonella enterica causes 19,000 hospitalizations each year, and has been associated with produce. Presently, chlorine based sanitizers are most often used, however organic matter reduces its antimicrobial activity. Bacteriophage treatments are an all-natural, alternative method for pathogen inactivation. The objective of this study was to determine the efficacy of a five-strain bacteriophage treatment against a S. enterica cocktail in simulated wash waters at different temperatures. Bacteriophage and S. enterica were enumerated in simulated wash water solutions. One set of experiments studied bacteriophage and S. enterica growth in TSB+vegetable solutions. Bacteriophage behavior was not statistically different (p < 0.05) in spinach, romaine, or iceberg lettuce across different concentrations of organic matter. S. enterica reduction was approximately 2 log over 135 minutes for vegetable solutions and for the TSB control. S. enterica reduction was only 0.5 log in water solutions. The next set of experiments studied bacteriophage and S. enterica growth in vegetable solutions. Spinach wash water and tryptone soy broth solutions (TSB) at 20 °C and 37 °C. S. enterica was not reduced in spinach solution studies at 20 °C and 37 °C or at broth solutions at 20 °C. However, S. enterica was effectively reduced 4 log in broth solutions at 37 °C up to 7.5 hours, but grew to high levels after 24 hours. These results indicate that bacteriophage could not effectively control bacteria levels in produce wash water, and may need to be optimized.

Because of increased commuter traffic volume on rural roadways collisions between motor vehicles and farm equipment have increased in frequency and severity over the last several years. This study investigated the effects of a multimedia narrative simulation program that taught hazard recognition and promoted defensive driving on rural roadways shared by farm equipment. A companion animated driving game allowed users to practice reaction/stopping time distances with a simulated automobile on a simulated highway when objects appeared suddenly in the path of the automobile. The program and game were delivered by an objective-oriented client/server computer program that also recorded and stored student pre-test, performance, and posttest data. Prior to the main study a user test and pilot study were conducted. Fifteen instructional systems design graduate students completed the user test to evaluate the study procedures and debug the program. Then, a pilot study sample of 17 rural high school students completed the narrative simulation exercise, the reaction/stopping time game, and the study measures that included a demographic survey, pre- and post measures of predicted reaction/stopping time, recognition of collisions hazard cues, numerical performance scores for the simulation exercise, and tracking logs of each student's performance during the animated raction/stopping time game. The main study sample included 123 students age 16 years and older who attended four rural and suburban county high schools. The schools were randome assigned to the four treatment conditions, one control and three treatment groups. The treatment group students completed either (a) the multimedia narrative simulation only, (b) the animated reaction time/stopping time game only, or (c) both the multimedia simulation and the reaction time/stopping time game. As hypothezised, students in the groups that completed the hazard recognition and defensive driving skills performed significantly better on posttests of those skills than students in groups that did not complete the simulation. Compared to students that did not complete the reaction/stopping time game, significantly more students that did complete the game became aware that they could not stop the simulated automobile before hitting an object in its pathway. Yet there was nodifference across the four groups in students' estimates of reaction/stopping time distances. Limitations of the study are noted and discussed. Recommendations for future studies are proposed.

Doctor of Philosophy (PhD) / Injuries to people living and working on farms in New South Wales continue to be a significant burden on the health system, Workers’ Compensation system, agricultural industries and farming families. Strategies to reduce the number and severity of injuries suffered by farmers and people working on farms rely on accurate information. Unfortunately there is no one dataset available to describe the circumstances surrounding farm injuries and the size of this burden in Australia. Hence, a number of different data sources are required to provide a picture of farm injuries. To date, there has been very little critical examination of what value each of these datasets provides to describing farm injuries. This Thesis aimed to: • Undertake surveillance of injuries occurring to people on farms or during agricultural production in NSW using data from an Emergency Department, NSW Hospital Separations information, NSW Workers’ Compensation Claims, and ABS Deaths data. • Critically examine the utility of Emergency Department, Hospital, Workers’ Compensation, and ABS Deaths Data for the surveillance of farm injuries in NSW. • Critically examine data classification systems used in Emergency Department, Hospital, Workers’ Compensation, and ABS Deaths data collections to describe the breadth of farm injuries in NSW. • Define the priority areas for farm injury prevention initiatives in NSW based on the information obtained from the examination of the data from Emergency Department, Hospital, Workers’ Compensation, and ABS Deaths. • Evaluate the effectiveness of the NSW Rollover Protective Structure (ROPS) rebate scheme and examine the utility of the data currently available in NSW to measure the performance of the program. Four datasets, Tamworth Emergency Department, Hospital Separations, Workers’ Compensation and the Australian Bureau of Statistics (ABS) Deaths data were used to provide information on the surveillance of farm injuries, describe the breadth of classifications used to describe farm injuries, and define priorities for the prevention of farm injuries. There were 384 farm-related injuries which presented to the Emergency Department at the Tamworth Base Hospital between 1 September 1997 and 31 August 1998. Emergency Department data collected in this study used the Farm Injury Optimal Dataset (FIOD) for classification, which allowed for a comprehensive picture of the circumstances surrounding the injury event. The three most common external causes of injury were related to horses, motorcycles, and animals. Commonly people were working at the time of injury. Children represented 21% of the people injured. The average number of injuries per 100 farms per annum was 34.7. An examination of hospital discharge data for NSW was undertaken for the period 1 July 1992 to 30 June 2000 where the location of the injury was a farm. Classification of cases in this dataset conformed to the International Classification of Disease (ICD) versions 9 and 10. There were 14,490 people who were injured on a farm during the study period. The three most common external causes of injury were motorcycles, animals being ridden and agricultural machinery. Children represented 17% of all farm injury cases. The rate per 1,000 farms ranged from 19 to 42 per annum. An examination of Workers’ Compensation claims for agricultural industries in NSW between 1 July 1992 and 30 June 2001 was undertaken. The ‘Type of Occurrence’ classification system was used to code the claims. There were 24,332 claims of which the majority were males (82%). The incidence of injury / disease in agriculture per annum varied from 37 per 1,000 workers to 73 per 1,000 workers. The rate per 1,000 agricultural establishments varied from 54 to 76. The average cost of a claim was $10,880 and the average time lost per claims was 9.2 weeks. There were 81 deaths and 3,158 permanent disabilities. The three most common agents were sheep / goats (5%), ferrous and non-ferrous metals (5%), crates / cartons / boxes / etc (5%). Using ABS deaths data to examine the deaths of people working and living on farms was limited to males whose occupation was recorded as ‘farmer and farm manager’ and ‘agricultural labourer and related worker’. There were 952 deaths over the period 1 January 1991 and 31 December 2000. The information provided a consistent series of cases over time. Areas where prevention should be directed included motor vehicle accidents; falls; agricultural machinery; other machinery; firearms; poisoning; and drowning. Using any one of the datasets alone to examine people injured on farms not only underestimates the number of people injured, but also misses particular types of agents involved in farm injuries. Each of the datasets used in this Thesis provides a different perspective of farm injury in NSW. By examining the information together, there are a number of areas which are consistently represented in each dataset such as falls and agricultural machinery. While no one dataset provided all the information that would be useful for the prevention of injuries, the available information does provide direction for the development of prevention strategies. The overall weakness of the information provided is that it misses a number of risk factors that contribute to farm injuries such as fatigue and training. The lack of appropriate denominator information also makes it difficult to directly compare the datasets and estimate the size of the problem. There are a number of additional coding categories that could be included in each dataset that would provide a better understanding of the different groups at risk of sustaining an injury on a farm or during agricultural work. These coding categories include activity at time of injury, admission to hospital, and occupation. An example of the use of data to determine the effectiveness of a farm injury prevention program is the ‘NSW Rollover Protective Structure (ROPS) Rebate Scheme’ evaluation. Tractor rollover deaths have been identified as an issue for prevention by Farmsafe Australia; however, such deaths were not identified in any of the datasets used in this Thesis due to coding limitations in the ABS data. In this Thesis information about the evaluation of the ‘NSW ROPS Rebate Scheme’ is presented. The scheme was successful in fitting 10,449 ROPS to tractors and the following lessons were learnt: when providing a rebate, the administration (i.e. sending the cheque) needs to be done well; advertising is important and should be co-ordinated, increase the awareness of the risk(s) the intervention is aiming to prevent and effectiveness of subsequent solution (s); the program should ensure there is an increased awareness of the outcome the intervention is aiming to prevent; if regulation is part of the program, enforcement needs to undertaken; and should address any barriers to uptake. The information provided in this Thesis highlights the substantial burden that farm injury places on the agricultural and rural sector of NSW. While there is no one data source that can describe the circumstances and the burden of farm injuries, the currently available datasets do provide an insight into the circumstances of farm injuries and the burden these injuries place on health, Workers’ Compensation, agricultural industries and farming families.

Doctor of Philosophy (PhD) / Injuries to people living and working on farms in New South Wales continue to be a significant burden on the health system, Workers’ Compensation system, agricultural industries and farming families. Strategies to reduce the number and severity of injuries suffered by farmers and people working on farms rely on accurate information. Unfortunately there is no one dataset available to describe the circumstances surrounding farm injuries and the size of this burden in Australia. Hence, a number of different data sources are required to provide a picture of farm injuries. To date, there has been very little critical examination of what value each of these datasets provides to describing farm injuries. This Thesis aimed to: • Undertake surveillance of injuries occurring to people on farms or during agricultural production in NSW using data from an Emergency Department, NSW Hospital Separations information, NSW Workers’ Compensation Claims, and ABS Deaths data. • Critically examine the utility of Emergency Department, Hospital, Workers’ Compensation, and ABS Deaths Data for the surveillance of farm injuries in NSW. • Critically examine data classification systems used in Emergency Department, Hospital, Workers’ Compensation, and ABS Deaths data collections to describe the breadth of farm injuries in NSW. • Define the priority areas for farm injury prevention initiatives in NSW based on the information obtained from the examination of the data from Emergency Department, Hospital, Workers’ Compensation, and ABS Deaths. • Evaluate the effectiveness of the NSW Rollover Protective Structure (ROPS) rebate scheme and examine the utility of the data currently available in NSW to measure the performance of the program. Four datasets, Tamworth Emergency Department, Hospital Separations, Workers’ Compensation and the Australian Bureau of Statistics (ABS) Deaths data were used to provide information on the surveillance of farm injuries, describe the breadth of classifications used to describe farm injuries, and define priorities for the prevention of farm injuries. There were 384 farm-related injuries which presented to the Emergency Department at the Tamworth Base Hospital between 1 September 1997 and 31 August 1998. Emergency Department data collected in this study used the Farm Injury Optimal Dataset (FIOD) for classification, which allowed for a comprehensive picture of the circumstances surrounding the injury event. The three most common external causes of injury were related to horses, motorcycles, and animals. Commonly people were working at the time of injury. Children represented 21% of the people injured. The average number of injuries per 100 farms per annum was 34.7. An examination of hospital discharge data for NSW was undertaken for the period 1 July 1992 to 30 June 2000 where the location of the injury was a farm. Classification of cases in this dataset conformed to the International Classification of Disease (ICD) versions 9 and 10. There were 14,490 people who were injured on a farm during the study period. The three most common external causes of injury were motorcycles, animals being ridden and agricultural machinery. Children represented 17% of all farm injury cases. The rate per 1,000 farms ranged from 19 to 42 per annum. An examination of Workers’ Compensation claims for agricultural industries in NSW between 1 July 1992 and 30 June 2001 was undertaken. The ‘Type of Occurrence’ classification system was used to code the claims. There were 24,332 claims of which the majority were males (82%). The incidence of injury / disease in agriculture per annum varied from 37 per 1,000 workers to 73 per 1,000 workers. The rate per 1,000 agricultural establishments varied from 54 to 76. The average cost of a claim was $10,880 and the average time lost per claims was 9.2 weeks. There were 81 deaths and 3,158 permanent disabilities. The three most common agents were sheep / goats (5%), ferrous and non-ferrous metals (5%), crates / cartons / boxes / etc (5%). Using ABS deaths data to examine the deaths of people working and living on farms was limited to males whose occupation was recorded as ‘farmer and farm manager’ and ‘agricultural labourer and related worker’. There were 952 deaths over the period 1 January 1991 and 31 December 2000. The information provided a consistent series of cases over time. Areas where prevention should be directed included motor vehicle accidents; falls; agricultural machinery; other machinery; firearms; poisoning; and drowning. Using any one of the datasets alone to examine people injured on farms not only underestimates the number of people injured, but also misses particular types of agents involved in farm injuries. Each of the datasets used in this Thesis provides a different perspective of farm injury in NSW. By examining the information together, there are a number of areas which are consistently represented in each dataset such as falls and agricultural machinery. While no one dataset provided all the information that would be useful for the prevention of injuries, the available information does provide direction for the development of prevention strategies. The overall weakness of the information provided is that it misses a number of risk factors that contribute to farm injuries such as fatigue and training. The lack of appropriate denominator information also makes it difficult to directly compare the datasets and estimate the size of the problem. There are a number of additional coding categories that could be included in each dataset that would provide a better understanding of the different groups at risk of sustaining an injury on a farm or during agricultural work. These coding categories include activity at time of injury, admission to hospital, and occupation. An example of the use of data to determine the effectiveness of a farm injury prevention program is the ‘NSW Rollover Protective Structure (ROPS) Rebate Scheme’ evaluation. Tractor rollover deaths have been identified as an issue for prevention by Farmsafe Australia; however, such deaths were not identified in any of the datasets used in this Thesis due to coding limitations in the ABS data. In this Thesis information about the evaluation of the ‘NSW ROPS Rebate Scheme’ is presented. The scheme was successful in fitting 10,449 ROPS to tractors and the following lessons were learnt: when providing a rebate, the administration (i.e. sending the cheque) needs to be done well; advertising is important and should be co-ordinated, increase the awareness of the risk(s) the intervention is aiming to prevent and effectiveness of subsequent solution (s); the program should ensure there is an increased awareness of the outcome the intervention is aiming to prevent; if regulation is part of the program, enforcement needs to undertaken; and should address any barriers to uptake. The information provided in this Thesis highlights the substantial burden that farm injury places on the agricultural and rural sector of NSW. While there is no one data source that can describe the circumstances and the burden of farm injuries, the currently available datasets do provide an insight into the circumstances of farm injuries and the burden these injuries place on health, Workers’ Compensation, agricultural industries and farming families.

There are an estimated 70,000 quadbikes (also called All-Terrain Vehicles, four-wheelers or ATV) in use within the farming community in New Zealand, and these have been linked to approximately seven deaths per year since 1997. ACC paid compensation in excess of $3.6 million on new and on-going quadbike-related claims in 2003-4, and the Department of Labour (OSH) have identified quadbike use as their single greatest area of concern in farm safety. The main aims of this series of studies were to: establish the scale and cost of LCE involving quadbikes as well as their context of use on New Zealand farms, develop an investigative methodology suitable for off-road application, identify risk factors for LCE and their interactions, and generate potential interventions. The research comprised a literature review and three studies. The first was an epidemiological analysis of ACC claims data for serious quadbike-related injuries in the year to July 2001, identified individually through the free narrative text in the ACC database. The second study explored the context of farm quadbike use through data gathering with multiple industry sources including users at the 55 farms visited for the third study. This final phase consisted of investigations at the scenes of 156 LCE. Data on the event sequences and contributory factors were captured and analysed using a modified event sequence chart developed from established methods in the ergonomics, forensics psychology and aviation literature. Environmental context-dependent cues and scale models of quadbikes and implements were used to assist the subjects recall event details. An interactive quadbike LCE model incorporating information processing features was developed as a further analysis tool. Potential interventions were identified acting at various system levels. These included changes in: farm management, work organisation, regulation, vehicle conceptualisation, detailed quadbike design and modification, rider awareness and training.

In the U.S., work-related death is an all too familiar occurrence on farms. Tractor overturns continue to be the most frequent cause of these fatalities. Efforts to alter farming’s ranking as one of the most deadly occupations in the country must provide proven strategies for the elimination of these preventable deaths. In the past, efforts to decrease the rate of overturn fatalities and injuries have largely focused onincreasing the proportion of tractors with a rollover protective structure (ROPS). These devices, in combination with seatbelts, are 99% effective in protecting the tractor operator from death or injury. Unfortunately, only 59% of U.S. tractors are currently equipped with ROPS. Due to the relative lack of political willpower to legislate ROPS installation and the less than encouraging response to education and awareness programs to date, it appeared necessary to explore alternative intervention strategies. The over-arching purpose of this thesis project has been to assess the utility of social marketing as a framework for developing effective health and safety interventions in the farm community. However, our specific objectives included; a more thorough understanding of the perceived barriers and motivators that influence farmer’s safety decisions, the design and evaluation of social marketing incentives developed to encourage safe behaviors and the evaluation of a social marketing campaign designed to positively impact farmer’s intentions and readiness to retrofit unsafe tractors. The research was by and large conducted in New York State and supported by grants from the National Institutes of Occupational Safety and Health (NIOSH). Previous research conducted in the New York farm community had indicated that small crop and livestock farmers would be an ideal intervention target for a social marketing tractor overturn intervention as their farms accounted for close to 85% of New York farms which lack or have only one ROPS protected tractor. A qualitative assessment of perceived barriers and motivators regarding retrofitting behaviors was performed with representatives of the small crop and livestock community. Grounded theory analysis of these in-depth interviews revealed several key categories which include: 1) risk becomes “normal”, 2) risk becomes part of a “farming identity”, and 3) risk becomes “cost-effective”. This information was used to design potential intervention incentives, such as toll-free assistance finding and purchasing ROPS, financial rebates, and campaign messages designed to address farmer’s stated concerns. Subsequent research included testing and revising messages and evaluating the effect of the different campaign incentives in a prospective quasirandomized controlled trial conducted in different regions of New York and Pennsylvania. The results indicate that social marketing offers a promising framework for the development of injury or fatality prevention programs in farm communities. Farmers in the social marketing region demonstrated the most significant changes in both behavioral intention and readiness to retrofit compared to farmers from other regions. Data also indicated that social norms strongly influence farmer’s decisions to work safely, as demonstrated by the strong correlations between behavioral intention measures and measures of social norms. As well as providing an assessment of the utility of social marketing as an intervention framework, the thesis provides a cogent example of how behavioral theories can be used in the design and evaluation of intervention programs. Both stages of change theory and the theory of planned behavior proved to be valuable for measuring dispositional and behavioral changes and for finetuning future interventions.