USF Libraries
USF Digital Collections

An analysis of workplace amputation injuries in Florida

MISSING IMAGE

Material Information

Title:
An analysis of workplace amputation injuries in Florida
Physical Description:
Book
Language:
English
Creator:
Hanna, Eve N
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
Industrial safety -- Management   ( lcsh )
Amputation   ( lcsh )
Amputation, Traumatic -- prevention & control   ( mesh )
Accidents, Occupational -- prevention & control   ( mesh )
Occupational Health   ( mesh )
safety training
prevention
occupational
industrial hazards
machine safety
Dissertations, Academic -- Public Health -- Masters -- USF   ( lcsh )
Genre:
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: Nationally, 10,852 workplace amputation injuries in the US were reported in 1997. This number has been gradually decreasing over time to 9,658 amputations in 2000. In Florida, the amputation injuries have been increasing from 272 in 1997 to 417 in 2000. The reason for this increase is unclear and deserves further study. The purpose of a portion of this study was to examine the Federal and Florida BLS data to determine the characteristics of the typical worker with an amputation injury. A worker with an amputation injury was most likely to be a white male between the ages of 35 and 44 with 1 to 5 years of employment. The amputated part was usually a finger which was injured in a machinery source. The purpose of the self-designed survey was to determine if there were any differences in the amputation prevention strategies between companies with and without amputation injuries. The survey also attempted to uncover any changes made to the safety training program after an amputation injury occurred. Using the Florida workers compensation data, surveys were sent to all companies with amputation injuries for the years 1999, 2000, and 2001. An equal number of surveys were sent to companies without amputation injuries during the same years of interest. A total of 840 surveys were mailed out. 146 were returned as undeliverable. 35 surveys were returned in the amputation group and 35 surveys were returned in the control group. All SIC codes were represented except mining. In the amputation group during the years investigated, 15 out of 35 (42.9%) companies claimed amputation injuries and 20 (57%) companies did not. The true error rate for the workers compensation data was a 13% misclassification of amputation injuries into that category. Comparison of the prevention strategies between the amputation and control groups revealed no significant differences. Of the companies claiming amputation injuries, 8 out of 15 (53%) made no changes to their safety training program after the injury occurred. Proactive techniques and global changes in safety culture mindset will be necessary before major reductions in amputation injuries can occur.
Thesis:
Thesis (M.S.P.H.)--University of South Florida, 2003.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Eve N. Hanna.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 76 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001430576
oclc - 52281643
notis - AJL4037
usfldc doi - E14-SFE0000094
usfldc handle - e14.94
System ID:
SFS0024790:00001


This item is only available as the following downloads:


Full Text

PAGE 1

AN ANALYSIS OF WORKPLACE AMPUTATION INJURIES IN FLORIDA by EVE N. HANNA, M.D. A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Public Health Department of Environmental and Occupational Health College of Public Health University of South Florida Major Professor: Stuart M. Brooks, M.D. Matthew A. Vuskovich, M.D. Phillip P. Roets, Sc.D. Date of Approval: April 8, 2003 Keywords: safety training, prevention, occupational, industrial hazard s, machine safety Copyright 2003, Eve N. Hanna, M.D.

PAGE 2

Acknowledgments I would like to acknowledge everyone at the USF Safety Florida Consultation Program for their technical support and expertise as well as their encouragement. Without this group, I would not have been able to undertake this project. I would also like to thank Tom Berish for his time and assistance with the behavior based safety portion of my thesis. Finally, I would like to say a word of appreciation for the guidance and vision fr om Dr. Brooks.

PAGE 3

i Table of Contents List of Tables iii List of Figures iv Abstract v Introduction 1 Study Purpose 2 Research Objectives 3 Thesis Hypothesis 3 Literature Review 4 Material and Methods 13 Analysis of Bureau of Labor Statistics Data 13 Source and Event Analysis of the Amputation Group 14 Amputation Prevention Strategy Survey 14 Survey Content 15 Question 1 15 Question 2 15 Question 3 16 Question 4 16 Question 5 16 Results 18 Summary Data for US and Florida Bureau of Labor Statistics 18 US Bureau of Labor Statistics 18 Florida Bureau of Labor Statistics 20 Comparison of Demographic and Case Date between US and Florida 21 Source and Event for the Amputation Group 22 Survey Results 23 Question 1 24 Question 2 25 Question 3 26 Question 4 29 Question 5 30

PAGE 4

ii Discussion 32 External Validity 33 Study Limitations 33 Conclusion 37 References 38 Bibliography 41 Appendices Appendix A: SAS Program 42

PAGE 5

iii List of Tables Table 1 Amputation Demographic and Case Data for 2000 20 Table 2 Source of Injury 23 Table 3 Event: Contact with object or Equipment 23 Table 3 SIC for Amputation and Control Group 25 Table 4 Frequency of Amputations for Study Years 26 Table 5 Prevention Strategy Responses by SIC 28 Table 6 Comparison of Amputation Prevention Strategies 29 Table 7 Post Amputation Safety Training Program Changes 30

PAGE 6

iv List of Figures Figure 1 Survey 17 Figure 2 Comparison of Number of Amputations: US vs. Florida 19 Figure 3 Comparison of Amputation Rates: US vs. Florida 19

PAGE 7

v An Analysis of Workplace A mputation Injuries in Florida Eve N. Hanna, M.D. ABSTRACT Nationally, 10,852 workplace amputation injuries in the US were reported in1997. This number has been gradually decreasing over time to 9,658 amputations in 2000. In Florida, the amputation inj uries have been increasing from 272 in 1997 to 417 in 2000. The reason for this increase is unclear and deserves further study. The purpose of a portion of this study was to examine the Federal and Florida BLS data to determine the characteristics of the typical worker with an amputation injury. A worker with an amputation injury was most likely to be a white male between the ages of 35 and 44 with 1 to 5 years of employment. The amputated part was usually a finger which was injured in a machinery source The purpose of the self designed survey was to determine if there were any differences in the amputation prevention strategies between companies with and without amputation injuries. The survey also attempted to uncover any changes made to the safety t raining program after an amputation injury occurred. Using the Florida workers compensation data, surveys were sent to all companies with amputation injuries for the years 1999, 2000, and 2001. An equal number of surveys were sent to companies without a mputation injuries during the same years of interest. A total of 840 surveys were mailed out. 146 were returned as undeliverable. 35 surveys were returned in the amputation group and 35 surveys were returned in the control group. All SIC codes were

PAGE 8

vi rep resented except mining. In the amputation group during the years investigated, 15 out of 35 (42.9%) companies claimed amputation injuries and 20 (57%) companies did not. The true error rate for the workers compensation data was a 13% misclassification o f amputation injuries into that category. Comparison of the prevention strategies between the amputation and control groups revealed no significant differences. Of the companies claiming amputation injuries, 8 out of 15 (53%) made no changes to their saf ety training program after the injury occurred. Proactive techniques and global changes in safety culture mindset will be necessary before major reductions in amputation injuries can occur.

PAGE 9

1 Introduction Amputations in the workplace are preventable yet persistent occupational injuries. The majority of workplace amputations involve one or more fingers. Although this loss may not seem serious, finger dexterity is a crucial factor for ma ny jobs. The Federal Bureau of Labor Statistics reported 10,852 workplace amputations in the US in 1997. 1 This number has been gradually decreasing over time to 9,658 amputations in 2000. In Florida, the amputation numbers have been increasing from 272 in 1997 to 417 in 2000. 2 The reason for this increase is unclear and deserves further investigation. Despite the continuing awareness of this injury, little is known about amputation risk factors, or the underlying causes. Before effective prevention s trategies can be implemented, these basic factors should be identified. Injury prevention strategies in the workplace always include hazard analysis, training, engineering controls, administrative controls, and as a last resort, PPE. It is well known tha t 90% of accidents are caused by unsafe behaviors and only 10% are caused by unsafe conditions. It would make sense to emphasize a behavioral based type of prevention program. A new approach to safety based on behavior techniques has been developed and i mplemented around the country beginning at Proctor and Gamble in the mid 1970s. This proactive program identifies critical safety related behaviors, develops preventive actions, measures compliance through observation, and provides feedback for improvemen t on an ongoing basis. 3 The prevalence of this technique in Florida is not known and should be uncovered.

PAGE 10

2 In 1990, NIOSH felt that industry would be interested in the frequency and severity of machine related injuries resulting in amputations and fatalit ies. The article, Machine Safety Research at NIOSH and the Future Directions led to research priorities toward alleviating problems of manufacturing machine related injuries and fatalities. 4 These conclusions were the basis for the OSHA National Emphasis Program on Amputations. 5 Initially, a National Emphasis Program on mechanical power presses was established and implemented in 1997. This program was expanded in 1999 and was updated in 2000 and renamed the National Emphasis Program on Amputations. M ajor changes included the addition of all types of power presses, press brakes, saws, shears, and slicers as well as the original mechanical power presses. The purpose of the program is to identify and reduce workplace machine hazards which are likely to cause amputations. It includes three activities which are outreach, targeting and selection, and inspection. OSHA believes that the failure to guard machinery is a primary cause of amputations. This program will target industries with recorded amputatio n injuries and violations of 29 CFR 1910.212 (all machines), 1910.213 (woodworking machinery), and 1910.217 (power presses) for inspections. Although this approach will pinpoint high risk industries with amputation injuries, it still does not address the specific behaviors that lead to unsafe working conditions and the potential for future amputations. Study Purpose The purpose of this study is to delineate the prevention strategies used by companies to prevent amputation injuries.

PAGE 11

3 Research Objectiv es 1. To delineate the demographic and nature of injury data for workers with amputation injuries in the US and Florida and to compare the two groups. 2. To uncover the prevention strategies used by companies to pre vent amputations. 3. To determine if there are any differences in prevention strategies between companies that had amputation injuries and companies who did not. 4. To discover if any changes were made in the safety training program after a company had an amputation injury. 5. To define the percentage of companies using behavior based safety training as their prevention strategy or as their post amputation safety program change. Thesis Hypothesis There will be no difference in the prevention strategies used by companies with amputation injuries and companies without amputation injuries.

PAGE 12

4 Literature Review The National Electronic Injury Surveillance System (NEISS), funded by NIOSH, reported on the nonfatal occupation injuries and illnesses tr eated in hospital emergency departments in the US. 6 For the year 1998, 27% of the injuries were lacerations, puncture wounds, avulsions, or amputations. Hands and fingers were the most commonly injured body parts. Amputations were not discussed separat ely. Younger workers had the highest rates of work related injuries and illnesses. The NIOSH publication, Injuries and Amputations Resulting from Work with Mechanical Power Presses, emphasized the amputation hazards of these machines. 7 Their research in dicated that 10% of amputations have occurred among power press operators. Statistics gathered by OSHA indicated that about 50% of the injuries from mechanical power presses resulted in amputations. NIOSH found that young male operators appeared to be at the greatest risk for injury. They developed recommendations for the safe use of mechanical power presses especially those operated by foot or dual palm button controls. On the state level, Minnesota has investigated work related amputations in the stat e and several articles are available on this topic. Minnesota is the only state with a surveillance program for work related amputation injuries. This state has the Minnesota Sentinel Event Notification System for Occupational Risks (SENSOR) program fund ed by NIOSH which has collected data on work related amputation injuries since 1992. One of their studies looked into the medical, personal, and occupational outcomes for 601

PAGE 13

5 amputation injuries. 8 66% reported that their amputation injury had adversely a ffected their participation in sports, hobbies, yard work, house work and other activities. The article on the Nature, Incidence, and Cause of Work related Amputations in Minnesota revealed that most of the amputations happened to young males with little advanced education. 9 Over half of the finger amputations were to the dominant hand. About three fourths of these injuries were associated with machinery. In many cases, the required safety devices for the machines were either not functioning properly or removed to do a specific job task. Michigan developed a five year strategic plan for reducing amputations in their state. 10 They plan to reduce amputation injuries by 15% in 2003. They identified high hazard activities and high risk industries. Then t hey provided consultations for industries in the form of hazard surveys and evaluation of Safety and Health programs. They coordinated state wide seminars and developed guidelines. Part of their strategy involved enforcement with focused and comprehensiv e inspections. 425 inspections yielded 4702 violations. It will be interesting to see how this intensive and focused approach worked to decrease amputation injuries at the end of this year. No articles relating to workplace amputations were found for th e state of Florida. OSHA published the bulletin, Safeguarding Equipment and Protecting Workers from Amputations as a prevention strategy in this high risk group. 11 The publication discussed how to recognize amputation hazards and how to safeguard against these hazards. It identified hazards and controls for specific types of machinery such as mechanical power presses, power press brakes, conveyors, printing presses, roll forming and roll bending machines, shearing machines, food slicers, meat grinders, me at cutting band saws, drill presses, milling machines, grinding machines, and slitters. OSHA also

PAGE 14

6 developed a Fact Sheet on Amputations. 12 It pointed out the sources of amputation injuries in the workplace. It discussed the types of machine components a nd kinds of mechanical motions that are hazardous. It recommended safe work practices, employee training and administrative controls to help prevent and control amputation hazards. It also suggested that machine safeguarding with guards or devices was th e best way to control amputations caused by stationary machinery. These publications delineated the most effective prevention strategies for decreasing amputation injuries in high risk situations relating to machinery. Education for Job Safety and Health w as a provocative article which tied together the importance of engineering controls with worker behavior in preventing injury and illness using job safety and health training. 13 Behavioral approaches to the prevention and control of work hazards need to b e thought of as complementing engineering and personal protective device (PPD) controls. Education and training programs must teach the worker to avoid behaviors that might reduce the effectiveness of existing engineering controls and PPDs. The worker mu st be taught to use existing engineering controls and PPDs to his advantage to enhance their effectiveness. The worker must be taught to engage in work practices and personal health practices which reduce contact with hazards and limit opportunities for i llness and injury. The article gave examples of the success of this behavior engineering approach in preventing injury and illness in the workplace. The article, Health and Safety Training in a Sample of Open shop Construction Companies, tried to compare safety outcomes such as accident rates between companies with safety and health training and those without it but were unable to due so because of the small sample size. 14 The authors felt that observation of on the job performance was

PAGE 15

7 probably the most objective method for evaluating training effectiveness. Most of the companies surveyed had safety and health training programs but spent little time assessing the quality or effectiveness of training in terms of actual knowledge gain or skills enhancement The study recommended that methods be developed to objectively quantify if the safety and health training increased productivity, work quality, morale, and job satisfaction. The descriptive study, Mechanical Equipment Injuries in Small Manufacturing Bus inesses: Knowledge, Behavioral, and Management Issues did not specifically address amputation injuries but it made several important points about the factors that impact on the high rate of injuries due to mechanical equipment. 15 Eighteen percent of the employees from these small businesses commented during the interview process that their managers were not aware of the potential dangers of the workplace machinery. However, the managers all stated that they were aware of the hazards resulting in a perception difference. Half of the mangers had never received any safety training and 22% of the employees had not received any training either. Language barriers were mentioned as an issue that may have contributed to safety problems. Employers that we re aware of workplace hazards felt they could be managed by having well trained competent staff who followed good work practices. The hazards were not seen as a problem that needed to be fixed. In these small businesses, only half had specific safety reg ulations and the enforcement of these regulations did not always occur when they existed. None of the companies interviewed analyzed the injuries to determine priorities for injury prevention. Although 83% of the managers knew that safety management was their responsibility, only 66% of the workers felt that the managers took responsibility for

PAGE 16

8 this area. The article suggested that enforcement, education, and training could improve the injury rate from mechanical equipment in these small businesses. Ant hony Bimonte from DuPont Company in his article, Win Empowered Workers Commitment to Exceeding Basic Safety Compliance, stated that employee empowerment and taking personal responsibility for safety can reduce incident rates in the workplace. 16 This conc ept compliments the employee involvement principle of an effective Safety and Health program and can be thought of as an injury and illness prevention strategy. There are several research papers on behavior based safety and its effectiveness as well as ca se reports from companies with successful behavior based safety training programs. The paper, Beyond Training: Organizational Performance Management Techniques, notes that in the safety arena, the science of behavior has been increasingly used to address the human aspect of injury prevention. 17 Despite training, workers have returned to earlier bad habits, failed to use, circumvented, or misused proper precautions. The applied branch of behavioral science known as performance management looks at these i ssues. Performance management sees behavior as a function of prior and current events that influence performance. Contextual factors and antecedents or happenings that precede the performances of concern are important but consequences or events that foll ow exert the heaviest influence on behavior. Consequences can be reinforcing or aversive. To use this information practically, performance managers undertake an ABC analysis (Antecedents Behavior and Consequences) of the unwanted and preferred behaviors on the job. Specifically, they look at which antecedents and consequences precede and follow each class of behavior. The consequences are classified as positive or

PAGE 17

9 negative, immediate or delayed, highly or minimally probable, and how meaningful they are to the worker at that time. To successfully change behavior, reinforcers should be primarily positive. Consequences should be immediate, and need to occur regularly. Consequences must be meaningful to the individual and specific feedback is an effectiv e example. This study lists 44 published papers of performance management safety interventions. The authors feel that performance management is a valuable tool within the total safety and health program. The use and effectiveness of behavior based safet y was evaluated in the article, Researching Behavior Based Safety: A Multi Method Assessment and Evaluation. 18 Advantages of a behavior based safety process are that it can be administered by an individual with little professional training. It can aff ect workers in the setting where the problem occurs. Management can be taught the behavior based technique most likely to work under their specific circumstances. It is cost effective because the tactics are straightforward and relatively easy to adminis ter. Progress can be assessed by personnel monitoring the target behavior without any special tools. A nationwide survey of safety professionals revealed that 80% believed that behavior based safety was a viable approach for reducing at risk work behavio rs. Most respondents indicated that they had an accurate but narrow awareness of what behavior based safety entailed. The majority saw behavior based safety as an observation and feedback tool rather than a general strategy to improve the human aspect of safety. Few survey participants used the percent safe behaviors to monitor the success of their programs. Other outcome measures such as OSHA recordables, lost time accidents or total recordable injury rates were used instead. These results point out t hat behavior based safety is in use and felt to be a credible

PAGE 18

10 technique but that there are some misconceptions and misapplications of this approach. In a companion article also by Geller et al titled, Critical Success Factors for Behavior Based Safety: A Study of Twenty Industry Wide Applications, his group reported on interviews and focus group meetings at twenty organizations that had implemented a behavior based process for at least one year. 19 The main method used by every company that reported redu ction of injuries with behavior based safety was observation and feedback. These tactics required employee participation for success. Five variables were predictive of employee involvement in the behavior based safety process. These were: perceptions t hat behavior based training was effective, trust in management abilities, accountability for behavior based safety through performance appraisals, education in behavior based safety, and tenure with the company. The article by Gilmore and Perdue stated in the title that behavior based safety is the next step in injury prevention. 20 They discussed how the behavioral approach to safety performance improvement had recently been developed in the last part of the 20 th century. It attempted to understand emplo yee behaviors in the context of their work culture. This enlightenment was used to reduce hazards for the employees. Behavior based safety principles have been used in the form of a behavioral observation and feedback process. This process works by incr easing the frequency of safe behaviors and by decreasing the number of at risk behaviors. These techniques are similar to those used in the field of performance management. In spite of the authors support of this process, they feel that true change in s afety performance will require more than a successful behavior based safety program. Individuals must work together for the safety of one another. All these behavioral change interventions need to have a positive impact on the

PAGE 19

11 safety culture of the orga nization. The ultimate goal is Total Safety Culture. This is defined as a culture in which the individuals hold safety as a value, feel a sense of responsibility for the safety of themselves and their fellow workers, and are willing and have the skills a nd tools to go beyond the call of duty for the safety of others. Concerning behavior based training, the paper by Groover discussed the need in industrial safety and health to engage personnel at all levels to motivate themselves and each other through ha nds on action planning for performance improvement. 21 The practice of behavior based training focuses on four main concepts. The first principle is that the goal of training is behavioral change. This is followed by the concept that some behaviors are h arder to change than others. Behavior change often requires system change. Finally, to change the wage workers behavior, expect to change managements behavior. The point of behavior based training is to help trainees reduce the number and frequency of their identified at risk behaviors and increase the number and frequency of safe behaviors before an incident or injury occurs. This is accomplished by analyzing the antecedents and consequences for the at risk behaviors and the antecedents and consequenc es for the safe behaviors and then formulating an action plan. The discussion by Roberts in the article, Employees Forgot Their Hard Hats Again? Seven Lessons from Behavior Based Safety for Increasing PPE Use gave a practical application for this strateg y. 22 The first critical concept for success of behavior based safety is to involve the employees in an observation and feedback process. It is necessary to reinforce the targeted PPE behaviors. The factors influencing PPE use must be considered. It is important to know which factors contribute to safe behaviors and which contribute to at risk behaviors. Workers need to be involved in determining

PAGE 20

12 appropriate interventions. The focus of feedback needs to be on the potential for injury, not the safety ru les. This can be done more effectively by using examples, images, and case studies not just injury statistics. Finally, consider the issues beyond PPE use. Although PPE use is under the control of the individual, there are multiple causes for injury tha t may not be under the control of the individual. Managements willingness to address the factors influencing at risk behaviors under their control will increase feelings of trust and teamwork. This sense of management commitment will increase employee p articipation in safety efforts under their control. In Behavioural Safety: A case study from ICI Autocolors, Stowmarket, this UK company chose the B Safe Programme. 23 Initially, the evaluation of 31 work areas revealed only 48% safe behaviors. After 4 3 weeks of this behavior based program, the areas increased to an average of 86% safe behaviors. This change in safety prevention strategy resulted in a reduction of their accident rate by 40% in the first year. The case report, Employee Driven Behaviora l Change Improves Safety described another behavior based safety success story. 24 The lost time injury rate went from 6.3 to 2.1 after implementing a behavior based safety program for one year at Olympic National Park in Washington state. The organizatio n used employee driven strategies to improve safety communication and positive reinforcement to motivate safe work behavior resulting in a total safety culture. There were no articles discovered that used behavior based techniques to specifically target th e reduction of amputation injuries.

PAGE 21

13 Materials and Methods Because it was necessary to maintain confidentiality of the workers compensation data from the Florida Bureau of Workers Compensation, demographic data from the workers in the amputation and control groups were not used. The data sets for the amputation injury group and the control group were accessed by the computer specialist for the USF Safety Florida Consultation Program and interfaced into the mail merge for the cover letter and survey o f the selected companies. Analysis of Bureau of Labor Statistics Data Data was obtained from the Federal and Florida Bureau of Labor Statistics for amputation injuries in the US and Florida. The Federal BLS collects data annually from a sample of appro ximately 200,000 private establishments. The Florida BLS collects data from 11,000 private establishments. The data include all recordable injuries or illnesses. The case and demographic data involve all cases with at least one lost workday. Age, gend er, occupation, and length of absence are required information but race, ethnicity and years employed are optional. The case data delineates the type of injury, the body part injured, the source of the injury, and the event that created the injury. This data comes from the OSHA logs for injury and illness for each company surveyed. The summary data were analyzed for the numbers and incidence rates, and trends in amputation injuries. The case and demographic data were analyzed for age, gender, race and e thnicity, years employed, occupation, major industry category, body part injured, source of injury and the event surrounding the injury. This data was also

PAGE 22

14 compared between the national and state groups. Source and Event for the Amputation Injury Each in jury or illness reported to the Florida Bureau of Workers Compensation has a separate field describing the events surrounding the injury. A random sample of this description of injury field for 50 of the amputation injuries from the cohort group was analyz ed for the source of injury and the method of injury. This was done to assure that the cohort group was similar to the Florida and national workers with amputation injuries and could be done without breaching confidentiality of the case or the company. Am putation Prevention Strategy Survey A cohort and control group of Florida companies were selected for participation in a self designed survey concerning amputation injuries and prevention strategies for workplace amputations. The workers compensation data from the Florida Bureau of Workers Compensation within the Department of Insurance was made available to the USF Safety Florida Consultation Program for research purposes. The data was queried for all the amputation injuries for the years 1999, 2000, and 2001. This was done using the 200 code for amputations and the 4700 code for severance. Amputation was defined as cut off extremity, digit, or protruding part of the body. Severance was defined as to separate, divide or to take off. This query yielded 420 companies that had reported one or more amputation injuries to the Florida Bureau of Workers Compensation. A control group of 420 Florida companies was randomly selected from the same data base for the same years of 1999, 2000, and 2001. This group c ould have any type of work related injury or illness except amputation or severance. The surveys were mailed to the company name and address listed on the workers compensation data base. The surveys

PAGE 23

15 were addressed generically to the owner/safety profess ional of that selected company. No specific contact names were obtained. The surveys from the amputation group were marked with a 47 to designate that group. Survey Content (Figure1) Question 1 The survey participants were asked to list the SIC (standard industrial code) for their company. The companies were then placed into the appropriate major industrial category based on their response. A category for unknown was added. The major categories used were agriculture, forestry, and fishing, mining, cons truction, manufacturing, transportation and public utilities, wholesale trade, retail trade, finance, insurance, and real estate, services, and public administration. The manufacturing category was subdivided to reflect the special areas designated by the OSHA Emphasis Program on amputations. These high risk categories were lumbar and wood products, furniture and fixtures, primary metal industries, fabricated metal products, and industrial machinery and equipment. Question 2 The survey participants were asked to fill in the number of amputation injuries, if any, for each of the years of interest. This information would be readily available from the company OSHA logs for injuries and illnesses since an amputation injury would be recordable in nearly 100% of the cases. Question 3 The participants were given a list of ten strategies that might be used to prevent amputation injuries. The participants were instructed to check as many strategies as were

PAGE 24

16 applicable for their company. The question was design ed to cover a wide range of elements of a Safety and Health Program such as job site analysis, management commitment, employee involvement, administrative controls, engineering controls, analysis of injury and illness trends and near misses, and behavior b ased as well as traditional safety training. Question 4 The participants were asked to evaluate any changes that were made to their safety training after an amputation injury occurred. The participants could choose to respond yes, no, or not applicable. If the participant chose yes, they could select up to four options such as behavior based safety training, additional traditional safety training, safety committee investigation, or other. Question 5 To determine if there was a prevention technique not listed in the survey that might be valuable in decreasing injury or illness rates, the participants were given the option to write in a response to this question and share their successful strategies.

PAGE 25

17 Figure 1 SURVEY 1. What is the Standard Industrial Clas sification (SIC) for your business? 2. Did your company report any amputation injuries in the following years? 1999 Yes_____ # of injuries_____ No_____ 2000 Yes_____ # of injuries_____ No_____ 2001 Yes_____ # of injuries_____ No_____ 3. What specific actions has your company taken to prevent amputation injuries? Please check as many of the following that apply. ____Performed job site analysis to identify amputation hazards (i.e. power presses, press brakes, saws, shears, and slicers, nip points, pinch points, shear points) ____Eliminated the amputation hazards or changed the work process to avoid amputation hazards ____Held employee training on amputation hazards ____Instituted any form of behavior based safety tra ining which identifies critical safety related behaviors, develops preventive actions, measures compliance through observation, and provides feedback for improvement that is soon, certain, and positive on an ongoing basis. ____Analyzed trends in injur y and illness logs (OSHA 200 and 300 logs) ____Performed preventive maintenance of hazardous machinery ____Emphasized administrative support for a safety culture (i.e. bonuses or promotions tied to a decrease in amputation or general injury rate) ____Encouraged employee involvement in prevention with a defined complaint or suggestion procedure with feedback from management ____Evaluated near miss incidents ____Instituted engineering controls for amputation hazards (i.e. ma chine guarding, light curtains, additional safety devices or warning signage) 4. In regard to safety training, did you make any specific changes to your safety training program after the amputation(s) injury occurred? Yes_____ No_____ NA_____ If ye s, what were the changes? Check as many of the following that apply. ____Held more frequent training sessions on amputation hazards ____Instituted any form of behavior based safety training program ____Assigned safety committee to investigate the amputati on hazard and make recommendations for training and prevention ____Other 5. Is there any program or method that you have instituted as part of your safety and health program that has been particularly effective in decreasing injury or illness rates?

PAGE 26

18 Resul ts Summary Data for US and Florida Bureau of Labor Statistics (Figures 2,3) The number of amputation injuries for the years 1997, 1998, 1999, and 2000 are plotted graphically to show the trends in injury numbers for the US and Florida. For the US, there was a 5.6% decrease between 1997 and 1998. There was a 2.5% decrease between 1998 and 1999. There was a 3.3% decrease in amputations between 1999 and the year 2000. In Florida, there was a 17% increase between 1997 and 1998 but a 19.4% decrease between 1998 and 1999. Then between 1999 and 2000, there was a 62.3% increase in amputations in the state of Florida. The incidence rates for amputation injuries represent the number of injuries per 10,000 full time workers. They are also plotted graphically an d used to show the trends in amputations as well as a visual comparison between the US and Florida. The downward trend in amputation injuries in the US and the recent increase in amputations in Florida is even more apparent on the incidence rate graph. U S Bureau of Labor Statistics (Table 1) For the amputation injuries in the year 2000, the mean age of the worker was 35 to 44. This worker was usually a white, non Hispanic male. He had been working with his company 1 to 5 years before the amputation inj ury occurred. The most common jobs for amputation injuries were operators, fabricators, and laborers in the manufacturing industry. The source of the injury was most often a machine. The body part which was usually a finger was caught in, compressed, o r crushed by the machine.

PAGE 27

19 Figure 2 Comparison of Number of Amputations: US vs Florida 272 319 257 417 10,852 10,243 9,985 9,658 0 2000 4000 6000 8000 10000 12000 1997 1998 1999 2000 Years Number of Amputations Florida US Figure 3 Comparison of Amputation Rates: US vs Florida 1.3 1.2 1.1 1.1 0.6 0.6 0.5 0.8 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1997 1998 1999 2000 Years Incidence Rates US Florida

PAGE 28

20 Florida Bureau of Labor Statistics (Table 1) The Florida Bureau of Labor Statistics gives us a snapshot of the demographic data and the case characteristics of amputation injuries in the state. For the ye ar 2000, most all of the workers were male between 35 and 44 years old. The majority of the workers were Hispanic and the rest were white, non Hispanic. 29% had less than 3 month length of service with their employer, and 35% had 1 to 5 years service p rior to their amputation injury. The most frequequent occupations of the workers were in the agriculture, forestry, and fishing industries, or in the precision production, craft or repair industries. Most all of the amputations involved the upper extrem ity with nearly all of the amputations occurring to the fingers. Machinery and parts and materials were responsible for the amputations. This happened by contact with an object or equipment. Most events involved being caught in the object, equipment, o r material and the rest occurred by being struck by the object. Table 1 Amputation Demographic and Case Data for 2000 Florida US Total 417 9,658 Major Industry Division Agriculture, Forestry, Fishing 151 703 Mining 127 Construction 1,143 Manufacturing 94 4,516 Transportation, Utilities 364 Wholesale Trade 956 Retail Trade 1,196 Finance, Insurance, Real Estate 121 Services 532 Gender Male 343 8,397 Female 1,171 Not reported 89 Race or Ethnic Ori gin White, non Hispanic 79 5,072 Black, non Hispanic 730

PAGE 29

21 Hispanic 244 1,872 Asian or Pacific Islander 189 American Indian, Alaskan Native 51 Not reported 82 1,744 Age 16 19 560 20 24 59 1,223 25 34 61 2,277 35 44 197 2, 728 45 54 1,661 55 64 59 1,041 65 and over 107 Not reported 61 Length of Service Less than 3 months 121 1,652 3 months to 11 months 72 1,842 1 year to 5 years 148 3,141 More than 5 years 65 2,415 Not reported 607 Occupations Groups Managerial/Professional 106 Tech, Sales, Adminstrative support 423 Service 72 814 Farming, Forestry, Fishing 151 734 Precision production, craft, repair 107 2,378 Operators, fabricators, laborers 58 5,184 Part of Body Affected Upp er Extremity 348 9,163 Finger 343 8,810 Hand, except finger 192 Lower Extremity 69 495 Foot, toe 68 374 Source of Injury Containers 302 Machinery 245 5,429 Parts & Materials 107 1,130 Floors, walkways or ground surfaces 54 Hand tools 1,203 Vehicles 739 All other sources 762 Event Leading to Injury Contact with objects 415 9,340 Struck by object 82 1,921 Struck against object 777 Caught in object, compressed, or crushed 311 6,559 Transportation ac cidents 102 Fires & explosions 115 All other events 83 Comparison of Demographic and Case Data between US and Florida Comparison of the demographic and case data between Florida workers with amputation injuries and US workers with amputatio n injuries revealed many similarities.

PAGE 30

22 The most prevalent gender and age categories were the same. The most frequent industry division represented in Florida was agriculture, forestry, and fishing as opposed to manufacturing for the US. This reflects th e prominence of agriculture in Florida with the large numbers of citrus and produce industries in the state. The most prevalent occupation group for amputations in Florida was farming, forestry, and fishing. This parallels the major industry division rep resented. The greatest occupation category for the US was operators, fabricators and laborers which reflects the major industry division of manufacturing. There were more Hispanic workers with amputations in Florida than whites. This mirrors the increas ed numbers of Hispanics in the workforce in Florida as well as their employment in agriculture. Florida has the fourth largest population of seasonal and migrant farm workers and 86% of them are Hispanic. The US data showed a preponderance of white, non Hispanic workers with amputations. Florida and the US had the most amputation injuries in the group with one to five years of service. Florida proportionally had more workers with amputations who had been working less than three months when compared to t he US workers. This could reflect a lack of training issue. In both groups, the most frequently affected body part was the finger. The greatest source of injury was machinery for the Florida and US groups. The event leading to the injury was contact wit h an object for the two groups. Most often, the affected part was caught in the object, compressed, or crushed followed by being struck by the object for the injured workers in Florida and the US. Source and Event for the Amputation Group (Table 2,3) F or the 50 random description of injury fields, the sources of injury were machinery and parts or materials, followed by handtools, and vehicles. The event was always contact

PAGE 31

23 with the object or equipment. Most of the amputations occurred because the body part was caught in the object, equipment, or material. Nine happened when the injured part struck against the object and 6 occurred when the affected part was struck by the object. Table 2 Source of Injury Hand tools 4 Machinery 22 Parts & Materials 22 Vehicle 2 n = 50 Table 3 Event: Contact With Object or Equipment Struck by 6 Struck against object 9 Caught in object, compressed, or crushed 35 n = 50 Survey Results 840 surveys were mailed out to the companies in Florida with and without amputat ion injuries for the years 1999, 2000, and 2001. 146 surveys were returned as undeliverable with no forwarding address. 86 were in the amputation group and 60 were in the control group. Two companies stated that they would be unable to participate in t he survey. One was in the amputation group and the other was in the control group. Out of the remaining 692 surveys, 35 surveys were returned in the amputation group for a response rate of 10.5%. In the control group, 35 surveys were returned for a resp onse rate of 9.7%.

PAGE 32

24 Survey results Question 1 (Table 3) Question one asked for the SIC code for the company. Of the 35 companies in the amputation group, all of the Standard Industrial Codes (SIC) were represented except mining and public administration. In the manufacturing subgroups, four of the high risk industries responded to the survey. There was one from the furniture and fixtures subgroup, two from primary metal industries, and one from fabricated metal products. In the control group, all the S IC groups were represented except agriculture, forestry, and fishing, and mining. One high risk industry responded to the survey in the control group. This was a lumber and wood products company. There were differences in the percentages of the SIC repr esentation between the amputation group and the control group. The amputation group had more unknowns than the controls. Manufacturing was highly represented in the amputation group (20% vs 8.6%). Adding the high risk manufacturing subgroups to the gene ral manufacturing category, increased the differences between these groups to 31.4% vs 11.5%. The control group had more companies responding from service industries (37.1% vs 14.3%). Public administration was represented in the control group by three co mpanies and by none in the amputation group.

PAGE 33

25 Table 3 SIC for Amputation and Control Group SIC Class Amputation % Control % PeRcent DIFFERENCE Unknown 6 17.1% 4 11.4% 5.7% Agriculture, Forestry, Fishing 1 2.9% 0 0% 2.9% Construction 4 11.4% 2 5.7% 5.7 % Manufacturing 7 20% 3 8.6% 11.4% High Risk Manufact. 4 11.4% 1 2.9% 8.5% Tranportation, Utilities 1 2.9% 2 5.7% 2.8 Wholesale trade 3 8.6% 2 5.7% 2.9% Retail trade 3 8.6% 4 11.4% 2.8% Finance, Insurance, Real Estate 1 2.9% 1 2.9% 0% Services 5 14. 3% 13 37.1% 22.8% Public Admin. 0 0% 3 8.6% 8.6% TOTAL 35 35 Question 2 (Table 4 ) For the year 1999, the amputation group listed 8 amputation injuries and the control group had none. For the year 2000, there were 5 injuries in the sturdy group an d none in the control group. For the year 2001, there were 6 amputation injuries in the amputation group and none in the control group. Three companies had more than one amputation injury. Of the 35 companies in the study group, only 15 or 42.9% claimed amputation injuries for any of the three years of interest. The other 20 companies or 57.1% did not list any amputation injuries even though their names were selected from the amputation group database. To determine the true error rate for the workers c ompensation data, the description of injury fields for all 420 selected amputation injuries were reviewed. There were 55 injuries that were misclassified in the amputation group for an error rate of 13%. The inaccuracy occurred mostly with the severence code and not the amputation code. The term severence had been used to describe a separated shoulder or separated ribs, for example.

PAGE 34

26 Table 4 Frequency of Amputations 1999 Amputations Controls Yes No Yes No 8 27 0 35 2000 Amputations Controls Yes N o Yes No 5 30 0 35 2001 Amputations Controls Yes No Yes No 6 29 0 35 1999 2001 # Companies Claiming Amputations Yes No 15 20 Question 3 (Table 5,6) The raw data for each prevention strategy reponse was tabulated by the corresponding SIC in Tabl e 5. Each of the 10 prevention strategy choices were compared separately between the amputation group and the control group using chi square. For each 2x2 table there was one degree of freedom. The critical chi square for 1 df was 3.841. The numbers li sted in Tables 5 and 6 represent the yes responses for both groups except for the none category. The no responses were calculated by subtracting the number of yes responses from 35 (# participants in each group) and then subtracting the number of none res ponses for that specific group (2 in the ampuation group and 7 in the control group). It was determined that the participants that chose not to select any choices (none category) would be not be analyzed. Since the purpose was to compare prevention strat egy choices between the study group and the control group, there would be nothing to compare among two groups that didnt select any prevention strategies.

PAGE 35

27 There were no significant differences between the prevention strategy choices among the two groups except for choice #5 which was analysis of the OSHA log trends. This was significant for an alpha of 0.05. Because multiple chi square tests were performed, this increased the probability of a Type I error and significance occurring by chance alone. A type I error occurs when a true null hypothesis is rejected. To correct for this problem, the Holm procedure was applied to the p values. The p i values were placed in increasing numerical order. P 1 is the smallest and P n is the largest. Then each p i value was compared with alpha/(n i + 1) for rejection. Starting with the smallest p value, one continues applying these comparisons until the first nonrejection is encountered. In this case, n equals 10 (from the 10 chi squares). The p value for stra tegy choice # 5 was the smallest so it was given #1 in the rank order. Therefore, the formula calculated as 0.05/(10 1 + 1). Solving this equation gave an adjusted alpha of 0.005. The obtained p value of 0.019 is larger than 0.005 and now the null hyp othesis is not rejected and this choice is no longer significant. There was no significant difference in the number of responses chosen between the amputation group and the control group. These results are presented in Table 6. A logistic regression w as also performed on the prevention strategy data. Logistic regression is used to obtain predicted probabilities that a unit of the study population will acquire the event of interest (dependent variable) as a linear function of one or more independent va riables. The logistic regression was not significant for any or all independent variables.

PAGE 36

28 Table 5 Prevention Strategy Reponses by SIC Amputations SIC 1 2 3 4 5 6 7 8 9 10 None Unknown 3 4 2 3 1 1 1 2 Agriculture, Forestry, Fishing 1 1 1 1 1 1 Construction 2 2 4 2 2 2 1 2 Manufacturing 4 3 6 4 3 4 2 4 4 4 High Risk Manufacturing 3 3 2 1 1 2 1 3 3 3 1 Transportation, Utilities 1 Wholesale Trade 1 1 1 2 1 2 1 1 1 2 1 Retail Trade 3 1 3 2 1 2 1 1 1 F inance, Insurance, Real Estate 1 1 1 1 Services 2 1 1 4 2 3 3 2 2 3 TOTAL 20 15 18 17 8 19 12 16 13 18 2 Controls SIC 1 2 3 4 5 6 7 8 9 10 None Unknown 2 1 3 2 2 2 2 2 1 Construction 2 1 1 2 1 2 1 1 1 Manufacturing 3 2 3 2 1 3 2 2 2 2 High Risk Manufacturing 1 1 Transportation, Utilities 2 2 1 1 2 2 2 1 1 2 Wholesale Trade 1 1 1 1 1 1 Retail Trade 1 2 1 1 1 1 2 Finance, Insurance, Real Estate 1 1 1 1 Services 6 5 5 8 8 8 2 8 5 6 2 Public Administration 1 1 1 2 TOTAL 17 12 11 20 15 19 11 18 12 14 7 Key 1=Identify amp hazards 2=Eliminate amp hazards 3=Hazard training 4=Behavior based training 5=OSHA log trends 6=Machinery maintenance 7=Admin. Support 8=E mployee Feedback 9=Evaluate near misses 10=Engineering controls

PAGE 37

29 Table 6 Comparison of Amputation Prevention Strategies strategy Amputation Control chi square P Value Identify amp hazards 20 17 0.0001 0.993 Eliminate amp hazard 15 12 0.041 0.839 Hazard training 18 11 1.421 0.234 Behavior based training 17 20 2.517 0.113 OSHA log trends 8 15 5.547* 0.019* Machinery maintenance 19 19 0.681 0.409 Admin. Support 12 11 0.055 0.815 Employee feedback 16 18 1.532 0.216 Evaluate near misses 13 12 0.075 0.7 84 Engineering controls 18 14 0.006 0.939 Total # of responses 156 149 0.28 No strategies chosen 2 7 NA Critical ? = 3.841 significant with a of 0.05 Question 4 (Table 7) This question focuses on the changes made after an amputation injury occurred so only the companies in the amputation group were analyzed in this case. The initial choices for this qu estion were yes, no, or not applicable. The 20 companies that did not claim any amputaion injuries were listed as not applicable because in their mind, they had no amputation injuries. Any changes made to their safety training program after an amputation injury would not apply. Of the remaining 15 companies, there were 21 responses to this question. Eight of the 15 companies or 42% made no changes to their safety training after the amputation injury occurred. Seven companies stated that they had made some changes to their safety training programs after the injury occurred. Three companies held more frequent training sessions on amputation hazards. Two companies instituted behavior based safety training. Five companies assigned the safety committee t o investigate the amputaion hazard and make recommendation for training and prevention. Under the other response choice, one company included 10 and 30 hour OSHA training, and one company improved their machine guards.

PAGE 38

30 Table 7 Post Amputation Safety Training Program Changes changes # of responses Frequency NA: no amputations claimed 20 20/35=57.1%* No 8 8/35=22.9%* Yes 7 (12 choices) 7/35=20%* More training 3 3/7=42.9%** Behavior based training 2 2/7=28.6%** S afety Committee investigation 5 5/7=71.4%** Other 2 2/7=28.6%** Frequency based on total # of participants in amputation group ** Frequency based on total number of participatants responding yes Question 5 The comments to this question varied but the rewere a few common themes. The amputation group had 16 responses to this question which concerned any program or method that was effective in decreasing injury or illness rates. Several companies commented on the use of a safety committee or scheduled sa fety meetings. A few participants mentioned the importance of management commitment and presence at the worksite. Others had variants of a safety committee that were successful such as a safety report card, safety improvement teams which focused on accid ents and prevention, and a safety advisory commmitttee that had budget resources to act on the recommendations of the safety committee. Two companies evaluated compentency before allowing employees to operate dangerous equipment. Two companies enforced s afety rules strictly with discipline or time off for non compliance with the rules. Two companies mentioned the importance of a drug free workplace and drug testing. Employee involvement in accident and illness investigation was listed as an effective te chnique. One company allowed employees to refuse to operate equipment without safety devices in place. Other successful strategies included behavior based safety training, safety culture mindset with constant reinforcement, and best practices for safety and work

PAGE 39

31 performance. One compny felt that video presentations for safety training were helpful. One respondant thought that the OSHA 10 and 30 hour training courses were effective. An ergonomics committee was a useful method at one company. The contr ol group had 18 responses to this question. The control group mentioned similar successful approaches to decreasing injury and illness rates. Safety meetings and regular training were mentioned by several companies. Two of theses involved the use of saf ety incentives. One company had a shop safety program jointly conducted by safety inspectors and the shop curriculum specialist. Another used annual inspections from risk management as an effective technique. Several participants focused on hazard recog nition, followed by elimination of the hazard or specific engineering controls. One company cited the importance of support from administration. One company felt that repetition in training was a successful method. Investigation of accidents and near m isses worked well for one company. Equipment training and providing state of the art equipment to improve the working conditions were two effective strategies listed. Ergonomic assessments and corrections were helpful in decreasing rates according to one respondant. One company with a knive hazard did not allow minors to use knives at the workplace. They also made sure the hazardous tasks were done during slow work periods. A quality improvement process worked well for one company. Overall, the concep ts of continual assessment, evaluation, and training along with making safety a priority resulted in positive outcomes for these companies.

PAGE 40

32 Discussion According to the literature search, there is little information on the topic of amputations in the wo rkplace. There are no studies on prevention strategies for amputation injuries other than those involving machinery. Most of the data indicates that the amputations occur when an employee is using machinery or parts or materials and is either struck by t he object or caught in, crushed or compressed by the object. This information led OSHA to develop its Emphasis Program on Amputations and to recognize high risk industries and sources of amputations such as power presses, machines, slicers, and saws. The main solution offered so far is engineering controls, especially machine guarding. This seems to have led to a gradual decrease in the amputation rate in the US although not in Florida. Traditionally, safety prevention has been done in the form of a pro gram training process for each element and education from post accident investigation. In the study by Vojtecky and Schmitz it pointed out that the people responsible for safety training arent even sure if it is effective. 25 Safety training and post a ccident investigation both heighten awareness at the time of the training or the accident. This awareness tapers off as time elapses. To prevent this, safety needs to be on everyones mind every day and this is difficult to accomplish. The article, Cont inuous Safety Progress Focuses on Upstream Factors in Analyses discussed the accident cycle and ways to break free of it. 26 An increase in incident rates triggers more attention to safety until performance improves and then the attention is directed els ewhere. Krause suggests the way to continuous safety progress is through behavior based safety systems. Perhaps the system needs to be

PAGE 41

33 adjusted so it wont allow accidents to happen. Investigation into medical errors has come to the conclusion that the individuals want to do the right thing but the system doesnt always allow that to occur. 27 Engineering controls such as machine guarding are one way to remove the hazards from the system. Another proactive approach is the use of a behavior based safety program. As previously reviewed, this technique involves management commitment to the program, employee involvement in the process as well as maintaining an awareness of safety and changing behavior by reinforcing safe behaviors and altering unsafe ones BEFORE an accident occurs. This concept appears to be in use in Florida. 48.6% of the companies with amputation injuries were using it as a prevention strategy and 57% of the companies without amputation injuries listed it as a prevention strategy. Exter nal Validity The source of injury and the event for the amputation study group were compared with the Bureau of Labor Statistics data for Florida and the US. The most frequent sources of injury for the study group were machinery and parts & materials. T his is comparable to the two largest groups for Florida and the first and third largest groups for the US. The most frequent group in the event category for the study population was caught in the object, compressed, or crushed. This was the same for the Florida and the US Bureau of Labor Statistics dataset. This analysis implies that the amputation study group is generalizable to other workers in Florida and the US at least for the source of injury and the event. Study Limitations The surveys were sent out using the addresses that the companies gave to the

PAGE 42

34 Department of Insurance when the injury was reported. Despite this, 146 were returned without being delivered. 58.9% were in the amputation group and 41% were in the control group. There were more unavailable companies in the amputation group and this factor may have been important. These unavailable companies may have gone out of business for a variety of reasons which could include poor safety habits and may have been different from the companies that actually received the surveys. The two companies that were unable to participate were equally represented between the control and study group. They stated that they were not allowed to participate for legal reasons and a corporate policy on surveys This number of non participation was too small to affect the study results. The number of non responders in each group was large. There were 334 non responders in the amputation group and 360 non responders in the control group. Not only did this re sult in small numbers for analysis in each group, the non responder group may have been different than the companies that chose to respond. The cover letter suggested that participation in the survey would result in information that could benefit their co mpany by decreasing injury rates and workers compensation costs. Perhaps this statement was not enough incentive to participate. Maybe other companies were concerned about the legal issues of giving out company information. The companies in the amputati on group may not have felt comfortable admitting that they had an amputation. It would be interesting to know why the companies chose not to return the surveys. There was a problem with misclassification among the amputation group. Over half of the co mpanies selected to the amputation group did not report an amputation

PAGE 43

35 injury for the years of interest. Taking into account the error rate of 13% for misclassification of the amputation injuries by the Department of Insurance, this still leaves 44.1% of amputations unaccounted for in the study group. This information should have been available from the OSHA logs for 1999, 2000, and 2001. Without violating the confidentiality of the study and making contact with the safety professionals for the companies there is no way to uncover the reason for this misclassification. Again, it is possible that the companies with amputation injuries were reluctant to admit this information to a voluntary and unknown source. By design, the participants were divided int o the amputation and control groups and then identified only by their Standard Industrial Classification. There were differences between the two groups based on this classification which could be viewed as a confounder. A confounder is a third factor tha t affects the outcome but isnt the independent variable. In this study, the independent variables are the prevention strategies and the outcome is the amputation injury. The amputation group had more representation from industries involved in manufactur ing including those at high risk for amputations. The control group had more representation from service industries and public administration which are low risk industries for amputation injuries. Since there were many choices for selection of the contro l group, this confounder could have been controlled for by matching the control group to the frequency of the SIC divisions in the amputation group. There was a difficulty with the validity of the results for the main survey question involving the prevent ion strategies amputation injuries because of the small sample size. There was an increased probability of a Type I error due to multiple chi square testing.

PAGE 44

36 This was adjusted for using the Holm Procedure. Although the differences between the study gr oup and the control group were not significant, there was not enough power to conclude that the non significant findings were truly not significant or due to a Type II error. There were 9 participants that chose to select no prevention strategies for that survey question. Two were in the amputation group and 7 were in the control group. Since it would be inappropriate to compare the use of no prevention strategies among the two groups, these numbers contributed to lowering the number of responses availab le for chi square calculation. If the survey return rate had been 40% instead of 10%, and the differences between the two groups had been at least 18, the results would have been significant. The near significant strategies for the control group were ana lysis of injury and illness trends from the OSHA logs, behavior based safety training, and employee involvement and feedback. The only near significant strategy for the amputation group was employee training for amputation hazards. The control group appe ars to be using more global safety approaches and the amputation group is focusing on more narrow, specific tactics. An area of concern was the 8 companies out of 15 who claimed an amputation injury but made no changes to their safety training program aft erwards. It seems that a significant injury like an amputation would trigger an investigation and perhaps some proactive changes to prevent any further injuries. One company wrote that their workers compensation insurance carrier investigated the amputat ion injury and found that it was due to the employees carelessness and not to any deficiencies in their safety program. There may have been similar circumstances with other companies.

PAGE 45

37 Conclusion The data clearly show that amputation injuries occur in all industries not just in the high risk manufacturing companies. Therefore, it would be important to find a universal strategy that could prevent these injuries and be applied in all types of industry. It was hoped that this study would uncover a spec ific successful strategy(s) that could be used by all companies in Florida to decrease the amputation rate in the state. Areas of promise include analysis of trends in injury from OSHA logs, a behavior based safety program, and employee involvement and fe edback. Once initiated in Florida, the OSHA Emphasis Program on Amputations may be effective for high risk industries. Since there was not evidence of any one significantly successful strategy, further research in this area would be beneficial. A nested case controlled study in a high risk amputation industry using one specific strategy such as a behavior based safety program could be used to evaluate the effect of this strategy on the companys amputation rate.

PAGE 46

38 Research References 1. United State s Department of Labor. Bureau of Labor and Statistics. Available at http:// www.bls.gov/iif Accessed July, 2002. 2. Florida Department of Labor and Employment Security. Bureau of Labor and Statistics. Available at http:// www@fdles.state.fl.us Accessed July, 2002. 3. Krause T. The Behavior Based Safety Process. New York, NY. Van Nostrand Reinhold Publishers. 1990. 4. Etherton JR, Myers ML. Machine Safety Researc h at NIOSH and the Future Directions, International Journal of Industrial Ergonomics. 1990, Volume 6, 163 174. 5. OSHA Instruction: National Emphasis Program on Amputations. US Dept of Labor, Occupational Safety and Health Administration. March 26, 200 2. Directive No. CPL 2 1.35. 6. CDC, 2001, Nonfatal Occupational Injuries Treated in Hospital Emergency Departments United States, 1998. MMWR. Morbid Mortal Wkly Report, 50(16): 313 7. 7. Injuries and Amputations Resulting from Work with Mechanical P ower Presses. Cincinnati, Ohio: Dept of Health and Human Services (NIOSH). May 22, 1987. Publication No. 87 107. 8. Boyle D, Larson C, Parker D, Pessoa Brandao L. Medical, Personal, and Occupational Outcomes for Work Related Amputations in Minneso ta. Amer J Ind Med. 37(5): 551 7, 2000 May. 9. Boyle D, Larson C, Parker D, Pessoa Brandao L. Nature, Incidence, and Cause of Work Related Amputations in Minnesota. Amer J Ind Med. 37(5): 542 50, 2000 May. 10. Cook S, Long L. MIOSHAs Strategic Pl an for Reducing Amputations in Michigan. Available at http://www.michigan.gov/cis Accessed August, 2002. 11. Safeguarding Equipment and Protecting Workers from Amputations. U.S. Dept of Labor, Occupational Safety and Health Administration. 2001. Publication OSHA 3170.

PAGE 47

39 12. OSHA Fact Sheet: Amputations. U.S. Dept. of Labor, Occupational Safety and Health Administration. 2002. 13. Vojtecky M. Education for Job Safety and Health. Health Education Quar terly. Vol. 15(3): 289 298. Fall 1988. 14. Goldenhar L, Moran S, Colligan M. Health and Safety Training in a Sample of Open Shop Construction Companies. Journal of Safety Research. Vol. 32 (2001) 237 252 15. Gardner D, Carlopio J, Fonteyn P, Cr oss J. Mechanical Equipment Injuries in Small Manufacturing Businesses. Knowledge, Behavioral, and Management Issues. International Journal of Occupational Safety and Ergonomics. Vol.5, No.1, 59 71. 1999. 16. Bimonte A. Win Empowered Workers Commit ment to Exceeding Basic Safety Compliance. Occupational Health & Safety. Vol. 63, No. 5. May, 1994. 17. Sulzer Azaroff, B Harris, T McCann, K. Beyond Training: Organizational Performance Management Techniques. Occupational Medicine: State of the A rt Reviews. Vol. 9, No. 2, April June 1994. 18. Geller ES, Boyce TE, Williams JH et al. Researching Behavior Based Safety: A Multi Method Assessment and Evaluation. Proceedings of the 37 th Annual Professional Development Conference and Exposition. Des Plaines, IL: American Society of Safety Engineers, 1998. 19. DePasquale J, Geller ES. Critical Success Factors for Behavior Based Safety: A Study of Twenty Industry Wide Applications. Journal of Safety Research. 30 (4), 237 249, 1999. 20. Gilmore MR, Perdue SR. Behavior Based Safety: The Next Step in Injury Prevention. Proceedings of the Society of Petroleum Engineers International Conference. Kuala Lumpur, Malaysia: Society of Petroleum Engineers, Inc., March 2002. 21. Groover D, Hodson S. Behavior Based Training for Safety & Health. Occupational Hazards. Vol. 60, Issue 9, 45 51, September 1998. 22. Roberts SD. Employees Forgot Their Hard Hats Again? Seven Lessons from Behavior Based Safety for Increasing PPE Use. Safety and Health. October, 2000. 23. Cooper D. Behavioural Safety: A Case Study from ICI Autocolors, Stowmarket. Available at http://www.behavioral safety.com Accessed August, 2002. 24. Valadez S, Williams JH, Rober ts S. Employee Driven Behavioral Change Improves Safety. People, Land, and Water. July/August, 1999.

PAGE 48

40 25. Vojtecky M, Schmitz, M. Program Evaluation and Health and Safety Training. Journal of Safety Research. Vol. 17, No. 2, pp. 57 83. 1986. 26. Krause TR. Continuous Safety Progress Focuses on Upstream Factors in Analyses. Occupational Health & Safety. Vol.63, No. 10 October 1994. 27. Sheff, R. Breakthrough Approaches to Reducing Medical Errors: A New Way of Thinking. Presented at CME conference for Tampa General Hospital; January 28, 2003; Tampa, Florida.

PAGE 49

41 Bibliography Geller ES. Behavior Based Safety: Confusion, Controversy, and Clarification. Occupational Health and Safety. 68 (1), 40 49, 1999. Geller ES, Krueger B, Fren ch A et al. The Human Factor. Utility Safety. April 2000. Marsh T. Behavioural Safety The Current Position. Training Solutions. Issue 9, summer 1999. Olson DK, Gerberich SG. Traumatic Amputations in the Workplace. Journal of Occupational Medici ne. 28(7): 480 5, 1986 July. Owens S. BAPP Safety Builds Improved Organization Functioning. Perspectives in Performance Improvement. April, 2002, 11 14. Sorock GS, Smith E, Hall N. Hospitalized Occupational Finger Amputations, New Jersey, 1985 and 1986. Amer J Ind Med. 23(3): 439 47, 1993 March. Williams JH, Geller ES. Behavior Based Intervention for Occupational Safety: Critical Impact of Social Comparison Feedback. Journal of Safety Research.

PAGE 50

42 Appendices

PAGE 51

43 Appendix A : SAS Program libname bob 'A: \ ; proc import out = bob.eve datafile = "A: \ eve.xls" dbms = excel2000 replace; getnames = yes; run ; proc print ; run ; proc freq ; tables jsa amputation/ chisq ; run ; proc freq ; tables elim amputation/ chisq ; run ; proc freq ; tables train amputation/ chisq ; run ; proc freq ; tables behave amputation/ chisq ; run ; proc freq ; tables oshalog amputation/ chisq ; run ; proc freq ; tables maint amputation/ chisq ; run ; proc freq ; tables cult amputation/ chisq ; run ; proc freq ; tab les employee amputation/ chisq ; run ; proc freq ; tables nearmiss amputation/ chisq ; run ; proc freq ; tables engineer amputation/ chisq ; run ; proc freq ; run ; proc sort data = bob.eve; by amputation; run ; proc freq ; by amputation; run ; proc logistic ; cl ass jsa elim train behave oshalog maint cult employee nearmiss engineer; model amputation =jsa elim train behave oshalog maint cult employee nearmiss engineer; run ;

PAGE 52

44 Appendix A: (Continued) The FREQ Procedure Table of jsa by amputation jsa(jsa) amputation(amputation) Frequency Percent Row Pct Col Pct 0 1 Total 0 11 13 24 18.03 21.31 39.34 45.83 54.17 39.29 39.39 1 17 20 37 27.87 32.79 60.66 45.95 54.05 60.71 60.61 Total 28 33 61 45.90 54.10 100.00 Frequency Missing = 9 Statistics for Table of jsa by amputation Statistic DF Value Prob Chi Square 1 0.0001 0.9931 Likelihood Ratio Chi Square 1 0.0001 0.9931 Continuity Adj. Chi Square 1 0.0000 1.0000 Mantel Haenszel Chi Square 1 0.0001 0.9932 Phi Coefficient 0.0011 Contingency Coefficient 0.0011 Cramer's V 0.0011 Fisher's Exact Test Cell (1,1) Frequency (F) 11 Left sided Pr <= F 0.6008 Right sided Pr >= F 0.6063 Table Probability (P) 0.2071 Two sided Pr <= P 1.0000 Effective Sample Size = 61 Frequency Missing = 9 WARNING: 13% of the data are missing.

PAGE 53

45 Appendix A (Continued) Th e FREQ Procedure Table of elim by amputation elim(elim) amputation(amputation) Frequency Percent Row Pct Col Pct 0 1 Total 0 16 18 34 26.23 29.51 55.74 47.06 52.94 57.14 54.55 1 12 15 27 19.67 24.59 44.26 44.44 55.56 42.86 45.45 Total 28 33 61 45.90 54.10 100.00 Frequency Missing = 9 The FREQ Procedure Statistics for Table of elim by amputation Statistic DF Value Prob Chi Square 1 0.0414 0.8387 Likelihood Ratio Chi Square 1 0.0414 0.8387 Continuity Adj. Chi Square 1 0.0000 1.0000 Mantel H aenszel Chi Square 1 0.0407 0.8400 Phi Coefficient 0.0261 Contingency Coefficient 0.0260 Cramer's V 0.0261 Fisher's Exact Test Cell (1,1) Frequency (F) 16 Left sided Pr <= F 0.6776 Right sided Pr >= F 0.5223 Table Probability (P) 0.1998 Two sided Pr <= P 1.0000 Effective Sample Size = 61 Frequency Missing = 9 WARNING: 13% of the data are missing.

PAGE 54

46 Appendix A (Continued) The FREQ Procedure Table of train by amputation train(train) amputation(amputation) Frequency Percent Row Pct Col Pct 0 1 Total 0 17 15 32 27.87 24.59 52.46 53.13 46.88 60.71 45.45 1 11 18 29 18.03 29.51 47.54 37.93 62.07 39.29 54.55 Total 28 33 61 45.90 54.10 100.00 Frequency Missing = 9 The FREQ Procedure Stat istics for Table of train by amputation Statistic DF Value Prob Chi Square 1 1.41 43 0.2343 Likelihood Ratio Chi Square 1 1.4212 0.2332 Continuity Adj. Chi Square 1 0.8686 0.3513 Mantel Haenszel Chi Square 1 1.3911 0.2382 Phi Coefficient 0.1523 Contingency Coefficient 0.1505 Cramer's V 0.1523 Fisher's Exact Test Cell (1,1) Frequency (F) 17 Left sided Pr <= F 0.9263 Right sided Pr >= F 0.1758 Table Probability (P) 0.1021 Two sided Pr <= P 0.3058 Effective Sample Size = 61 Frequency Missing = 9 WARNING: 13% of the data are missing.

PAGE 55

47 Appendix A (Continued) The FREQ Procedure Table of behave by amputation behave(behave) amputation(amputation) Frequency Percent Row Pct Col Pct 0 1 Total 0 8 16 24 13.11 26.23 39.34 33.33 66.67 28.57 48.48 1 20 17 37 32.79 27.87 60.66 54.05 45.95 71.43 51.52 Total 28 33 61 45.90 54.10 100.00 Frequency Missing = 9 The FREQ Procedure Statistics for Table of behav e by amputation Statistic DF Value Prob Chi Square 1 2.5170 0.1126 Likelihood Ratio Chi Square 1 2.5516 0.1102 Continuity Adj. Chi Square 1 1.7517 0.1857 Mantel Haenszel Chi Square 1 2.4757 0.1156 Phi Coefficient 0.2031 Contingency Coefficient 0.1991 Cramer's V 0.2031 Fisher's Exact Test Cell (1,1) Frequency (F) 8 Left sided Pr <= F 0.0924 Right sided Pr >= F 0.9686 Table Probability (P) 0.0610 Two sided Pr <= P 0.1256 Effective Sample Size = 61 Frequency Missing = 9 WARNING: 13% of the data are missing.

PAGE 56

48 A ppendix A (Continued) The FREQ Procedure Table of oshalog by amputation oshalog(oshalog) amputation(amputation) Frequency Percent Row Pct Col Pct 0 1 Total 0 13 25 38 21.31 40.98 62.30 34.21 65.79 46.43 75.76 1 15 8 23 24.59 13.11 37.70 65.22 34.78 53.57 24.24 Total 28 33 61 45.90 54.10 100.00 Frequency Missing = 9 The FREQ Procedure Statistics for Table of oshalog by amputation Statistic DF Value Prob Chi Square 1 5.5473 0.0185 Likelihood Ratio Chi Square 1 5.6094 0.0179 Continuity Adj. Chi Square 1 4.3689 0.0366 Mantel Haenszel Chi Square 1 5.4564 0.0195 Phi Coefficient 0 .3016 Contingency Coefficient 0.2887 Cramer's V 0.3016 Fisher's Exact Test Cell (1,1) Frequency (F) 13 Left sided Pr <= F 0.0180 Right sided Pr >= F 0.9958 Table Probability (P) 0.0138 Two sided Pr <= P 0.0330 Effective Sample Size = 61 Frequency Missing = 9 WARNING: 13% of the d ata are missing.

PAGE 57

49 Appendix A (Continued) The FREQ Procedure Table of maint by amputation maint(maint) am putation(amputation) Frequency Percent Row Pct Col Pct 0 1 Total 0 9 14 23 14.75 22.95 37.70 39.13 60.87 32.14 42.42 1 19 19 38 31.15 31.15 62.30 50.00 50.00 67.86 57.58 Total 28 33 61 45.90 54.10 100.00 Frequency Missing = 9 The FREQ Procedure Statistics for Table of maint by amputation Statistic DF Value Prob Chi Square 1 0.6817 0.4090 Likelihood Ratio Chi Square 1 0.6854 0.4077 Continuity Adj. Chi Square 1 0.3142 0.5751 Mantel Haenszel Chi Square 1 0.6705 0.4129 Phi Coefficient 0.1057 Contingency Coefficient 0.1051 Cramer's V 0.1057 Fisher's Exact Test Cell (1,1) Frequency (F) 9 Left sided Pr <= F 0.2883 Right sided Pr >= F 0.8624 Table Probability (P) 0.1507 Two sided Pr <= P 0.4398 Effective Sample Size = 61 Frequency Missing = 9 WARNING: 13% of the data are missing.

PAGE 58

50 Appendix A (Continued) The FREQ Procedure Table of cult by amputation cult(cult) amputation(amputation) Frequency Percent Row Pct Col Pct 0 1 Total 0 17 21 38 27.87 34.43 62.30 44.74 55.26 60.71 63.64 1 11 12 23 18.03 19.67 37.70 47.83 52.17 39.29 36.36 Total 28 33 61 45.90 54.10 100.00 Frequency Missing = 9 The FREQ Procedure Statistics for Table of cult by amputation Statistic DF Value Prob Chi Square 1 0.0551 0.8145 Likelihood Ratio Chi Square 1 0.0550 0.8145 Continuity Adj. Chi Square 1 0.0000 1.0000 Mantel Haenszel Chi Square 1 0.0542 0.8160 Phi Coefficient 0.0300 Contingency Coefficient 0.0300 Cramer 's V 0.0300 Fisher's Exact Test Cell (1,1) Frequency (F) 17 Left sided Pr <= F 0.5114 Right sided Pr >= F 0.6916 Table Probability (P) 0.2030 Two sided Pr <= P 1.0000 Effective Sample Size = 61 Frequency Missing = 9 WARNING: 13% of the data are missing.

PAGE 59

51 Appendix A (Continued) The FREQ Procedure Table of employee by amputation employee(employee) amputation(amputation) Frequency Perce nt Row Pct Col Pct 0 1 Total 0 10 17 27 16.39 27.87 44.26 37.04 62.96 35.71 51.52 1 18 16 34 29.51 26.23 55.74 52.94 47.06 64.29 48.48 Total 28 33 61 45.90 54.10 100.00 Frequency Missing = 9 The FREQ Procedure Statistics for Table of employee by amputation Statistic DF Value Prob Chi Square 1 1.5329 0.2157 Likelihood Ratio Chi Square 1 1.5431 0.2142 Continuity Adj. Chi Square 1 0.9594 0.3273 Mantel Haenszel C hi Square 1 1.5078 0.2195 Phi Coefficient 0.1585 Contingency Coefficient 0.1566 Cramer's V 0.1585 Fisher's Exact Test Cell (1,1) Frequency (F) 10 Left sided Pr <= F 0.1637 Right sided Pr >= F 0.9332 Table Probability (P) 0.0970 Two sided Pr <= P 0.3017 Effective Sample Size = 61 Frequency Missing = 9 WARNING: 13% of the data are missing.

PAGE 60

52 Appendix A (Continued) The FREQ Procedure Table of nearmiss by amputation nearmiss(nearmiss) amputation(amputation) Frequency Percent Row Pct Col Pct 0 1 Total 0 16 20 36 26.23 32.79 59.02 44.44 55.56 57.14 60.61 1 12 13 25 19.67 21.31 40.98 48.00 52.00 42.86 39.39 Total 28 33 61 45.90 54.10 100.00 Frequency Missing = 9 The FREQ Proce dure Statistics for Table of nearmiss by amputation Statistic DF Value Prob Chi Square 1 0.0751 0.7840 Likelihood Ratio Chi Square 1 0.0751 0.7841 Continuity Adj. Chi Square 1 0.0002 0.9897 Mantel Haenszel Chi Square 1 0.0739 0.7858 Phi Coefficient 0.0351 Contingency Coefficient 0.0351 Cramer's V 0.0351 Fisher's Exact Test Cell (1,1) Frequency (F) 16 Left sided Pr <= F 0.4944 Ri ght sided Pr >= F 0.7039 Table Probability (P) 0.1982 Two sided Pr <= P 0.8001 Effective Sample Size = 61 Frequency Missing = 9 WARNING: 13% of the data are missing.

PAGE 61

53 Appendix A (Continued) The FREQ Procedure Table of engineer by amputation engineer(engineer) amputation(amputation) Frequency Percent Row Pct Col Pct 0 1 Total 0 13 15 28 21.31 24.59 45.90 46.43 53.57 46.43 45.45 1 15 18 33 24.59 29.51 54.10 45.45 54.55 53.57 54.55 Tot al 28 33 61 45.90 54.10 100.00 Frequency Missing = 9 The FREQ Procedure Statistics for Table of engineer by amputation Statistic DF Value Prob Chi Square 1 0.0058 0.9394 Likelihood Ratio Chi Square 1 0.0058 0.9394 Continuity Adj. Chi Square 1 0.0000 1.0000 Mantel Haenszel Chi Square 1 0.0057 0.9399 Phi Coefficient 0.0097 Contingency Coefficient 0.0097 Cramer's V 0.0097 Fisher's Exact Test Cell (1,1) Frequency (F) 13 Left sided Pr <= F 0.6308 Right sided Pr >= F 0.5718 Table Probability (P) 0.2025 Two sided Pr <= P 1.0000 Effective Sample Size = 61 Frequency Missing = 9 WARNING: 13% of the data are missing.

PAGE 62

54 Appendix A (Continued) The FREQ Procedure SIC Cumulative Cumulative SIC Frequency Percent Frequency Percent 10 14.29 10 14.29 aff 1 1.43 11 15.71 con 6 8.57 17 24.29 finance 2 2 .86 19 27.14 hrman 5 7.14 24 34.29 man 10 14.29 34 48.57 pub 3 4.29 37 52.86 rtrade 7 10.00 44 62.86 serv 18 25.71 62 88.57 trans 3 4.29 65 92.86 w strade 5 7.14 70 100.00 F2 Cumulative Cumulative F2 Frequency Percent Frequency Pe rcent 0 62 88.57 62 88.57 1 8 11.43 70 100.00 F3 Cumulative Cumulative F3 Frequency Percent Frequency Percent 0 66 94.29 66 94.29 1 4 5.71 70 100.00 F4 Cumula tive Cumulative F4 Frequency Percent Frequency Percent 0 65 92.86 65 92.86 1 5 7.14 70 100.00

PAGE 63

5 5 Appendix A (Continued) The FREQ Procedure jsa Cumulative Cumulative jsa Frequency Percent Frequency Percent 0 24 39.34 24 39.34 1 37 60.66 61 100.00 Frequency Missing = 9 elim Cumulative Cumulative elim Frequency Percent Frequency Percent 0 34 55.74 34 5 5.74 1 27 44.26 61 100.00 Frequency Missing = 9 train Cumulati ve Cumulative train Frequency Percent Frequency Percent 0 32 52.46 32 52.46 1 29 47.54 61 100.00 Frequency Missing = 9 behave Cumulative Cumula tive behave Frequency Percent Frequency Percent 0 24 39.34 24 39.34 1 37 60.66 61 100.00 Frequency Missing = 9 oshalog Cumulative Cumulative oshalog Frequency Percent Frequency Percent 0 38 62.30 38 62.30 1 23 37.70 61 100.00 Frequency Missing = 9 maint Cumulative Cumulative maint Frequency Percent Frequency Percent 0 23 37.70 23 37.70 1 38 62.30 61 100.00 Frequency Missing = 9

PAGE 64

56 Appendix A (Continued) cult Cumulative Cumulative cult Frequency Percent Frequency Percent 0 38 62.30 38 62.30 1 23 37.70 61 100.00 Frequency Missing = 9 employee Cumulative Cumulative employee Frequency Percent Frequency Percent 0 27 44.26 27 44.26 1 34 55.74 61 100.00 Frequency Missing = 9 nearmiss Cumulative Cumulative nearmiss Frequency Percent Frequency Percent 0 36 59.02 36 59.02 1 25 40.98 61 100.00 Frequency Missing = 9 engineer Cumulative Cumulative engineer Frequency Percent Frequency Percent 0 28 45.90 28 45.90 1 33 54.10 61 100.00 Frequency Missing = 9

PAGE 65

57 Appendix A (Continued) change Cumulative Cum ulative change Frequency Percent Frequency Percent na 55 78.57 55 78.57 no 8 11.43 63 90.00 yes 7 10.00 70 100.00 freqtrain Cumulative Cumulative freqtrain Frequency Percent Frequency Percent 62 100.00 62 100.00 Frequency Missing = 8 anyform Cumulative Cumulative anyform Frequency Percent Frequency Percent 63 100.00 63 100.00 Frequency Missing = 7 committee Cumulative Cumulative committee Frequency Percent Frequency Percent 63 100.00 63 100.00 Frequency Missing = 7 other Cumulative Cumulative other Frequency Percent Frequency Percent 63 100.00 63 100.00 Frequency Missing = 7 amputation Cumulative Cumulative amp utation Frequency Percent Frequency Percent 0 35 50.00 35 50.00 1 35 50.00 70 100.00

PAGE 66

58 Appendix A (Continued) ----------------------------------------amputation=0 ---------------------------------Th e FREQ Procedure SIC Cumulative Cumulative SIC Frequency Percent Frequency Percent 4 11.43 4 11.43 con 2 5.71 6 17.14 finance 1 2.86 7 20.00 hrman 1 2.86 8 22.86 man 3 8.57 11 31.43 pub 3 8.57 14 40.00 rtrade 4 11.43 18 51.43 serv 13 37.14 31 88.57 trans 2 5.71 33 94.29 wstrade 2 5.71 35 100.00 F2 Cumulative Cumulative F2 Frequency Percent Frequency Percen t 0 35 100.00 35 100.00 F3 Cumulative Cumulative F3 Frequency Percent Frequency Percent 0 35 100.00 35 100.00 F4 Cumulative Cumulative F4 Frequency Percent Frequency Percent 0 35 100.00 35 100.00

PAGE 67

59 Appendix A (Continued) ----------------------------------------amputation=0 --------------------------------The FREQ Procedure jsa Cumulative Cumulative jsa Frequency Per cent Frequency Percent 0 11 39.29 11 39.29 1 17 60.71 28 100. 00 Frequency Missing = 7 elim Cumulative Cumulative elim Frequency Percent Frequency Percent 0 16 57.14 16 57.14 1 12 42.86 28 100.00 Frequency Missing = 7 train Cumulative Cumulative train Frequency Percent Frequency Percent 0 17 60.71 17 60.71 1 11 39.29 28 100.00 Frequency Missing = 7 behave Cumulative Cumulative behave Frequency Percent Frequency Percent 0 8 28.57 8 28.57 1 20 71.43 28 100.00 F requency Missing = 7 oshalog Cumulative Cumulative oshalog Frequency Percent Frequency Percent 0 13 46.43 13 46.43 1 15 53.57 28 100.00 Frequency Missing = 7 maint Cumulative Cumulative maint Frequency Percent Frequency Percent 0 9 32.14 9 32.14 1 19 67.86 28 100.00 Frequency Missing = 7

PAGE 68

60 Appendix A (Continued) cult Cumulative Cumulative cult Frequency Percent Frequency Percent 0 17 60.71 17 60.71 1 11 39.29 28 100.00 Frequency Missing = 7 employee Cumulative Cumulative employee Frequency Percent Frequency Percent 0 10 35.71 10 35.71 1 18 64.29 28 100.00 Frequency Missing = 7 nearmiss Cumulative Cumulative nearmiss Frequency Percent Frequency Percent 0 16 57.14 16 57.14 1 12 42.86 28 100.00 Frequency Missing = 7 engineer Cumulative Cumulative engineer Frequency Percent Frequency Percent 0 13 46.43 13 46.43 1 15 53.57 28 100.00 Frequency Missing = 7

PAGE 69

61 Appendix A (Continued) change Cumulative Cumulative change Frequency Percent Frequency Percent na 35 100.00 35 100.00 freqtrain Cumulative Cumulative freqtrain Frequen cy Percent Frequency Percent 34 100.00 34 100.00 Frequency Missing = 1 anyform Cumulative Cumulative anyform Frequency Percent Frequency Percent 35 100.00 35 100.00 committee Cumulative Cumulative committee Frequency Percent Frequency Percent 35 100.00 35 100.00 other Cumulative Cumulative other Frequency Percent Frequency Percent 35 100.00 35 100.00

PAGE 70

62 Appendix A (Continued) ----------------------------------------amputation=1 --------------------------------The FREQ Procedure SIC Cumulative Cumulative SIC Frequency Percent Frequency Percent 6 17.14 6 17.14 aff 1 2.86 7 2 0.00 con 4 11.43 11 31.43 finance 1 2.86 12 34.29 hrman 4 11.43 16 45.71 ma n 7 20.00 23 65.71 rtrade 3 8.57 26 74.29 serv 5 14.29 31 88.57 trans 1 2.86 32 91.43 wstrade 3 8.57 35 100.00 F2 Cumulative Cumulative F2 Frequency Percent Frequency Percent 0 27 77.14 27 77.14 1 8 22.86 35 100.00 F3 Cumulative Cumulative F3 Frequency Percent Frequency Percent 0 31 88.57 31 88.57 1 4 11.43 35 100.00 F4 Cumulative Cumulative F4 Frequency Percent Frequency Percent 0 30 85.71 30 85.71 1 5 14.29 35 100.00

PAGE 71

63 Appendix A (Continued) ----------------------------------------amputation=1 --------------------------------The FREQ Procedure jsa Cumulative Cumulative jsa Frequency Percent Frequency Percent 0 13 39.39 13 39.39 1 20 60.61 33 100.00 Frequency Missing = 2 elim Cumulative Cumulative elim Frequency Percent Frequency P ercent 0 18 54.55 18 54.55 1 15 45.45 33 100.00 Frequency Missing = 2 train Cumulative Cumulative train Frequency Percent Frequency Percent 0 15 45.45 15 45.45 1 18 54.55 33 100.00 Frequency Missing = 2 behave Cumulative Cumulative behave Frequency Percent Frequency Percent 0 16 48.48 16 48.48 1 17 51.52 33 100.00 Frequenc y Missing = 2 oshalog Cumulative Cumulative oshalog Frequency Percent Frequency Percent 0 25 75.76 25 75.76 1 8 24.24 33 100.00 Frequency Miss ing = 2 maint Cumulative Cumulative maint Frequency Percent Frequency Percent 0 14 42.42 14 42.42 1 19 57.58 33 100.00 Frequency Missing = 2

PAGE 72

64 Appendix A (Continued) cult Cumulative Cumulative cult Frequency Percent Frequency Percent 0 21 63.64 21 63.64 1 12 36.36 33 100.00 Frequency Missin g = 2 employee Cumulative Cumulative employee Frequency Percent Frequency Percent 0 17 51.52 17 51.52 1 16 48.48 33 100.00 Frequency Missing = 2 nearmiss Cumulative Cumulative nearmiss Frequency Percent Frequency Percent 0 20 60.61 20 60.61 1 13 39.39 33 100.00 Frequency Missing = 2 engineer Cumulative Cumulative engineer Frequency Percent Frequency Percent 0 15 45.45 15 45.45 1 18 54.55 33 100.00 Frequency Missing = 2

PAGE 73

65 A ppendix A (Continued) change Cumulative Cumulative change Frequency Percent Frequency Percent na 20 57.14 20 57.14 no 8 22.86 28 80.00 yes 7 20.00 35 100.00 freqtrain Cumulative Cumulative freqtrain Frequency Percent Frequency Percent 28 100.00 28 100.00 Frequency Missing = 7 anyfor m Cumulative Cumulative anyform Frequency Percent Frequency Percent 28 100.00 28 100.00 Frequency Missing = 7 committee Cumulative Cumulative committee Frequency Percent Frequency Percent 28 100.00 28 100.00 Frequency Missing = 7 other Cumulative Cumulative other Frequency Percent Frequency Percent 28 100.00 28 100.00 Frequency Missing = 7

PAGE 74

66 Appendix A (Continued) The LOGISTIC Procedure Model Information Data Set BOB.EVE Response Variable amputation amputation Number of Response Levels 2 Number of Observations 61 Model binary logit Optimization Technique Fisher's scoring Response Profile Ordered Total Value amputation Frequency 1 0 28 2 1 33 Probability modeled is amputation=0. NOTE: 9 observations were deleted due to missing values for the response or explanatory variables. Clas s Level Information Design Variables Class Value 1 jsa 0 1 1 1 elim 0 1 1 1 t rain 0 1 1 1 behave 0 1 1 1 oshalog 0 1 1 1 maint 0 1 1 1 cult 0 1 1 1 employee 0 1 1 1 nearmiss 0 1 1 1 engineer 0 1 1 1

PAGE 75

67 Appendix A (Continued) Model Convergence Status Convergence criterion (GCONV=1E 8) satisfied. Model Fit Statistics Intercept Intercept and Criterion Only Covariates AIC 86.154 96.122 SC 88.265 119.342 2 Log L 84.154 74.122 Testing Global Null Hypothesis: BETA=0 Test Chi Square DF Pr > ChiSq Likelihood Ratio 10.0315 10 0.4377 Score 9.4158 10 0.4931 Wald 8.2151 10 0.6078 Type III Analysis of Effects Wald Effect DF Chi Square Pr > ChiSq jsa 1 0.0538 0.8166 elim 1 0.0071 0.9327 train 1 1.8048 0.1791 Behave 1 0.7107 0.3992 oshalog 1 2.7155 0.0994 maint 1 0.7874 0.3749 cult 1 0.0028 0.9581 employee 1 1.0308 0.31 00 nearmiss 1 0.2285 0.6327 engineer 1 0.2543 0.6141 Analysis of Maximum Likelihood Estimates Standard Wald Parameter DF Estimate Error Chi Square Pr > ChiSq Intercept 1 0.2894 0.3614 0.6411 0.4233 jsa 0 1 0.0853 0.3677 0.0538 0.8166 elim 0 1 0.0288 0.3407 0.0071 0.9327 train 0 1 0.4399 0.3275 1.8048 0.1791 behave 0 1 0.2723 0.3230 0.710 7 0.3992 oshalog 0 1 0.5279 0.3204 2.7155 0.0994 maint 0 1 0.3094 0.3486 0.7874 0.3749 cult 0 1 0.0168 0.3209 0.0028 0.9581 employee 0 1 0.3228 0.3180 1.0308 0.3100 nearmiss 0 1 0.1632 0.3414 0.2285 0.6327 engineer 0 1 0.1843 0.3654 0.2543 0.614 1

PAGE 76

68 Appendix A (Continued) Odds Ratio Estimates Point 95% Wald Effect Estimate Confidence Limits jsa 0 vs 1 0.843 0.200 3.563 elim 0 vs 1 1.059 0.279 4.027 train 0 vs 1 2.411 0.668 8.701 behave 0 vs 1 0.580 0.164 2.058 oshalog 0 vs 1 0.348 0.099 1.221 maint 0 vs 1 0.539 0.137 2.113 cult 0 vs 1 1.034 0.294 3.638 employee 0 vs 1 0.524 0.151 1.824 nearmiss 0 vs 1 1.386 0.364 5.283 engineer 0 vs 1 1.446 0.345 6.055 The LOGISTIC Procedure Associat ion of Predicted Probabilities and Observed Responses Percent Concordant 70.7 Somers' D 0.425 Percent Discordant 28.1 Gamma 0.430 Percent Tied 1.2 T au a 0.215 Pairs 924 c


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001430576
003 fts
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 031007s2003 flua sbm s000|0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0000094
035
(OCoLC)52281643
9
AJL4037
b SE
SFE0000094
040
FHM
c FHM
049
FHME
090
RA425
1 100
Hanna, Eve N.
3 245
An analysis of workplace amputation injuries in Florida
h [electronic resource] /
by Eve N. Hanna.
260
[Tampa, Fla.] :
University of South Florida,
2003.
502
Thesis (M.S.P.H.)--University of South Florida, 2003.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 76 pages.
520
ABSTRACT: Nationally, 10,852 workplace amputation injuries in the US were reported in 1997. This number has been gradually decreasing over time to 9,658 amputations in 2000. In Florida, the amputation injuries have been increasing from 272 in 1997 to 417 in 2000. The reason for this increase is unclear and deserves further study. The purpose of a portion of this study was to examine the Federal and Florida BLS data to determine the characteristics of the typical worker with an amputation injury. A worker with an amputation injury was most likely to be a white male between the ages of 35 and 44 with 1 to 5 years of employment. The amputated part was usually a finger which was injured in a machinery source. The purpose of the self-designed survey was to determine if there were any differences in the amputation prevention strategies between companies with and without amputation injuries. The survey also attempted to uncover any changes made to the safety training program after an amputation injury occurred. Using the Florida workers compensation data, surveys were sent to all companies with amputation injuries for the years 1999, 2000, and 2001. An equal number of surveys were sent to companies without amputation injuries during the same years of interest. A total of 840 surveys were mailed out. 146 were returned as undeliverable. 35 surveys were returned in the amputation group and 35 surveys were returned in the control group. All SIC codes were represented except mining. In the amputation group during the years investigated, 15 out of 35 (42.9%) companies claimed amputation injuries and 20 (57%) companies did not. The true error rate for the workers compensation data was a 13% misclassification of amputation injuries into that category. Comparison of the prevention strategies between the amputation and control groups revealed no significant differences. Of the companies claiming amputation injuries, 8 out of 15 (53%) made no changes to their safety training program after the injury occurred. Proactive techniques and global changes in safety culture mindset will be necessary before major reductions in amputation injuries can occur.
590
Adviser: M.D., Stuart M. Brooks
0 650
Industrial safety
x Management.
Amputation.
653
safety training.
2
Amputation, Traumatic
prevention & control.
Accidents, Occupational
prevention & control.
Occupational Health.
prevention.
occupational.
industrial hazards.
machine safety.
690
Dissertations, Academic
z USF
Public Health
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.94