1. UNION PACIFIC’S FATIGUE RISK MANAGEMENT SYSTEM (FRMS) Dennis W. Holland, Ph.D. Jackie Keenan, M.B. A.

3. Fatigue Risk Management System Model Level One (L1) Level Two (L2) Continuous Improvement Process FAID Analysis Measurement Behavioral Symptoms Screening Tools Peer Identification Corporate Responsibility Ensuring Adequate Sleep Opportunity FAID Analysis /Action Plans Concept Taken From “Managing The Risks Of Organizational Accidents” by James Reason Level Three (L3) Critical Incident!! Level Four (L4) Individual Responsibility Using Time off for Rest

4. UNION PACIFIC’S FRMS Broad comprehensive plan to manage the human resource. Integrated and Interdisciplinary approach to managing manpower, fatigue and quality of life issues. Safety & Health Priorities. Evidence-based – Theory and research

5. FRMS Risk Management Model  Identify, measure and prioritize risk and develop and implement controls Addresses Critical Challenges  24/7 Operations & 24/7 Society  Unsupervised TE&Y workforce  Aging/new workforce  Unhealthy workforce Scientifically based “toolbox” approach

6. KEY ELEMENTS OF FRMS Policy  Corporate & Local Policies Training and Education Ensuring adequate average sleep opportunity  Company responsibility  Software analysis & Measurement (FAID)

7. KEY ELEMENTS OF FRMS Ensuring employee preparedness: ensure that individuals who received an adequate average sleep opportunity have achieved sufficient sleep to ensure safe level of alertness [SLA]  Education on signs & symptoms  Education & policy regarding minimal sleep and reporting  Shared Responsibility partnerships with employee, labor organizations and government

8. KEY ELEMENTS OF FRMS Additional Countermeasures (Toolbox)  Planned Nap Program  Lodging  Families  Alternative to Discipline  Sleep Disorder Screening  Measurements  Technology Review Process Research - ongoing process of pertinent research to ensure FRMS is evidence-based.

9. IMPLEMENTING FRMS UNDERSTANDING THE FRMS LEVELS OF CONTROL

10. Fatigue Risk Management System Model

11. Qualitative Level One Controls 10-hour Undisturbed Rest/AM Mark-Up  Completed Proactive Notification  Completed  Ongoing Enhancement Work/Rest Cycles & Call Windows  Magic bullet mentality not justified -- Comprehensive and integrated process  More than just Crew Scheduling  Example - Seattle  Example - North Platte/Marysville

12. Quantitative Level One Controls FAID Process  FAID Analysis -- Software package to identify the extent of any fatigue problems (i.e. locations or times of reduced average sleep opportunity)  System identification of where reduced sleep opportunity may lead to work-related fatigue  Assess whether interventions have increased sleep opportunity  Used in conjunction with the Board Game software for simulating the impact of changes at the board/pool level

13. Fatigue Scores are Indicators Only Fatigue scores are indicators only of the impact of sleep deprivation. They are based on a statistical analysis of research performed into fatigue levels over a broad sample of population and provide guidance on the fatigue of an individual.

14. Level Two Controls Level 2 controls are designed to ensure that individuals who received an adequate average sleep opportunity have achieved sufficient sleep to ensure safe level of alertness [SLA] This requires a high level of shared responsibility This will require significant cultural change and time and is likely to be the most difficult element to implement effectively Labor will play an important role in shaping the employee attitudes. Labor/management consultation will be essential Implementation: Employee, Management and Family education and awareness.

15. Level Two Implementation Plan Stage 1 - Initial Awareness  Minimum Sleep/Wake Needs - 5/12 rule  Commenced -- Q1 2004 Stage 2 - Personal Management tool  Introduce Prior Sleep/Wake Model and behavioral scale  Used to manage personal behavior and ensure SLA  No reporting requirement  Proposed Completion Stage 3 - Organizational Integration  Introduce Scoring System  Reporting of non-zero scores  Develop local decision matrices  Mark off

16. Additional Level Two Controls Proposed Industry collection of US Sleep/Wake Data  Ensure Level One Models are valid  Link L1 -> L2 -> L3 frequencies to ensure evidence-based policy  Establish industry benchmarks/policy guidelines for Sleep/Wake behavior Research projects  Projects collecting Work/Rest, Sleep/Wake, Behavioral data  Kansas City [WR,SW, Actigraphy, PVT]  Des Moines/St. Paul [WR,SW, PVT]  San Antonio [Diaries, Actigraphy]

17. Level Three Controls These reflect a general organizational shift in philosophy and policy  Increased focus on employee involvement in fatigue hazard identification and self-management of fatigue  i.e. self and peer education leading to self assessment  Clear policy guidelines on managing pre-defined unacceptable levels of fatigue-related behavior)  Example: Alternative to Discipline Discipline vs. Peer Intervention

18. Level Three Controls Physiological Monitoring Systems  Ongoing internal technological review process  No current systems with acceptable Cost/Benefit Analysis Self and Peer Identification of Fatigue-Related Behaviors  Generic symptom checklists (developed)  Task-specific symptom checklists Sleep Disorders Screening  Certification process  FRA/UP Assessment Research Project  Voluntary Program (OHNS)  Education and Awareness  Related Health Issues Overall Research Program

19. Level Four & Five Controls Identifying fatigue as a cause of errors or incidents  Evidence of L1-3 data consistent with fatigue  Nature of incident is consistent with fatigue-related error Needs to be an industry initiative to ensure level playing field  UP willing to collaborate with industry initiatives Model and Boardgame Analysis  CIT

21. FATIGUE, SLEEP DISORDERS AND RELATED HEALTH ISSUES UNION PACIFIC RESEARCH UPDATE

22. RESEARCH AGENDA Assess the relationship between safety and risk factors such as: fatigue, sleep disorders, stress, depression, obesity, etc. Part of UP FRMS Create evidence-based interventions Partnerships with Labor and Regulators Bottom-line Implications: Improve employee health and safety while reducing medical costs, absenteeism and presenteeism.

23. INTRODUCTION Behavioral Health: 2003 WHO study (Wang, et al) suggests that depression most highly significant factor in accident and injury. Depression expected to become #1 cause of disability world-wide. Wellness Inventory (Pfizer) UP studies indicate behavioral issues (depression, stress, anxiety) are most costly in presenteeism, absenteeism and medical costs. Fatigue: Sleep Deprivation & Sleep Disorders are a societal issue impacting the work place: NSF reports 74% of Americans do not get enough sleep each night.

24. FATIGUE & SLEEP DISORDERS Sleep Disorders - Uservices – Objectives: Assess percentage of TE&Y at risk for excessive daytime sleepiness Demonstrate viability of sleep assessment as part of re- certification process – Methods N= 437 Epworth Sleepiness Scale Scores > 10 categorized at risk

25. SLEEP (cont’d) – Results At risk employee is > 35% (95% CI 35.4-44.8%) Willingness to participate (confidentiality maintained) Confirmation Study – Scores > 10 offered opportunity to wear a device to determine presence of OSA – 20 tests ordered, 10 completed – 9 confirmed positive for OSA Recommendations – Additional research – Continued voluntary screening – Continued education and awareness

26. SLEEP (cont’d) Actigraph - DU  Objectives  Determine fatigue levels  Behavioral implications of feedback  Data for biomathematical validation  Methods  Sleep Diaries/Self Report  Two Types of Actigraphs  Epworth Sleepiness Scale & PSQI  N (Questionaire) = 180 (60.4% response rate)  N (Actigraph) = 36

27. SLEEP (cont’d) – Results Lower mean sleepiness score than other sites (match approximate national average) Behavioral outcomes (actigraph): – No statistical difference between those with feedback actigraph and those with non-performance actigraph – Those with self-efficacy more likely to use data to change behavior – Those without self-efficacy (external locus of control) unwilling to change behaviors Implications for changing behaviors and providing concrete change

28. Behavioral Health Joint Projects and Funding UPRR/FRA  Coping Mechanisms – HIS  Depression – DU  Obesity – HIS  Integrated Health Improvement Initiative - HIS BAA – FRA Additional proposals currently under consideration.

29. CONCLUSION Pilots assessed for system implementation Evidence-based development of programs and interventions to reduce or eliminate risks Informed discussions, communications and programs Implications for the industry Safety, health and productivity focus Additional safety research needed in all associated areas.