Repeat workers compensation claims A prospective analysis in Ontario, Canada and Victoria, Australia Etches J 1, Mustard CA 1,2, Ruseckaite R 3 and Collie A 3,4 1. Institute for Work & Health, Toronto, Canada 2. Dalla Lana School of Public Health, University of Toronto 3. Institute for Safety, Compensation and Recovery Research, Melbourne 4. School of Public Health and Preventive Medicine, Monash University, Melbourne RAACWI Forum, Toronto, March 30, 2012
Conclusions A large fraction of lost-time claims are repeat claims Risk of second claim varies by industry, occupation as for first claims Risk of second claim varies little by nature of injury A large fraction of repeat claims are with the same employer A large fraction of repeat claims are for a different nature of injury Risk of repeat claim depends on age & duration of previous claim The pattern of repeat claims in Ontario, Victoria and Alberta are similar Long-duration claims have lower re-injury risk Long-duration claims followed by re-injury had lower wage replacement and medical costs 2
Introduction WC claims often assumed independent events Alberta Claim = LT + NLT, injury + disease 54.0% of men filed second claim within 0-10 years 37.9% of women filed second claim within 0-10 years 67.6% of 1995 claimants filed second claim by end of 2004 241,000 second claims (not re-opens) Victoria First claim = LT (>10 days) + NLT (>$500), injury only 40.5% of men filed second claim within 9-14 years 31.3% of women filed second claim within 9-14 years 63,000 second claims (often re-opens) 3
Adj. HR of second c 1.4 1.3 1.2 Risk of second claim by industry Manufacturin Transportation Wholesale/retail Municipal/educ/health Construction 1.1 1.0 Agriculture/forestry Mining/petroleum Provincial gov Business/personal/prof services 0 10000 20000 30000 40000 50000 Number of first claims 4
Probability of secon 0.50 0.45 0.40 0.35 0.30 0.25 0.20 Risk of second claim by industry Other Manufactu Mining Transport/postal Utilities Construction Administrative Heatlh care/social assistance Wholesale Media/telecomm Arts/recreation Rental/real Public administration estate Education Missing Agr/Forestry/Fishing Retail Accommodation/food Professional Financial 0 10000 20000 30000 40000 Number of first claims 5
Adj. HR of second c 2.5 2.0 1.5 1.0 Processing Forestry/logging Social sciences Artistic Farming Crafts/equipment literary Mining Teaching Natural science Managers Food/bevera Machining Medicine/health Material handling Transport Risk of second claim by occupat Sales Fabrication Clerical Construction Service 0 10000 20000 30000 40000 50000 Number of first claims 6
Probability of secon 0.40 0.35 0.30 0.25 Risk of second claim by occupat Community Machinery service Technicians Labo Professionals Missing Managers Sales Clerical 10000 20000 30000 40000 50000 Number of first claims 7
Adj. HR of second c 0.6 Risk of second claim by injury ty 1.0 Other Surface wounds Sprains/s 0.8 Symptoms Traumatic Burns Infectious Systemic disesase Multiple Fractures/dislocations Open wounds 0.4 0.2 Neoplasms 0 50000 100000 150000 Number of first claims 8
Probability of secon 0.40 0.38 0.36 0.34 Intracranial Burns Other Risk of second claim by affliction Wounds Traumatic so 0.32 0.30 0.28 Nerves Fractures 0 20000 40000 60000 80000 Number of first claims 9
Adj. HR of second c Risk of second claim by accident 1.0 Assaults/violence Contact (obje 0.9 0.8 Substances Bodily reaction/e Falls 0.7 Transportation Fires/explosions 0.6 0 20000 40000 60000 80000 100000 120000 14000 Number of first claims 10
Adj. HR of second c 1.7 1.6 1.5 1.4 1.3 1.2 1.1 Risk of second claim by age at fi 18-24 25-3 35-44 45-54 1.0 55-65 40000 60000 80000 100000 120000 Number of first claims 11
First and second claims of repeat claimants (Victoria) Different affliction + different bodily location: 45.5% Same industry: 74.8% Same occupation: 45.6% Same employer: 65.6% Dates of changes unknown 12
Claim costs and duration (Victoria) Impact Single First Second Mean Std Mean Std Mean Std Total cost $16,731 $87,011 $8,813 $50,938 $13,147 $65,100 Total wage replacement $5,249 $30,877 $2,631 $18,501 $4,875 $27,204 Total medical $3,447 $29,128 $2,034 $9,996 $2,886 $13,025 Days lost 67 278 42 196 66 244 Why are the durations and costs so much less for the first claims of repeat claimants? 13
Ontario All lost-time claims 2000-01-01 through 2009-12-31 Injury + disease Design 1: 5 year inception cohort 2000-2004, 5 year follow-up from first accident date Similar to Victoria study: Ruseckaite & Collie (2011) 23.6% of men filed second claim within 5 years 21.0% of women filed second claim within 5 years 96,000 second claims (mostly not re-opens) Mostly trivial difference between post-accident date and post-benefit definition of 5 year follow-up 14
Probability of secon 0.30 0.25 0.20 0.15 Risk of second claim by industry Communication/Utility Missing (Schedule 2) Government Transportation/Storage Fishing/Trapping Health/Social Service Wholesale Trade Manufactu Educational Retail Trade Other Service Real Estate Agriculture Business Accomm/Food/Beverage Logging/ForestryConstruction Mining Finance/Insurance 0 20000 40000 60000 80000 100000 120000 Number of first claims 15
Probability of secon 0.30 0.25 0.20 Health Bus/Fin/Admin Risk of second claim by occupat Processing/Manuf/Utilities Sales, Service Trades/Transp/Operato 0.15 Natural/Applied Sci Management Social Sci/Educ/Gov/Re Primary Industry 0.10 Art/Culture/R 0 20000 40000 60000 80000 100000 120000 14000 Number of first claims 16
Probability of secon 0.24 0.22 0.20 0.18 Burns Risk of second claim by affliction Musculoskele Traumatic soft tissue Other Wounds 0.16 Fracture Missing 0 50000 100000 150000 20000 Number of first claims 17
First and second claims of repeat claimants (Ontario) Ontario Victoria Different nature + different bodily location 47.1% 45.5% Same industry 83.9% 74.8% Same occupation 69.7% 45.6% Same employer 71.3% 65.6% Dates of changes unknown Ontario: 5 year follow-up Victoria: 9-14 year follow-up 18
19
Measures of agreement between Part of Body codes on first and second claims (Ontario) Kappa: 0.11 (0.10,0.11) Test Chi-square DF p Independence 4095 36 < 0.0001 Quasi-independence 285 29 < 0.0001 Symmetry 241 21 < 0.0001 Quasi-symmetry 34 15 0.003 Marginal homogeneity 207 6 < 0.0001 Number of claim pairs = 96,000 20
Measures of agreement between Nature of Injury codes on first and second claims (Ontario) Kappa: 0.11 (0.10,0.11) Test Chi-square DF P Independence 3825 25 < 0.0001 Quasi-independence 313 19 < 0.0001 Symmetry 523 15 < 0.0001 Quasi-symmetry 34 10 0.0002 Marginal homogeneity 489 5 < 0.0001 Number of claim pairs = 96,000 21
Claim costs and duration (Ontario) Impact Years* Single First Second Mean Std Mean Std Mean Std Wage replacement ($) 5 8,523 26,401 2,641 9,821 7,499 23,628 1 3,164 6,373 1,770 3,794 3,127 6,352 Medical benefits ($) 5 2,125 16,893 715 2,286 1,810 10,003 1 1,171 6,411 545 1,424 1,081 4,843 Days of benefits (days) 5 176 452 73 247 150 397 1 70 117 43 85 66 113 * Dollar amounts accrued during the first year or first 5 years of the claim. 22
Why are first claims far shorter for repeat claimants? 1. Immortal time bias No, bias is small 2. Labour-force exit Maybe, we don t have the data to test this 3. Effect of prior claim history No, Victoria saw the same pattern using true first claims 4. Long-duration claims have lower repeat claim risk Yes, costs and durations for claims < 365 days similar for repeat and non-repeat claimants Labour force exit? Changed occupation/task? 23
First year claim costs and duration (Ontario) Single First Median Mean Std Median Mean Std Duration < 365 Benefit days 8 33.6 59.1 8 29.7 53.5 Wage replacement 418 1,611 3,136 413 1,431 2,775 Medical 143 562 2,071 138 443 1,023 Duration > 365 Benefit days 365 365 0 365 365 0 Wage replacement 15,380 15,964 10,607 8,335 11,074 9,901 Medical 3,176 6,249 18,049 2,223 3,227 4,609 Probability of second claim within 5 years: Duration < 365 = 0.23 Duration > 365 = 0.12 24
Ontario Design 2: 10 years of time-to-event data Two alternating time-to-event processes Time-to-RTW Time-to-next-claim Unmeasured competing risk: insured labour market exit School Home keeping Non-WC disability benefit Retirement Self-employment Emmigration Similar to Alberta study: Cherry et al. (2011) 25
26
27
28
Hazard ratio relative to manufacturing 2.0 1.5 1.0 0.5 Accommodation/Food/Be Retail Trade Risk of lost-time claim by industry of previous claim and Construction Health/Social Service Transportation/Storage Wholesale Trade 2.0 1.5 1.0 0.5 1 2 3 4 1 2 3 4 1 2 3 4 Number of previous claims 29
Hazard ratio relative to sales and service 1.6 1.4 1.2 1.0 0.8 1.6 1.4 1.2 1.0 0.8 Risk of lost-time claim by occupation of previous claim a Business, Finance, Administration Health Management Processing, Manufacturing, Utilities 1.6 1.4 1.2 1.0 0.8 Social Science, Education, Governm Trades, Transport, Equipment Operat 1 2 3 4 1 2 3 4 Number of previous claims 30
Conclusions A large fraction of lost-time claims are repeat claims Risk of second claim varies by industry, occupation as for first claims Risk of second claim varies little by nature of injury A large fraction of repeat claims are with the same employer A large fraction of repeat claims are for a different nature of injury Risk of repeat claim depends on age & duration of previous claim The pattern of repeat claims in Ontario, Victoria and Alberta are similar Long-duration claims have lower re-injury risk Long-duration claims followed by re-injury had lower wage replacement and medical costs 31
Jacob Etches jetches@iwh.on.ca
33