Does Size Count? Incidence and Reporting of Occupational Disease by Size of Company

1
Does Size Count? Incidence and Reporting of
Occupational Disease by Size of Company

• Tim Morse, Ph.D.
• ErgoCenter
• UConn Health Center

2
Collaborators

• Charles Dillon, NHANES
• Joseph Weber, CT Labor Dept.
• Nick Warren, UCHC
• Heather Bruneau, UCHC
• Rongwei Fu, UCHC

3
NIOSH/OSHA Report higher rates for larger
companies
4
Reasons for Correlation?

• Increased risk vs. better reporting
• Increased risk?
• Biersner and Winn, 1998
• More repetition in larger companies?
• Connected to industry segment or other co-variate
  such as worker age?
• Better reporting?
• Oleinick, et al. 1995
• MSD is under-reported
• Occupational disease is primarily MSD
• Better recordkeeping?
• Less fear of reporting?

5
Why do we care?

• How do you best target industries?
• Grants for small employer training
• Prioritize OSHA inspections
• Other policy issues
• Recordkeeping by small employers
• What is source of problem?
• Repetition, stress, other risk factors
• Need for and focus of internal training for
  companies
• Under-estimate of Occupational Disease if
  under-reporting

6
Under-reporting

• CUSP (CT Upper-Extremity Surveillance Project)
  Data
• 9.1 of population with likely work-related
  prevalent MSD
• 0.78 (95 CI 0.58-1.24) doctor-called incident
  cases
• 10.6-21.0 had filed workers compensation claims

7
Correlates of under-reporting (CUSP)

• Severity of MSD
• Surgery (OR 3.5)
• Time off work (OR 4.5)
• Doctor diagnosis (OR 13.7)
• Psycho-social factors
• Management cares (OR 2.0)
• Fear of reporting
• Union (OR 4.1)
• Industry/Occupation
• Manufacturing, transport, trade higher
• Hourly wage workers (OR 2.8)

8
Population-based telephone survey (CUSP)

• Random sample of 3,200 CT workers
• 78 interview response rate
• with likely work-related MSD
• of cases reported to workers comp
• Compare to BLS MSD figures by size of company
• Size of company coded by CT Labor Dept
  additional coding by InfoUSA

9
Statistical methods

• Data reduction of risk factors by factor analysis
• Tabular analysis of MSD by size of company
• Partial correlations and Logistic regression

10
ConnOSHA/BLS Survey

• Connecticut, 1996
• Repetitive Trauma
• 61.6 of occupational illnesses
• 3.6 of all injuries and illnesses
• 3,711 cases of repetitive trauma
• 28.8 per 10,000 workers

11
CT BLS Repetitive trauma rates also increase by
size of business
12
Results Coding for Size

• Only 64 of respondents could be coded for size
• No major differences between coded and uncoded
  for gender, age, and ethnicity
• Minor differences in education
• 33 (uncoded) vs. 27 (coded) High school grad
• 13 vs 20 for post-graduate
• Differences in industry
• government (5.2 uncoded vs. 20.1 coded)
• service (60.2 vs. 50.7)
• construction (8.1 vs. 4.1)

13
Demographic characteristics by company size

• No difference in gender distribution
• Higher education in larger companies
• chi-square110.3, siglt.001
• Blacks and Hispanics over-represented in larger
  companies
• chi-square39.6, sig.006
• Older workers in very large and very small
  companies
• chi-square72.7, siglt.001

14
Risk Factors

• Factor analysis
• Physical risk factor (push/pull,reach above,
  wrist bent, tool use)
• Stress/computer factor (job stress, computer use)
• Correlations with business size
• physical risk factor (r -.14)
• stress/computer factor (.14)

15
Partial correlations

• Controlling for gender, race, marriage, age, and
  education.
• Physical risk factor and Business size
• -.078 (p.001)
• Stress/computer risk factor and business size
• .120 correlation (plt.001)

16
MSD Prevalence by Company Size, CUSP, CT, 1996
17
MSD by Company Size, CUSP, CT, 1996Chi-square9.4
, sig.052
18
(No Transcript)
19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
(No Transcript)
23
Physical risk by MSD prevalence, by firm size,
CUSP, CT, 1996
24
Logistic Regression

• MSD case on Size
• OR0.91
• CI 0.83-1.00
• Doctor called MSD on Size
• OR0.88
• CI 0.78-0.99

25
Logistic Regression

• Entered Company size, gender, age, industry,
  occupation, married, race
• Backward conditional regression

26
Logistic regression

• MSD
• Stay in equation Gender, age, race, occupation
• Size marginally significant (OR0.90 0.81-1.00)
• Larger companies have lower rates
• Doctor called
• Stay in regression occupation, gender, race
• Size not significantly related to MSD

27
Cautions and limitations

• Self-reported data
• Prevalence, not incidence
• Just MSD
• Only 64 could be coded for size
• Likely that sample under-represented smaller
  companies
• Demographics similar between coded and uncoded
• Not likely to systematically affect rate of MSD
  by size

28
Conclusions

• Business size is only weakly related to MSD, in
  negative direction (in contrast to BLS reports)
• Risk factors vary somewhat by size largest
  companies have
• Lowest physical risks,
• Highest stress and computer risks

29
Under-reporting

• Strong positive correlation in BLS reports
  between MSD and company size most likely due to
  better reporting in larger companies
• Appears to be large under-reporting for smaller
  companies