Medical Care Utilization for Work-related Injury Health Conditions in the United States 2002-2006

On September 16, 2009, Terceira Berdahl made this presentation at the 2009 Annual Conference. Select to access the PowerPoint® presentation (894 KB) (Plugin Software Help).

Slide 1

Medical care utilization for work-related injury health conditions in the United States 2002-2006

By Terceira A. Berdahl, PhD Agency for Healthcare Research and Quality Email: terceira.berdahl@ahrq.hhs.gov
and
Marc Zodet, MS.Agency for Healthcare Research and Quality Email: marc.zodet@ahrq.hhs.gov

Slide 2

Focus of our talk

• Workplace injuries and medical care
  • Provide descriptive estimates of medical care for work injuries
  • Examine racial-ethnic/gender disparities in work injuries
  • Exploring racial-ethnic/gender disparities in health care utilization

Slide 3

Background

• 24 percent of workers experience work injuries at some point in their mid-careers.
• Work injuries are common and account for 30% of medically treated injuries in the United States.
• Of the millions of workers who are injured on the job each year, many seek healthcare.

Slide 4

Background

• Prior research on occupational health disparities finds inconsistent patterns of racial-ethnic and gender differences.
• Minority workers earn lower wages, disproportionately employed in low skilled jobs, over-represented in the most dangerous occupations.

Slide 5

Medical Care for Work Injuries

• A large body of prior research documents racial-ethnic/gender disparities in healthcare access and utilization.
• Little research on healthcare access/use disparity patterns for injured workers.

Slide 6

Current study

• We examine work injuries and healthcare utilization.
• Nationally representative survey data from the Medical Expenditure Panel Survey (2002-2006).
• Sample includes broad age ranges, occupations, and industries.

Slide 7

Research Questions: Current Study

• What kinds of healthcare do individuals use when they suffer from work-related health problems?
• Do racial-ethnic/gender disparities exist for reporting a work-injury related health problem?
• Do racial-ethnic/gender disparities exist for treatment seeking among individuals who report being bothered by a work-related injury?

Slide 8

Data and Sample

• Medical Expenditure Panel Survey (MEPS).
• National probability sample survey that is representative of the U.S. civilian non-institutionalized population.
• Workers aged 18 and older.
• To enhance the statistical power of the study we pooled cross-sectional MEPS data from years 2002-06. Thus, our findings represent the average annual odds of reporting a workplace injury and seeking treatment. We also present estimates of the total dollars spent on healthcare associated with these work-related injuries across all 5 years.

Slide 9

Outcome measures

• Descriptive
  • Medical Expenditures. Total expenditures and types of treatment.
• Modeling
  • Workplace Injury. A dichotomous measure of having any workplace accident/injury related condition in the past year.
  • Healthcare Utilization. A dichotomous measure that is coded "1" if the person had any office-based, outpatient, emergency department, or inpatient services for a work injury and coded "0" if otherwise.

Slide 10

Independent Variables

• Race-ethnicity/sex subgroups. Race, ethnicity, and sex were used to create eight analysis subpopulations: White non-Hispanic men; black non-Hispanic men; other non-Hispanic men; Hispanic men; white women; black women; other race women; Hispanic women.
• Age. Four age categories: 18-24 years, 25-44 years, 45-64 years, and 65+ years.
• Education. Three categories: less than high school, high school, or more than high school.
• Occupation & Industry. Occupation consists of US Census 9-category occupation codes and 14-category industry codes.
• Insurance Status. Individuals are coded as having private insurance, public insurance, or no health insurance.

Slide 11

Analysis

• We perform descriptive and multivariate statistical analysis using STATA 10.0.
• Descriptive statistics on healthcare use, types of use, total expenditures for work injuries.
• Two sets of logistic regression models predicting
  • 1) the odds of work injury
  • 2) the odds of seeking treatment at selected points of service
• All estimates are weighted using the appropriate year sample weight and all estimates include standard errors adjusted for complex survey design.

Slide 12

Descriptive Findings

• Work related injuries comprised approximately 17% of all injury related medical conditions.
• Of workers aged 18 and over, 6% reported being bothered by a work-related health problem.
• Individuals with work injury-related conditions spend $1,843 on average per year for those medical expenses.

Slide 13

Descriptive findings.

• During the years 2002-2006, a total of $85.9 billion were spent on office-based, outpatient, emergency department, and inpatient services for workplace injuries.
• Among people with work injuries, nearly 40% of their total healthcare expenses were for work-related conditions.

Slide 14

What kinds of healthcare do individuals use when they suffer from work-related health problems?

Slide 15

Multivariate Findings

• Logistic regression models predicting
  • The odds of reporting a work injury
  • The odds of seeking treatment among those with work injuries

Slide 16

Do racial-ethnic/gender disparities exist for reporting a work injury related health problem?

Figure 1: Predicted probability of injury in adjusted models (models control for age, education, occupation, and industry).
 

Slide 17

Variables associated with work injury

• Education: higher educated workers had lower risk.
• Age: middle age categories are significantly more likely to be injured compared to young adults (age 18-24).
• Occupation: decreased risk for manager, services, sales and professional, increased risk for construction.
• Industry: increased risk for natural resources, construction, decreased risk for finance, professional and service work.

Slide 18

Do racial-ethnic/gender disparities exist for treatment seeking among individuals who report being bothered by a work-related injury?

Figure 2: Predicted probability of seeking treatment adjusted models (models controlling for age, education, occupation, and industry).
 

Slide 19

Variables associated with healthcare use

• Age: the odds of seeking treatment oldest workers (age +65) were 46% lower compared to young adult workers (age 18-24) to seek treatment.
• Education: had no significant association with treatment seeking
• Occupation: had no significant association with treatment seeking
• Industry: had no significant association with treatment seeking
• Health Insurance: The odds of seeking treatment for uninsured workers were 33% lower compared to privately insured workers. We found no statistically significant difference between publicly and privately insured workers.

Slide 20

Discussion

• We find some evidence of racial-ethnic/gender disparities in reporting work-related health conditions.
  • White men have greatest risk of injury.
  • • Black and Hispanic men had significantly lower risk.
    • Women had significantly lower risk.

Slide 21

Discussion

• Health care use for workplace injury conditions.
  • Less evidence of racial-ethnic gaps.
  • • White women more likely to see a doctor compared to white men.
  • Older workers with work injuries less likely to seek treatment.
  • Uninsured workers do not seek treatment at comparable rates-this may be an indicator of health care access barriers for workers without health insurance.

Slide 22

Conclusion

• Work injuries are costly to the health care system.
• Most workers with work-related injuries seek some form of healthcare treatment.
• Our utilization findings are exploratory first step. They suggest that access disparities may influence how people seek treatment for work-injury health problems.
• Given persistent disparities in employment outcomes and healthcare access, it is important to examine how racial-ethnic/gender differences shape work injury risk and healthcare use.