Terry L. Bunn

Disparities in work‐related injuries associated with worker compensation coverage status

BACKGROUND This exploratory study addresses patterns of injury in an emerging population of contingent workers who are not covered by either workers compensation (WC) or health insurance. The primary purpose is to improve the information base regarding the entire population of uninsured, injured workers. Because Latino workers are over-represented in the uninsured group, we include additional characterization of their patterns of injury. Recent studies have found that worker compensation claims and reports address a shrinking proportion of occupational injury and exposure, and about two-thirds of occupational injuries are not captured in the U.S. national surveillance system. METHODS Following the NEISS methodology, a work-relatedness indicator was retrieved for emergency department (ED) visits to an academic health center in fiscal year 2005. RESULTS Twenty percent of self-declared work-related injuries were not associated with self-reported WC coverage. Parametric and non-parametric statistical analysis found several significant disparities in workers without WC. These disparities included a higher proportion of Latinos, workers under age 25, and construction workers. In the uninsured group, Latino workers had a higher proportion of moderate and severe injuries. Nearly all (92 percent) workers without WC also lacked health insurance. Injured low-income workers who lack access to both WC and employer-sponsored health insurance comprise an increasing percentage of the occupationally injured. Our exploratory study found this to be particularly true in high-risk populations. CONCLUSIONS Work-relatedness indicators collected routinely in ED and outpatient settings should be incorporated into standard reporting systems to facilitate more accurate and comprehensive surveillance and better-targeted interventions.

The effect of poison control center consultation on accidental poisoning inpatient hospitalizations with preexisting medical conditions.

In 2005, the Kentucky Poison Control Center (PCC) recorded 46,625 poisoning calls; 27% received hospital treatment. Probabilistic data linkage of accidental poisoning inpatient hospital (IPH) discharge data and PCC data (years 2000–2004) was performed. This study compared IPH with/without preexisting medical conditions and IPH with/without PCC consultation, examining total length of stay and total hospitalization charges. When compared to the IPH reference group with no preexisting medical conditions and who did not consult the PCC (mean charges =

Heroin and fentanyl overdoses in Kentucky: Epidemiology and surveillance

8748, mean length of stay = 3.2 d), PCC consultation without a preexisting medical condition was significantly associated with decreased total hospitalization charges and decreased length of stay (mean charges =

Surveillance of methadone-related poisonings in Kentucky using multiple data sources

4999, mean length of stay = 1.9 d). When the patient had a preexisting medical condition, PCC consultation was still associated with decreased total hospitalization charges and length of stay (mean charges =

Motor vehicle injuries among semi truck drivers and sleeper berth passengers

8145, mean length of stay = 2.4 d) compared to those patients with a preexisting medical condition who did not consult the PCC (mean charges =

Drug Overdose Surveillance Using Hospital Discharge Data

10,607, mean length of stay = 3.6 d). These results suggest that after accounting for a patients age and gender, consultation with the PCC is significantly associated with reduced total hospitalization charges and reduced length of stay for IPH, and this association holds for patients with and without a preexisting medical condition.

Characterization of fatal occupational versus nonoccupational motor vehicle collisions in Kentucky (1998-2000)

BACKGROUND The study aims to describe recent changes in Kentuckys drug overdose trends related to increased heroin and fentanyl involvement, and to discuss future directions for improved drug overdose surveillance. METHODS The study used multiple data sources (death certificates, postmortem toxicology results, emergency department [ED] records, law enforcement drug submissions, and prescription drug monitoring records) to describe temporal, geographic, and demographic changes in drug overdoses in Kentucky. RESULTS Fentanyl- and heroin-related overdose death rates increased across all age groups from years 2011 to 2015 with the highest rates consistently among 25-34-year-olds. The majority of the heroin and fentanyl overdose decedents had histories of substantial exposures to legally acquired prescription opioids. Law enforcement drug submission data were strongly correlated with drug overdose ED and mortality data. The 2016 crude rate of heroin-related overdose ED visits was 104/100,000, a 68% increase from 2015 (62/100,000). More fentanyl-related overdose deaths were reported between October, 2015, and September, 2016, than ED visits, in striking contrast with the observed ratio of >10 to 1 heroin-related overdose ED visits to deaths. Many fatal fentanyl overdoses were associated with heroin adulterated with fentanyl; <40% of the heroin overdose ED discharge records listed procedure codes for drug screening. CONCLUSIONS The lack of routine ED drug testing likely resulted in underreporting of non-fatal overdoses involving fentanyl and other synthetic drugs. In order to inform coordinated public health and safety responses, drug overdose surveillance must move from a reactive to a proactive mode, utilizing the infrastructure for electronic health records.

Crash and burn? Vehicle, collision, and driver factors that influence motor vehicle collision fires

The methadone poisoning death rate for Kentucky in the year 2005 was the sixth highest in the US and increased 17‐fold from 1999 to 2005. The purpose of this study was to identify and characterize methadone related poisonings in Kentucky using multiple data sources: inpatient hospitalization discharge data, poison control center data, vital statistics mortality data, and Kentucky All Schedule Prescription Electronic Reporting (KASPER) data.

The delayed type hypersensitivity assay using protein and xenogeneic cell antigens.

INTRODUCTION Injuries and fatalities due to large truck and other vehicle crashes have decreased over the last decade, but motor vehicle injuries remain a leading cause of death for both the working and general populations. The present study was undertaken to determine semi truck driver and sleeper berth passenger injury risk in a moving semi truck collision using a matched-pair cohort study. METHOD Study data were obtained from the Kentucky Collision Report Analysis for Safer Highways (CRASH) electronic files for 2000-2010. A matched-pair cohort study was used to compare the odds of injury of both drivers and sleeper berth passengers within the same semi truck controlling for variables specific to the crash or the semi truck. The crude odds ratio of injury was estimated and a statistical model for a correlated outcome using generalized estimating equations was utilized. RESULTS In a moving semi truck collision, the odds for an injury were increased by 2.25 times for both semi truck drivers and sleeper berth passengers who did not use occupant safety restraints compared to semi truck drivers and sleeper berth passengers who used occupant safety restraints at the time of the collision. The driver seat or sleeper berth position in the vehicle was not a significant factor (p-value=0.31) associated with a moving semi truck collision injury. CONCLUSION Nonuse of occupant safety restraints by either drivers or sleeper berth passengers significantly increased the odds of an injury in a moving semi truck collision; semi truck seating position (drivers seat or sleeper berth) did not increase the odds for an injury in moving collisions. IMPACT ON INDUSTRY Trucking companies should include the mandatory use of occupant safety restraints by both semi truck drivers and sleeper berth passengers in their company safety policies.

The Effects of Semi Truck Driver Age and Gender and the Presence of Passengers on Collisions with Other Vehicles

Objectives. We compared three methods for identifying drug overdose cases in inpatient hospital discharge data on their ability to classify drug overdoses by intent and drug type(s) involved. Methods. We compared three International Classification of Diseases, Ninth Revision, Clinical Modification code-based case definitions using Kentucky hospital discharge data for 2000–2011. The first definition (Definition 1) was based on the external-cause-of-injury (E-code) matrix. The other two definitions were based on the Injury Surveillance Workgroup on Poisoning (ISW7) consensus recommendations for national and state poisoning surveillance using the principal diagnosis or first E-code (Definition 2) or any diagnosis/E-code (Definition 3). Results. Definition 3 identified almost 50% more drug overdose cases than did Definition 1. The increase was largely due to cases with a first-listed E-code describing a drug overdose but a principal diagnosis that was different from drug overdose (e.g., mental disorders, or respiratory or circulatory system failure). Regardless of the definition, more than 53% of the hospitalizations were self-inflicted drug overdoses; benzodiazepines were involved in about 30% of the hospitalizations. The 2011 age-adjusted drug overdose hospitalization rate in Kentucky was 146/100,000 population using Definition 3 and 107/100,000 population using Definition 1. Conclusion. The ISW7 drug overdose definition using any drug poisoning diagnosis/E-code (Definition 3) is potentially the highest sensitivity definition for counting drug overdose hospitalizations, including by intent and drug type(s) involved. As the states enact policies and plan for adequate treatment resources, standardized drug overdose definitions are critical for accurate reporting, trend analysis, policy evaluation, and state-to-state comparison.