Workers’ compensation claim counts and rates by injury event/exposure among state-insured private employers in Ohio, 2007-2017

Abstract

Introduction:

This study analyzed workers’ compensation (WC) claims among private employers insured by the Ohio state-based WC carrier to identity high-risk industries by detailed cause of injury.

Methods:

A machine learning algorithm was used to code each claim by US Bureau of Labor Statistics (BLS) event/exposure. The codes assigned to lost-time (LT) claims with lower algorithm probabilities of accurate classification or those LT claims with high costs were manually reviewed. WC data were linked with the state’s unemployment insurance (UI) data to identify the employer’s industry and number of employees. BLS data on hours worked per employee were used to estimate full-time equivalents (FTE) and calculate rates of WC claims per 100 FTE.

Results:

140,780 LT claims and 633,373 medical-only claims were analyzed. Although counts and rates of LT WC claims declined from 2007-2017, the shares of leading LT injury event/exposures remained largely unchanged. LT claims due to Overexertion and Bodily Reaction (33.0%) were most common, followed by Falls, Slips, and Trips (31.4%), Contact with Objects and Equipment (22.5%), Transportation Incidents (7.0%), Exposure to Harmful Substances or Environments (2.8%), Violence and Other Injuries by Persons or Animals (2.5%), and Fires and Explosions (0.4%). These findings are consistent with other reported data. The proportions of injury event/exposures varied by industry, and high-risk industries were identified.

Conclusions:

Injuries have been reduced, but prevention challenges remain in certain industries. Available evidence on intervention effectiveness was summarized and mapped to the analysis results to demonstrate how the results can guide prevention efforts.

Practical Applications:

Employers, safety/health practitioners, researchers, WC insurers and bureaus can use these data and machine learning methods to understand industry differences in the level and mix of risks, as well as industry trends, and to tailor safety, health, and disability prevention services and research.

1: Introduction

Workers’ compensation (WC) in the United States (US) involves large, state-governed, administrative data collection systems. WC claims data includes coded fields for the injured worker’s occupation/industry, injury cause, part of body, and nature of injury, as well as unstructured narratives that describe how the incident occurred and the worker’s occupation. States use a variety of coding systems, and codes for each claim are manually generated by a combination of employers, claims administrators, insurers and/or state WC bureaus based on free text descriptions of how the incident occurred, diagnoses, and other claims information.

WC systems have been successfully utilized by several states for occupational sentinel surveillance, research, and employer worksite follow-up. Recent studies demonstrated that large state datasets of WC claims can be successfully linked to state employment data to examine claim counts and rates by industry and cause of injury.^1–6 Several of these studies were developed through cooperative agreements funded by the National Institute for Occupational Safety and Health (NIOSH) awarded to the states of California, Massachusetts, Michigan, Ohio, and Tennessee.

Although there are challenges with WC claim data quality and completeness, other studies have determined that machine learning techniques^7–13 can be successfully applied to the incident narratives in WC claims and other similar data to gather prevention insights. For example, Liberty Mutual uses such techniques to publish their popular Safety Index.¹⁴ The US Bureau of Labor Statistics (BLS) also uses these approaches to analyze data from the Survey of Occupational Injuries and Illnesses (SOII). Researchers from NIOSH and the Ohio Bureau of Workers’ Compensation (OHBWC) applied these techniques to identify ergonomic and safety priorities within many specific industries.¹⁵ Machine learning approaches are constantly evolving, and NIOSH recently sponsored a crowd-sourcing competition to improve methods for coding free text narratives in WC claims and other similar data.¹⁶ NIOSH and other partners have shared machine learning programs with various WC bureaus, insurers and employers, but these approaches are not yet in widespread use in WC systems. The current analyses fill a gap in the literature to demonstrate how the next iteration of machine learning methods can enable the coding of claims to a more detailed causation level in large WC datasets to encourage wider use.

The main purpose of this study was to analyze patterns in OHBWC-insured private employer WC claims to identify specific high-risk industries by detailed cause of injury. Another purpose was to summarize the leading examples of safety/health programs and interventions by cause types for higher-risk industries identified in the data. This demonstrated the opportunities for use and potential impact of the detailed analysis of claims data.

2: Methods and Materials

The OHBWC is the largest state-run WC system in the US. From 2007-2017, the OHBWC provided WC insurance for two-thirds of Ohio’s workers. The remaining one-third of workers were either employed by a small number of larger employers (usually > 500 employees) authorized to self-insure or worked at other employers (e.g. sole proprietorships) exempt from OHBWC coverage. Employers can self-insure if they demonstrate the following: strong financial stability, an organizational plan for the administration of the WC law, and risk and claims management procedures for establishing a safe and more cost-effective workplace.¹⁷ Sole proprietorships are exempt from WC coverage but can electively purchase policies.¹⁸ In 2017, 54% of Ohio employers with more than 500 employees were estimated to be self-insured (692 employers), and 38% of the sole proprietorships did not purchase OHBWC coverage.

WC claims for 2007-2017 from private employers insured by the OHBWC were analyzed by industry and cause of injury. The dataset included accepted lost-time (LT) claims (8 or more days away from work, DAW) and medical-only (MO) claims (medical care only and/or 7 or fewer days DAW). From 2007-2011, OHBWC manually coded only LT claims using a proprietary cause coding system and began using the ICD- External cause code system for all claims in 2013. Because the codes were incomplete and used different systems, a NIOSH-developed machine learning algorithm⁸ was used to auto-code each OHBWC claim to the US BLS Occupational Injury and Illness Classification System (OIICS version 2.01) 2-digit event/exposure categories.¹⁹ The OIICS system was chosen because it includes an extensive coding hierarchy, instructions, and training systems and is widely recognized by employers and public health users because the BLS SOII uses this system. The machine learning algorithm was generated based on computer analysis of word frequencies and combinations in a “training set” of 7,200 WC claims that were manually coded by experts.

The algorithm assigned a probability of accuracy score to each of its code assignments. To increase the accuracy for coding the most severe injuries, LT claims with bottom quartile algorithm probabilities or high costs (95^th percentile) were manually reviewed. Because the algorithm relies on observed probabilities of certain words occurring in the training datasets, and workplace violence claims were relatively rare in these training sets, claims with relatively high workplace violence probabilities and claims in industries (NAICS 11, 71, and 92) with lower than expected number of claims coded as violence were also manually reviewed. MO claims were not manually reviewed due to resource constraints.

Primarily, 1-digit event/exposure level (a.k.a. Divisions in BLS terms) counts/rates are presented in this paper. This was done to allow the results of this study to be readily compared to recent data from other state WC systems. This 1-digit codes include 1: Violence and other Injuries by Persons or Animals; 2: Transportation Incidents; 3: Fires and Explosions; 4: Falls, Slips, Trips; 5: Exposure to Harmful Substances or Environments; 6: Contact with Objects and Equipment; and 7: Overexertion and Bodily Reaction, and 9: Non-classifiable.

OHBWC claims data were linked via the employer’s federal tax identification number with Ohio’s unemployment insurance (UI) data to identify the employer’s industry code (North American Industry Classification System, NAICS) and number of employees. BLS National Labor Productivity and Costs (LPC) data²⁰ on hours worked per employee by 4-digit NAICS industry were used to adjust UI data and estimate full-time equivalents (FTE). Rates of WC claims per 100 FTE were then calculated. Methods differed slightly for single-location versus multiple-location employers, and these methods were previously developed and described in more detail in a prior study.⁵ Methods in the current paper differ slightly in that BLS LPC data now represent hours worked rather than hours paid.²⁰ Industry sectors were defined by NIOSH National Occupational Research Agenda (NORA) classifications.²¹

NAICS codes for each National Occupational Research Agenda (NORA) sector are as follows: Agriculture, Forestry, Fishing/Hunting (AFF) = 11; Construction (CON) = 23; Health Care and Social Assistance (HCSA) = 62, 54194, 81291; Manufacturing (MNF) = 31, 32, 33; Mining (MIN) = 21; Oil & Gas (OIL) = 211, 213111, & 213112; Public Safety (PSA) = 92212, 92214, 92216, 62191; Services, except Public Safety (SRV) = 51–56, 61, 71, 72, 81, 92; Transportation, Warehousing, Utilities (TWU) = 22, 48, 49; Wholesale Trade/Retail Trade (WRT) = 42, 44, 45. Only the private sector portion of the PSA sector is included in the current analyses (62191, Ambulance Services). The sector for Support Activities for Mining, 21311, has been labelled as MIN/OIL in 5-digit NAICS level analyses because at this level it spans both the MIN and OIL sectors.

Prevention indices (PI)^6,22 were calculated by averaging the count and rate ranks for each 5-digit NAICS code. PIs were developed by WA State researchers as a simple method to rank and prioritize industries for prevention emphasis by giving equal weighting to the total count and rate of claims. To be included in the indices, 5-digit NAICS industries needed to report hours in 7 or more years of the study period, have an average of at least 100 FTE per year and at least 55 claims for 2007-2017, which are proportional to the inclusion criteria developed by WA State.⁶ To clarify, these criteria do not limit the inclusion of small employers, but do limit the inclusion of relatively small industries. NAICS code versions for 2007, 2012, and 2017 were harmonized to present the PI ratings over the 2007-2017 period.

Overall findings were compared to the BLS SOII and available WC data. WC data were often reported using codes for injury cause, nature, and part of body developed by the Workers’ Compensation Insurance Organizations (WCIO).²³ The WCIO codes are more limited in the level of detail than BLS OIICS codes, and do not have a defined coding hierarchy or instructions. NIOSH, the MA Department of Public Health, and others have developed crosswalks between WCIO cause and BLS OIICS event/exposure codes at the basic 1-digit level,⁴ and these crosswalks were applied prior to comparisons. More detailed crosswalks between WCIO and OIICS are problematic because the WCIO codes are not hierarchal, are less detailed, and may map to more than one OIICS code at the 2-digit level.

All analyses were conducted using SAS version 9.4.²⁴ This study was internally reviewed by NIOSH, and it was determined that it did not constitute human subjects research. Rather, the study involved the analysis of coded and previously collected WC administrative claims data.

3: Results

3.1: Demographics

From 2007 to 2017, 140,780 accepted LT claims and 633,373 accepted MO claims (94.7% of the 817,103 accepted claims) for OHBWC-insured private employers could be reliably matched to UI data on NAICS and employee count and were included in subsequent event/exposure rate analyses. For a proportion of claims (11,098 claims, 1.3%), a reliable employee count could not be determined for the corresponding policy/year or quarter, and these claims are not included in reported analyses. For another proportion of claims, a reliable 5- or 6-digit NAICS code could not be assigned for certain multiple-location policies (5,677 policies, 0.3%; 31,852 claims, 3.9%), and these claims are also not included in analyses reported here. During this period, approximately 87% of WC claims received for OHBWC-insured private employers were accepted. The mean annual number of covered FTE for this period in the analyzed population of employers was 2,028,561. The mean annual number of employers with covered policies for this period was 166,263. Number of FTEs and policies by NORA sector are provided in Table 1.

Table 1:

Demographics for state-insured private employers by industry sector, 2007-2017

Demographic	Metric	NORA a Industry Sector
		AFF	CON	HCSA	MIN	MNF	OIL	PSA	SRV	TWU	WRT	ALL
FTE Count (full time equivalents)	Count	147,486	1,757,873	3,040,615	44,708	3,638,072	48,568	52,056	9,141,110	831,781	3,611,898	22,314,167
	% of all FTE	0.66%	7.88%	13.63%	0.20%	16.30%	0.22%	0.23%	40.97%	3.73%	16.19%	100.00%
Policy Count	Count	12,063	203,880	201,690	1,864	135,605	4,372	1,102	870,824	55,591	341,900	1,828,891
	% of all policies	0.66%	11.15%	11.03%	0.10%	7.41%	0.24%	0.06%	47.61%	3.04%	18.69%	100.00%

^aNORA = National Occupational Research Agenda

NORA Sectors by North American Industry Classification System (NAICS) codes include: Agriculture, Forestry, Fishing/Hunting (AFF) = 11; Construction (CON) = 23; Health Care and Social Assistance (HCSA) = 62, 54194, 81291; Manufacturing (MNF) = 31, 32, 33; Mining (MIN) = 21; Oil & Gas (OIL) = 211, 213111, & 213112; Public Safety (PSA) = 92212, 92214, 92216, 62191; Services, except Public Safety (SRV) = 51–56, 61, 71, 72, 81, 92; Transportation, Warehousing, Utilities (TWU) = 22, 48, 49; Wholesale Trade/Retail Trade (WRT) = 42, 44, 45. Only the private sector portion of the PSA sector is included in the current analyses (62191, Ambulance Services).

3.2: Event/Exposure Time Trends

All claims were initially coded for event/exposure using the algorithm⁸ described previously and 28,239 of 140,780 LT claims (20%) from 2007-2017 were manually reviewed. For LT claims, the estimated accuracy after the manual review was 87% at the 1-digit event/exposure level and 78% at the 2-digit level. For each OIICS event/exposure, Table 2 presents private sector LT claim count, rate, and percentage of total claims, for each year 2007-2017. For LT claims, Overexertion and Bodily Reaction represented the largest share of claims from 2007 to 2017, followed by Falls, Slips, and Trips and Contact with Objects and Equipment. LT counts/rates declined for all event/exposure types. Among the top three events/exposures for aggregate counts and rates, Overexertion and Bodily Reaction declined the most in both rates and counts, followed by Falls, Slips, and Trips, and Contact with Objects and Equipment. The shares of each OIICS event/exposure in the total LT count remained largely unchanged from 2007 to 2017.

Table 2:

Lost-time ^a claim counts, proportions, and rates ^b by event/exposure, 2007-2017 ^c

Year	Metric	Violence and Other Injuries by Persons or Animals	Transportation Incidents	Fires and Explosions	Falls, Slips, Trips	Exposure to Harmful Substances or Environments	Contact with Objects and Equipment	Overexertion and Bodily Reaction	Non-classifiable	Total
2007	Count	447	1108	80	5555	542	4183	6414	81	18410
	% of all LT claims	2.43%	6.02%	0.43%	30.17%	2.94%	22.72%	34.84%	0.44%	100.00%
	Rate	0.02	0.05	0.00	0.26	0.03	0.20	0.30	0.00	0.86
2008	Count	424	1051	76	5252	412	3641	5332	36	16224
	% of all LT claims	2.61%	6.48%	0.47%	32.37%	2.54%	22.44%	32.86%	0.22%	100.00%
	Rate	0.02	0.05	0.00	0.25	0.02	0.18	0.26	0.00	0.78
2009	Count	349	753	48	3882	312	2312	3971	24	11651
	% of all LT claims	3.00%	6.46%	0.41%	33.32%	2.68%	19.84%	34.08%	0.21%	100.00%
	Rate	0.02	0.04	0.00	0.21	0.02	0.12	0.21	0.00	0.62
2010	Count	345	890	62	4343	333	2767	4595	22	13357
	% of all LT claims	2.58%	6.66%	0.46%	32.51%	2.49%	20.72%	34.40%	0.16%	100.00%
	Rate	0.02	0.05	0.00	0.23	0.02	0.15	0.24	0.00	0.71
2011	Count	297	898	63	3998	336	2759	4210	38	12599
	% of all LT claims	2.36%	7.13%	0.50%	31.73%	2.67%	21.90%	33.42%	0.30%	100.00%
	Rate	0.02	0.05	0.00	0.21	0.02	0.14	0.22	0.00	0.66
2012	Count	287	837	50	3364	330	2719	3997	37	11621
	% of all LT claims	2.47%	7.20%	0.43%	28.95%	2.84%	23.40%	34.39%	0.32%	100.00%
	Rate	0.01	0.04	0.00	0.17	0.02	0.14	0.21	0.00	0.60
2013	Count	282	812	38	3708	338	2827	3961	62	12028
	% of all LT claims	2.34%	6.75%	0.32%	30.83%	2.81%	23.50%	32.93%	0.52%	100.00%
	Rate	0.01	0.04	0.00	0.19	0.02	0.14	0.20	0.00	0.60
2014	Count	243	887	49	3898	361	2774	3783	46	12041
	% of all LT claims	2.02%	7.37%	0.41%	32.37%	3.00%	23.04%	31.42%	0.38%	100.00%
	Rate	0.01	0.04	0.00	0.19	0.02	0.14	0.19	0.00	0.59
2015	Count	271	952	44	3785	326	2684	3557	68	11687
	% of all LT claims	2.32%	8.15%	0.38%	32.39%	2.79%	22.97%	30.44%	0.58%	100.00%
	Rate	0.01	0.05	0.00	0.18	0.02	0.13	0.17	0.00	0.56
2016	Count	282	863	22	3386	293	2611	3415	63	10935
	% of all LT claims	2.58%	7.89%	0.20%	30.96%	2.68%	23.88%	31.23%	0.58%	100.00%
	Rate	0.01	0.04	0.00	0.16	0.01	0.12	0.16	0.00	0.51
2017	Count	284	863	31	3081	312	2422	3171	63	10227
	% of all LT claims	2.78%	8.44%	0.30%	30.13%	3.05%	23.68%	31.01%	0.62%	100.00%
	Rate	0.01	0.04	0.00	0.14	0.01	0.11	0.14	0.00	0.47
2007-2017	Count	3511	9915	563	44257	3895	31703	46411	540	140780
	% of all LT claims	2.49%	7.04%	0.40%	31.44%	2.77%	22.52%	32.97%	0.38%	100.01%
	Rate	0.016	0.04	0.003	0.20	0.017	0.14	0.21	0.00	0.63
2007-2017 Count % Change		−36.47%	−22.11%	−61.25%	−44.54%	−42.44%	−42.10%	−50.56%	−22.22%	−44.45%
2007-2017 Rate % Change		−38.23%	−24.28%	−62.33%	−46.08%	−44.04%	−43.71%	−51.94%	−24.39%	−45.99%

^aEight or more days away from work

^bRate per 100 full-time equivalents

^cAmong state-insured private employers

3.3: Event/Exposures by Sector and Detailed Industry

Table 3 presents LT claim counts and rates by 1-digit OIICS event/exposure for private employers in each NORA sector for 2007 to 2017. Similar information is provided for total claims (LT and MO claims combined) in Table 4. The top three event/exposures by percentage of LT and total claims for most sectors were Overexertion and Bodily Reaction; Falls, Slips, and Trips; and Contact with Objects and Equipment, although the rank order of these three divisions did vary by sector (Tables 3, 4). There were some exceptions to this pattern for LT claims, including HCSA, where Violence and Other Injuries by Persons or Animals had the third highest proportion by event/exposure type, and PSA and TWU, where Transportation Incidents had the second highest (tied with Falls, Slips, Trips) and third highest proportions, respectively.

Table 3:

Lost-time ^a claim counts, proportions, and rates ^b by event/exposure by industry sector, 2007-2017 ^c

Event/Exposure Type	Metric	NORA d Industry Sector
		AFF	CON	HCSA	MIN	MNF	OIL	PSA	SRV	TWU	WRT	ALL
All Event/Exposures	Count	861	16186	18213	502	30823	523	867	39769	11875	21161	140780
	% of all LT claims	0.61%	11.50%	12.94%	0.36%	21.89%	0.37%	0.62%	28.25%	8.44%	15.03%	100.00%
	Rate	0.584	0.921	0.599	1.123	0.847	1.077	1.666	0.435	1.428	0.586	0.631
Violence and Other Injuries by Persons or Animals	Count	105	67	1898	4	87	1	10	1027	79	233	3511
	% of all LT claims	2.99%	1.91%	54.06%	0.11%	2.48%	0.03%	0.28%	29.25%	2.25%	6.64%	100.00%
	% of sector LT claims	12.20%	0.41%	10.42%	0.80%	0.28%	0.19%	1.15%	2.58%	0.67%	1.10%	2.49%
	Rate	0.071	0.004	0.062	0.009	0.002	0.002	0.019	0.011	0.009	0.006	0.016
Transportation Incidents	Count	70	1018	1131	46	671	62	146	2916	2301	1553	9914
	% of all LT claims	0.71%	10.27%	11.41%	0.46%	6.77%	0.63%	1.47%	29.41%	23.21%	15.66%	100.00%
	% of sector LT claims	8.13%	6.29%	6.21%	9.16%	2.18%	11.85%	16.84%	7.33%	19.38%	7.34%	7.04%
	Rate	0.047	0.058	0.037	0.103	0.018	0.128	0.280	0.032	0.277	0.043	0.044
Fires and Explosions	Count	3	106	10	5	157	16	1	121	31	113	563
	% of all LT claims	0.53%	18.83%	1.78%	0.89%	27.89%	2.84%	0.18%	21.49%	5.51%	20.07%	100.00%
	% of sector LT claims	0.35%	0.65%	0.05%	1.00%	0.51%	3.06%	0.12%	0.30%	0.26%	0.53%	0.40%
	Rate	0.002	0.006	0.000	0.011	0.004	0.033	0.002	0.001	0.004	0.003	0.003
Falls, Slips, Trips	Count	287	5753	6684	120	6257	142	146	14346	3992	6525	44252
	% of all LT claims	0.65%	13.00%	15.10%	0.27%	14.14%	0.32%	0.33%	32.42%	9.02%	14.75%	100.00%
	% of sector LT claims	33.33%	35.54%	36.70%	23.90%	20.30%	27.15%	16.84%	36.07%	33.62%	30.84%	31.43%
	Rate	0.195	0.327	0.220	0.268	0.172	0.292	0.280	0.157	0.480	0.181	0.198
Exposure to Harmful Substances or Environments	Count	21	501	250	8	1125	12	3	1529	114	332	3895
	% of all LT claims	0.54%	12.86%	6.42%	0.21%	28.88%	0.31%	0.08%	39.26%	2.93%	8.52%	100.00%
	% of sector LT claims	2.44%	3.10%	1.37%	1.59%	3.65%	2.29%	0.35%	3.84%	0.96%	1.57%	2.77%
	Rate	0.014	0.029	0.008	0.018	0.031	0.025	0.006	0.017	0.014	0.009	0.017
Contact with Objects and Equipment	Count	239	3954	1264	193	10741	173	38	8536	1717	4844	31699
	% of all LT claims	0.75%	12.47%	3.99%	0.61%	33.88%	0.55%	0.12%	26.93%	5.42%	15.28%	100.00%
	% of sector LT claims	27.76%	24.43%	6.94%	38.45%	34.85%	33.08%	4.38%	21.46%	14.46%	22.89%	22.52%
	Rate	0.162	0.225	0.042	0.432	0.295	0.356	0.073	0.093	0.206	0.134	0.142
Overexertion and Bodily Reaction	Count	129	4730	6917	122	11673	114	521	11132	3595	7473	46406
	% of all LT claims	0.28%	10.19%	14.91%	0.26%	25.15%	0.25%	1.12%	23.99%	7.75%	16.10%	100.00%
	% of sector LT claims	14.98%	29.22%	37.98%	24.30%	37.87%	21.80%	60.09%	27.99%	30.27%	35.31%	32.96%
	Rate	0.087	0.269	0.227	0.273	0.321	0.235	1.001	0.122	0.432	0.207	0.208
Non-classifiable	Count	7	57	59	4	112	3	2	162	46	88	540
	% of all LT claims	1.30%	10.56%	10.93%	0.74%	20.74%	0.56%	0.37%	30.00%	8.52%	16.30%	100.00%
	% of sector LT claims	0.81%	0.35%	0.32%	0.80%	0.36%	0.57%	0.23%	0.41%	0.39%	0.42%	0.38%
	Rate	0.005	0.003	0.002	0.009	0.003	0.006	0.004	0.002	0.006	0.002	0.002

^aEight or more days away from work

^bRate per 100 full-time equivalents

^cAmong state-insured private employers

^dNORA = National Occupational Research Agenda; NORA Sectors by North American Industry Classification System (NAICS) codes include: Agriculture, Forestry, Fishing/Hunting (AFF) = 11; Construction (CON) = 23; Health Care and Social Assistance (HCSA) = 62, 54194, 81291; Manufacturing (MNF) = 31, 32, 33; Mining (MIN) = 21; Oil & Gas (OIL) = 211, 213111, & 213112; Public Safety (PSA) = 92212, 92214, 92216, 62191; Services, except Public Safety (SRV) = 51–56, 61, 71, 72, 81, 92; Transportation, Warehousing, Utilities (TWU) = 22, 48, 49; Wholesale Trade/Retail Trade (WRT) = 42, 44, 45. Only the private sector portion of the PSA sector is included in the current analyses (62191, Ambulance Services).

Table 4:

Total (medical-only and lost-time ^a) claim counts, proportions, and rates ^b by event/exposure by industry sector, 2007-2017 ^c

Event/Exposure Type	Metric	NORA d Industry Sector
		AFF	CON	HCSA	MIN	MNF	OIL	PSA	SRV	TWU	WRT	ALL
All Events/Exposures	Count	4901	72416	113982	2356	194117	1901	4658	227273	36769	115780	774153
	% of all LT claims	0.63%	9.35%	14.72%	0.30%	25.07%	0.25%	0.60%	29.36%	4.75%	14.96%	100.00%
	Rate	3.323	4.120	3.749	5.270	5.336	3.914	8.948	2.486	4.421	3.206	3.469
Violence and Other Injuries by Persons or Animals	Count	258	913	15763	15	841	28	144	6073	535	1639	26209
	% of all LT claims	0.98%	3.48%	60.14%	0.06%	3.21%	0.11%	0.55%	23.17%	2.04%	6.25%	100.00%
	% of sector LT claims	5.26%	1.26%	13.83%	0.64%	0.43%	1.47%	3.09%	2.67%	1.46%	1.42%	3.39%
	Rate	0.175	0.052	0.518	0.034	0.023	0.058	0.277	0.066	0.064	0.045	0.117
Transportation Incidents	Count	128	2208	3348	103	1527	127	527	6619	4014	4073	22674
	% of all LT claims	0.56%	9.74%	14.77%	0.45%	6.73%	0.56%	2.32%	29.19%	17.70%	17.96%	100.00%
	% of sector LT claims	2.61%	3.05%	2.94%	4.37%	0.79%	6.68%	11.31%	2.91%	10.92%	3.52%	2.93%
	Rate	0.087	0.126	0.110	0.230	0.042	0.261	1.012	0.072	0.483	0.113	0.102
Fires and Explosions	Count	4	116	19	7	191	16	8	164	40	135	700
	% of all LT claims	0.57%	16.57%	2.71%	1.00%	27.29%	2.29%	1.14%	23.43%	5.71%	19.29%	100.00%
	% of sector LT claims	0.08%	0.16%	0.02%	0.30%	0.10%	0.84%	0.17%	0.07%	0.11%	0.12%	0.09%
	Rate	0.003	0.007	0.001	0.016	0.005	0.033	0.015	0.002	0.005	0.004	0.003
Falls, Slips, Trips	Count	993	15076	27555	438	25136	424	645	52469	9825	23631	156192
	% of all LT claims	0.64%	9.65%	17.64%	0.28%	16.09%	0.27%	0.41%	33.59%	6.29%	15.13%	100.00%
	% of sector LT claims	20.26%	20.82%	24.17%	18.59%	12.95%	22.30%	13.85%	23.09%	26.72%	20.41%	20.18%
	Rate	0.673	0.858	0.906	0.980	0.691	0.873	1.239	0.574	1.181	0.654	0.700
Exposure to Harmful Substances or Environments	Count	240	3079	3682	78	10552	110	82	16578	911	3535	38847
	% of all LT claims	0.62%	7.93%	9.48%	0.20%	27.16%	0.28%	0.21%	42.68%	2.35%	9.10%	100.00%
	% of sector LT claims	4.90%	4.25%	3.23%	3.31%	5.44%	5.79%	1.76%	7.29%	2.48%	3.05%	5.02%
	Rate	0.163	0.175	0.121	0.174	0.290	0.226	0.158	0.181	0.110	0.098	0.174
Contact with Objects and Equipment	Count	2386	36056	31131	1258	110864	833	784	103481	11325	54646	352764
	% of all LT claims	0.68%	10.22%	8.82%	0.36%	31.43%	0.24%	0.22%	29.33%	3.21%	15.49%	100.00%
	% of sector LT claims	48.68%	49.79%	27.31%	53.40%	57.11%	43.82%	16.83%	45.53%	30.80%	47.20%	45.57%
	Rate	1.618	2.051	1.024	2.814	3.047	1.715	1.506	1.132	1.362	1.513	1.581
Overexertion and Bodily Reaction	Count	862	14715	32146	447	44397	353	2449	41102	9996	27797	174264
	% of all LT claims	0.49%	8.44%	18.45%	0.26%	25.48%	0.20%	1.41%	23.59%	5.74%	15.95%	100.00%
	% of sector LT claims	17.59%	20.32%	28.20%	18.97%	22.87%	18.57%	52.58%	18.08%	27.19%	24.01%	22.51%
	Rate	0.584	0.837	1.057	1.000	1.220	0.727	4.705	0.450	1.202	0.770	0.781
Non-classifiable	Count	30	253	338	10	609	10	19	787	123	324	2503
	% of all LT claims	1.20%	10.11%	13.50%	0.40%	24.33%	0.40%	0.76%	31.44%	4.91%	12.94%	100.00%
	% of sector LT claims	0.61%	0.35%	0.30%	0.42%	0.31%	0.53%	0.41%	0.35%	0.33%	0.28%	0.32%
	Rate	0.020	0.014	0.011	0.022	0.017	0.021	0.036	0.009	0.015	0.009	0.011

^aEight or more days away from work

^bRate per 100 full-time equivalents

^cAmong state-insured private employers

^dNORA = National Occupational Research Agenda; NORA Sectors by North American Industry Classification System (NAICS) codes include: Agriculture, Forestry, Fishing/Hunting (AFF) = 11; Construction (CON) = 23; Health Care and Social Assistance (HCSA) = 62, 54194, 81291; Manufacturing (MNF) = 31, 32, 33; Mining (MIN) = 21; Oil & Gas (OIL) = 211, 213111, & 213112; Public Safety (PSA) = 92212, 92214, 92216, 62191; Services, except Public Safety (SRV) = 51–56, 61, 71, 72, 81, 92; Transportation, Warehousing, Utilities (TWU) = 22, 48, 49; Wholesale Trade/Retail Trade (WRT) = 42, 44, 45. Only the private sector portion of the PSA sector is included in the current analyses (62191, Ambulance Services).

Table 5 presents the LT-claim PI for the 5-digit NAICS codes with the lowest PIs. Low PIs correspond to high prevention priority, i.e. a high ranking. The 5-digit NAICS industries are listed in ascending order of LT PI for all events/exposures combined, and for each industry, the PI for each specific event/exposure is also reported separately. The top five 5-digit NAICS industries for LT claims for all event/exposures combined as ranked by PI included General Freight Trucking, Long-Distance (48412), Other Motor Vehicle Parts Manufacturing (33639), Ambulance Services (62191), Specialized Freight (except Used Goods) Trucking, Long-Distance (48423), and Specialized Freight (except Used Goods) Trucking, Local (48422). Many of the industries with the highest rankings overall had high injury rates/counts for several specific types of event/exposures. PIs for specific event/exposures that are among the five lowest are highlighted, but note that while Table 5 includes industries with the lowest PIs overall, the table is not long enough to include all industries with a PI among the five lowest for any event/exposure. Table 6 presents LT and LT+MO claim counts and rates by 2-digit event/exposure, for all industries combined.

Table 5:

Lost-time ^a claim counts, rates ^b, and prevention indices ^c by event/exposure by detailed industry sector, 2007-2017^d

NAICS e Code	NAICS Description	NORA f Sector	Mean Annual FTE	All Event/Exposures Combined						Violence and other injuries by persons or animals	Transportation incidents	Fires and explosions	Falls, slips, trips-	Exposure to harmful substances or environments	Contact with objects and equipment	Overexertion and bodily reaction
				LT Claim Rate	LT Claim Rate Rank	LT Claim Count	LT Claim Count Rank	LT PI	LT PI Rank	LT PI Rank	LT PI Rank	LT PI Rank	LT PI Rank	LT PI Rank	LT PI Rank	LT PI Rank
48412	General Freight Trucking, Long-Distance	TWU	17938	1.89	7	3728	4	5.5	1	95	2	46	1	94.5	37	6
33639	Other Motor Vehicle Parts Manufacturing	MNF	9080	1.52	31	1518	11	21.0	2	120	99	143	57	31	17	1
62191	Ambulance Services	PSA	4732	1.67	14	867	32	23.0	3	40.5	6	133	64.5	235	234.5	2
48423	Specialized Freight (except Used Goods) Trucking, Long-Distance	TWU	2845	2.18	4	683	47	25.5	4	62	7	19	5	18	87.5	10
48422	Specialized Freight (except Used Goods) Trucking, Local	TWU	7125	1.53	30	1198	22	26.0	5	49.5	1	37	3	103	105	48
42393	Recyclable Material Merchant Wholesalers	WRT	5102	1.57	23	879	30	26.5	6	99.5	33.5	1	24	5	2	70.5
33151	Ferrous Metal Foundries	MNF	2865	2.10	5	661	49	27.0	7.5	224	206	2	86	2	5	8
33152	Nonferrous Metal Foundries	MNF	3404	1.85	8	693	46	27.0	7.5	233	248.5	8	85	1	7	9
49311	General Warehousing and Storage	TWU	8122	1.41	38	1264	19	28.5	9	84	24	256	27	83.5	20.5	3.5
33211	Forging and Stamping	MNF	7959	1.39	40	1218	21	30.5	10.5	251	157	71	77	27	1	11
48411	General Freight Trucking, Local	TWU	8675	1.36	44	1294	17	30.5	10.5	114	4	75	4	156	99.5	29
56211	Waste Collection	SRV	3228	1.77	10	630	53	31.5	12	70	8	41	21	96.5	32.5	12
42481	Beer and Ale Merchant Wholesalers	WRT	2761	1.96	6	595	58	32.0	13	89	47	120	13	208	86	5
56132	Temporary Help Services	SRV	46587	1.16	71	5920	2	36.5	14	55	57.5	48	43.5	14	6	16
48541	School and Employee Bus Transportation	TWU	1794	2.24	2	443	75	38.5	15	24	4	106	11	331	259	62
62311	Nursing Care Facilities (Skilled Nursing Facilities)	HSA	40133	1.13	75	5002	3	39.0	16	8	196	119	8	51	140	3.5
42351	Metal Service Centers and Other Metal Merchant Wholesalers	WRT	7646	1.25	53	1048	26	39.5	17	78	127	40	58	58.5	3	37
23829	Other Building Equipment Contractors	CON	4283	1.36	43	639	50	46.5	18	60.5	71	86	9	12	36	26
33231	Plate Work and Fabricated Structural Product Manufacturing	MNF	5805	1.23	57	785	37.5	47.3	19	305.5	176.5	43	78	39.5	4	44
56173	Landscaping Services	SRV	15047	1.08	91	1786	8	49.5	20	63	15	91.5	36	86.5	8	67
48421	Used Household and Office Goods Moving	TWU	1979	1.74	12	379	87.5	49.8	21	210	49.5	256	23	331	58	17
23816	Roofing Contractors	CON	5397	1.21	61	719	45	53.0	22	77	84.5	95	2	30	80	105.5
21311	Support Activities for Mining	MIN/OIL	3961	1.32	48	574	60	54.0	23	237	16.5	4	50	72	11	110
33142	Copper Rolling, Drawing, Extruding, and Alloying	MNF	2021	1.64	16	365	92.5	54.3	24	116	108	32	106	55	22	14
33131	Alumina and Aluminum Production and Processing	MNF	1864	1.64	15	337	99	57.0	25	305.5	213	30	84	29	12.5	46
23812	Structural Steel and Precast Concrete Contractors	CON	1967	1.61	19	349	96.5	57.8	26	305.5	218.5	31	22	145	24	60.5
62331	Continuing Care Retirement Communities and Assisted Living Facilities for the Elderly	HSA	12064	1.04	102	1376	14	58.0	27	14	264	256	20	80	187.5	7
62321	Residential Intellectual and Developmental Disability Facilities	HSA	12945	1.00	108	1427	13	60.5	28	2.5	36	145	19	210	170	39.5
32732	Ready-Mix Concrete Manufacturing	MNF	2004	1.56	24	345	98	61.0	29	75	14	54	18	46	92	103
23814	Masonry Contractors	CON	4739	1.16	70	605	55	62.5	30	126	114.5	55	15	75	77	24
23731	Highway, Street, and Bridge Construction	CON	5950	1.11	84	727	43	63.5	31.5	189	11	25	47	34	23	109
53111	Lessors of Residential Buildings and Dwellings	SRV	6872	1.07	92	810	35	63.5	31.5	33	205	66	7	53	111.5	20
62431	Vocational Rehabilitation Services	HSA	4505	1.16	69	577	59	64.0	33	23	53	256	6	186	152.5	59
32629	Other Rubber Product Manufacturing	MNF	5011	1.13	76	622	54	65.0	34.5	132	320	256	121.5	7	25	13
33122	Rolling and Drawing of Purchased Steel	MNF	1809	1.57	22	313	108	65.0	34.5	205	270	52	134.5	93	12.5	47
33281	Coating, Engraving, Heat Treating, and Allied Activities	MNF	9150	1.01	107	1021	28	67.5	36	122	130	13	95	3.5	15.5	33.5
44411	Home Centers	WRT	3607	1.17	66	466	71	68.5	37	105	109	126	101.5	221	26	19
32621	Tire Manufacturing	MNF	1032	2.18	3	248	139	71.0	38	305.5	242	39	238	26	30.5	18
42441	General Line Grocery Merchant Wholesalers	WRT	3095	1.20	64	409	80	72.0	40	145	31.5	256	41	160	110	27.5
48599	Other Transit and Ground Passenger Transportation	TWU	1535	1.61	20	272	124	72.0	40	47	4	256	163	331	286	192
56133	Professional Employer Organizations	SRV	10137	0.97	120	1078	24	72.0	40	82	35	256	30	62.5	65.5	35.5
32721	Glass and Glass Product Manufacturing	MNF	2128	1.37	41	320	104	72.5	42	305.5	279	256	137	81	28	21
32619	Other Plastics Product Manufacturing	MNF	12736	0.94	130	1315	16	73.0	43	162	197.5	42	79	15	32.5	15
32192	Wood Container and Pallet Manufacturing	MNF	2735	1.23	58	369	91	74.5	45	222	116	256	186	185	9	102
33637	Motor Vehicle Metal Stamping	MNF	6427	1.02	105	723	44	74.5	45	194	203	138	112	79	14	23
71121	Spectator Sports	SRV	1442	1.62	18	257	131	74.5	45	1	92	256	225	331	206.5	121
62322	Residential Mental Health and Substance Abuse Facilities	HSA	2365	1.28	51	332	101	76.0	47	2.5	80	256	56	285	255.5	167
23891	Site Preparation Contractors	CON	8268	0.96	125	871	31	78.0	48	163	21	9	40	43	29	129.5
48841	Motor Vehicle Towing	TWU	1662	1.49	33	272	124	78.5	49	106	9	51	54	222	103	160
56291	Remediation Services	SRV	2686	1.18	65	349	96.5	80.8	50	87	93	62	32	16	67	124
23811	Poured Concrete Foundation and Structure Contractors	CON	4203	1.06	95	489	67	81.0	51	117	94	256	28	48	65.5	76
33299	All Other Fabricated Metal Product Manufacturing	MNF	6906	0.96	124	730	42	83.0	52.5	195	189	36	199	101	10	31
56179	Other Services to Buildings and Dwellings	SRV	1659	1.44	37	262	129	83.0	52.5	68	31.5	50	38	28	133	89
81111	Automotive Mechanical and Electrical Repair and Maintenance	SRV	10956	0.87	146	1050	25	85.5	54	129	54	5.5	99	92	43	45
23711	Water and Sewer Line and Related Structures Construction	CON	5191	0.99	110	565	62	86.0	55	138	39	33.5	43.5	104	30.5	141
31151	Dairy Product (except Frozen) Manufacturing	MNF	1898	1.30	49	272	124	86.5	56	305.5	216	107	49	6	90	74
23899	All Other Specialty Trade Contractors	CON	7817	0.91	137	785	37.5	87.3	57	133.5	44	68.5	75.5	9	47.5	85
62412	Services for the Elderly and Persons with Disabilities	HSA	15830	0.83	164	1451	12	88.0	58	7	26	148	16	139	208.5	72
23822	Plumbing, Heating, and Air-Conditioning Contractors	CON	28215	0.81	172	2500	5	88.5	59	151.5	67	23.5	37	22	71	57
62399	Other Residential Care Facilities	HSA	2171	1.21	62	288	115.5	88.8	60	4	89	256	48	284	250	201
23611	Residential Building Construction	CON	14787	0.84	163	1360	15	89.0	61	144	95.5	70	12	76	68.5	91
23819_ 23831_ 23833	Other Foundation, Structure, and Building Exterior Contractors (23819), Drywall and Insulation Contractors (23831), and Flooring Contractors (23833)	CON	6608	0.94	131	680	48	89.5	62	115	150.5	139	26	89	91	35.5
23821	Electrical Contractors and Other Wiring Installation Contractors	CON	20722	0.80	174	1820	7	90.5	63	109	63	21	17	9	109	79
48111	Scheduled Air Transportation	TWU	1699	1.35	46	252	135.5	90.8	64	197.5	162	256	52.5	278	175	25
42449	Other Grocery and Related Products Merchant Wholesalers	WRT	2338	1.15	72	296	110.5	91.3	65	123	120	256	46	238	167	30
32611	Plastics Packaging Materials and Unlaminated Film and Sheet Manufacturing	MNF	2997	1.08	88	357	95	91.5	66.5	305.5	290	67	239.5	107	15.5	41
56172	Janitorial Services	SRV	19390	0.80	173	1710	10	91.5	66.5	74	45	115	14	32.5	111.5	86
42443	Dairy Product (except Dried or Canned) Merchant Wholesalers	WRT	1491	1.46	35	239	149	92.0	68	186	174	256	69	273	61	51
56221	Waste Treatment and Disposal	SRV	2398	1.12	78	296	110.5	94.3	69	218	51	10	39	62.5	144	104
62221	Psychiatric and Substance Abuse Hospitals	HSA	999	1.76	11	193	178	94.5	70	5	179	256	114	256	327	94.5
42447	Meat and Meat Product Merchant Wholesalers	WRT	1237	1.55	25	211	164.5	94.8	71	305.5	37	256	81	147	54.5	97.5
23713	Power and Communication Line and Related Structures Construction	CON	3173	1.05	99	365	92.5	95.8	72	148.5	48	73	83	45	64	88
48819	Other Support Activities for Air Transportation	TWU	1373	1.46	36	220	158	97.0	73.5	179	69.5	256	59	331	142	53
62161	Home Health Care Services	HSA	26920	0.75	188	2230	6	97.0	73.5	17	19	153	35	127	196	52
42448	Fresh Fruit and Vegetable Merchant Wholesalers	WRT	1601	1.35	45	238	150	97.5	75	103	42.5	256	51	131.5	120.5	78
23622	Commercial and Institutional Building Construction	CON	14514	0.79	176	1258	20	98.0	76.5	209	101	146	33	71	59.5	94.5
44132	Tire Dealers	WRT	3984	0.97	118	425	78	98.0	76.5	305.5	81	3	147	122	63	49

^aEight or more days away from work

^bRate per 100 full-time equivalents

^cPI = Prevention Index (Mean LT Claim Rate and Count Rank)

^dAmong state-insured private employers

^eAll North American Industry Classification System (NAICS) codes are 2017 version terms. Certain NAICS were impacted by code changes and were combined across 2007, 2012, and 2017 versions. For this table, this includes 23819_23831_23833.

^fNORA = National Occupational Research Agenda; NORA Sectors by North American Industry Classification System (NAICS) codes include: Agriculture, Forestry, Fishing/Hunting (AFF) = 11; Construction (CON) = 23; Health Care and Social Assistance (HCSA) = 62, 54194, 81291; Manufacturing (MNF) = 31, 32, 33; Mining (MIN) = 21; Oil & Gas (OIL) = 211, 213111, & 213112; Public Safety (PSA) = 92212, 92214, 92216, 62191; Services, except Public Safety (SRV) = 51–56, 61, 71, 72, 81, 92; Transportation, Warehousing, Utilities (TWU) = 22, 48, 49; Wholesale Trade/Retail Trade (WRT) = 42, 44, 45. Only the private sector portion of the PSA sector is included in the current analyses (62191, Ambulance Services).

Shaded cells identify industries in the top-5 PI ranking for either all event/exposures combined or an individual event/exposure. Due to space limitations, entire top 5 PI lists for all individual event/exposures are not provided in this table. For Fires and Explosions, this includes Iron and Steel Mills and Ferro Alloy Manufacturing (33111), which had a PI rank of 5.5. For Exposure to Harmful Substances or Environments, this includes Other Basic Organic Chemical Manufacturing (32519), which had a PI rank of 3.5.

Table 6:

Lost-time ^a and total (medical-only and lost-time) claim counts and rates ^b by 2-digit event/exposure, 2007-2017 ^c

Event/Exposure	Event/Exposure Description	Lost-time claims				Medical-only and lost-time
		Count	% among all 2-digit codes	% within 1-digit code	Rate	Count	% among all 2-digit codes	% within 1-digit code	Rate
10	Violence Or Other Injuries By Persons Or Animals, Unspecified	3	0.00%	0.09%	0.0000	3	0.00%	0.01%	0.000
11	Intentional Injuries By Person	1770	1.26%	50.41%	0.0079	8050	1.04%	30.71%	0.036
12	Unintentional Or Intent Unknown Injuries By Person	1011	0.72%	28.80%	0.0045	3443	0.44%	13.14%	0.015
13	Animal Or Insect Related Incidents	727	0.52%	20.71%	0.0033	14713	1.90%	56.14%	0.066
20	Transportation Incident, Unspecified	2	0.00%	0.02%	0.0000	2	0.00%	0.01%	0.000
21	Aircraft Incident	49	0.03%	0.49%	0.0002	50	0.01%	0.22%	0.000
22	Rail Vehicle Incident	17	0.01%	0.17%	0.0001	17	0.00%	0.07%	0.000
23	Animal And Other Non-Motorized Vehicle Transportation Incidents	34	0.02%	0.34%	0.0002	36	0.00%	0.16%	0.000
24	Pedestrian Vehicular Incidents	1226	0.87%	12.37%	0.0055	1758	0.23%	7.75%	0.008
25	Water Vehicle Incidents	12	0.01%	0.12%	0.0001	12	0.00%	0.05%	0.000
26	Roadway Incidents, Motorized Land Vehicles	7762	5.51%	78.29%	0.0348	19847	2.56%	87.53%	0.089
27	Non-Roadway Incidents, Motorized Land Vehicles	810	0.58%	8.17%	0.0036	950	0.12%	4.19%	0.004
29	Transportation Incident, Nec	2	0.00%	0.02%	0.0000	2	0.00%	0.01%	0.000
31	Fires	173	0.12%	30.73%	0.0008	230	0.03%	32.86%	0.001
32	Explosions	390	0.28%	69.27%	0.0017	470	0.06%	67.14%	0.002
40	Falls, Slips, Trips, Unspecified	93	0.07%	0.21%	0.0004	94	0.01%	0.06%	0.000
41	Slip Or Trip W/O Fall	9100	6.46%	20.56%	0.0408	33362	4.31%	21.36%	0.150
42	Falls On Same Level	21969	15.61%	49.65%	0.0985	86774	11.21%	55.56%	0.389
43	Falls To Lower Level	12766	9.07%	28.85%	0.0572	35209	4.55%	22.54%	0.158
44	Jumps To Lower Level	316	0.22%	0.71%	0.0014	745	0.10%	0.48%	0.003
45	Fall Or Jump Curtailed By Personal Fall Arrest System	8	0.01%	0.02%	0.0000	8	0.00%	0.01%	0.000
50	Exposure To Harmful Substances Or Environments, Unspecified	4	0.00%	0.10%	0.0000	6	0.00%	0.02%	0.000
51	Exposure To Electricity	291	0.21%	7.47%	0.0013	1363	0.18%	3.51%	0.006
52	Exposure, Radiation And Noise	52	0.04%	1.34%	0.0002	1320	0.17%	3.40%	0.006
53	Exposure To Temperature Extremes	2162	1.54%	55.51%	0.0097	18774	2.43%	48.33%	0.084
54	Exposure To Air And Water Pressure Changes	4	0.00%	0.10%	0.0000	4	0.00%	0.01%	0.000
55	Exposure To Other Harmful Substances	1374	0.98%	35.28%	0.0062	17372	2.24%	44.72%	0.078
56	Exposure To Oxygen Deficiency	2	0.00%	0.05%	0.0000	2	0.00%	0.01%	0.000
57	Exposure To Traumatic Or Stressful Event	4	0.00%	0.10%	0.0000	4	0.00%	0.01%	0.000
59	Exposure To Harmful Substances Or Environments, Nec	2	0.00%	0.05%	0.0000	2	0.00%	0.01%	0.000
60	Contact With Objects And Equipment, Unspecified	735	0.52%	2.32%	0.0033	6197	0.80%	1.76%	0.028
61	Needlestick Without Exposure To Harmful Substance	50	0.04%	0.16%	0.0002	8088	1.04%	2.29%	0.036
62	Struck By Objects Or Equipment	17945	12.75%	56.61%	0.0804	219851	28.40%	62.32%	0.985
63	Struck Against Object Or Equipment	5181	3.68%	16.34%	0.0232	77182	9.97%	21.88%	0.346
64	Caught In Or Compressed By Equipment Or Objects	7633	5.42%	24.08%	0.0342	38127	4.92%	10.81%	0.171
65	Struck, Caught, Or Crushed In Collapsing Structures, Equipment Or Materials	70	0.05%	0.22%	0.0003	77	0.01%	0.02%	0.000
66	Rubbed Or Abraded By Friction Or Pressure	63	0.04%	0.20%	0.0003	3218	0.42%	0.91%	0.014
67	Rubbed, Abraded Or Jarred By Vibration	19	0.01%	0.06%	0.0001	19	0.00%	0.01%	0.000
69	Contact With Objects And Equipment, Nec	3	0.00%	0.01%	0.0000	5	0.00%	0.00%	0.000
70	Overexertion And Bodily Reaction, Unspecified	697	0.50%	1.50%	0.0031	1595	0.21%	0.92%	0.007
71	Overexertion Involving Outside Sources	37083	26.34%	79.91%	0.1662	144065	18.61%	82.67%	0.646
72	Repetitive Motions Involving Microtasks	1754	1.25%	3.78%	0.0079	3569	0.46%	2.05%	0.016
73	Other Exertions Or Bodily Reactions	6716	4.77%	14.47%	0.0301	24843	3.21%	14.26%	0.111
74	Bodily Conditions, Nec	56	0.04%	0.12%	0.0003	58	0.01%	0.03%	0.000
78	Multiple Types Of Overexertions And Bodily Reactions	98	0.07%	0.21%	0.0004	132	0.02%	0.08%	0.001
79	Overexertion And Bodily Reaction And Exertion, Nec	2	0.00%	0.00%	0.0000	2	0.00%	0.00%	0.000
99	Non-classifiable	540	0.38%	NA	0.0024	2503	0.32%	NA	0.011
	TOTAL	140780	1		0.6309	774153	100%		3.469

^aEight or more days away from work

^bRate per 100 full-time equivalents

^cAmong state-insured private employers

Violence and Other Injuries by Persons or Animals

The sectors with the highest proportions of LT claims for Violence and Other Injuries by Persons or Animals were AFF (12.2%), HCSA (10.4%), and SRV (2.6%) (Table 3). The top five, 5-digit NAICS industries for LT claims for event/exposures of this type, as ranked by PI, included Spectator Sports (71121), Residential Mental Retardation Facilities (62321), Residential Mental Health and Substance Abuse Facilities (62322), Other Residential Care Facilities (62399), and Psychiatric and Substance Abuse Hospitals (62221). Veterinary Services (54194) were also highly ranked (Table 5). Across all industries, half of LT Violence and Other Injuries by Persons or Animals claims were due to Intentional Injuries by Person (50%), followed by Unintentional or Intent Unknown Injuries by Person (29%) and Animal or Insect Related Incidents (21%) (Table 6).

Transportation Incidents

The sectors with the highest proportions of LT claims for Transportation Incidents were TWU (19.4%), PSA (16.8%), and OIL (11.9%) (Table 3). Most sectors had at least one 5-digit NAICS industry with very high LT claims for Transportation Incidents as ranked by PIs. The top five 5-digit NAICS industries for LT claims as ranked by PI for Transportation Incidents included Specialized Freight (except Used Goods) Trucking, Local (48422), General Freight Trucking, Long-Distance (48412), General Freight Trucking, Local (48411), School and Employee Bus Transportation (48541), and Other Transit and Ground Passenger Transportation (48599). Ambulance Services (62191) were also highly ranked (Table 5). The majority of LT Transportation Incidents claims across all industries were due to Roadway Incidents, Motorized Land Vehicles (78%), followed by Pedestrian Vehicular Incidents (12%) (Table 6).

Fires and Explosions

The sectors with the highest proportions of LT claims for Fires and Explosions claims were OIL (3.1%) and MIN (1.0%) (Table 3). Most sectors had at least one 5-digit NAICS industry with a high LT PI for Fires and Explosions. The top five 5-digit NAICS industries for LT claims as ranked by PI for Fires and Explosions included Recyclable Material Merchant Wholesalers (42393), Ferrous Metal Foundries (33151), Tire Dealers (44132), Support Activities for Mining (21311), and Iron and Steel Mills and Ferro Alloy Manufacturing (33111) (data not shown). Automotive Mechanical and Electrical Repair and Maintenance (81111) was also highly ranked (Table 5). Across all industries, the majority of LT Fires and Explosions claims were due to Explosions (69%) followed by Fires (31%) (Table 6).

Falls, Slips, and Trips

The sectors with the highest proportions of LT Falls, Slips, and Trips claims were HCSA (36.7%), SRV (36.1%), and CON (35.5%) (Table 3). Almost every sector had at least one 5-digit NAICS industry that was very high in LT claims as ranked by PI for Falls, Slips, and Trips. The top five 5-digit NAICS industries for LT claims as ranked by PI for Falls, Slips, and Trips included General Freight Trucking, Long-Distance (48412), Roofing Contractors (23816), Specialized Freight (except Used Goods) Trucking, Local (48422), General Freight Trucking, Local (48411), and Specialized Freight (except Used Goods) Trucking, Long-Distance (48423) (Table 5). Across all industries, most LT Falls, Slips, and Trips claims were Falls on Same Level (50%), followed by Falls to Lower Level (29%), and Slip or Trip without Fall (21%) (Table 6).

Exposure to Harmful Substances or Environments

The sectors with the highest proportions of LT claims for Exposure to Harmful Substances or Environments were SRV (3.8%), MNF (3.7%), CON (3.1%), and OIL (2.3%) (Table 3). The top five 5-digit NAICS industries for LT claims as ranked by PI for Exposure to Harmful Substances or Environments included Nonferrous Metal Foundries (33152), Ferrous Metal Foundries (33151), Other Basic Organic Chemical Manufacturing (32519) (data not shown), Coating, Engraving, Heat Treating, and Allied Activities (33281), and Recyclable Material Merchant Wholesalers (42393) (Table 5). The majority of LT Exposure to Harmful Substances or Environments claims across all industries were due to Exposure to Temperature Extremes (56%), followed by Exposure to Other Harmful Substances (35%) (Table 6).

Contact with Objects and Equipment

The sectors with the highest proportions of LT claims for Contact with Objects and Equipment were MIN (38.4%), MNF (34.9%), and OIL (33.1%) (Table 3). However, two of the top 5-digit NAICS for Contact with Objects and Equipment were in the WRT sector. The top five 5-digit NAICS industries for LT claims as ranked by PI for Contact with Objects and Equipment included Forging and Stamping (33211), Recyclable Material Merchant Wholesalers (42393), Metal Service Centers and Other Metal Merchant Wholesalers (42351), Plate Work and Fabricated Structural Product Manufacturing (33231), and Ferrous Metal Foundries (33151). Temporary Help Services (56132) was also highly ranked (Table 5). The majority of LT Contact with Objects and Equipment claims across all industries were due to Struck by Objects or Equipment (57%), followed by Caught in or Compressed by Equipment or Objects (24%) and Struck Against Object or Equipment (16%) (Table 6).

Overexertion and Bodily Reaction

The sectors with the highest proportions of LT Overexertion and Bodily Reaction claims were PSA (60.1%), HCSA (38.0%), and MNF (37.9%) (Table 3). However, every NORA sector had at least one 5-digit NAICS industry that was highly ranked for Overexertion and Bodily Reaction. The top five 5-digit NAICS industries for LT claims as ranked by PI for Overexertion and Bodily Reaction included Other Motor Vehicle Parts Manufacturing (33639), Ambulance Services (62191), General Warehousing and Storage (49311), Nursing Care Facilities (62311), and Beer and Ale Merchant Wholesalers (42481) (Table 5). The majority of LT Overexertion and Bodily Reaction claims across all industries were due to Overexertion Involving Outside Sources (80%) followed by Other Exertions or Bodily Reactions (14%) (Table 6).

3:4: Comparisons to BLS SOII and other WC Systems

Overall proportions of injury events/exposures from OHBWC insured private employers were similar to those among private employers in data from the US national and Ohio BLS SOII.²⁵ In 2017, Overexertion and Bodily Reaction represented the largest share of LT injuries in all three datasets (BLS National, 33.6%; BLS Ohio, 33.5%; OHBWC, 31.0%), followed by Contact with Objects and Equipment (BLS National, 26.0%; BLS Ohio, 29.3%; OHBWC, 23.7%), Falls, Slips, Trips (BLS National, 25.8%; BLS Ohio, 25.8%; OHBWC, 30.1), Transportation Incidents (BLS National, 5.5%; BLS Ohio, 4.5%; OHBWC, 8.4%), Violence and Other Injuries by Persons or Animals (BLS National, 4.5%; BLS Ohio, 3.9%; OHBWC, 2.8%), Exposure to Harmful Substances or Environments (BLS National, 4.3%; BLS Ohio, 2.8%; OHBWC, 3.1%), and Fires and Explosions (BLS National, 0.1%; BLS Ohio private, 0.0%; OHBWC, 0.3%).

Similar overall proportions of injury events/exposures for 2014-2017 were also found in other state WC systems in CA, MA, MI, TN, and WA, and these are summarized in Table 7. The definition of LT varied in these states and there are a number of other system differences that preclude direct comparison. Although the WA WC system data are perhaps most comparable to OHBWC in terms of the type of administrative data, size, and timeframe, there are still notable differences. For example, the WA data include public employers, and the event/exposure data fields are manually coded. Also, the WA LT cases are defined as having four or more DAW. Despite these differences, shares of injury event/exposures types in total (both LT and MO) claims in OHBWC data from 2007-2017 are similar to those reported for state-insured employers in WA State from 2007-2017 among all accepted WC claims (both LT and MO).²⁶ The count ranking of event/exposure type is the same, although there are differences in their shares, especially for Contact with Objects and Equipment and Overexertion and Bodily Reaction (Table 7). These differences in reported events/exposures could reflect differences in the mix of employer types and industries within each dataset. As noted, this study did not include public employers, and overall there were a greater proportion of MNF and HCSA employers in the OHBWC dataset.²⁷

Table 7:

Comparison of event/exposure proportions in selected WC systems 2014-2017

Event/Exposure a	Lost-time b + medical-only count %						Lost-time b claim count %
	OH 2014-2016	OH 2015-2017	OH 2007-2017	WA 2007-2017	CA 2015-2017	TN 2014-2016	OH 2007-2017	WA 2007-2017	MA 2014-2016	NCCI 2016
Violence and Other Injuries by Persons or Animals	4%	4%	3%	4%	3%	5%	2%	3%	4%
Transportation Incidents	3%	3%	3%	3%	2%	4%	7%	4%	4%	5%
Fires and Explosions	0.08%	0.07%	0.09%	<1%	<1%	0.14%	0.4%	<1%	0.03%
Falls, Slips, Trips	21%	20%	20%	17%	15%	20%	31%	23%	22%	28%
Exposure to Harmful Substances or Environments	5%	5%	5%	5%	3%	5%	3%	3%	3%
Contact with Objects and Equipment	46%	46%	46%	37%	25%	35%	23%	19%	15%	31%
Overexertion and Bodily Reaction	22%	22%	23%	31%	37%	27%	33%	45%	30%	31%
Non-classifiable	0.33%	0.38%	0.32%			3%	0.38%		22%
Other				2%	9%			3%		6%
Multiple					6%

^aNIOSH cross-walked the event/exposure from WCIO codes in CA, TN, and NCCI

^bLost-time definition: OH (8 or days away from work, DAW); WA (4 DAW), MA (6 DAW), NCCI states (4-8 DAW)

The National Council on Compensation Insurance (NCCI) recently published 2016 data for LT claims for over 30 states combined. The definition of LT in these states ranged from 4 to 8 DAW.²⁸ NCCI reported WC LT claim frequencies using WCIO codes for injury cause, nature, and part of body²³ and NIOSH applied the WCIO-OIICS crosswalks to the published NCCI data report. The event/exposure proportions for LT claims as reported by NCCI for 2016 were highly correlated (0.92) with the proportions in the OHBWC insured private employer data for the same period. Comparisons are limited because NCCI did not report WCIO code proportions for several smaller OIICS event/exposure types (Table 7).

4: Discussion

4.1: Machine Learning Comparisons

This study used an algorithm based on logistic regression⁸ to determine 1- and 2-digit OIICS event/exposures (up to forty-six categories). Overall, findings were consistent with a prior study¹⁵ that used an earlier algorithm based on Bayesian analyses to determine three broad categories of causation.¹¹ Additionally, more recent published data²⁹ indicated that shares of the three broad causation categories [musculoskeletal disorders due to overexertion and bodily reaction (31%), slips, trips, falls (33%), and other causes (35%)] were very similar to shares found in this study for 1-digit event/exposure divisions [Overexertion and Bodily Reaction (33%) and Falls, Slips, and Trips (31%) and all other 1-digit events/exposures (35%)] for the same time period and population. Proportions were also similar at NORA industry levels (data not shown).

These findings demonstrate the consistency between results, despite different statistical approaches, and this provides some support for the validity of both methods. The results also show that the new algorithm builds on the prior research primarily by expanding the ability to code subcategories of causation within Falls, Slips, Trips (for example falls from height versus on level) and further defining other event/exposures by sub-type (Contact with Objects and Equipment, Transportation Incidents, Exposure to Harmful Substances or Environments, Violence and Other Injuries by Persons or Animals, and Fires and Explosions). This is an important advance. Compared to the three broad categories, the additional levels of detail help to develop prevention insights and the OIICS categories are being used by a growing number of comparable data sources (such as BLS SOII and other state WC data) with results coded at this level of causation and more detailed levels.

Currently, the coding of WC claims is a manually intensive process that is conducted by a number of parties within the WC system (insurers, claims administrators, employers, and state bureaus) using a variety of coding systems (WCIO, OIICS, etc.). This study provides additional evidence that the use machine learning approaches can likely be expanded in WC and related systems to improve the efficiency and accuracy of claims coding and harmonize the coding of narratives to utilize the BLS OIICS coding system in a manner similar to the SOII. NIOSH and other partners are continuing to share open source algorithms and technical assistance to WC bureaus and others interested in applying the methods.

4.2: Comparisons to BLS SOII and other WC Systems

Comparison of OHBWC data to SOII are limited because event/exposure divisions for the BLS SOII data are only available for cases with one or more DAW. By contrast, the OHBWC LT cases are defined to have eight or more DAW. There are also several other differences between the data collection systems, such as the sizes and industry types of employers included, because the OHBWC data only includes insured employers, which generally have 500 or fewer employees.⁵ SOII data are collected from a national sample of employers via surveys, whereas OHBWC data are population-level, collected from all insured employers. Despite these system differences, proportions and rates of LT OHBWC claims by 1-digit injury event/exposure divisions were largely similar to those among private employers in data from the US national and Ohio BLS SOII.²⁵ Differences in proportions are likely a function of the higher relative severity of OHBWC LT claims, differences in injury patterns nationally versus Ohio, the absence of claims from self-insured employers in the OHBWC data, and other noted differences compared with the BLS SOII and state WC systems.

Other WC data systems may contain data that are more similar to OHBWC data than SOII, but generally these data are not publicly available for the same time period (2007-2017) except for data from WA State and NCCI. A NIOSH surveillance grant resulted in such data becoming more readily available in several states^1–4 for more recent years 2014-2017 and moving forward. Despite differences in state WC systems, industry mixes, and covered employer populations (private employers with 500 or fewer employees in some states, and all employers in other states), shares of events/exposure types in total claim counts in data published by NCCI and other individual states for 2014-2017 were similar to those in OHBWC data. However, OHBWC total claim (LT+MO) data did tend to have a higher proportion of Contact with Objects and Equipment compared to other states (Tables 4, 6, and 7).

In summary, although there are many differences between the BLS SOII and state WC systems, similar overall proportions of injury event/exposures for 2014-2017 were generated by these data sources, especially when comparing injury measures using similar severity criteria.

4.3: Event/Exposure Time Trends

Counts and rates of WC LT claims among state-insured private employers in OH declined substantially from 2007-2017 (46% decline in rate per 100 FTE), and downward trends for this time period have also been noted nationally in WC systems and the BLS SOII for all claims overall and LT claims specifically. Overall US private industry SOII LT rates (1 or more DAW or restricted duty) for private industries dropped from 1.22 per 100 FTE in 2007 to 0.89 in 2017 (−26.8% decline).^25,30 NCCI noted a 19% decline in the number of LT claims per unit of earned premium for over thirty states (with the LT definition ranging from 4 to 8 DAW) from 2011-2016.²⁸ Reasons for this national trend may include safety and technological improvements,^28,31,32 underreporting,³³ and economic factors such as unemployment trends.^31,34 Understanding these national trends is a complex task that requires additional discussion and research beyond the scope of the current analyses.

Despite declining counts and rates, the ranking of leading injury event/exposures remained largely unchanged among OHBWC-insured private employer LT claims from 2007-2017 (Table 2). Trends of leading injury event/exposures did differ by NORA industry sector (Tables 3 and 4) and specific 5-digit NAICS industries (Table 5), and this is important to understand to develop tailored prevention programs.

4.4: Industry and Prevention Insights by Event/Exposure

Statistics that point to high numbers and rates of certain types of injuries in certain industries are only useful if they assist in targeting actions to reduce those injuries. The downward trends we have seen may indicate that some progress has been made in Ohio and nationally in developing effective OSH programs and interventions to reduce injuries. But challenges remain, especially in certain industries, to develop new interventions and perhaps increase adoption of existing approaches for specific event/exposures. This section of the discussion provides a brief overview of available methods of injury reduction and relates them to the needs of industries with high rates of major injury types that we have identified in Ohio. The purpose is to help point the way to more detailed matching of interventions to industries that still more detailed analysis of workers compensation data could support in the future.

On the broadest level, research has indicated that OSH programs that include several integrated elements (management commitment, employee participation, hazard identification, hazard control, training, and program evaluation) can be effective in reducing workplace exposure and injuries.^35–40 These programs and their elements can be evaluated by using leading indicators to measure performance,⁴¹ develop specific program improvement plans, and eliminate future injuries.

The OSH principle of the hierarchy of controls⁴² also suggests that hazard elimination and control is most effective through fundamental design and process changes, followed by engineering controls and equipment to improve physical and cognitive aspects of the worksite. If these approaches are not sufficiently protective or practical by themselves, personal protective equipment (PPE) and administrative controls to reduce an individual worker’s exposure through practices such as job rotation and job enlargement are the next most effective approach. Specific examples of OSH programs and interventions are discussed below for industries with high risks for each major event/exposure type.

4.4.1: Overexertion and Bodily Reaction

Overexertion and Bodily Reaction represented the highest proportion of LT claims in all sectors combined (33.0%) among OHBWC-insured private employers (Table 3). The majority of these injuries were due to Overexertion Involving Outside Sources (Table 6). Many of the top 5-digit NAICS industries for Overexertion and Bodily Reaction LT claims (Table 5) involve heavy manual material handling in manufacturing or warehousing/delivery or patient handling activities. Two of these industries were also highlighted in recent OHBWC-NIOSH claim analyses for Ambulance Services⁴³ and Nursing Care Facilities.⁴⁴

There is some evidence that broad ergonomic programs are effective at reducing overexertion-related injuries.^45–47 Specific equipment interventions can also be effective. A number of studies have demonstrated the effectiveness of patient handling equipment in reducing Overexertion and Bodily Reaction injuries in healthcare settings,^48–53 and material handling equipment in construction⁵⁴ and manufacturing settings.^55–58 Patient-handling intervention examples include powered cots (to move patients to/from vehicles), gurney lifts (to move patients to/from beds), and transfer sheets. Construction intervention examples include electrical cable feeding/pulling systems, concrete sawing equipment, skid steer attachments for concrete breaking, and boom lifts. Manufacturing material handling intervention examples include lift and tilt tables, powered hand trucks, jibs and cranes, and vacuum lift assists.^. Model ergonomic programs,⁴⁰ methods for ergonomic risk factor identification,⁵⁹ and compilations of control interventions and best practices are available online for specific industries including construction,^60–62 agriculture, manufacturing,⁶³ mining, healthcare,⁶⁴ wholesale retail trade,^62,65 and public employers.⁶⁶

4.4.2: Falls, Slips, and Trips

Falls, Slips, and Trips represented the second highest proportion of LT claims in all sectors (31.4%) among OHBWC-insured private employers (Table 3). Falls to Lower Level represented 29% of these injuries (Table 6). The number of transportation-related industries in the top 5 LT PI rankings (Table 5) highlights the importance of Falls, Slips, Trips prevention in this sector. Many of these injuries occur while workers are entering/exiting vehicles, especially in wet and icy conditions. WA State has developed a special emphasis program called Trucking Injury Reduction Emphasis (TIRES)⁶⁷ and maintains a specific website⁶⁸ to highlight prevention activities. Example interventions include slip-resistant footwear and improved vehicle access.

Construction industries also have known hazards for Falls, Slips, Trips including the use of ladders and scaffolding equipment. Fall prevention in construction (especially among roofers) has been an emphasis area for many years⁶⁹, and specific fall protection standards and awareness campaigns have been developed.^{70, 71} Example Falls, Slips, Trips interventions include a reduction in ladder use by building stairways early in the construction process,⁷² improved ladder/scaffolding designs to improve stability and detect unsafe conditions, reduction of worksite clutter, and use of slip-resistant footwear.⁶⁹

Falls, Slips, Trips prevention programs are also important in healthcare industries and involve several elements including slip-resistant footwear and flooring, and best practices to reduce slippery conditions and trip hazards.^73–75 Slip resistant footwear has been shown to be effective in variety of industries, including food service^76–78 and healthcare settings service.⁷⁵ Slip-resistant flooring has also been shown to be effective in reducing the frequency and cost of WC claims.⁴⁹

4.4.3: Contact with Objects and Equipment

Contact with Objects and Equipment represented the third highest proportion of LT claims in all sectors (22.5%) among OHBWC-insured private employers (Table 3). The majority of these injuries were due to Struck by Objects or Equipment, followed by Caught in or Compressed by Equipment or Objects (Table 6). Many of the top 5-digit NAICS industries for Contact with Objects and Equipment LT claims (Table 5) involve the usage of machinery with known hazards for Contact with Objects and Equipment injuries.

The use of standardized machine guarding review checklists is encouraged in metal-working and similar industries to identify hazards as well as general safety program assessments to identify larger system deficiencies.^79–81 Such programs are important in all industries but may be especially critical in industries involving the work of temporary help services, where workers have less safety training,⁸² shorter tenure, and higher injury rates for Contact with Objects and Equipment claims^83,84 compared to peers performing similar work.

Once hazards are recognized, a number of specific interventions have been shown to be effective in reducing Contact with Objects and Equipment injuries. This includes machine guarding, safety saws,⁴⁹ and active suspension seats to reduce whole body vibration among truck drivers.⁸⁵ PPE, including safety glasses,^86,87 gloves,^88,89 and general awareness campaigns⁹⁰ has also been shown to help reduce Contact with Objects and Equipment injuries when integrated as part of an overall safety program. The NIOSH Manufacturing Sector NORA Council recently developed a Hazardous Energy Control (Lockout and Other Means) Resource Guide to help summarize approaches to prevent Contact with Objects and Equipment injuries in the manufacturing industry.⁹¹

4.4.4: Transportation Incidents

Transportation Incidents claims represented the fourth highest proportion of LT claims in all sectors (7.0%) among OHBWC-insured private employers (Table 3). The majority of these injuries were due to Roadway Incidents, Motorized Land Vehicles (Table 6). Many of the top 5-digit NAICS industries for Transportation Incidents LT claims (Table 5) require workers to spend the majority of their work shift driving vehicles, which increases opportunities and hazards for Transportation Incidents. Several prevention programs and interventions have been developed to address these hazards.^92–94 For example, the WA SHARP TIRES program⁶⁷ highlights many interventions designed to prevent transportation incidents. Certain interventions such as in-vehicle monitoring systems^95–97 have been shown to reduce Transportation Incidents from occurring. Other interventions have focused on improving survivability after incidents through improved vehicle design and engineering.⁹⁸ Newer technologies are rapidly developing, including systems to monitor driver fatigue and proximity sensors to detect where workers are located throughout worksites to alert when possible Transportation Incidents are imminent.⁹⁹

4.4.5: Exposure to Harmful Substances or Environments

Exposure to Harmful Substances or Environments claims represented the fifth highest proportion of LT claims in all sectors (2.8%) among OHBWC-insured private employers (Table 3). Most of these injuries were due either to Exposure to Temperature Extremes or Exposure to Other Harmful Substances (Table 6). Although the top LT PI for Exposure to Harmful Substances or Environments was dominated by MNF industries, several SRV industries (restaurants, dry cleaning/laundry services and janitorial services) and one WRT industry were also represented. Many of these industries involve exposures to temperature extremes as well as chemical, electrical, and non-ionizing radiation hazards (Table 5).

WC data systems predominantly capture injuries versus long-term illnesses.¹⁰⁰ It follows that WC claims frequencies due to Exposure to Harmful Substances or Environments are undercounted to a greater extent than other event/exposure types. However, it is still useful to conduct analyses of WC claims data because other data sources are lacking. This helps to highlight industries such as restaurants and janitorial services that might be typically overlooked.

Heat stress prevention among outdoor workers and those indoors in manufacturing and other hot environments has been a focus of NIOSH, OSHA, and others for some time.¹⁰¹ These groups recently developed a smartphone app that features real-time, location specific, heat indices and hourly forecasts as well as tailored recommendations.¹⁰² Additional detailed WC claims analyses are also providing heat stress prevention insights. ¹⁰³

Reducing chemical exposures has also been one of the main prevention focus areas of NIOSH, OSHA and other organizations for many years.^104–108 General prevention approaches should follow the hierarchy of controls, where process changes or substitution of materials are used first to eliminate or reduce some exposures. Industrial ventilation controls are then typically used next to reduce hazards^109–112 PPE, such as respirators,^113,114 and administrative controls, such as reducing exposure times through job rotation and enlargement, are used as last steps when exposures cannot be further reduced through the other more effective methods. OSHA also provides online guidance for prevention of Exposure to Harmful Substances or Environments for specific industries such as restaurants.¹¹⁵

Despite known control methods, respiratory protection standard violations are still some of the most frequent OSHA citations.¹¹⁶ and occupational exposure information is lacking in the US.^117–119 Standardized data collection forms may enable future pooling of data in OSHA and WC systems,¹²⁰ and new direct reaching sensor technologies may be used to understand more fully workplace exposures.¹²¹

4.4.6: Violence and Other Injuries by Persons or Animals

Claims related to Violence and Other Injuries by Persons or Animals represented the sixth highest proportion of LT claims in all sectors (2.5%) among OHBWC-insured private employers (Table 3). Most of these injuries were due to Intentional Injuries by Person (Table 6). Many of top 5-digit NAICS industries for Violence and Other Injuries by Persons or Animals LT claims involve potential violence from patients, hazards for animal related injuries, or sports-related violence (Table 5).

Violence directed at caregivers in HCSA settings is a recognized concern,^122–124 although it is likely underreported.^125,126 Workers in nursing homes and residential mental health and substance abuse facilities may be especially vulnerable because those receiving care may lack the ability to control behaviors that lead to violence towards caregivers.^123,127 Staff shortages may also increase the possibility that the caregiver is working alone, which may increase the risk of violent incidents.

Integrated violence-prevention programs can be effective in preventing violence-related incidents in HCSA facilities.^127,128 Such programs include engineering controls (controlled access areas within facilities and security systems to alert additional staff of situations), standing health and safety committees, and training for staff and management.^129–131

Prevention of bites and other animal-related injuries in the veterinary and pet care industries is also a priority.^132,133 Interventions include engineering controls (such as passive restraints and animal handling equipment) as well as specialized training to recognize situations that may trigger aggressive behavior in pets, body language that precedes attacks, and methods for de-escalation.¹³⁴

4.4.7: Fires and Explosions

Fires and Explosions represented the lowest proportion of LT claims in all sectors (0.4%) among OHBWC-insured private employers (Table 3). The majority of these injuries were due to Explosions (Table 6). The number of WRT and SRV industries in the top rankings (Table 5) points to the importance of applying Fire and Explosion prevention principles beyond heavy manufacturing industries. As an example, Fires and Explosions have become a growing concern among recycling operations, and a recent fire prevention and management program has been developed specifically for this industry.¹³⁵ As another example, a number of explosions in WRT and SRV are related specifically to tire explosions, and prevention approaches for these incidents have been developed and are available online.¹³⁶

Although the relative frequency of Fires and Explosions was low, these events clearly can be high severity incidents that result in disabling injuries or death. Process safety is a focused OSH field that deals with the prevention of large-scale, catastrophic events such as building fires and plant explosions.¹³⁷ OSHA has developed a fire-prevention plan standard and numerous regulations have been designed to identify electrical, mechanical, and chemical circumstances that could lead these events.¹³⁸

5: Limitations

As with other data sources, there are several limitations associated with the use of WC claims data for OSH surveillance. This includes underreporting, which differs by industry, especially for illnesses.^139–147 There are also still a limited (although growing) number of comparable datasets because there are several differences between state-based WC systems and the BLS SOII. This analysis also was focused on OHBWC-insured private employers and does not include self-insured or public employers. There are also limitations to the FTE estimation methods used as described previously.⁵ Finally, this study employed machine learning algorithms to determine event/exposure for most claims. Despite these limitations, several states and organizations^1–6 have recently demonstrated that WC data are useful in augmenting traditional surveillance sources such as BLS SOII and the Census of Fatal Occupational Injuries (CFOI) to provide insight into injury causes, identify higher risk industries, and develop prevention approaches.

6: Conclusions

Although counts and rates of WC LT claims declined for all injury types among OHBWC-insured private employers from 2007-2017, the relative ranking of leading injury event/exposures remained largely unchanged. The majority of claims were due to three main event/exposure types: Overexertion and Bodily Reaction; Falls, Slips, and Trips; and Contact with Objects and Equipment. The other event/exposures in decreasing order of LT frequency include Transportation Incidents; Exposure to Harmful Substances or Environments; Violence and Other Injuries by Persons or Animals; and Fires and Explosions. Despite many differences between the BLS SOII and state WC systems, these systems all reported similar proportions of injury event/exposures for 2014-2017 and overall declines in injury counts and rates. However, the relative ranking of leading injury event/exposures differed by NORA industry sector and specific NAICS industries in OH and other states, and this is important to understand to target prevention efforts and develop tailored prevention programs. These patterns and time trends may indicate that progress has been made in developing effective OSH programs and interventions to reduce injuries.

Challenges remain, especially in certain industries, to develop new approaches and perhaps increase adoption of known prevention approaches for specific event/exposures. Available evidence on intervention effectiveness was summarized and mapped to the analysis results to demonstrate how the results can be used to guide future prevention efforts. As next steps, several focused industry type/exposure WC claims studies ^43,44 are being conducted to help identify actionable interventions for specific hazards.

This study and others^1–6 have shown that WC data are a useful and complementary source to the SOII and other survey data. Because WC data are a census, analyses can be conducted on a more detailed level, both by injury type and by industry within states. This study also demonstrated that machine learning approaches can be applied to large WC datasets to code claims narratives to detailed levels of causation. Such methods can be expanded and used by WC insurers, bureaus, and employers to improve the accuracy and efficiency of using WC systems to identify prevention priorities. NIOSH and other partners will continue to share open source coding algorithms and offer technical assistance. In conclusion, this study highlights the need for states to continue to use both SOII and WC as well as other data sources to identify specific prevention needs and leverage machine learning approaches to optimize and harmonize case coding processes.

7: Practical Applications

Employers can use these data to benchmark their safety and health performance against industry peers and develop data-driven plans for prevention. OSH practitioners and researchers can also use these data to understand industry differences in the level and mix of risks, as well as industry trends, and to tailor safety, health, and disability prevention services and research. WC bureaus, regulators, insurers, and employers can use the open source machine learning algorithms and methods to code narrative incident information within injury and illness tracking systems.