The World's Most Referenced Occupational Injury Dataset & What It Reveals About Hand Exposure
Fingers, hands and wrists remain among the most frequently injured body regions in U.S. occupational injury reporting systems, appearing consistently across manufacturing, construction, mining, logistics and offshore energy datasets. The United States maintains the largest and most cited body of federal occupational injury data in the Observatory — drawing on five independent regulatory and research agencies, each covering distinct industrial populations. This profile examines what those datasets collectively suggest about where hand exposure occurs and what reduction may be possible.
The United States does not operate a single unified occupational injury reporting system. Instead, hand and upper limb injury data is distributed across five federal agencies, each with distinct mandates, covered populations, reporting thresholds and data collection methods. Understanding this architecture is essential to interpreting U.S. injury data correctly — and to recognising where it provides reliable signal and where it has structural gaps.
The five-agency structure means that no single U.S. dataset covers the full working population with identical definitions. BLS SOII excludes federal government workers, self-employed individuals, farms with fewer than 11 workers, and some other categories. MSHA covers only mining. BSEE covers only offshore OCS. Cross-agency comparison requires awareness of these population boundaries.
The following statistics are drawn from publicly available U.S. federal datasets. They are presented as reported — without extrapolation beyond what the source data states. Where figures change annually, the pattern is described rather than a single year's number cited as definitive.
BLS, OSHA, MSHA and BSEE use different reporting systems, different covered populations and different injury thresholds. The fact that hand and finger injuries appear prominently across all of them — in manufacturing, mining, construction and offshore — is more significant than any single number from any single source.
A finding that appears in one dataset may reflect that dataset's design. A pattern that appears consistently across five independent systems is more likely to reflect an underlying reality about where hand exposure occurs in U.S. industry.
On underreporting: Multiple academic studies — including research published by NIOSH — have concluded that BLS SOII systematically undercounts actual injury incidence, with estimates of undercapture ranging from 30% to 70%. The true frequency of hand injuries in U.S. industry is therefore likely to be substantially higher than published statistics indicate. This does not invalidate the data for pattern analysis — but it should be borne in mind when using SOII figures for absolute comparisons.
BLS SOII, OSHA records and MSHA data all identify sector-level concentrations of hand and finger injuries. The following analysis draws on those datasets to describe which industries show elevated hand injury rates and what industrial characteristics appear to contribute to that concentration.
What is notable about U.S. hand injury data is not that it appears in one or two sectors but that it appears prominently across every major industrial sector tracked by federal agencies. Manufacturing, construction, mining, oil and gas, warehousing and agriculture all show elevated hand injury patterns in their respective regulatory datasets. The consistency of this finding across independent reporting systems suggests a structural characteristic of industrial work rather than a sector-specific problem.
U.S. federal injury datasets record outcomes. They do not record the specific task context that produced each injury. The following likely exposure drivers are inferences from available data patterns, sector characteristics and published NIOSH and OSHA investigation findings. The language used — appears to be, consistent with, a probable contributor, the data suggests — reflects the inferential nature of this analysis. These are not definitive conclusions about causation.
Observatory methodology: The drivers listed below are identified independently from the Australia profile. They are not mapped against a predefined framework. Where a driver that appeared in the Australia profile also appears here, that represents a data-driven convergence — not a template being applied. Cross-country comparison of drivers will occur in the separate Comparative Analysis publication.
OSHA severe injury data and BLS SOII both identify machinery contact — specifically "caught in or compressed by" events — as the leading cause of work-related amputations in U.S. manufacturing. This pattern, combined with the high volume of hand injuries in manufacturing, suggests that machine interaction during operation, servicing and jam clearing is a probable major contributor to hand exposure in this sector.
OSHA National Emphasis Programs on amputations have repeatedly identified inadequate machine guarding as a cited violation across manufacturing sub-sectors. The probable task contexts include jam clearing, adjustment during production, and maintenance activities where the hand enters the machine's operating zone.
Overexertion and contact with objects are the two leading event types for hand and upper limb injuries in BLS SOII data across multiple sectors. In construction, warehousing and manufacturing, the probable task context is manual positioning — guiding, placing, steadying or securing components by hand during installation, assembly or storage operations. The hand's presence at the point where two surfaces converge is a probable contributor to the caught-between and struck-by injuries recorded in this category.
BSEE offshore incident data and available OSHA records for onshore oil and gas operations consistently document hand and finger injuries in drilling and well-servicing contexts. The probable task contexts are tubular handling — running, making up and breaking out drill pipe and casing — where the hand is used to guide, align and steer pipe sections under crane or iron roughneck operations, and tong and slips operations where the hand is close to rotating or moving equipment.
OSHA crane and derrick standards (29 CFR 1926 Subpart CC) and enforcement data identify suspended load operations as a significant risk area in construction. While OSHA enforcement citations relate to equipment and procedural requirements, the injury data pattern — elevated struck-by and caught-between events in construction — is consistent with hand exposure during load guidance and final positioning. The probable task context is manual load control during the terminal phase of crane-assisted lifts in structural and industrial construction.
NIOSH research and OSHA enforcement data have consistently identified meat, poultry and food processing as among the highest-risk environments for severe hand injuries in U.S. manufacturing. The probable task contexts include manual feeding of cutting and slicing equipment, hand positioning during high-speed processing operations, and clearing and cleaning activities on cutting machinery. OSHA has cited inadequate machine guarding in this sector in multiple enforcement actions.
BLS SOII data consistently shows hand tools — powered and non-powered — as a significant source of hand and finger injuries across multiple sectors including construction, manufacturing and mining. The probable task contexts include impact tool operations where a stabilising hand is within the strike zone, cutting operations where the guiding hand is close to the cutting edge, and fastening operations where tool slip or recoil brings the tool into contact with the hand. This driver appears to be broadly distributed rather than concentrated in a single sector.
The six likely exposure drivers identified for the United States are derived from cross-agency data pattern analysis and published NIOSH and OSHA investigation findings. They represent probable contributors based on available evidence — not a definitive classification of U.S. hand injury causation. The cross-country comparison publication will examine where these drivers converge with or diverge from patterns observed in Australia, Canada and the United Kingdom.
The United States does not have a single unified hand safety standard. Regulatory requirements for hand protection and machine safety are distributed across OSHA standards, MSHA regulations and voluntary ANSI/ASSE standards. The following describes the primary instruments relevant to hand exposure in U.S. industrial settings.
Unlike Australia's WHS Act, which explicitly codifies the hierarchy of controls — placing elimination above PPE as a legal obligation — OSHA's General Industry and Construction standards do not establish an equivalent explicit hierarchy in most contexts. OSHA's PPE standard for hand protection (29 CFR 1910.138) requires employers to provide appropriate hand protection when workers' hands are exposed to hazards — but does not require employers to first demonstrate that engineering controls cannot eliminate the exposure.
This regulatory distinction is relevant to understanding why PPE-dependence persists in U.S. industry even where engineering controls might be practicable.
OSHA's most frequently cited standards by year consistently include 1910.147 (Lockout/Tagout), 1910.212 (Machine Guarding) and 1926.503 (Fall Protection Training). The persistence of machine guarding and lockout/tagout in OSHA's top citation list — year after year — suggests that the regulatory requirement exists but that compliance at the task level remains incomplete. This is consistent with the injury patterns recorded in BLS and OSHA severe injury data.
Voluntary vs. mandatory: A significant portion of the U.S. machinery safety standards architecture is voluntary rather than mandatory. ANSI B11 series, ANSI/RIA R15.06 (robot safety) and related standards are not directly enforceable by OSHA in most circumstances. This creates variability in machine safety practice across U.S. industry that does not exist in regulatory environments with mandatory machinery safety standards.
The U.S. federal injury data system is extensive and publicly accessible, but it carries structural limitations that are important to acknowledge before drawing conclusions from it. These limitations do not invalidate the data — they define its boundaries.
Multiple peer-reviewed studies — including research published by NIOSH — have found that BLS SOII substantially undercounts actual occupational injury incidence. Estimates of undercount range from 30% to over 60% depending on sector and injury type. Probable causes include employer incentives to minimise reported injuries (affecting safety programme metrics and insurance premiums), worker reluctance to report minor injuries, and misclassification of work-related conditions as non-occupational. For hand injuries specifically, minor lacerations and crush injuries that do not result in days away from work are likely to be particularly undercounted.
BLS SOII excludes self-employed workers. In construction — one of the highest hand-injury sectors — self-employed workers represent a significant share of the workforce, particularly in residential and specialty contracting. Agriculture has a small-farm exemption that similarly excludes a large proportion of agricultural workers. These exclusions mean that two of the highest-risk contexts for hand exposure are partially invisible in the primary national dataset.
The United States has no federal near-miss reporting requirement. OSHA's Severe Injury Reporting captures amputations and hospitalisations — the most severe end of the injury spectrum — but there is no national system for recording events where a hand entered the hazard zone without resulting in a recordable injury. This means the denominator of hand exposure events is unknown, and sites cannot use federal data to understand whether they are reducing exposure frequency or merely improving injury severity management.
BLS SOII, MSHA and BSEE each cover different worker populations with different reporting thresholds. A mining worker injured on an MSHA-covered site appears in MSHA data but may not appear in BLS SOII. An offshore worker on a BSEE-regulated platform appears in BSEE statistics but may not be separately identifiable in BLS data. Constructing a true national total for U.S. hand injuries requires combining data across agencies — each with different definitions, collection methods and injury classification systems.
BLS SOII event-or-exposure categories — "contact with objects and equipment," "overexertion and bodily reaction," "falls, slips, trips" — describe the physical event but not the task that produced it. A finger amputation during "caught in or compressed by" machinery could arise during jam clearing, during routine feeding of a machine, during maintenance with inadequate isolation, or during adjustment. The aggregate category does not distinguish between these task contexts, making task-level exposure analysis impossible from aggregate SOII data alone.
BLS SOII publishes annual data with approximately a 12–18 month lag from the reference year. The most recent published data therefore describes workplace conditions from the prior year. In fast-changing sectors — large-scale construction projects, new offshore developments, expanding e-commerce fulfilment — current exposure patterns may differ from the most recently published statistics. OSHA Severe Injury Reporting provides more current data for the most severe injuries but does not cover the full injury spectrum.
The following opportunities are identified from the intersection of U.S. injury data patterns, sector characteristics and available OSHA and NIOSH investigation findings. They are stated in technology-neutral terms. The direction of improvement is described — not the specific method. Site-level assessment is required to determine which approaches are practicable in each operating context.
The persistence of machine guarding and lockout/tagout in OSHA's most-cited violations list — alongside consistent amputation rates in manufacturing — suggests that the regulatory requirement to isolate machines before servicing is frequently not met. The exposure reduction opportunity is not primarily procedural. It is structural: designing machines so that service points are accessible without requiring hand entry into the operating zone, and creating physical separation between the hand and the machine's energy.
Reduction direction: increase physical separation between personnel and machinery operating zones during servicing; reduce the requirement for hand entry through design; improve isolation verification before any intervention.
Across construction, manufacturing and oil and gas, hand-guided positioning of heavy or moving components is a recurring context in the injury data. The exposure reduction opportunity is to reduce the dependency on hand placement for guiding, steadying and aligning components — particularly at the point where two surfaces are converging and the hand is between them.
Reduction direction: introduce remote handling where practical; reduce manual alignment requirements during component installation; increase physical separation between personnel and converging surfaces during make-up operations.
NIOSH research and OSHA enforcement data identify food processing — particularly meat and poultry — as one of the most consistently documented contexts for severe hand injuries in U.S. manufacturing. The exposure reduction opportunity is to reduce the frequency with which the hand must enter the zone adjacent to cutting, slicing and high-speed processing equipment during both production and servicing operations.
Reduction direction: introduce remote feeding and clearing where practical; reduce dependency on hand placement for product positioning on cutting lines; improve physical separation between the hand and cutting equipment during clearance operations.
Drill floor and tubular handling operations in U.S. oil and gas are documented in both BSEE offshore data and OSHA onshore records as elevated-risk contexts for hand injuries. The exposure reduction opportunity is to reduce the frequency and duration of manual hand contact with tubulars and drill string components during running, making-up and racking operations.
Reduction direction: reduce dependency on hand placement for pipe alignment and stabilisation; increase physical separation between personnel and rotating or moving tubulars; improve remote handling capability where operationally practical.
BLS SOII source-of-injury data consistently shows hand tools — powered and non-powered — as a significant source of hand injuries across multiple sectors. The exposure reduction opportunity is to reduce the frequency with which the non-operating hand must stabilise the workpiece or a secondary tool within the operating zone of the primary tool.
Reduction direction: reduce dependency on hand placement for workpiece stabilisation during impact operations; introduce fixtures or clamping where practical; improve task sequencing to pre-secure components before applying impact forces.
BLS SOII struck-by data and OSHA crane enforcement records both point toward suspended load operations as a probable contributor to hand injuries in construction and industrial work. The exposure reduction opportunity is to reduce the requirement for manual hand contact with loads during the approach and final positioning phases of crane-assisted lifts.
Reduction direction: increase physical separation between personnel and suspended loads; reduce manual alignment requirements during load landing; improve remote load control capability during terminal positioning phases.
Index methodology note: Exposure reduction opportunities are identified from data pattern analysis and do not constitute prescriptions. Specific methods, tools, systems or products are not named at the country profile level. The direction of reduction is described in task-structural terms. Site-level assessment — including task analysis, risk assessment and feasibility review — is required before any reduction approach can be evaluated for a specific operation.
The U.S. data raises a question that injury statistics alone cannot answer. BLS, OSHA and MSHA collectively represent one of the most extensive injury reporting systems in the world. Yet hand and finger injuries persist as a leading category across manufacturing, construction, mining and oil and gas — sectors with well-established regulatory frameworks. The question is not whether the data is adequate. It is whether outcome data alone is sufficient to address a problem that is structural in origin.
A site with ten recorded hand injuries per year and ten thousand hand-at-hazard events per year has a fundamentally different exposure profile from a site with ten injuries and one hundred such events. Current U.S. data systems — including BLS SOII, OSHA 300 logs and MSHA Form 7000-1 — record outcomes. None records how frequently hands entered hazard zones without resulting in injury. The denominator of exposure frequency is unknown from any public U.S. dataset.
OSHA's Severe Injury Reporting captures amputations and hospitalisations. BLS SOII captures days-away-from-work cases. Neither captures the frequency of hand-at-hazard events that did not produce a recordable outcome. Managing hand safety from outcome data alone addresses the tail of the distribution — the events serious enough to trigger a reporting obligation — without addressing the conditions that generate that distribution in the first place.
The HSF Exposure Elimination Framework™ is a conceptual framework that addresses this gap directly. Its central principle, derived from the cross-country evidence gathered by the Observatory, is:
Exposure elimination does not prescribe a single solution. Depending on the task and operating environment, it may involve task redesign, physical separation between the worker and the hazard, remote handling, mechanisation, automation, engineered safeguards, process modification or equipment redesign. No single approach is universally applicable. The appropriate pathway depends on the specific task interface and operational constraints.
Before selecting any approach to exposure reduction, the task interface must be understood. Where does this task require the hand to enter the hazard zone? How frequently? For how long? Is that requirement intrinsic to the task — or intrinsic to the current method of performing it?
The distinction matters. A task that genuinely cannot be performed without hand contact at the hazard interface presents a different challenge from a task that has simply always been performed that way. The U.S. data suggests both types are present — and that the two are not always distinguished in current safety investment.
Further detail: handsafetyfirst.in/hsf-exposure-elimination-framework