Review of Fall Accidents in Construction

1.1 Introduction and Literature Review

Of the construction workers that suffer fatal injuries, more are involved in falls than any other single cause. This has been a consistent statistic for many years. In a study of construction worker fatalities occurring between 1985 to 1989, it was determined that falls accounted for 33% (Department of Labor, 1990; Hinze, 1997).

Other researchers have examined the prevention of falls by various methods. Singh (2000) investigated fall accidents occurring on low-rise roofs and evaluated some innovative fall protection measures. He determined that there was no single method of fall prevention that would prevent all falls on low-rise roofs; however, prefabrication was determined to be the most promising method, followed closely by the personal fall arrest system (PFAS) and its variants. Duncan and Bennett (1991) reviewed the performance of various fall protection systems, and concluded that both active measures (those that prevent workers from falling, e.g., guardrails) and passive measures (those that protect workers after falling, e.g., safety belts) are useful in reducing fall injuries. Vargas et al. (1996a, b) developed an expert system for construction falls, which analyzed the causes of falls by using fault-tree methods. They concluded that guardrails, safety nets, and PFAS can all be inadequate, under differing circumstances.

Weisgerber et al. (1999) examined the "safety through design" approach which has strong practical implications for construction. They provided an outline of a comprehensive program to prevent falls at the design phase. Gambatese and Hinze (1996), developed a software program to assist designers in addressing safety in design. The software offered many design suggestions, with nearly a third devoted to fall prevention.

Fall prevention was also addressed by OSHA (1998) which has suggested several methods to control fall hazards. These methods include the elimination or substitution of the operation which can lead to falls, the use of engineering controls to guard against falls, informing/reminding workers-at-risk to avoid fall hazards (through warnings and administrative controls such as training and inspections), and the appropriate use of personal protective equipment (PPE).

OSHA has addressed fall prevention in its regulations in recent years. The OSHA revisions to the regulations of particular interest are the following:

(1) 1915.159 Personal fall arrest systems (PFAS). The 1996 revised regulations stipulated that it is no longer acceptable to use body belts as a personal fall arrest system. Body harnesses were mandated for PFAS to provide proper protection to workers to prevent falls.

(2) 1915.160 Specifications on positioning device systems. This revision, which became effective on January 1, 1998, stated that a positioning device system was not to be used for fall prevention. As a result of this change, only properly tied-off body harnesses are regarded as qualified personal fall arrest systems.

These changes to the OSHA regulations were intended to drastically impact the incidence of fall accidents. Similar conclusions were reached by NIOSH safety researchers in which they found several opportunities for employers and employees to be more vigilant in fall prevention. NIOSH concluded that “it is essential for employers to develop and implement comprehensive, written fall protection programs where workers are exposed to fall hazards.” (NIOSH 2000).

1.2 Research Methodology

The purpose of this research was to examine the causes of construction fall accidents and to identify any trends related to fall accidents. A better understanding of the causes of accidents can provide insight into accident prevention. Since falls have been the focus of recent OSHA changes in the regulations, it was also of interest to determine if these modifications might have impacted fall injuries in the construction industry.

The database used in this study is one maintained by OSHA. The initial intent was to utilize the data that are already available on the Internet on the OSHA homepage. While some of the data were contained in this database, the database had not been fully updated since 1996. Since the more recent data were of particular interest in this study, OSHA was contacted and informed of our research needs. The researchers were informed that the updating of the data in the website would not occur in the immediate future. A special request was made to obtain the files directly from OSHA and this was granted. The data were provided in Microsoft ACCESS format and were easily converted to files that could be input into the Statistical Package for the Social Sciences (SPSS).

The data provided by OSHA included all reported OSHA investigations of fatalities and serious injuries from January 1990 through October 2001. The analysis of the data examined all falls in the construction industry in that time period. Subsequent analysis was focused on the most recent years, the period in which the new fall standards were implemented and the period when additional information was being included in the injury records.

1.3 RESEARCH FINDINGS

1.3.1 Description of Fall Accidents

The data for the time period from January 1990 through October 2001 included a total of 7,543 OSHA-investigated accidents. Falls (both from elevation and from the same level) accounted for 34.6% of the injuries (See Figures 1.1 and 1.2). It is obvious that the proportion of falls has increased with time in the past 12 years. The average proportion of falls was 34.1% during the years before 1996, and it increased to 38.4%

in the following years. The total number of OSHA investigated construction accidents is relatively constant during the years. A simple analysis reveals that the Pearson’s correlation between the proportion of the falls and the year is 0.841, which provides strong evidence that the proportion of fall accidents increases during the past 12 years.

In the analysis of the other main types of accidents, the Pearson’s correlations between the years and the proportions of accidents are: -0.492 for struck-by, -0.232 for shock, 0.469 for caught-in-and-between, and -0.672 for the other accidents. The data may suggest that the proportion of caught-in-and-between accidents has also increased during the years, while proportions of struck-by and shock accidents have decreased. For some unknown reason, the proportion of “other” accidents has decreased.

Among the 7,543 construction accidents investigated by OSHA between January 1990 to October 2001, 2,741 were falls, with 2,687 falls being from elevation and 54 were from the same level. These resulted in 2,955 OSHA recordable fall injuries, with some accidents involving multiple workers. In the analysis, records were found that lacked some information. For this reason, the total number of cases analyzed for different descriptors may vary.

1.3.2 The time of occurrence of falls

The occurrence of falls was examined in regard to the distribution of the timing of their occurrence. July is the month when the occurrence of accidents reaches a peak, while February is the month with the least accidents. This pattern probably reflects the heightened amount of construction activity occurring in the summer and the reduced level of activity in the winter. The distribution of fall accidents seems to be consistent with the occurrence distribution of other accidents as well.

The distribution of fall accident occurrence was also examined in terms of the day of the week. This did not show a clear trend of occurrence, with an even distribution during the workweek, and an expected drop in injuries over the weekend. The analysis also considered fall occurrence in terms of the hour of the workday. The distribution of falls by hour of the day is similar to the pattern observed for other construction accidents. The least number of accidents occur during the lunch hour between noon and 13:00 o’clock and most accidents occur between 10:00 and 11:00 in the morning and between 13:00 and 14:00 o’clock in the afternoon. This is also similar to the distribution of accidents as noted in other research studies (Hinze, 1997).

1.3.3 Types of construction projects and falls

More detailed information was available from the OSHA investigations that took place between January 1997 and October 2001. While some earlier reports may have contained the information about the types of projects, this was not done either consistently or systematically. This results in the more recent years of data providing a richer resource about the information related to falls.

While the information about the types of projects does not implicitly describe the root causes of falls, the information may indicate areas where the focus on fall protection is perhaps most needed. This project information was examined. Findings show that fall accidents occurred more frequently on certain types of projects. Table 1.1 summarizes this information and shows that nearly half of the fall accidents occur on projects involving commercial buildings and single family or duplex dwellings. Whether commercial buildings are multistory or single story, the potential of falls remains a potential hazard. While exceptions exist, many single residential buildings are constructed by small contractors, these firms are known to generally provide relatively informal safety training and inadequate PPE (Glenn, 2000).

Table 1.1 Distribution of injuries by type of facility being constructed

Type of Facility	Falls		All Injuries
	Count	Percent	Count	Percent
Commercial buildings	404	33.4%	715	22.9%
Other buildings	212	17.5%	412	13.2%
Single family or duplex	211	17.4%	503	16.1%
Multi-family	113	9.3%	183	5.9%
Manufacturing plant	79	6.5%	168	5.4%
Tower, tank, storage elevator	71	5.9%	103	3.3%
Bridge	28	2.3%	94	3.0%
Other	92	7.6%	941	30.2%
Subtotal	1210	100.0%	3119	100.0%
Not Known	5		23
Total	1215		3142

Project costs were also examined in this study. The projects with the lower construction costs accounted for the largest number of falls, with nearly one half of the falls occurring on projects costing less than $250,000 (see Table 1.2). This might be expected as the number of smaller projects comprises the larger volume of the construction industry’s effort. The type of construction work performed was also examined. This showed that nearly 60% of the falls occurred on projects involving new construction, namely new projects or additions (see Table 1.3). This may also be an expected result as most construction work tends to be new construction.

Table 1.2 Distribution of injuries by cost of project

Project Cost	Falls		All Injuries
	Count	Percent	Count	Percent
Under $50,000	341	28.2%	990	31.7%
$50,000 to $250,000	229	18.9%	601	19.3%
$250,000 to $500,000	119	9.8%	289	9.3%
$500,000 to $1 million	134	11.1%	341	10.9%
$1 million to $5 million	188	15.5%	464	14.9%
$5 million to $20 million	117	9.7%	244	7.8%
$20 million and over	83	6.9%	191	6.1%
Subtotal	1211	100.0%	3120	100.0%
Not Known	4		22
Total	1215		3142

Table 1.3 Distribution of injuries by type of construction work

Project Cost	Falls		All Injuries
	Count	Percent	Count	Percent
New project or new addition	721	59.5%	1640	52.6%
Alteration or rehabilitation	219	18.1%	565	18.1%
Maintenance	189	15.6%	531	17.0%
Demolition	41	3.4%	101	3.2%
Other	41	3.4%	283	9.1%
Subtotal	1211	100.0%	3120	100.0%
Not Known	4		22
Total	1215		3142

1.3.4 Fall height

A total of 2,741 fall accidents were examined and of these, 1,018 provided information on the height of the projects and the number of stories. On projects where falls were involved, 807 projects (81%) were either one, two or three stories. The average facility height (whether a building or other type of structure) was 37.4 feet. From this it can be concluded that most falls happened on projects that were not particularly high. Perhaps fall prevention is viewed as a less serious concern on these lower elevation projects.

The data recorded since January 1997 also provided additional information of interest that was not available for accidents occurring in earlier years. In particular, the elevation from which the falls originated and the fall distances of the workers has been consistently documented since January 1997. Of the 1,215 falls that occurred since January 1997, more than 70% occurred at elevations of less than 70 feet. The distribution of the fall heights is shown in Figure 1.3. The average elevations of the fall height (the elevation where the fall originated) and fall distance are 35.4 and 34.9 feet, respectively. Findings show that more than 70% of the fall accidents occur at elevations below 30 feet. The elevation of a project can be regarded as one of the most hazardous aspects of the construction site. According to the OSHA regulations (CFR1926 Subpart M), fall prevention must be implemented at all elevations above six feet. It might be inferred that the implementation of fall prevention techniques might be lacking or inadequate on these lower elevation projects.

1.4 Injuries Resulting From all Accidents

The 2,741 fall accidents resulted in 2,955 injuries. The predominant occupations of the injured workers included construction laborers, roofers, carpenters, structural metal workers, painters, bricklayers and stonemasons, electricians, supervisors, drywall installers, plumbers and pipefitters. The types of fall injuries that are most common include fracture, concussion, and bruise/contusion/abrasion. Half of the injuries were head injuries, and about one third were multiple injuries. Fall injuries frequently involve injury to the chest, neck, back, abdomen, and legs. Regarding the severity of injuries, two-thirds Regarding the severity of injuries, two thirds of the workers involved in falls were killed, emphasizing the serious nature of fall accidents.

The ages of workers most frequently involved in falls are between 31 and 40, with the overall average age being 38.3. This is similar to the distribution of the ages of workers involved in other types of injuries. If the frequency of falls is calculated for different age groups, it is apparent that the proportions of age groups below 35 are lower than for those above 35. This may suggest that experience in construction may not necessarily lead to a decrease in fall accidents or that younger workers tend to be more alert and flexible when fall hazards occur.

1.5 Causes of Fall Accidents

1.5.1 Work tasks and fall occurrence

Falls may be associated with the particular work tasks being performed. This was examined in this study. Table 1.4 shows the most frequent types of tasks performed when fall accidents occurred. It was noted that the most common work tasks included roofing, erecting structural steel and exterior carpentry. Most of the tasks were conducted at elevation or on temporary structures, where fall hazards are often present. Falls occurred at relatively low elevations. About 75% of the roofing worker falls occurred at elevations of less than 30 feet, and 45% were less than 20 feet. Workers on these projects may underestimate or ignore the fall hazards at these seemingly low elevations. To avoid falls, any work done above 6 feet should be performed with care and with the appropriate fall protection.

Table 1.4 Distribution of injuries by type of work being performed

Work Activity	Falls
	Count	Percent
Roofing	252	21.6%
Erecting structural steel	99	8.5%
Exterior carpentry	89	7.6%
Exterior masonry	57	4.9%
Installing equipment (HVAC, etc.)	54	4.6%
Demolition	53	4.5%
Other	561	48.2%
Subtotal	1165	100.0%
Not Known	50
Total	1215

1.5.2 Location of falls

Over half of the fall injuries are related to environmental factors involving the working surface or facility layout conditions. Falls from roofs are the most frequent accidents, especially in commercial buildings and single family or duplex dwelling projects. Most falls took place from roofs, from/with structures, from/with scaffolds, from/with ladders, and through openings (see Table 1.5). These account for about 80% of all construction fall accidents. Provision of adequate preventive equipment in these locations is essential to avoid falls.

Table 1.5 Distribution of injuries by type of work being performed

Work Activity	Falls
	Count	Percent
Fall from roof	333	28.4%
Fall from/with structure (other than roof)	227	19.3%
Fall from/with scaffold	153	13.0%
Fall from/with ladder	133	11.3%
Other	328	27.9%
Subtotal	1174	100.0%
Not Known	41
Total	1215

1.5.3 Human errors resulting in falls

Analysis of human errors showed that misjudgment of the hazardous situation is the most frequent type of human error involving falls, accounting for about one third of all the accidents. The distribution of misjudgment for different ages is roughly similar to the age distribution of fall injuries, except for the age group between 21 and 25 where a disproportionate number of falls are the result of misjudgment (see Fig. 1.4). Human errors categorized as “misjudgment of hazardous situation” account for about 30 percent of the falls. It is possible that age or experience do not noticeably improve judgment where hazardous situations are concerned. Results suggest that workers between the ages of 21 to 25 should be trained and educated about fall hazards. Of course, it must be recognized that the “misjudgment of hazardous situation” is a subjective assessment and is difficult to quantify or verify.

Further analysis revealed that most falls associated with human errors occurred at lower elevations. For example, more than half of falls related to ‘misjudgment of hazardous situation’ were less than 20 feet in elevation, and 23.5% of them were less than 10 feet. Falls were also associated with ‘insufficient or lack of protective work clothing and equipment’ with 40% occurring at elevations of less than 20 feet.

Inadequate or inappropriate use of fall protection (PPE), and removed and inoperative safety equipment contributed to more than 30% of the falls. This situation does not change significantly after 1996, when the OSHA regulations on PPE were significantly revised. Typical examples include work being performed by workers who are not tied-off. There are also several falls that occurred when a body harness was unhooked to facilitate movement to a different location. Therefore, the adequate provision and proper use of fall protective PPE are necessary to ensure worker safety.

1.5.4 Immediate sources of falls

A contributing factor in one-third of the fall accidents was the working surface. The accidents included typical situations where workers slipped on sloped roofs and fell to the ground, workers fell through floor openings, and workers slipped on the walking surface of scaffolds and fell. Inadequate fall preventive equipment in buildings/structures, and/or failure of buildings/structures also caused some workers to fall. These falls can be effectively prevented by the use of the appropriate fall preventive equipment.

1.6 Distribution of Fall Accidents and OSHA Inspections

The states where falls occurred were also examined. The results showed that Texas had the most fall accidents, followed by Florida, New York, California, and Illinois. The Standard Industrial Classification (SIC) codes associated with falls were also examined. The primary SIC codes associated with falls were 1761 (Roofing, Siding, and Sheet Metal Work), 1791 (Structural Steel Erection), 1542 (General Contractors-nonresidential Buildings, Other Than Industrial), 1799 (Special Trade Contractors, Not Elsewhere Classified), 1731 (Electrical Work), and 1711 (Plumbing, Heating and Air-conditioning). The occurrence of falls is clearly related to the type of work performed.

When comparing fall accidents with all construction accidents, the average penalty on fall injuries is higher than on the aggregate of other accidents. For example, the average and median value of the imposed penalty for fall-related accidents were $8,917 and $2,250, respectively. This is in contrast to $7,757 and $1,800 for the overall accidents. It was also found that there are more serious violations in fall inspections than in overall inspections. The Bureau of Labor Statistics (BLS) website (http://www.bls.gov/iif) provided the most frequent citations to Construction Special Trade Contractors (SIC 17), that account for most (about 80%) falls. Nearly half of the citations are associated with fall hazards on the jobsite or the lack of training on falls. Clearly, emphasis on construction fall protection is warranted: scaffolding, ladder, fall protection training, training for scaffolding, and so on.

1.7 CONCLUSIONS AND RECOMMENDATIONS

Falls are the most frequent accidents that occur on construction sites. Results show that fall accidents in the construction industry are the cause of many serious injuries and fatalities. At the same time, the data analysis shows that falls have certain properties, which may help to devise preventive approaches.

Many parties have sought means to prevent falls, including OSHA, the construction industry, and various researchers. Some measures do not work as well as others. For example, since the OSHA regulations on PPE for fall prevention were revised in 1996, neither the quantity nor pattern of falls on construction sites has changed significantly. In fact, the proportion of accidents that are caused by falls has actually increased. This may stem from the strong economy that the U.S. construction industry has enjoyed in the years following 1995. The strong economic growth has

resulted in the hiring of many workers, a large proportion of whom may be inadequately trained. Clearly a continued focus on falls by OSHA is well warranted and more training of the workforce is needed as well. It is not clear what the incident rate of fall injuries would have been had there not been a clear focus on fall prevention in recent years.

For the construction industry, fall hazards analysis and communication of related findings are necessary to ultimately impact the occurrence of fall accidents. It was noted that falls commonly occur on projects that can be characterized as being small and relatively low in cost, and involve new construction of commercial buildings and residential projects.

It is evident that many fall hazards go unnoticed or that efforts to prevent falls are not effectively implemented. Fall hazards on sites should be detected through rigorous inspections of construction sites and eliminated through effective preventive approaches. The accumulation of information on previous accidents can disclose the most common hazards on construction sites. Operations particularly susceptible to falls include roofing, erecting structural steel and exterior carpentry. Falls are often associated with workers on roofs, scaffolds, ladders, and on floors with openings. Occupations such as construction laborers, roofers, carpenters, and structural metal workers are commonly involved in falls and should be specifically addressed through fall prevention efforts. Fall hazards mapping, as suggested by Gambatese and Stewart (1999), can serve as a very useful technology to indicate where fall hazards exist.

Through the analysis of past fall accidents, documented fall-related near misses, as well as fall-related citations, the most hazardous locations on sites can be identified. Providing fall preventive equipment to workers, including full-body harnesses, along with the proper training, should reduce the number of falls. The lack of safety training is often the root cause for many falls. According to the analysis, misjudgment of workers may account for about one third of the construction worker falls. Fall prevention training can be effective in addressing numerous causes of accidents. Traditional safety training, restricted to the verbal and manual descriptions of the OSHA regulations, may not be sufficient to enable the workers to detect and eliminate all fall hazards. Innovative training approaches should be explored.

Allan St John Holt (2001, page 159) stated: “fall prevention is far more effective than fall protection, which often involves personal protective equipment. Reliance on people to make the ‘right’ decision about wearing personal protective equipment has been shown by events to be unsatisfactory – they forget, decide not wear it in view of the expected short exposure time, or do not wear or use it correctly. The first stage in fall prevention is during the design process, which influences the construction method.” While addressing worker safety in the design phase is a relatively new concept, it is clear that safer plans, developed through thoughtful design decisions, can decrease the occurrence of falls.

For safety researchers, many topics related to falls need to be investigated in greater detail. For example, the current personal fall arrest systems (PFAS) can effectively protect workers after they fall from elevation. While these may constrain the movement of workers, as with steel erection operations, such approaches should be considered. Motivating workers to fully utilize the safety equipment at all times is perhaps the greatest challenge. Some workers fell because they did not tie-off their body harnesses, either because they felt it was troublesome to be tied off to a fixed anchorage or when they unhooked the lanyards to change their positions. Such practices must change. More flexible PFAS might be able to save more lives. Different kinds of new technology, which can help prevent falls and protect workers from injury by falls, should be developed.

1.8 Acknowledgements

Special gratitude is extended to Dr. J. DuBois of OSHA, who has provided considerable assistance during the collection and analysis of the data. Gratitude is also extended to the Construction Safety Alliance for its financial support of this research effort.

1.9 REFERENCES

Department of Labor, Occupational Safety and Health Administration, 1990, Analysis of Construction Fatalities – The OSHA Data Base 1985-1989. (OSHA, Washington, D.C.).

Duncan C. W., and Bennett III, R., 1991, Fall Protection and Debris Containment During Construction. In Proceedings of the ASCE Construction Congress (second), Preparing for Construction in the 21st Century, (American Society of Civil Engineers, New York, N. Y.) pp. 97-102.

Gambatese J. A., and Stewart, P. J., 1999, Application of risk mapping to construction project jobsites. In Proceedings of the 2nd International Conference, CIB W99, Implementation of Safety and Health on Construction Sites (Balkema Publishers, Rotterdam, Netherlands) pp. 785-792.

Glenn, W., 2000, Safety in Residential Construction. Master Thesis, 2000. (The University of Florida).

Hinze, J., and Gambatese, J., 1996, Addressing Construction Worker Safety in the Project Design. Construction Industry Institute, research report 101-11, (The University of Texas at Austin).

Hinze, J., 1997, Construction Safety. (Prentice Hall, New Jersey).

Holt, A. J., 2001, Principles of Construction Safety. (Blackwell Science, Malden, MA.)

OSHA, 1998, Fall Protection in Construction. OSHA Publication 3146, (OSHA, Washington, D.C.).

NIOSH, 2000, Worker Deaths by Falls: A Summary of Surveillance Findings and Investigative Case Reports, Publication No. 2000-116, (DHHS, NIOSH).

Singh, A., 2000, Innovative Fall Protection for Construction Workers on Low-Rise Roofs. In Construction Safety and Health Management. (Prentice Hall, New Jersy), pp. 87-114.

Vargas, C., Hadipriono, F., and Larew, R., 1996a, A Fault Tree Model for Falls from a Form Scaffolding. In Proceedings of the 1st International Conference, CIB W99, Implementation of Safety and Health on Construction Sites, (Balkema Publishers, Rotterdam, Netherlands), pp. 291-301.

Vargas, C., Hadipriono, F., and Larew, R., 1996b, Causes of Falls from Floor Openings and Edges. In Proceedings of the 1st International Conference, CIB W99, Implementation of Safety and Health on Construction Sites, (Balkema Publishers, Rotterdam, Netherlands), pp. 303-313.

Weisgerber, F. E., and Wright, M. C., 1999, Elements of a fall safety through design program. In Proceedings of the 2nd International Conference, CIB W99, Implementation of Safety and Health on Construction Sites, (Balkema Publishers, Rotterdam, Netherlands), pp. 867-874.