In Table 4, we present crude and multivariate rate ratios for variables included in the final model. Increased rates were found for those veterinarians who had 5 or fewer years of experience (RR = 3.1, 95% CI = 1.4–6.8) and 6–10 years (RR = 1.9, 95% CI = 0.9–4.0), compared with more than 20 years; smoked currently (RR = 4.1, 95% CI = 1.8–9.1); incurred prior injuries (RR = 1.7, 95% CI = 1.1–2.6); slept 6 or fewer hours per night on average (RR = 1.8, 95% CI = 1.0, 3.3); sat less than 1 hour per day (RR = 3.2, 95% CI = 1.6–6.3); stood less than 1 hour per day (RR = 1.7, 95% CI = 0.9–3.0); and participated in any type of sports (RR = 1.7, 95% CI = 1.0–2.6).
We found a dose-response effect in weight of patients lifted without assistance. Compared with lifting ≤40 lb, increasing rates were seen for lifting 41–75 lb (RR = 3.1, 95% CI = 1.6–5.9), 76–100 lb (RR = 3.2, 95% CI = 1.6–5.9), and >100 lb (RR = 6.1, 95% CI = 2.5–15.0). Those who did not use hydraulic lifts experienced a slight increase in risk (RR = 1.3, 95% CI = 0.7–2.5), and those who said the lifts were not needed experienced a large increase (RR = 5.9, 95% CI = 2.3–15.0). A decreased rate of injury was seen for the perception of low risk of injury in the work, compared with the perception of a high risk of injury (RR = 0.4, 95% CI = 0.2–0.9).
Availability of technician assistance was also considered an important potential risk factor to examine. There were higher rates for those who answered “frequently” or “sometimes” compared with those who said they always had assistance when working with animals (RR = 1.8, 95% CI = 0.9–3.9, and RR = 1.9, 95% CI = 0.9–4.1, respectively). A similar result was seen when veterinarians were asked about the availability of sharps boxes used for disposal of needles in their usual work environment. Those who responded “not available” had increased rates of injury (RR = 1.8, 95% CI = 1.0–3.2); those who responded “not applicable” had an even greater rate (RR = 4.8, 95% CI = 2.2–10.4).
Decreased rates were identified for increasing age (36–45 years, RR = 0.6, 95% CI = 0.4–0.9; 46–55 years, RR = 0.4, 95% CI = 0.3–1.0; 56–80 years, RR = 0.3, 95% CI = 0.2–1.3), male gender (RR = 0.5, 95% CI = 0.3–0.8), use of alcohol (RR = 0.6, 95% CI = 0.4–0.9), sitting from 1 to <4 hours per day vs ≥4 hours (RR = 0.6, 95% CI = 0.3–1.2), and participation in aerobic activities (RR = 0.6, 95% CI = 0.4–1.0).
To evaluate potential selection bias, we sent a brief questionnaire to all nonresponding cases (N = 91) and controls (N = 168). The return rate for this brief survey was 35% for cases and 37% for controls. A sample of the original cohort that had not responded in the initial phase of the study was also contacted (N = 110). In Table 5, results for one variable (hours of sleep) are shown from this substudy. This information was used to evaluate the potential effect of nonresponse bias. Specifically, this bias may influence the effect estimates if it is nonrandom with respect to injury and exposure status (for example, if exposed cases were more likely to respond than the other groups). It appears that exposed cases, and unexposed controls to a lesser extent, were more likely to participate in the study than were unexposed cases and exposed controls. This selection would appear to exaggerate the effect estimates that we report.
Limitations of the study include the fact that data were collected retrospectively and required the participants to recall events that had occurred in the past. Cases had the opportunity to be reminded of the injury they had reported in the initial comprehensive survey and reply regarding the month before this event. Controls were randomly assigned a month and thus did not have a marker event to stimulate their memories. In many veterinary environments, the activities and species of animals treated vary throughout the year, so it was important to account for the seasonal variation by focusing on a 1-month period, even though results in the initial survey found slightly increased proportions of injuries only in December and comparable numbers of veterinarians working in each month. Participants who were selected more than once as a control were allowed to indicate “same” on questionnaires if they felt the exposures were similar to those identified in the first questionnaire they completed. This prompting may have resulted in imprecision in measuring the differences among months. Few participants took advantage of this option and completed each questionnaire separately. Because cases who were also picked as controls were allowed to indicate “same” on these questionnaires if they perceived that the information did not vary, this option may have decreased our ability to distinguish differences in exposures between the cases and controls.
Because the relevant hazard period for injuries is usually very short, immediately preceding the injury, asking about average exposure history for the prior month may miss important short-term changes. For exposures that are more stable through time, such as type of practice or demographic characteristics, this issue should be less problematic. There is also a potential for information bias when asking about exposures several months after the incident; those injured may recall more details, or remember specific incidents, whereas the controls may have less specific memory triggers.
It was not possible to control for some important confounders, especially the behavior of a veterinarian toward a particular animal involved in an injury or toward animals in general and the behavior of the animal involved in the injury. For the veterinarian’s behavior, there is no good, reliable measure of this item; for the animal’s behavior, comparable control information cannot be readily obtained.
Veterinarians may change jobs or tasks fairly often, which may blur the determination of past exposures. There also is a difference in exposure assessment between the small-animal practitioner, who is in the same setting each day, and the large-animal practitioner, who spends much of the day traveling among a variety of farms and sales barns and also works in a clinic.
The finding of an increased rate for current smoking behavior has some substantiation from previous studies, including increased injury risk from motor vehicle crashes 7,8 and other sources. 9,10 Potential behavioral differences between smokers and nonsmokers and the possibility of carbon monoxide effects 8,9,11,12 have been reported.
Although the association of decreased alcohol use and injury appears contrary to the known relation between alcohol use and injury outcomes, 13 most of the available data are case-based and do not consider the overall population exposures. The finding from the current effort is consistent with two case-control studies that addressed farming-related injuries 14 and dairy operation-related injuries. 15
Increasing years of experience were associated with decreasing rates of injury. This effect may reflect the position within the practice, owner vs associate, and learning to work better with patients and their owners over time. It may also reflect the type of work done, as the more experienced veterinarian may be involved in more specialty-type practices. The increased rate of injury with 6 or fewer hours of sleep per night is consistent with earlier data from Belloc and Breslow, 16 who examined the relation between amount of sleep and general health.
We thank George Maldonado, Division of Environmental and Occupational Health, School of Public Health, University of Minnesota, for his important contributions to design, analysis, and interpretation.
1. Langley R, Pryor W Jr, O’Brien K. Health hazards among veterinarians: a survey and review of the literature. J Agromed 1995; 2: 23–52.
2. Landercasper J, Cogbill T, Strutt PJ, Landercasper BO. Trauma and the veterinarian. J Trauma 1988; 28: 1255–1259.
3. Hashemi K, Brown R, Buckley A. Accidents in practice. Vet Rec 1993; 133: 580.
4. Poole A. Survey of occupational hazards in companion animal practices. J Am Vet Med Assoc 1998; 212: 1386–1388.
5. Gabel C. A study of risk factors
for injuries among Minnesota veterinarians (Doctoral thesis). Minneapolis: University of Minnesota, 2000.
6. SAS Institute. SAS Software Release 6.12. Cary, NC: SAS Institute, 1996.
7. Maldonado G, Greenland S. Simulation study of confounder-selection strategies. Am J Epidemiol 1993; 138: 923–935.
8. Brison R. Risk of automobile accidents in cigarette smokers. Can J Public Health 1990; 81: 102–106.
9. Tsai S, Cowles S, Ross C. Smoking and morbidity frequency in a working population. J Occup Med 1990; 32: 245–249.
10. Reynolds K, Heckel H, Witt C, et al
. Cigarette smoking, physical fitness, and injuries in infantry soldiers. Am J Prev Med 1994; 10: 145–150.
11. DiFranza J, Winters T, Goldberg R, Cirillo L, Biliouris T. The relationship of smoking to motor vehicle accidents and traffic violations. NY State J Med 1986; 86: 464–467.
12. Liddell F. Motor vehicle accidents (1973–1976) in a cohort of Montreal drivers. J Epidemiol Community Health 1982; 36: 140–145.
13. Baker D. Interpersonal trauma and animal-related experiences in female and male military veterans: implications for program development. Mil Med 1998; 163: 20–25.
14. Elkington J. A case-control study of farmwork related injuries in Olmsted County, Minnesota (Doctoral thesis). Minneapolis: University of Minnesota, 1990.
15. Boyle D. Case-control study of dairy operation injuries (Doctoral thesis). Minneapolis: University of Minnesota, 1995.
16. Belloc N, Breslow L. Relationship of physical health status and health practices. Prev Med 1972; 1: 409–421.
Keywords:© 2002 Lippincott Williams & Wilkins, Inc.
injury; veterinarian injuries; risk factors; occupational injuries