Associations of Unscavenged Anesthetic Gases and Long Working Hours With Preterm Delivery in Female Veterinarians

Shirangi, Adeleh MPH, PhD; Fritschi, Lin MBBS, PhD; Holman, C D’Arcy J. MBBS, PhD

Obstetrics & Gynecology:
doi: 10.1097/AOG.0b013e31819fe996
Original Research

OBJECTIVE: To examine whether occupational hazards such as anesthetic gases, radiation, pesticides, and working hours in veterinary practice are associated with preterm delivery (before 37 weeks of gestation) in female veterinarians.

METHODS: The Health Risks of Australian Veterinarians project was conducted as a questionnaire-based survey of all graduates of Australian veterinary schools during the 40-year period 1960–2000. Approximately 1,200 female veterinarians participated in the survey. Pregnancy was defined as the unit of analysis. We restricted analyses to pregnancies of those women who reported being employed when the pregnancy began and were working only in clinical practice. Of 1,355 pregnancies in total in the file, 744 pregnancies were eligible for the final analysis.

RESULTS: The prevalence of preterm birth in women exposed to unscavenged anesthetic gases was 7.3% compared with 5.7% in the general population. In a Cox proportional hazards model controlling for the potential confounders, there was a significant 2.5-fold increase (hazard ratio 2.56, 95% confidence interval [CI] 1.33–4.91) in the risk of preterm delivery in women exposed to unscavenged anesthetic gases for 1 or more hours per week compared with the unexposed group (women who did not perform surgery and those who performed surgery in the presence of a scavenger system). A dose–response relation also was found within this group when exposure was divided into finer categories. There was also a strong and monotonic increase in risk of preterm delivery and the number of hours worked per week with veterinarians working more than 45 hours a week (hazard ratio 3.69, 95% CI 1.40–9.72) compared with those working fewer than 45 hours per week.

CONCLUSION: Long working hours and performing surgery in the absence of a scavenger system for anesthetic gases are important risk factors for preterm birth in female veterinarians.


In Brief

Long working hours and performing surgery in the absence of a scavenger system for anesthetic gases are risk factors for preterm birth in female veterinarians.

Author Information

From the 1School of Population Health, Faculty of Medicine, Dentistry and Health Sciences, and the 2Western Australian Institute for Medical Research, the University of Western Australia, Perth, Australia.

The authors thank National Health and Medical Research Council of Australia (grant ID: 303297), The Cancer Foundation of Western Australia, and The University of Western Australia Research Grants for their financial support.

The authors thank Professor Carol Bower and Dr. Lesley Day, who assisted with the development of the Health Risks of Australian Veterinarians survey.

Corresponding author: Adeleh Shirangi, BSc, MPH, PhD, Department of Epidemiology and Public Health, Faculty of Medicine, Imperial College London, St Mary’s Campus, Norfolk Place, London, W2 1PG; e-mail:

Financial Disclosure The authors did not report any potential conflicts of interest.

Article Outline

Veterinarians have occupational exposure to several known reproductive hazards, such as radiation, anesthetic gases, pesticides, and long working hours. Concerns have been expressed over the risk of adverse reproductive outcomes in female veterinarians based on a small number of U.S. studies.1–4 In a recent cross-sectional survey of the same cohort population, we addressed the prevalence of exposure to potentially harmful occupational situations including radiation, anesthetic gas, and pesticides and showed that considerable variability of exposure still exists within the profession.5 We also have investigated maternal occupational exposures and risk of spontaneous abortion in veterinary practice.6

Despite improvements in antenatal care, the occurrence of preterm delivery (less than 37 weeks of gestation) remains high (5–10%) in developed countries and even more so in developing countries. Identifying occupation-related risk factors for preterm delivery is of particular importance because these risk factors are amenable to change through policies granting leaves of absence or modifying working conditions during pregnancy. However, studies on the relation between occupational factors and preterm delivery have yielded contradictory results.7–11 With regard to veterinary work and preterm delivery, there has been only one study that did not find any strong associations.1 This lack of findings is likely to have been a reflection of limitations of the study design. The authors of that study combined unemployed women’s pregnancies with employed, nonexposed women’s pregnancies to form the reference group owing to the small numbers in each category.1 They recommended further study of reproductive outcomes among female veterinarians, with intensive efforts directed to more complete ascertainment of exposures and stricter criteria for the reference group. The aim of the present study is to estimate the association between occupational hazards and preterm delivery in female veterinarians.

Back to Top | Article Outline


The Health Risks of Australian Veterinarians study was the first complete national survey of health aspects of veterinary practice in Australia and was conducted in 2002 (Shirangi A, Occupational hazards in veterinary practice and possible effects on reproductive outcomes in female veterinarians [doctoral thesis]. School of Population Health, Faculty of Medicine, Dentistry and Health Sciences The University of Western Australia, 2006).5,6,12,13 The study consisted of a questionnaire-based survey of all graduates from Australian veterinary schools during the 40 year period 1960–2000. Of 2,028 eligible female veterinarians who were sent the questionnaires, 1,197 replied (response rate 59%). Of 1,197 female veterinarians who participated in the survey, 631 had never been pregnant and 566 had had at least one pregnancy. We have used different sample populations and different analytical approaches to assess each different health outcome. The study was approved by the Human Research Ethics Committee of The University of Western Australia.

The mailed, self-administered questionnaire contained items about demographic details, a section on professional history of all veterinary jobs held for more than 6 months since graduation, history of smoking, and reproductive history for women. Information obtained for each job included start date, end date, job type, practice type, work hours, number of X-ray films taken per week, restraint of animals during X-rays, number of hours doing surgery per week, use of scavenger equipment to remove waste anesthetic gases (yes, no, don’t know, not applicable), driving (number of hours per week), and use of pesticides including antitick agents, anthelminthics, and flea treatments (daily, weekly, rarely, never). Reproductive history was assessed by asking, for each pregnancy (up to six), the number of months trying to become pregnant, the year the pregnancy ended, the number of weeks the woman was pregnant, whether it was a single, twin, or multiple birth, and sex of each child. Reproductive outcomes were assessed by using information about live births, stillbirths, terminations (induced abortions), miscarriage (spontaneous abortions), and birth defects. Preterm delivery was defined as all singleton liveborn neonates delivered before completion of 37 weeks of gestation.

We reorganized the structure of the data file such that pregnancy became the unit of observation. For the assessment of occupational exposures during pregnancy, we needed to evaluate each exposure for each pregnancy for each reported gestational week (from week 22 until week 37). We had information about the work history of each woman in the study after graduation, and we had the date of the end of the women’s pregnancies (actual delivery date) in our database. Therefore, we reshaped our database by cross-referencing work history data to the year before the end of the woman’s pregnancy date to assess occupational exposure during pregnancy. We evaluated exposure for each pregnancy for the time from the few months immediately before conception to the end of the pregnancy, which covered the critical times of vulnerability (weeks 22 to 37).

We excluded the following pregnancies: those current at the time of the survey (n=23), those that commenced before graduation from veterinary college (n=48), twin pregnancies (n=14), those where the mother had not worked up to the time of conception and during the pregnancy (n=163), those where the mother was not employed exclusively as a clinical practitioner at the time of conception and during the pregnancy (n=167), and pregnancies that ended in miscarriage (n=146), termination (n=24), or stillbirth (n=6). Among the remaining 764 singleton live births, we did not have an answer to the question regarding weeks of pregnancy for 20 pregnancies. Therefore, after all these restrictions, of 1,355 pregnancies in the file (566 women), 744 pregnancies in 399 women were eligible for the final analysis.

Crude preterm delivery risk overall and crude and adjusted relative risks (RRs) of preterm delivery by clinical practice type and occupational exposures were estimated using RR and 95% confidence interval (CI) formulas.14 Preterm delivery data also were analyzed using a person-week model, and the occurrence of a preterm delivery in each gestational week was treated as a dichotomous outcome. Pregnancies were entered into the risk sets starting at the estimated 22nd week of pregnancy, and the analyses were truncated at 37 weeks of gestation. We used a Cox life table and estimated Kaplan-Meier cumulative risks15of all pregnancies ending as preterm delivery for each reported gestational week from week 22 until week 37 for occupational exposures to anesthetic gases and working hours. A multivariable semiparametric Cox proportional hazards procedure using STATA 9 software (StataCorp, LP, College Station, TX) also was adopted to estimate confounder-adjusted RRs.15 Relative risks expressed as hazard ratios with associated 95% CI were derived from the Cox model. We used a Cox proportionate hazard model to see whether the distribution of gestational age was different for the veterinarians with and without occupational exposures. We used the Huber-White sandwich estimator of variance to take into account the potential autocorrelation created by multiple pregnancies in a single woman.16,17

The following occupational exposures were examined: clinical practice type (small, mixed, large), years in job (less than 2 years, 2 years or longer), working hours (less than 35 hours, 35–45 hours, more than 45 hours), X-ray exposure (none, at least one film/wk), pesticide (never or rarely, at least once/wk), surgery (none, surgery in the presence of a scavenger, surgery in the absence of a scavenger). We further divided surgery in the absence of a scavenger and below this point to understand the exposure–dose response relationship in the data. The following variables were treated as potential confounders: maternal age (35 years old or younger, older than 35 years), graduation year (continuous), university of graduation, history of smoking, age at the time of survey, animal restraint (at least once/wk, never, rarely), driving (at least 1 h/wk, none), birth order, gravidity, parity, number of previous spontaneous abortions, number of previous induced abortions, and time to pregnancy in months. The associations of potential confounders with exposure variables were assessed by cross-tabulations, χ2 tests, and also using the Mantel-Haenzel adjusted odds ratio (OR). If the difference between the crude OR and the adjusted OR was less than 10%, the covariate was not considered a confounder. After confirming the confounders, a semiparametric Cox proportional hazard model was implemented, including all the exposure variables and adding each of the variables that changed the hazard ratio by 10% or more (maternal age, graduation year, history of smoking, driving, restraint of animals, birth order). Adding each reproductive history variable (gravidity, parity, number of previous spontaneous abortions, number of previous induced abortions, and time to pregnancy in months) in the model did not change the hazard ratio by 10% or more; therefore, reproductive history was not included in the model. We also performed a test of the constant proportional hazards assumption based on Schoenfeld and scaled the Schoenfeld residuals and another postestimation method for the Cox model using STATA’S Plots (StataCorp, LP, College Station, TX) to evaluate the proportional hazards assumption in both models with and without adjustment for reproductive history variables. Interaction effects of factors related to preterm delivery were tested in the Cox model, but none contributed significantly to the model. Analysis was repeated for the subgroup of pregnancies reported by veterinarians employed exclusively in small-animal practices, those who graduated between 1980 and 2000, and for first pregnancies. We also repeated the analysis, including reproductive history to comment on the differences using logistic regression.

Back to Top | Article Outline


Respondents were born between 1938 and 1976, with a median age of 39.8 years in 2002, and 95% of respondents were more than 30 years old. Nearly 86% of respondents were born in Australia, and 6% percent were born in the United Kingdom. Approximately 72% of them graduated after 1980, with a median graduation year of 1984 and graduation age of 23 years. Of 744 singleton live births, 54 (7.3%) were classified as preterm (less than 37 completed weeks of gestation) and 11 (1.5%) were classified as very preterm (less than 32 completed weeks of gestation).

Risk of preterm delivery was highest in women who graduated from Sydney University and for those who had four pregnancies or more (Table 1). The RR was nonsignificantly higher in veterinarians who worked in small- and mixed-animal practices, those who smoked, and those who had four children or more (Tables 1 and 2). In the crude analyses of exposures, compared with those unexposed to that specific factor, risks were highest in veterinarians who worked 2 years or more in clinical practice, worked longer hours, and performed more surgery, especially in the absence of a scavenger (Table 2). A dose–response effect of anesthetic gas exposure and long working hours is apparent even in the life table graphs in Figures 1 and 2. In Figure 1, the no-exposure group consisted of women who did not do surgery and those who did surgery in the presence of a scavenger (418 pregnancies with 17 preterm deliveries). The low-exposure group consisted of women who did 1 to 15 hours of surgery in the absence of a scavenger (227 pregnancies with 21 preterm deliveries). The high-exposure group consisted of women who did more than 15 hours of surgery in the absence of a scavenger (42 pregnancies with 12 preterm deliveries). Those who answered “don’t know” or “not applicable” to the question of scavenger use were regarded as missing values and were not included in the analysis. There was no association of preterm delivery with exposure to radiation, use of pesticides, or driving more than 1 hour per week (Table 2).

In the multivariable Cox regression analysis using semiparametric models (Table 3), there was a strong and monotonic increasing association between risk of preterm delivery and the number of hours worked per week. There was a more than 3.5-fold higher risk of preterm delivery in veterinarians who worked more than 45 hours per week. Compared with veterinarians who did not perform surgery and those who performed surgery in the presence of a scavenger system, there was a 2.5-fold increase in the risk of preterm delivery in those exposed to unscavenged gases for 1 or more hours per week. In a separate multivariate Cox regression analysis adjusted for all variables indicated in Table 3, we also explored a possible nonlinear association between anesthetic gas exposure and preterm delivery. When unscavenged anesthetic gas exposure (performing surgery in the absence of a scavenger) was divided into two levels (1–15 hours and more than 15 hours per week), an elevated risk of preterm delivery was found in each of the two highest groups compared with the unexposed group (women who did not perform surgery and those who performed surgery in the presence of a scavenger system). The hazard ratio for the 1–15 h/wk group was 2.31 (95% CI 1.13–4.71), and there was a 4.16-fold increase (95% CI 1.96–10.24) in the risk of preterm delivery among 42 pregnancies exposed to unscavenged anesthetic gases for more than 15 hours per week. A dose–response anesthetic gas effect is apparent even in the unmodeled data using the Kaplan-Meier method. (Figure 1).

The Cox models did not violate the proportional hazards assumption except for those models that were adjusted for reproductive history. Examination of various solutions for the breach of proportional hazards (eg, to include the time-dependent variable for the nonproportional predictors or to stratify on the nonproportional predictors) did not solve this problem. Therefore, the analyses with adjustment for reproductive history were undertaken using logistic regression. The results for the logistic models were also very similar to those from the Cox models. For example, the key results from the logistic model for all eligible pregnancies with adjustment for reproductive history for the key variables such as surgery in the absence of a scavenger (OR 2.80, CI 1.33–5.91), working 35–45 h/wk (OR 2.84, CI 1.14–7.05), and working more than 45 h/wk (OR 4.07, CI 1.22–13.52) indicated very similar results.

In a separate multivariable Cox analysis adjusted for the same confounding variables restricted to: 1) veterinarians who graduated between the years 1980 and 2000, 2) those employed exclusively in small-animal practices, and 3) women in their first pregnancies, similar results were obtained for all exposure variables (Table 3). Similar statistically significant results were observed for the dose–response model of the variable surgery in the absence of a scavenger in all three subgroups of veterinarians after adjusting for the same confounding variables indicated in Table 3. For example, the hazard ratios of 2.43 (95% CI 1.04–5.70) in the 1–15 h/wk group and 5.36 (95% CI 1.45–19.82) in the more than 15 h/wk group were found for those veterinarians who graduated between 1980 and 2000 and similar results were found for those veterinarians employed exclusively in small-animal practice and for first pregnancies.

Back to Top | Article Outline


These results suggest that longer working hours, exposure to unscavenged anesthetic gases, length of time in veterinary practice, and working in a small-animal practice are associated with preterm delivery in female veterinarians. The overall proportions of liveborn preterm and very preterm neonates were 7.3% and 1.5%, respectively, compared with 5.7% and 1.1% from 1997–200118and 6.5% in 200519 in the general population of Australia. This rate also exceeded the 5.9% reported in a cohort study of 2,997 female graduates from U.S. veterinary colleges between 1970 and 1980.1

The risk of preterm delivery increased in women exposed to unscavenged gas in practices that cared exclusively for small animals. This may reflect the greater levels of hazardous exposures in this type of clinical practice, an observation that we have reported in earlier results from the survey.5 We found working hours to be significantly associated with preterm delivery. Although a previous study of female veterinarians did not find an association between risk of preterm delivery and the number of hours worked per week,1 our results are consistent with those reported for other occupations. Luke et al, in their case-control study of U.S. nurses, report an adjusted OR of 1.6 (95% CI 1.00–2.86) for preterm delivery in women working more than 36 hours per week.9 Similar results were reported by other investigators.8,10,20,21 However, a recent meta-analysis found no overall excess risk of preterm delivery associated with long working hours, although a subgroup analysis of the six studies of highest quality did show a significant statistical association between this exposure and preterm delivery.7

Several studies also have examined an increased risk of preterm delivery in women whose jobs involve other occupational factors such as standing for long durations, walking, heavy lifting, and other aspects of physical exertion, although the results have been inconsistent.7,8,10,22–24 Female veterinarians also engage in physically strenuous work such as standing for long hours, lifting, and restraining animals. We had information only about restraining animals and found no association between this activity and preterm delivery. There could be biologic mechanisms by which ergonomic stressors or long working hours contribute to preterm delivery. These factors may cause stress, leading to the release of catecholamines. Catecholamines may, in turn, increase blood pressure and uterine contractility and decrease placental function.25 However, the findings should be interpreted with caution because of the limitations of observational studies.

The data also showed that the risk of preterm delivery was higher in women exposed to anesthetic gases in the absence of scavenging equipment than it was in women who were unexposed or protected by scavenging equipment. There was also a dose–response relation within this group when exposure was divided into finer categories. A previous survey of female veterinarians in the United States did not find a strong association between exposure to halothane and preterm delivery.1 A 1970 study of operating room nurses in Denmark26 reports an increase in the risks of spontaneous abortion and premature neonates. If the new finding in the present study is a true reflection of the risk, the capacity to measure it may have arisen because the present study addressed some limitations of the earlier research and also was able to adjust more completely for confounding factors. For example, one of the limitations from a recent study on female veterinarians was that the authors were using unemployed women’s pregnancies combined with employed, nonexposed women’s pregnancies to form the reference group owing to the small numbers in each category. We addressed this limitation by using only those pregnancies during which the women were employed in clinical practice and using unexposed pregnancies as the reference group. How anesthetic gas might cause preterm delivery is unknown, although McGregor27 and Rowland et al28 have discussed biologic mechanisms by which occupational exposures to waste anesthetic gases might contribute to adverse reproductive outcomes.

There were several major limitations to this study. However, we tried to reduce as many as possible of the potential biases of epidemiological studies. Two of the most important limitations are the self-reported nature of the data and a possible selection bias owing to a suboptimal but not unusual response rate (59%). A similar survey achieved a response rate of approximately 42%.9 Other similar studies conducted in the past 15 years also have reported similar response rates: 56%,29 60%,30 and 69%.28,31 We included a large cohort of female veterinarians from all types of clinical practice and from all ages in Australia, suggesting that our study was generally representative of female veterinarians. In addition, we performed an analysis to identify the main reasons for nonparticipation from 195 veterinarians who acknowledged the receipt of questionnaires but refused to participate. The reasons included working for a postgraduate qualification (for example in research area or in teaching), being outside the profession for a long time, and working temporarily overseas, suggesting that those not in clinical practice or veterinary work were less likely to participate. Nonparticipants were also more likely to be older graduates than were the respondents. Another limitation was the method used to estimate exposure, which was based on questionnaire responses rather than on actual measurements. The detailed occupational exposure information collected from these highly educated women may have minimized this type of information bias. In addition, previous surveys have supported the use of exposure data from retrospective self-reports.31 Joffe32 found recall regarding gestational age to be accurate for 20 years or more. Moreover, information on personal habits, lifetime working histories, and fertility can be obtained only through questionnaire-based studies.

Halothane, nitrous oxide, enflurane, and methoxyflurane are the most commonly used anesthetic gases in Australian veterinary practice. Details of exposure to specific gases were not available and would have been difficult to interpret given that normal veterinary practice involves exposure to a combination of these agents over time. The National Institute for Occupational Safety and Health recommends that no worker be exposed to 8-hour time-weighted average concentrations greater than 25 ppm during anesthetic administration (nitrous oxide) and concentrations of halothane and methoxyflurane agents greater than 2 ppm for a period not to exceed 1 hour.33 We found that 92% of veterinarians were exposed to waste anesthetic gases for at least 1 hour per week, and, among those who were exposed, 39% did not use a waste anesthetic gas scavenging system. However, another limitation was that, being retrospective in its account of anesthetic gases, the study could not measure average intensity of exposure such as through use of biological exposure markers.

The potential confounding effects of alcohol, use of contraceptives, marital status, and social background could not be evaluated because the relevant information was unavailable. The relatively homogeneous social background of the veterinarians probably partly controlled for the effects of lifestyle factors. There are other known occupational risk factors for preterm delivery, such as job-related physical stress and physical exertion, such as walking, standing, and heavy lifting. There are also nonoccupational factors, such as infections, but we were unable to control for any of these because they were unavailable in our database. However, we were able to examine and control for important potential confounders including maternal age and occupational exposures such as radiation and pesticides.

The strength of the study is that the data set was large and included female veterinarians of all ages in Australia. In addition, the findings with respect to the relationship between working hours and preterm delivery were consistent with other studies, thus lending support to our results. Internal validity of our results was strengthened by the information about outcomes and exposure being obtained in the same way using a structured questionnaire. There was no mention of preterm delivery in the questionnaire, and the outcome of preterm delivery was ascertained using reported weeks of gestation. We also were able to examine potential dose–response relationships, which allowed us to report on differences according to durations of exposure. We also analyzed exposure to anesthetic gases with modification for safety practices such as use of a scavenging system. We also used only those pregnancies during which the women were employed in clinical practice and used unexposed pregnancies as the reference group, thus improving validity. The justification for including only those women who worked in clinical practice has been discussed in our previous article.6

The results of this study revealed that long working hours and undertaking surgery in the absence of an anesthetic gases scavenger system are important risk factors for preterm delivery in female veterinarians. These results have implications for veterinarians and those in other professions who encounter similar hazards in the workplace, such as veterinary nurses, dentists, dental assistants, surgical nurses, surgeons, anesthetists, and other medical personnel working with these agents. These factors are preventable causes of preterm delivery and can be mitigated by reducing working hours per week and using scavenging equipment in veterinary clinics. Further research to establish clinical evidence on the association of anesthetic gases and preterm birth is recommended.

Back to Top | Article Outline


1.Wilkins JR 3rd, Steele LL. Occupational factors and reproductive outcomes among a cohort of female veterinarians. J Am Vet Med Assoc 1998;213:61–7.
2.Steele LL, Wilkins JR 3rd. Occupational exposures and risks of spontaneous abortion among female veterinarians. Int J Occup Environ Health 1996;2:26–36.
3.Schenker MB, Samuels SJ, Green RS, Wiggins P. Adverse reproductive outcomes among female veterinarians. Am J Epidemiol 1990;132:96–106.
4.Johnson JA, Buchan RM, Reif JS. Effect of waste anesthetic gas and vapor exposure on reproductive outcome in veterinary personnel. Am Ind Hyg Assoc J 1987;48:62–6.
5.Shirangi A, Fritschi L, Holman CD. Prevalence of occupational exposures and protective practices in Australian female veterinarians. Aust Vet J 2007;85:32–8.
6.Shirangi A, Fritschi L, Holman CD. Maternal occupational exposures and risk of spontaneous abortion in veterinary practice. Occup Environ Med 2008;65:719–25.
7.Mozurkewich EL, Luke B, Avni M, Wolf FM. Working conditions and adverse pregnancy outcome: a meta-analysis. Obstet Gynecol 2000;95:623–35.
8.Saurel-Cubizolles MJ, Zeitlin J, Lelong N, Papiernik E, Di Renzo GC, Bréart G, et al. Employment, working conditions, and preterm birth: results from the Europop case-control survey. J Epidemiol Community Health 2004;58:395–401.
9.Luke B, Mamelle N, Keith L, Munoz F, Minogue J, Papiernik E, et al. The association between occupational factors and preterm birth: a United States nurses’ study. Research Committee of the Association of Women’s Health, Obstetric, and Neonatal Nurses. Am J Obstet Gynecol 1995;173:849–62.
10.Henriksen TB, Hedegaard M, Secher NJ, Wilcox AJ. Standing at work and preterm delivery. Br J Obstet Gynaecol 1995;102:198–206.
11.Zhu JL, Hjollund NH, Olsen J, National Birth Cohort in Denmark. Shift work, duration of pregnancy, and birth weight: the National Birth Cohort in Denmark [published erratum appears in Am J Obstet Gynecol 2004;191:2184]. Am J Obstet Gynecol 2004;191:285–91.
12.Fritschi L, Day L, Shirangi A, Robertson I, Lucas M, Vizard A. Injury in Australian veterinarians. Occup Med (Lond) 2006;56:199–203.
13.Fritschi L, Shirangi A, Robertson ID, Day LM. Trends in exposure of veterinarians to physical and chemical hazards and use of protection practices. Int Arch Occup Environ Health 2008;81:371–8.
14.Szklo M, Nieto J. Epidemiology: beyond the basics. Gaithersburg (MD): Aspen Publishers; 2000.
15.Hosmer DW, Lemeshow S. Applied survival analysis. New York (NY): John Wiley & Sons; 1999.
16.White H. Estimation, inference and specification analysis. New York (NY): Cambridge University Press; 1994.
17.Williams RL. A note on robust variance estimation for cluster-correlated data. Biometrics 2000;56:645–6.
18.Australian babies: their health and well being. Australian Government: Australian Institute of Health and Welfare Bulletin 21; 2004. Available at: Retrieved February 29, 2009.
19.Laws P, Abeywardana S, Walker J, Sullivan EA. Australia’s mothers and babies 2005. AIHW National Perinatal Statistics Unit. 2007;20:49.
20.McDonald AD, McDonald JC, Armstrong B, Cherry NM, Nolin AD, Robert D. Prematurity and work in pregnancy. Br J Ind Med 1988;45:56–62.
21.Mamelle N, Laumon B, Lazar P. Prematurity and occupational activity during pregnancy. Am J Epidemiol 1984;119:309–22.
22.Marbury MC. Relationship of ergonomic stressors to birthweight and gestational age. Scand J Work Environ Health 1992;18:73–83.
23.Launer LJ, Villar J, Kestler E, de Onis M. The effect of maternal work on fetal growth and duration of pregnancy: a prospective study. Br J Obstet Gynaecol 1990;97:62–70.
24.Tuntiseranee P, Geater A, Chongsuvivatwong Kor-anantakul O. The effect of heavy maternal workload on fetal growth retardation and preterm delivery. A study among southern Thai women. J Occup Environ Med 1998;40:1013–21.
25.Luke B, Papiernik E. The effects of lifestyle on prematurity. In: Elder, MG, Lamont RF, Romero R, editors. Preterm labor. New York (NY): Churchill Livingstone; 1997. p. 127–52.
26.Askrog V, Harvald B. Teratogen effekt af inhalationsanaestetika. Nord Med 1970;83:498–500.
27.McGregor DG. Occupational exposure to trace concentrations of waste anesthetic gases. Mayo Clin Proc 2000;75:273–7.
28.Rowland AS, Baird DD, Shore DL, Weinberg CR, Savitz DA, Wilcox AJ. Nitrous oxide and spontaneous abortion in female dental assistants. Am J Epidemiol 1995;141:531–8.
29.Hourani L, Hilton S. Occupational and environmental exposure correlates of adverse live-birth outcomes among 1032 US Navy women. J Occup Environ Med 2000;42:1156–65.
30.Cone JE, Vaughan LM, Huete A, Samuels SJ. Reproductive health outcomes among female flight attendants: an exploratory study. Journal of Occupational & Environmental Medicine 1998;40:210–6.
31.Eskenazi B, Pearson K. Validation of a self-administered questionnaire for assessing occupational and environmental exposures of pregnant women. Am J Epidemiol 1988;128:1117–29.
32.Joffe M. Male- and female-mediated reproductive effects of occupation: the use of questionnaire methods. Journal of Occupational Medicine 1989;31:974–9.
33.National Institute for Occupational Safety and Health. Occupational exposure to waste anesthetic gases and vapors: criteria for a recommended standard. Cincinnati (OH): U.S. Dept of Health, Education, and Welfare, Public Health Service, Centers for Disease Control, National Institute for Occupational Safety and Health, Publication DHEW (NIOSH). 1977:77–140.
© 2009 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.