Work Pattern Differences Between Male and Female Orthodontists in Canada

The number of women entering and working in formerly male-dominated health professions has increased significantly in recent decades.^1,2 As more women pursue professional careers, the number specializing within their profession is expected to increase as well, a trend that has been observed in the dental specialty of orthodontics.

In the initial Journal of Clinical Orthodontics American Practice Study³ in 1981, 0.6% of responding orthodontists were women; in repeated practice studies during 2005–2011,^4-7 women constituted 12–14% of the respondents. The number of women specializing in orthodontics is expected to continue to increase. In 1999, 34% of orthodontic residents in the United States were women,⁸ whereas in 2010, the proportion of female residents had increased to 39%.⁹ Similar trends have been observed in Canada: in 2006, 36% of orthodontic residents were women¹⁰; in 2013, the proportion was 47%.

Men and women receive the same education and training in their specialty programs; however, they have historically assumed different roles and responsibilities with respect to work and raising a family. Therefore, there is speculation that men and women will practise the same profession differently. As the proportion of women specializing in orthodontics has increased and is expected to continue to increase, speculations have been made concerning the potential effect this may have on the profession.^11-13

The purpose of this study was to assess the current work patterns of male and female Canadian orthodontists to determine whether any sex-specific differences exist. We examined personal and practice demographics, family structure, work patterns and practice characteristics and conducted sex-specific comparisons to identify factors that influence practice and work pattern characteristics, in an effort to consider whether the increased proportion of female orthodontists in Canada will affect the future delivery of orthodontic care.

Methods

A questionnaire was developed based on a previously published study comparing differences in practice patterns among male and female orthodontists in the United States.¹⁴ Following a pilot study, in which a revised questionnaire was given to a group of 5 local orthodontists to check for question error and relevancy, minor amendments were made and the final questionnaire was produced. The questionnaire and all correspondence were translated professionally from English to French to minimize potential language barriers for respondents. Three of the orthodontists involved in the pilot study were bilingual; they completed the survey in both French and English to ensure accuracy of the translation.

The estimated number of orthodontic specialists in Canada is 799, of which it is estimated that 191 are women (24%) and 608 are men (76%). A sample size of 378 was determined to be adequate for statistical power.¹⁵ To ensure quota sampling and prevent overrepresentation of either sex, this sample size was divided into a target sample of 89 (24%) female and 289 (76%) male orthodontists.

Participants were selected by numbering the lists of orthodontists, which had been stratified according to region of primary practice address and sex, and using a random number generator (SPSS version 20.0, IBM, Armonk, NY) to determine who to approach with the survey.

For survey implementation and data analysis, 2 distinct databases were created to ensure participant confidentiality. Orthodontists selected for participation received a copy of the survey package through regular mail, addressed to their primary practice address and, where email addresses were available, through email with a URL link to the online version (Survey Monkey, Palo Alto, CA). The survey package consisted of two letters; a letter of introduction and a hand-signed letter with a URL address and instructions for accessing the online questionnaire, a copy of the questionnaire and a self-addressed stamped return envelope. Participants in Quebec received all correspondence in both English and French. In addition, the online version was available in English or French. Each survey was linked to a blinded identification marker in the upper right hand corner of the questionnaire, body of the email and information letter for online survey access to differentiate respondents from non-respondents.

All selected survey participants were mailed and/or emailed an initial survey package by 17 April 2013, and a second “reminder” package was sent to non-respondents by 27 May 2013. In addition, an email information package was sent to all female orthodontists practising in the eastern region of Canada (n = 6), that were not included in the original random sample, on 3 July 2013, in an attempt to obtain a representative sample; this increased the total number of surveys sent to 384.

For mail-based surveys, data were entered manually, while web-based surveys were automatically compiled into Excel 2011 spreadsheets (Microsoft, Redmond, Wash.). Manually entered data were checked twice to ensure accuracy and web-based surveys were inspected to ensure that recorded data were relevant to questions asked (i.e., numerical versus text responses). The two spreadsheets were then combined. All data analyses were performed using SPSS version 20.0. Before completing testing, model assumptions were evaluated; when they were not satisfactorily met, further analyses were completed, which is described below. For all tests, statistical significance was set at α = 0.05.

Descriptive statistics were generated for each variable, including means, standard deviation, standard error, medians, ranges and total number of respondents. Cross-tabulations, with sex as the independent variable, were created, when applicable.

When appropriate, contingency tables were formulated and Pearson χ² values and probabilities were computed. In the comparison of means, ANOVA was used. When multiple means were compared, a one-way ANOVA in conjunction with Bonferroni post-hoc test was used. When equal variances between the 2 populations were not satisfied, the data were compared using either Tamhane’s post-hoc test or log-linear transformation.

The work patterns of Canadian orthodontists were evaluated, using the number of hours worked per week, number of patients seen per workday and number of new case starts per year (in 2012) as the response variables. The effects of sex on work patterns were evaluated, applying age and number of children or children living at home as covariates in the analysis using multivariate analysis of covariance (MANCOVA) in conjunction with Bonferroni post-hoc test.

In the evaluation of work-pattern differences, the number of female orthodontists in the survey sample was significantly lower than that of male orthodontists (27 and 134, respectively). After completing an overall analysis, a random sample of 40 male respondents was selected to maintain the number of male respondents at 1.5 times the number of female respondents to increase statistical power. Analysis of a random sample was repeated 20 times and compared with an analysis of the entire population (27 women and 134 men). Because similar trends resulted from the repeated random samples and the overall analysis, the overall population analysis was used in the discussion of results as this includes all of the information collected in the survey.

Approval from the University of Alberta’s Research Ethic Board was granted for this research (Pro00036677).

Results

The final sample size for the survey was 371 (280 men, 91 women). As of 28 July 2013, we received 207 responses (53.9% response rate), of which 94 (45.4%) were completed online and 113 (54.6%) were mailed. Of the respondents, 160 (77.3%) were men and 42 (20.3%) were women, 5 (2.4%) did not specify their sex.

Demographics

The average age of all respondents was 51 years. The age range for men was 29–77 years (mean 52.3, median 52); the age range for women was 32–65 years (mean 46.4, median 45). Average age differed significantly between the sexes: men were, on average, 5.9 years older than women (p = 0.002) (Table 1).

Respondents had graduated primarily from Canadian dental schools and orthodontic training programs, with no significant differences between the sexes (p > 0.05) (Table 2). Age at graduation from dental school and orthodontic training was similar for men and women. The average age at graduation from dental school was 25.6 years (p = 0.900), while the average age at graduation from orthodontic training was 31.4 years (p = 0.335) (Table 1).

Most male and female orthodontists were married, with no significant differences between the sexes (p = 0.212) (Table 2).

Spouses of male orthodontists were less likely to work full time than spouses of female orthodontists. Of the married female respondents, 94% reported full-time spousal employment; 25% of the married male respondents reported having spouses who were employed full time, 45% were employed part time and 30% were not currently employed. Female respondents’ spouses were most likely to be dentists, including dental specialists. Most male respondents’ spouses were employed in a “non-health other occupation,” the most common being office manager/administrative duties and bookkeeper (Table 2).

The number of children of male respondents ranged between 0 and 7, while the range for female respondents was 0–5. The average number of children for both men and women was 2.2 (median 2), with no significant difference between the sexes (p = 0.189) (Table 2).

The mean age at which both male and female orthodontists had their first and second children did not differ significantly (p = 0.976): 30.8 and 33.6 years for men and 31.8 and 33.6 years for women (Table 1).

Practice Information

Men were most likely to practise solo (65%), followed by group practice limited to orthodontics (29%). The most common arrangements for women were working as a solo practitioner (48%) and in a group practice limited to orthodontics (48%). Analysis of the data for “solo practitioner” versus “other,” including all other forms of practice, showed weak evidence to suggest that men are slightly more likely to work as a solo practitioner than their female colleagues (p = 0.061) (Table 3).

Location of main office was similar for both sexes. Both men and women were most likely to work in a metropolitan area. The second most common office location for both sexes was in a large city, followed by a small city and a rural area. There was no significant difference in the mean number of offices worked in for men and women (p = 0.241), with both sexes most commonly working in 1 office (Table 3).

Ownership status was similar for both sexes. Both men and women most commonly owned an orthodontic practice (77% of men; 71% of women). The second most common status was owning part of an orthodontic practice, followed by non-owner (Table 3). Comparing owning an orthodontic practice versus non-owner revealed no significant differences between the sexes (p = 0.588).

There was evidence to suggest a significant difference between the sexes for previously working as an orthodontic associate (p = 0.083): 47% of men and 62% of women (Table 3). However, the number of years worked as an associate did not differ significantly between the sexes (p = 0.545): the average length of associateship was 4.4 years for men and 5.1 years for women (Table 4).