The analysis presented here refers to Pap smear use and determinants of time-inappropriate cervical cancer screening among Inuit women aged 21–69 in Nunavik, Quebec. Overall, there was a sizeable high-risk group who had not been screened in the three years prior to cohort entry (25%), and 4% of the sample had never been screened for cervical cancer. Although Aboriginal populations in Canada have traditionally had lower screening rates than the general population, we found that the cervical cancer screening coverage in this population was comparable to the screening coverage reported for the Canadian population [3, 4, 23]. However, there is still cause for concern, given that there is a high prevalence of high-risk HPV (HR-HPV) infection and incidence of cervical cancer in this population [9, 10, 12]. If high-risk subsets of the population are systematically underscreened for cervical cancer, then it is likely that incidence and mortality reductions will not be observed, despite screening rates that are comparable with the general population of Canada. Understanding, the predictors of screening underuse will help determine opportunities to adapt or enhance health promotion activities for screening among underserved groups.
In this population, older age and no history of childbirth were significant predictors of time-inappropriate Pap smear use. We found that older women had a significantly higher odds of not being screened appropriately compared to younger women. A similar effect of Pap smear rates decreasing with age was found in other Canadian populations [4, 24]. It is worrisome that older women are at risk of not being screened appropriately, as cervical cancer mortality increases with age and cervical cancer incidence is higher among women 40 years and older . Menopausal or post-menopausal women may not always realize that screening is still necessary after reproductive years , as younger women may be more aware of the need for periodic screening.
Our results also suggest that screening may be related to reproductive care use, given that women who had no history of child birth were more likely to have had a time-inappropriate Pap smear. In Nunavik, Pap smear is offered as part of the prenatal and well-baby visits, so women who access these services would likely have multiple opportunities to be screened. Similarly, women from Ontario without a prenatal visit during a three year interval were also less likely to have been screened for cervical cancer than those who did attend a prenatal visit . Although we were unable to include use of birth control in the final model due to collinearity, in the univariate analysis we found that women who did not use birth control at baseline had an increased odds of having time-inappropriate screening. This is consistent with a previous study which found that women participated in screening because it was linked with the annual renewal of their oral contractive prescription . Indeed, the majority of women in our study (78%) who used birth control used a hormonal birth control type (oral contraceptive or Depo-Provera injections), which might explain some of this relationship.
We were not able to replicate the finding that single or less educated women were less likely to have time-appropriate screening . Although we were unable to investigate this finding further, differences in demographic and family structure characteristics of Nunavik, compared to the Canadian population in general may explain this null result [28, 29].
Geographic variation in screening rates was found in the univariate analysis, with one community being highly predictive of time-inappropriate cervical cancer screening, but this association did not maintain significance in the multivariate analysis. Although this study was not able to uncover the factors that contribute to these variations, previous studies suggest that community size , access to female health providers , perceived importance of screening by health providers , and promotion of screening in the community  may play a role. The population size of each community is small (under 2500 people) and health centres are easily accessible by foot in all communities. In the univariate model women from the largest community were found to have a significantly lower odds of time-appropriate Pap smears, suggesting that community size may contribute to the variation by community. The two other communities also showed an elevated, but non-significant odds of time-inappropriate Pap smear use compared to the referent community in both the univariate and multivariate analysis. However, communities 2, 3 and 4 had similar population sizes and therefore size may not be the only factor contributing to the effect of community. Factors such as attitudes around screening in the community and characteristics of the health centre staff are likely to also contribute to these differences. It has been previously shown that Inuit women have a strong preference for having screening performed by a female provider . Although, we were unable to assess the gender ratio of providers in these communities during data collection, the presence of a male provider may have led to a delay in screening acceptance among some women, which could help explain some of our findings. In the 2004 Nunavik Health Survey, 41% of women who had a Pap smear two or more years ago or never reported that cervical cancer screening was not offered by their doctors . The factors related to the relationship between women and their health care providers are especially important because women can only access health care from their community’s health centre.
A limitation of this study was that women were recruited non-randomly, while attending any visit at the health centre that necessitated a Pap smear. Given that our sample was comprised of women who consented to the cohort study, it is likely that we underestimated the proportion of women who have never had a Pap test. It is difficult to assess the extent of this underestimation given that there is no cervical screening registry in Quebec; however, one randomly-selected population based study found that 6.8% of women aged 18–29 years across Nunavik were never screened , whereas a convenience-sample of women 18–63 years in two communities of Nunavik found that only 3.4% had not previously received a Pap test . Our results are within the range reported among other samples in Nunavik, but caution must be exercised in comparison of these results due to the reported age-differences and the self-reported nature of cervical cancer screening history in these studies. Also, our sample was comparable to the general population of women in Nunavik, with the exception of an underrepresentation of women ages 45 and older . The sample size was small, but we had a fairly high overall coverage of the population, given that our study sample represented 50.5% of women aged 20–69 in the four recruitment communities. Furthermore, we assessed time-inappropriate screening as not having a Pap smear within the previous 3 years, but if we had a larger sample it would have been important to assess screening compliance in situations where a shorter interval is needed, such as the follow-up of an equivocal or borderline abnormality in the cytology result, where it is recommended that women have a repeat cytology within 6 to 12 months .
A major strength of this study was the use of a retrospective medical chart review to determine women’s cervical cancer screening history. Given that women tend to over-report their screening history, by obtaining this variable through chart review, we have reduced bias due to measurement error of the outcome variable. As this analysis was part of a larger research program on HPV and cervical cancer screening, we were able to utilize baseline questionnaire data and thus investigate the association with sexual health and reproductive history factors. This study is one of the few studies to be able to analyse the association between these behaviours and Pap screening. Despite the sensitive nature of many of these covariates, we had relatively low levels of missing data (range: 0-8%).
Understanding screening patterns and groups noncompliant with screening is highly relevant to future efforts to reduce the higher cervical cancer incidence and mortality among Inuit women in Nunavik. Although we found screening rates in Nunavik to be comparable to that of the Canadian population, there is cause for concern given the higher incidence of cervical cancer. Further investigation is needed to determine if factors that take place after screening, such as time to treatment, contribute to the differential risk of cervical cancer mortality between these populations.