Cancer Screening Behaviors Among Canadian Women Living With Physical Disabilities

Methods 

Questionnaire Development 

The questionnaire consisted of 12 sections addressing the following areas: level of disability or long-term health condition(s), sociodemographic information, use of personal assistants, use of disability-related aids, cancer screening behaviors, and general health.

The questionnaire incorporated health-related questions from existing population health surveys including the National Population Health Survey (NPHS)14 from which we took the questions regarding ever having a Pap test or pelvic examination (table 1). Questions related to limitation of activity were also adapted from the NPHS. Questions related to the barriers that women face in having these tests were developed for this survey.

Table 1. Variables From the NPHS for Pap Test and Pelvic Examination

	Question	Responses
	Have you ever had a Pap smear test? (this involves the physician inserting an instrument into your vagina to collect cells from your cervix) The Pap smear test checks for cancer of the cervix.	Yes No Don’t know
	If yes, when was the last time you had a Pap smear test?	Less than 6 months ago
		6 months to less than 1 year ago
		1 year to less than 3 years ago
		3 years to less than 5 years ago
		5 years or more ago
	Have you ever had a pelvic examination? (the physician places two gloved fingers into your vagina and presses on your lower abdomen with the other hand)	Yes No Don’t know
	If yes, when was the last time you had a pelvic exam?	Less than 6 months ago
		6 months to less than 1 year ago
		1 year to less than 3 years ago
		3 years to less than 5 years ago
		5 years or more ago

Focus groups occurred within the community to ensure that questionnaire items were relevant and to elicit additional questions. Forty-five women with a wide range of physical disabilities between the ages of 18 and 80 participated in the 6 focus groups in Ontario, Canada (2 in rural centers, 4 in urban centers).

Finally, the questionnaire was pilot tested in Ontario, Canada, to 2 convenience samples of socially and ethnically diverse women with physical disabilities (n=11, n=14) to determine clarity of the questions. Results of the pilot studies were used to revise the content, length, format, and language of the questionnaire.

Statistical Analysis 

We used descriptive statistics to describe the sample. Bivariate analyses were run between the dependent variables of ever having a Pap test or a pelvic examination and sociodemographic variables, disability measure variables, and barrier variables. All applicable variables were first examined on bivariate analysis with the dependent variable. If during bivariate analysis it was determined that there were no subjects within any cells of the 2×2 framework, the variable was then not included in the multivariate analysis. All variables were assessed for colinearity. A multivariate logistic regression was used to determine odds ratios (ORs) for the dependent variables of having a Pap test or pelvic examination.

We developed a model for the logistic regressions for ever having a Pap test and ever having a pelvic examination (fig 1). Variables selected for the model were based on the literature and on previous works in the area of disability and preventive health, as well as the bivariate analysis. No variables displayed colinearity and so they were all retained for multivariate analyses. Blocks of variables were added to the logistic regression sequentially. The first regression included sociodemographic variables. The second regression included the sociodemographic variables and the disability-related variables, and the third regression added the barrier-related variables. For each of the logistic regressions, the same sample of subjects was used for each of the 3 blocks of regressions.

View Large Image Download to PowerPoint

Fig 1. Pap and pelvic analysis framework. Legend: …, model 1: sociodemographic variables with dependent variables; ----, model 2: sociodemographic and disability variables with dependent variables; —, model 3: sociodemographic, disability and barrier variables with dependent variables.

Results 

Using the Abilities mailing list, we mailed out 2352 surveys, of which 926 eligible surveys were returned, 427 ineligible surveys were returned, and 999 surveys were not returned. The conservative response rate using the mailing list of Abilities is 48.1%. An additional 250 surveys were mailed out to women recruited from other sources and 225 eligible surveys were returned with a 90% response rate. The overall response rate was 53% from both sources. From the 1151 respondents, additional surveys were excluded if the participants stated no limitation of activities and/or reported no medical condition or physical disability. The resultant sample was 1095 surveys eligible for analysis.

The majority of women in this sample were between the ages of 30 to 59, with a mean age of 49±14.3 years. Forty-eight percent of the women were married or partnered. Overall the sample was well educated, with 31% having completed university education which may have accounted for the high levels of income with 20% having a household income of $60,000 or more per year. Table 2 provides the breakdown of sociodemographic information on the total sample.

Table 2. Sociodemographic-Related Variables (N=1095)

	Variable	Frequency	Percent
	Age (y)
	18–29	87	8.1
	30–39	211	19.6
	40–49	293	27.1
	50–59	248	23.0
	60–69	132	12.2
	70 and over	108	10.0
	Marital status
	Never married	308	28.6
	Partner/married	519	48.1
	Separated/divorced/widowed	251	23.3
	Education
	Some secondary or less	142	13.2
	Secondary	146	13.5
	Some college	126	11.7
	College	182	16.9
	Some university	151	14.0
	University	331	30.7
	Residence
	Urban	929	88.0
	Rural	129	12.0
	Total household income ($)
	<7,000	47	4.8
	7,000–11,999	154	15.7
	12,000–19,999	126	12.9
	20,000–29,999	135	13.8
	30,000–39,999	123	12.6
	40,000–49,999	114	11.6
	50,000–59,999	83	8.5
	≥60,000	198	20.2
	Ethnicity
	Canadian only	475	43.4
	Canadian and other	432	39.4
	Non-Canadian only	188	17.2

Frequencies and percentages of the disability variables are provided in table 3. The most frequently reported health conditions were 44% musculoskeletal, 17% neurologic, 13% sensory, and 10% cardiorespiratory. Sixty-nine percent of the women surveyed acquired their disability versus the 31% who were born with their disability. The majority of women reported having 3 or fewer health conditions and 63% had assistance with their activities of daily living. The majority of women (44%) did not report having pain that limited their activity.

Table 3. Disability-Related Variables (N=1095)

	Variable	Frequency	Percent
	Born vs acquired
	Born with disability	335	30.7
	Acquired disability	757	69.3
	No. of reported health conditions
	1	286	26.2
	2	239	21.9
	3	214	19.6
	4	135	12.4
	5	91	8.3
	6+	128	11.7
	Use of personal assistance
	Yes	682	62.9
	No	402	37.1
	Use of assistive device
	Yes	878	80.6
	No	212	19.4
	Activities prevented by pain
	No pain	467	43.7
	None	36	3.4
	A few	162	15.2
	Some	270	25.3
	Most	134	12.5

Barriers to having a Pap test or pelvic examination were identical within the questionnaire and were collapsed into 3 categories of barriers based on type of barrier prior to analysis. Three categories of barriers included individual barriers, disability-related barriers, and structural barriers. A barrier was considered an individual barrier if the characteristics of the person prevented the screening behavior (eg, “Lack of time/too busy”). Disability-related barriers were those where the disability itself prevented the screening tests (eg, “It is not easy to get into the position because of pain, spasms, etc”). Structural barriers were those where the environment or the interface between the disability and the environment was the reason for not having the procedures (eg, “I had problems finding transportation to get me there”).

The prevalence of ever having a Pap test for this sample (N=1080) is 90.2% (n=975), with 42% (n=400) of those reporting having the Pap test in the last year and 75% (n=708) within the last 3 years. The prevalence of ever having a pelvic examination (n=1044) was similar at 90.5% (n=944) and 43.8% (n=394) having had the test in the last year and 73% (n=659) within the last 3 years. Only those who had never had a Pap test or pelvic exam or had not had the test in question within the last year went on to answer the barrier-related questions.

As seen in table 4, the most frequently reported barriers to having a Pap test or pelvic examination were: individual barriers such as “not sexually active” (Pap, 19%; pelvic, 21%); disability-related barriers, for example, “my doctor told me that I do not need one” (Pap, 10%; pelvic, 11%) and structural barriers, for example “I have problems getting onto the exam table or the exam table is too high/narrow” (Pap, 12%; pelvic, 12%).

Table 4. Reported Barriers to Having a Regular (once a year) Pap Test or Pelvic Examination

	Barriers	Pap Frequency (%) (n=680)	Pelvic Frequency (%) (n=651)
	Individual barriers
	I am not sexually active	211(19.5)	203(20.7)
	I no longer have a cervix or ovaries	116(10.7)	90(9.2)
	Exams are painful	70(6.5)	53(5.4)
	Lack of time/too busy	69(6.4)	66(6.7)
	I cannot find a doctor I am comfortable with	55(5.1)	49(5.0)
	I wish to avoid hearing bad news	34(3.1)	24(2.5)
	I have a history of sexual abuse and I’m not comfortable with the procedure	23(2.1)	23(2.3)
	I use alternatives to traditional health care	6(0.6)	9(0.9)
	Disability-related barriers
	My doctor told me I do not need one	108(10.0)	103(10.5)
	It is not easy to get into the position because of pain, spasms, etc.	78(7.2)	65(6.6)
	Because of bladder/bowel incontinence I’m too embarrassed to go	31(2.9)	33(3.4)
	I do not need them because of my disability	18(1.7)	22(2.2)
	I cannot find a doctor who knows about my disability	12(1.1)	11(1.1)
	Structural barriers
	I have problems getting onto the table, the table is too high/narrow	128(11.8)	118(12.0)
	Inconvenient to schedule visits	42(3.9)	32(3.3)
	I had problems finding an accessible doctor’s office or clinic	37(3.4)	34(3.5)
	I had problems finding transportation to get me there	25(2.3)	28(2.9)
	The exam room is too cold	22(2.0)	18(1.8)

The results of the logistic regressions for the Pap test are provided in table 5 with a sample of 564 used within the models. Only the participants who responded to the combination of all the sociodemographic, disability-related, and barrier-related variables in addition to not having a Pap tests within the last year were included in the logistic regressions. The same participants were then included in each of the 3 modeled logistic regressions. Table 6 provides the results of the logistic regressions with ORs for the pelvic examination, using a sample of 544 participants. The same participants were included in the 3 modeled logistics and all participants were required to have responded to the sociodemographic, disability-related, and barrier-related variables in addition to not having a pelvic exam within the last year.

Table 5. ORs for Pap Test With 95% CI⁎

	Variables	Model 1: Sociodemographic	Model 2: Sociodemographic and Disability	Model 3: Sociodemographic, Disability, and Barriers
	Sociodemographic variables
	Age	1.37(1.08–1.74)	NS	NS
	Marital status
	Single/never married	1.0	1.0	1.0
	Married/partner	7.43(3.53–15.65)	7.74(3.31–18.09)	6.11(2.28–16.38)
	Separated/divorce/widowed	6.55(2.43–17.64)	6.51(2.19–19.36)	6.48(1.93–21.70)
	Disability variables
	CP	†	0.26(0.08–0.80)	0.17(0.05–0.64)
	Depression	†	2.97(1.10–8.04)	3.93(1.32–11.67)
	Barrier variables
	Not sexually active	†	†	0.31(0.14–0.71)
	Problems getting onto the table	†	†	3.05(1.05–8.86)
	Exam too painful	†	†	6.37(1.59–25.54)

Abbreviation: NS, not significant.

⁎ P<.05(n=564). † Did not enter the model.

Table 6. ORs for Pelvic Exam(95% CI)⁎

	Variables	Model 1: Sociodemographic	Model 2: Sociodemographic and Disability	Model 3: Sociodemographic, Disability, and Barriers
	Sociodemographic variables
	Age(in decades)	NS	NS	1.38(1.02–1.87)
	Marital status
	Single/never married	1.0	1.0	1.0
	Married/partner	6.36(3.15–12.84)	6.50(2.94–14.37)	4.12(1.68–10.10)
	Separated/divorce/widowed	10.20(3.58–29.10)	8.56(2.76–26.59)	7.57(2.29–25.01)
	Disability variables
	Assistance with daily activities	‡	2.18(1.09–4.36)	2.59(1.25–5.37)
	Barrier variables
	Not sexually active	‡	‡	0.34(0.16–0.73)
	I do not need them due to my disability	‡	‡	0.14(0.02–1.03)†

⁎ P<.05(n=544). † P=.054. ‡ Did not enter the model.

Regarding the Pap test and pelvic examination and the sociodemographic variables, those who were married or partnered, separated, divorced, or widowed had much higher odds of ever having had a Pap test or pelvic exam compared with those who were single and had never been partnered or married (for ORs, see Table 4, Table 5). Age was not related to ever having a Pap test in the final model, but age was associated with higher odds of having a pelvic exam (OR ± standard deviation, 1.38±0.21; 95% confidence interval [CI], 1.02–1.87). Turning to the disability-related variables, for the Pap test, those with the health condition of cerebral palsy (CP) had decreased odds of ever having a Pap test (OR=0.17±0.12; 95% CI, 0.05−0.64). Conversely those with depression as a condition were more likely to have had the Pap test (OR=3.93±2.18; 95% CI, 1.32−11.67). For pelvic exams, the only disability-related variable was that participants who used personal assistance were more likely to have had a pelvic exam (OR=2.59±0.96; 95% CI, 1.25−5.37).

When we examined the barriers to having had a Pap test, 3 barriers had increased or decreased odds of ever having had a Pap test: not sexually active (OR=0.31±0.13; 95% CI, 0.13−0.71), problems getting onto the examination table (OR=3.05±1.66; 95% CI, 1.05−8.86), and the examinations are too painful (OR=6.37±4.51; 95% CI, 1.59−25.54). The barriers to having a pelvic exam included; not sexually active (OR=0.34±0.13; 95% CI, 0.16−0.73) and a trend for I do not need them due to my disability (OR=0.14±0.14; 95% CI, 0.02−1.03).

Discussion 

This study found that the prevalence of women ever having a Pap test was 90%, which is very similar to data from the United States where, in 1994, 91% of the women living with functional limitations had ever had a Pap test.7 The prevalence of ever having had a pelvic examination was 91%. Information is also provided on the barriers that these women face in having cancer screening tests performed.

Within this model the overall logistic regressions for the Pap test and pelvic examination indicate that for women with physical disabilities those who are married, with a partner, or have previously been married compared with single or never married have increased odds of having had a Pap test or pelvic examination. These findings are consistent with the results of the 1996−1997 NPHS in Canada,1 but are inconsistent with reports of those with mobility impairments in a national study in the United States.16 Age, however, was not uniformly related to increased odds of having the Pap test or pelvic exam. Interestingly, increases in education or income did not have statistically significant change in ORs for the Pap test or pelvic examination, as had been previously reported in the literature for Canadian women.1

With respect to the disability-related variables, the Pap test and pelvic examination logistic regressions are very different. The ORs for barrier-related questions also have distinct features for the Pap test and pelvic examination. The following discussion will first examine the findings related to having ever had a Pap test followed by the pelvic examination.

The only 2 specific disabilities that displayed a relationship to ever having had the Pap test were those who reported having CP and those reporting depression. As previously discussed, other studies had examined the prevalence of screening among the deaf community12 and those with MS10; however, within this study those specific disabilities did not display a relationship with having ever had a Pap test.

Women who reported having CP had decreased odds of ever having a Pap test within the model used. This may indicate that their health condition takes priority when they see their physicians and that they never address the issue of cancer screening tests during their physician visits. This has been reported within subjective narratives by Thomas17 and by Canadian women with physical disabilities in a report from British Columbia.18

Conversely, having reported depression as a disability condition greatly increases the odds of ever having a Pap test. Possible explanations for the increased odds may be that women with depression undergo more clinical screening tests prior to their diagnoses. Alternatively, these women may be more aware of their health and are more likely to have the cancer screening tests. Given that we do not know the history of depression with these women it is impossible to determine the direct relationship between having had a Pap test and self-reporting the health condition of depression.

Women that reported not being sexually active as a barrier or reason for not having a Pap test were very unlikely to have ever had the test. This is consistent with findings from the British Columbia report.18 Cancer screening guidelines from various institutes, however, all report that screening should begin once a women is sexually active or at the age of 18.5 Thus, these women who feel that not having sex obviates a Pap test need further education on the necessity of being screened.

The barriers of “the exam is too painful” and “I have problems getting onto the exam table or table is too narrow” have increased odds of having a Pap test. This indicates that these women have had the test and may even realize the importance of the test but have individual barriers or disability-related barriers to having the test again. This also confirms the inaccessibility of examination tables quantitatively that has been reported qualitatively in studies examining the barriers to preventive services among those living with physical disabilities.9, 16, 19, 20

Regarding pelvic examinations, no health conditions exhibited increased or decreased odds of ever having a pelvic examination within this model. The use of personal assistance was the only disability-related variable that had increased odds of having ever had a pelvic examination. Women with personal assistance may use their assistance when going to the physician appointments, and thus may have fewer physical barriers to receiving a pelvic examination (eg, help with getting onto the table). Personal assistance in the context of this study could be from 3 major sources. First, a woman may have a family member or friend as an informal or formal personal assistant. Second, personal assistance may be government supported. Some government supported programs have personal assistance that is portable, such that the person living with a disability can get their needs met outside of the home (eg, school or health care). Third, a woman may pay for attendant services strictly out of pocket. In Canada, people with more activities of daily living needs will be more likely to have some form of government supported personal assistance. In this sample of women there was a mix of personal assistance and women may use a combination of informal and formal sources of assistance.

Similarly to the Pap test, the barrier of “not being sexually active” resulted in decreased odds of ever having had a pelvic examination. Again these women may need further education on the need and benefits of regular screening tests.

The barrier “I do not need the test due to my disability,” although not statistically significant, did display a trend for decreased odds in ever having a pelvic exam. This explanation for not having a pelvic examination needs further investigation. Is it that these women have been told that they do not need the test due to their disability? And if so who told them they did not require the test? Or do these women believe that they do not require the test, in which case they need further education on the benefits of pelvic examinations.

Conclusions 

Further studies are required to examine if women with physical disabilities are having cancer screening tests on a regular basis. This study only investigated if these women had ever had a Pap test or pelvic examination and because it was a cross-sectional design, it did not assess if the women were meeting suggested guidelines for frequencies of the tests. Different barrier patterns may be evident among women with physical disabilities who have only ever had 1 Pap test or pelvic examination compared with those who continue with the screening tests on a regular basis. In addition, there is no information available to report on the mortality or morbidity of cervical cancer for women living with physical disabilities.

Qualitative work may also help to determine the specific reasons that these women feel they do not require a Pap test or pelvic examination. As discussed earlier, is the lack of screening due to the fact that someone told them they do not require the test, and if so who provided them with that information? Or do these women believe that their disability precludes them from having cancer and thus they do not require any cancer screening tests?

Additional work is also required to examine women with physical disabilities living in institutions, because they may have very different responses to having ever had the cancer screening tests and more importantly may have many different barriers to receiving preventive health screening tests. Also French-speaking Canada needs to be surveyed to determine if there are cultural differences among women with physical disabilities in Canada.

This study, though, does provide us with information regarding the prevalence of Pap tests and pelvic examination among this sample of women with physical disabilities within Canada. Additionally, information related to barriers is important, because the specific barrier of “not needing the tests due to sexual inactivity” or believing that cancer screening tests are “not required due to their disabilities” indicates the need for further education of Canadian women with physical disabilities and their health care providers on the benefits of cancer screening tests.