Objective
The objective of the study was to investigate baseline knowledge and demographic factors associated with a lack of knowledge about urinary incontinence (UI) and pelvic organ prolapse (POP).
Study Design
This study was a community-based, cross-sectional survey of 431 racially and socioeconomically diverse women aged 19-98 years. The Prolapse and Incontinence Knowledge Questionnaire was used to assess participants’ knowledge. Primary endpoints were the total number of correct responses on the UI and POP scales, respectively. Percentages of individuals answering each item or group of items correctly were explored as secondary outcomes.
Results
All women lacked knowledge proficiency about UI and POP, although knowledge about UI was slightly greater than knowledge about POP. Overall, 71.2% of subjects lacked UI proficiency (<80% correct), whereas 48.1% lacked proficiency in POP knowledge (<50% correct). Black women demonstrated significantly less knowledge about UI and POP than white women, both before and after adjustment for age, education, and household income. When combined into 1 group, Asian, Hispanic, and other women also showed significantly less UI and POP knowledge than white women. Most women who reported symptoms of UI had not received treatment for their problems.
Conclusion
There is a global lack of knowledge about UI and POP among community-dwelling women, with more pronounced knowledge gaps among nonwhite women. UI and POP are chronic medical conditions that should be included in routine screening questions for well-woman care. Further studies are needed to explore how best to educate and improve women’s awareness of these prevalent pelvic floor disorders.
Pelvic floor disorders (PFDs) are a broad category of disorders in the gynecological, lower urinary, and gastrointestinal tract that affect the structure and/or function of organs in the female pelvis. The 3 most prevalent PFDs are urinary incontinence (UI), fecal incontinence (FI) and pelvic organ prolapse (POP). Approximately 25% of women 20 years old or older in the United States suffer from at least 1 of the 3 most prevalent PFDs, with projections of an exponential increase in prevalence over the next 40 years.
Women suffer significant physical and emotional distress from PFDs, including depression, loss of self-esteem, and social isolation. Additionally, UI and FI are common motivating factors for placing family members in nursing homes. US expenditures exceed 12 billion dollars annually for the management and treatment of UI alone, making PFD a major public health concern. Unfortunately, many women never seek care and instead suffer silently with their symptoms.
It is estimated that only 18-50% of women with PFDs seek care from health care providers. Insufficient knowledge and misperceptions about PFDs are thought to represent the largest barriers to seeking care. Compared with women who seek care, non–care-seeking women are more likely to believe that PFDs are a normal part of aging and to report that they are unaware of available treatment options or that they fear the need for invasive procedures. Moreover, many individuals do not consider UI symptoms to be a medical problem and thus may not provide information about their symptoms to their providers. Collectively, these findings suggest that women are grossly undereducated about PFDs and that the lack of education is having a negative impact on their ability to seek care.
Our current understanding of women’s knowledge of PFDs obtained using validated tools is limited. Most studies focus only on UI or on women already seeking care. The aim of our study was to assess knowledge proficiency of PFDs in a racially, ethnically, and socioeconomically diverse population of community-dwelling women and to identify demographic factors associated with lack of knowledge proficiency.
Materials and Methods
Study population and design
We conducted a cross-sectional written survey of women’s knowledge of UI and POP in New Haven County, Connecticut, from February 2010 to August 2011. Women were invited to participate by 1 of 5 study investigators at events free and open to the public. Potential venues were identified from advertisements in newspapers, local banners, billboards, and word of mouth. Venues were selected based on event location (within New Haven County), anticipated number of participants (>50), and racial and ethnic diversity of attendees as well as ease of access (walking or public transportation). These included 2 women’s health conferences (n = 81, 18.8%), 5 summer outdoor concerts or arts events (n = 213, 49.4%), 1 outdoor sports event (n = 33, 7.7%), lunch socials at 2 senior centers (n = 58, 13.5%), and other or missing location data (n = 46, 10.7%).
Women were approached and asked to complete a survey on women’s health. Each packet included a cover sheet explaining that this was a research study about an important women’s health care issue, and a copy of the unmodified Prolapse and Incontinence Knowledge Questionnaire (PIKQ) was included. Those agreeing to participate were asked to place completed surveys in a covered box or opaque envelope. The study was limited to English-speaking women because a validated Spanish version of the questionnaire was not available. Age was confirmed by direct inquiry at the time of distribution. Women reporting ever having completed the survey were excluded. Questionnaires were self-administered and no personal identifiers were required.
Study measurements and outcomes
Participants’ knowledge about UI and POP was assessed by the previously validated PIKQ, a 24 item questionnaire that includes 12 questions focused on UI knowledge (UI scale) and 12 questions focused on POP knowledge (POP scale). Each question had 3 possible responses: yes, no, and I do not know. The PIKQ is illustrated in the Figure . As per the original scoring criteria, 1 point was given for each correct response, and no points were given for incorrect responses, blank responses, or for the response of “I don’t know.”
The primary endpoints were the total UI and POP scale scores, defined as the sum of the total number of correct responses, with a maximal score of 12 for each scale. UI and POP questions were also evaluated as individual items as well as by the type of knowledge reflected ( Figure ). The percentages of individuals answering each item or group of items correctly were explored as secondary outcomes.
Demographic information obtained included age; race, categorized as Hispanic, white (non-Hispanic), African-American, Asian, and other; parity; menopausal status; marital status; yearly household income; and highest level of completed education. Questions about whether participants had a job in the medical field, had ever experienced prior problems with UI and/or POP, had ever seen a urologist or urogynecologist, and whether they had ever received treatment(s) for UI or POP were also included ( Table 1 ). This information was used to assess factors that could positively or negatively influence total UI and POP scale scores.
| Variable | Values |
|---|---|
| Age, y | |
| Mean ± SD | 49.2 ± 17.9 |
| 19-29, n (%) | 61 (14.7) |
| 30-39 | 74 (17.8) |
| 40-49 | 82 (19.8) |
| 50-59 | 88 (21.2) |
| 60-69 | 45 (10.8) |
| 70-79 | 34 (8.2) |
| >79 | 31 (7.5) |
| Race, n (%) | |
| White | 266 (63.9) |
| Hispanic | 26 (6.3) |
| African-American | 93 (22.4) |
| Asian | 13 (3.1) |
| Other | 18 (4.3) |
| Marital status, n (%) | |
| Not currently married | 224 (53.5) |
| Currently married | 195 (46.5) |
| Parity | |
| Mean ± SD | 1.6 ± 2.1 |
| 0, n (%) | 152 (36.5) |
| 1 | 71 (17.1) |
| 2 | 100 (24.0) |
| 3 | 53 (12.7) |
| >3 | 40 (9.6) |
| Menstrual status, n (%) | |
| Still having periods | 203 (50.1) |
| Near menopausal | 38 (9.4) |
| Menopausal | 164 (40.5) |
| Work in a medical field, n (%) | |
| No | 297 (71.7) |
| Yes | 117 (28.3) |
| Household income, n (%) | |
| <$10,000 | 33 (8.8) |
| $10,000-49,000 | 100 (26.5) |
| $50,000-100,000 | 156 (41.4) |
| >$100,000 | 88 (23.3) |
| Education, n (%) | |
| Eighth grade or less | 10 (2.4) |
| High school | 82 (19.6) |
| College | 191 (45.7) |
| Graduate school | 135 (32.3) |
| Ever seen a urologist/urogynecologist, n (%) | |
| No | 328 (78.7) |
| Yes | 89 (21.3) |
| Ever had a problem with urine leakage, n, (%) | |
| No | 282 (67.8) |
| Yes | 134 (32.2) |
| Ever been treated for leakage of urine, n (%) | |
| No | 397 (95.4) |
| Yes | 19 (4.6) |
| Ever had a problem with POP, n (%) | |
| No | 389 (94.0) |
| Yes | 25 (6.0) |
| Ever been treated for POP, n (%) | |
| No | 401 (95.9) |
| Yes | 17 (4.1) |
Proficiency was defined as scores of 80% or greater on the UI scale and 50% or greater on the POP scale, based on frequency data from the original authors showing that this represented higher-than-usual knowledge for the stated scales. The study was reviewed and exemption status was granted by the Yale University Human Subjects Committee, prior to initiation.
Statistical analysis
Questionnaire data were manually entered into a Microsoft Excel database (Microsoft Excel 2007; Microsoft Corp, Redmond, WA) with verification and variables examined for implausible values, missing data and deviations from normality. Demographic data were presented as means ± SD and/or percentages. Unadjusted and adjusted odds ratios (ORs) with 95% confidence intervals (CIs) were calculated for the odds of lacking UI or POP proficiency (<80% and <50% correct scale scores, respectively).
For the adjusted ORs, multivariate models were constructed using logistic regression and backward variable elimination, starting from a model containing all covariates as predictors (except for percentage of UI correct and percentage of POP correct), and using P ≥ .05 as the criterion for elimination from the model. Because age, race, household income, and education may be potentially important confounders, they were retained in the models, even if they were not significantly associated with the outcomes. The adjusted ORs are presented for variables that were retained in the final model, adjusted for each of the other variables that were maintained. Calculations were performed using SAS 9.1 (SAS Institute, Cary, NC). Statistical significance was defined as P < .05.
Results
Knowledge about UI and POP among the study population
Overall, 431 subjects were enrolled. Table 1 provides demographic data for the study participants. One subject (0.2%) did not complete the UI scale, whereas 13 subjects (3.0%) did not complete the POP scale. These subjects were excluded from the respective UI or POP analyses. All subjects, however, completed at least the UI or POP questionnaire and thus were retained in the overall study. Due to the community outreach design of the study, a precise response rate was indeterminable.
In total, 71.2% of subjects lacked UI proficiency (<80% correct), whereas 48.1% lacked proficiency in POP knowledge (<50% correct). The average scale score (number of correct responses) for UI knowledge among all subjects was 6.7 ± 3.7% (55.8 ± 30.8%); the median and interquartile range (IQR) were 7 (58.3%) and 6 (50.0%), respectively. For POP knowledge, the average scale score was 5.5 ± 3.5% (45.8 ± 29.2%); the median and IQR were 6 (50.0%) and 5 (41.7%), respectively.
Association between question type and baseline knowledge
The percentage of correct responses did not vary significantly between groups categorized according to the type of knowledge they represented ( P > .05, for all comparisons). For UI, the lowest scores were for questions regarding diagnosis (mean percentage correct, 55 ± 40%), whereas the highest scores were for questions regarding treatment (58 ± 38%). For POP, the lowest scores were for questions regarding etiology (mean percentage correct, 45 ± 32%) and treatment (45 ± 35%), whereas the highest scores were for questions regarding diagnosis (50 ± 40%).
Factors associated with UI or POP knowledge
To assess whether UI or POP knowledge varied by subject demographics, univariate analyses were conducted ( Table 2 ). The odds of lacking UI proficiency were significantly higher among subjects who lacked POP proficiency, were of age older than 79 years (vs 19-29 years), were African-American (vs white), had more than 3 children (vs none), had a household income of $10,000-49,000 (vs >$100,000), and had a high school or college education (vs graduate school education). In contrast, subjects who worked in a medical field had significantly lower odds of lacking UI knowledge proficiency, compared with those who did not.
| Stratum | Lack of proficiency on UI scale (<80% correct) | Lack of proficiency on POP scale (<50% correct) | ||||
|---|---|---|---|---|---|---|
| % | OR | 95% CI | % | OR | 95% CI | |
| Total | 71.2 | N/A | 48.1 | N/A | ||
| UI percentage correct | ||||||
| <80% | 61.3 | 8.97 | 5.16–15.58 a | |||
| ≥80% | 15.0 | 1.0 | Ref | |||
| POP percentage correct | ||||||
| <50% | 91.0 | 8.97 | 5.16–15.59 a | |||
| ≥50% | 53.0 | 1.0 | Ref | |||
| Age, y | ||||||
| 19-29 | 68.8 | 1.0 | Ref | 50.8 | 1.0 | Ref |
| 30-39 | 73.0 | 1.22 | 0.58–2.58 | 46.6 | 0.84 | 0.43–1.67 |
| 40-49 | 69.1 | 1.01 | 0.49–2.08 | 49.4 | 0.94 | 0.49–1.84 |
| 50-59 | 59.1 | 0.65 | 0.33–1.30 | 37.6 | 0.58 | 0.30–1.34 |
| 60-69 | 80.0 | 1.81 | 0.73–4.49 | 36.6 | 0.56 | 0.25–1.26 |
| 70-79 | 73.5 | 1.26 | 0.49–3.20 | 47.1 | 0.86 | 0.37–1.99 |
| Older than 79 | 93.5 | 6.56 | 1.42–30.35 a | 79.3 | 3.71 | 1.33–10.38 a |
| Race | ||||||
| White | 64.5 | 1.0 | Ref | 42.5 | 1.0 | Ref |
| Hispanic | 80.8 | 2.31 | 0.84–6.32 | 61.5 | 2.17 | 0.95–4.96 |
| African-American | 83.9 | 2.86 | 1.56–5.25 a | 55.1 | 1.66 | 1.02–2.70 a |
| Asian | 100.0 | N/E | N/E | 53.8 | 1.58 | 0.52–4.83 |
| Other | 77.8 | 1.92 | 0.62–6.01 | 66.7 | 2.71 | 0.99–7.44 |
| Marital status | ||||||
| Not married | 75.8 | 1.0 | Ref | 48.2 | 1.0 | Ref |
| Married | 67.2 | 0.65 | 0.43–1.00 | 48.4 | 1.01 | 0.69–1.49 |
| Children, n | ||||||
| 0 | 65.8 | 1.0 | Ref | 42.6 | 1.0 | Ref. |
| 1 | 67.1 | 1.06 | 0.58–1.94 | 41.4 | 0.95 | 0.54–1.70 |
| 2 | 72.0 | 1.34 | 0.77–2.32 | 52.6 | 1.50 | 0.89–2.50 |
| 3 | 77.4 | 1.78 | 0.86–3.67 | 54.7 | 1.63 | 0.87–3.07 |
| >3 | 87.5 | 3.64 | 1.35–9.85 a | 54.0 | 1.59 | 0.77–3.27 |
| Menstrual status | ||||||
| Still having periods | 70.3 | 1.0 | Ref | 48.8 | 1.0 | Ref |
| Near menopausal | 65.8 | 0.81 | 0.39–1.70 | 50.0 | 1.05 | 0.53–2.10 |
| Menopausal | 74.4 | 1.23 | 0.77–1.95 | 45.5 | 0.88 | 0.58–1.34 |
| Work in a medical field | ||||||
| No | 78.4 | 1.0 | Ref | 57.2 | 1.0 | Ref |
| Yes | 53.8 | 0.32 | 0.20–0.51 a | 23.9 | 0.24 | 0.14–0.38 a |
| Household income | ||||||
| <$10,000 | 72.7 | 1.85 | 0.77–4.43 | 48.5 | 1.51 | 0.67–3.40 |
| $10,000-49,000 | 80.8 | 2.92 | 1.51–5.62 a | 59.8 | 2.39 | 1.32–4.33 a |
| $50,000-100,000 | 67.9 | 1.47 | 0.85–2.52 | 42.5 | 1.19 | 0.69–2.04 |
| >$100,000 | 59.1 | 1.0 | Ref | 38.4 | 1.0 | Ref |
| Education | ||||||
| Eighth grade or less | 70.0 | 1.56 | 0.39–6.28 | 60.0 | 2.13 | 0.34–13.22 |
| High school | 89.0 | 5.41 | 2.50–11.71 a | 66.2 | 2.07 | 1.08–3.97 a |
| College | 71.6 | 1.68 | 1.05–2.68 a | 44.4 | 1.02 | 0.63–1.65 |
| Graduate school | 60.0 | 1.0 | Ref | 40.5 | 1.0 | Ref |
| Ever seen a urologist/urogynecologist | ||||||
| No | 70.6 | 1.0 | Ref | 48.6 | 1.0 | Ref |
| Yes | 73.0 | 1.13 | 0.67–1.90 | 44.7 | 0.86 | 0.53–1.38 |
| Ever had a problem with urine leakage | ||||||
| No | 73.4 | 1.0 | Ref | 50.5 | 1.0 | Ref |
| Yes | 66.9 | 0.73 | 0.47–1.15 | 42.2 | 0.71 | 0.47–1.09 |
| Ever been treated for leakage of urine | ||||||
| No | 71.2 | 1.0 | Ref | 47.7 | 1.0 | Ref |
| Yes | 73.7 | 1.13 | 0.40–3.22 | 47.4 | 0.99 | 0.39–2.39 |
| Ever had a problem with POP? | ||||||
| No | 71.4 | 1.0 | Ref | 49.3 | 1.0 | Ref |
| Yes | 64.0 | 0.71 | 0.31–1.67 | 20.8 | 0.27 | 0.10–0.74 a |
| Ever been treated for POP? | ||||||
| No | 71.0 | 1.0 | Ref | 48.5 | 1.0 | Ref |
| Yes | 76.5 | 1.33 | 0.42–4.16 | 29.4 | 0.44 | 0.15–1.28 |
The odds of lacking POP proficiency were significantly higher among subjects who lacked UI proficiency, were of age older than 79 years (vs 19-29 years), were African-American (vs white), had a household income of $10,000-49,000 (vs >$100,000), and had a high school education (vs graduate school education). Significantly lower odds of lacking POP proficiency were found among subjects who worked in a medical field and among subjects who had ever had a problem with POP.
After adjustment for potential confounders, multivariate analysis ( Table 3 ) of factors associated with a lack of proficiency in UI and POP knowledge showed that African-American subjects remained significantly less likely to be proficient in both UI and POP knowledge compared with white women. Asian subjects reported lower UI knowledge than white women, but an OR could not be calculated for this group because of all women having scores reflective of a lack of proficiency. Although racial backgrounds other than African American were not significantly associated with UI or POP proficiency individually, they showed relatively strong effect sizes (adjusted ORs greater than 2 or 3, except for other category in the case of UI proficiency). When these racial groups (Hispanic, Asian, and other) were combined together in an analysis, the ORs (adjusted for the other factors listed in Table 3 ) for this combined group (vs whites) became 3.37 (95% CI, 1.41–8.09) for lack of UI proficiency, and 2.74 (95% CI, 1.29–5.83) for lack of POP proficiency.
