A cross-sectional design was used to describe the prevalence, bother, beliefs, and help-seeking behavior of pregnant women. The Medical Ethics Committee of the Maastricht University Medical Centre (MUMC+) was consulted. It was stated that ethical approval was not necessary because of the non-invasive character of the study (MECC 019–1320). Pregnant women ≥ 18 years old, regardless of parity and weeks of gestation, and able to fill in a digital survey, were eligible to participate. Based on an overall expected prevalence of UI of 41%, a Z statistic of 1.96, and precision of 0.05, a minimal sample size of 371 women was estimated to fill in the survey . Nationwide midwifery and pelvic physiotherapy practices were among others asked to share a social media message (using Facebook and LinkedIn) containing brief information on the study (goal, eligibility) and a link to the patient information letter and digital survey. Before proceeding to the anonymized digital survey, eligible women signed informed consent forms electronically, in agreement with ethical regulations. The survey took 10 to 15 min to complete.
The survey consisted of four parts: (1) demographic variables such as age, trimester of pregnancy, educational level, and parity, (2) International Consultation on Incontinence Questionnaire-Urinary Incontinence Short Form (ICIQ-UI SF) , (3) International Consultation on Incontinence Questionnaire Lower Urinary Tract Symptoms Quality of Life (ICIQ-LUTSqol) , and (4) questions on beliefs and help-seeking behavior regarding UI.
The ICIQ-UI SF provides an indication of UI severity and consists of four questions. The first question assesses frequency of UI, with a score of 0 (never losing urine) to 5 (losing urine all the time). The second question describes the amount of urine loss, with four response categories ranging from 0 (no loss) to 6 (large amount). The third question assesses the impact of UI on daily life, ranging from 0 (not at all) to 10 (a great deal). The total score ranges from 0 (no impact of UI on quality of life) to 21 (very severe problem). The total score is divided into four severity categories: slight (1–5), moderate (6–12), severe (13–18), and very severe (19–21) . A fourth question on the occurrence of symptoms of UI was used to indicate SUI or MUI . A respondent was considered to have SUI when leaking urine with a cough or a sneeze and/or when physically active/exercising, but not before getting to the toilet. UUI is considered when the respondent leaks, because of irresistible need to void, before getting to the toilet. A respondent with MUI experiences both SUI and UUI.
The ICIQ-LUTSqol is a condition-specific health-related quality of life questionnaire (20 questions) adapted for use within the ICIQ structure from the King’s Health Questionnaire . It contains 19 questions that can be scored on life restrictions, emotional aspects, and preventive measures. It is scored on a four-point Likert scale ranging from 1 (not at all) to 4 (a lot). Three questions on relationships, sex life, and family life included additionally ‘not applicable.’ ‘Not applicable’ was considered as not affecting daily life . The sum score ranges between 19 and 76. A higher score indicates a higher impact on quality of life. Every question is accompanied by a question regarding experienced bother [ranging from 0 (no bother) to 10 (extreme bother)]. It is arbitrarily decided that a score of at least 5 indicates significant bother on a specific item. The 20th question is on how much urinary symptoms interfere with daily life. This is scored between 0 to 10 (similar to experienced bother). Both the ICIQ-UI SF and ICIQ-LUTSqol are rated as ‘high-quality’ questionnaires and are recommended by the ICI . The ICIQ-UI SF and the ICIQ-LUTSqol were provided in the Dutch language by the Bristol Urological Institute .
All respondents at least filled in the demographic variables and ICIQ-UI SF. Answering ‘never losing urine’ at the frequency item of the ICIQ-UI SF indicated continence and consequently the survey was finished. When reporting UI, women completed the remaining two parts on quality of life and help-seeking behavior.
The questions on beliefs and help-seeking behavior were self-constructed. Selection of question and answer options was based on models explaining help-seeking behavior and discussion with experts in the field (epidemiologists and obstetrician/gynecologist) and modified accordingly [16, 17]. Moreover, questions were reviewed by an expert for readability and comprehensiveness, followed by field testing. Ultimately, six questions were developed including four topics on health-seeking behavior [actual help-seeking, reason(s) to (not) seek help, reason to seek help in the future, and consulted health care provider(s)] and two topics on beliefs (self-perceived prognosis and self-perceived best intervention to treat UI in general).
Data were analyzed using descriptive statistics presented as proportions [frequency and means (SD)]. An independent sample t-test was conducted to compare help-seekers and non-help-seekers regarding UI severity (ICIQ-UI-SF total score), bother (ICIQ-LUTSqol total score), and interference in daily life. A chi-square test was used to test relationships between categorical variables. The effect size is estimated with Cohen’s d. Cohen’s d presents the difference between groups (help-seekers and non-help-seekers) in standard deviation units. To interpret the strength of the effect size, we follow the guidelines proposed by Cohen: 0.2 = small, 0.5 = medium, and 0.8 = large. An alpha of 0.05 is considered significant. Analyses were done using IBM Statistical Package for Social Sciences (SPSS), version 26.0 (New York, NY, USA).