Urinary incontinence (UI) is defined as an involuntary urinary leakage [1] and has been affecting up to 57.7% of women in low-income countries [2]. Pregnancy affects hormonal changes, which may add an additional risk of UI among adult women [3]. Additionally, other factors, such as damage to the periurethral, paraurethral, and pubo-urethral connective tissues, may occur during pregnancy, labor and delivery, or with obesity, which may impact the urethra and bladder neck position at rest, leading to UI [4]. Another recent meta-analysis, which included studies from 1998 to 2018, stated that about half of pregnant women (41.0%) experienced UI, and it negatively affects their quality of life (QoL) [5].

Women with UI had an unpredictable demand to use the bathroom and became more aware of the bathroom location when shopping and traveling. They felt fear and easily frustrated when they were unable to get to the bathroom on time [6]. During the nighttime, they visited the bathroom a few times, and their sleep was affected significantly. Interrupted sleep led to problems in occupational functioning, as well as psychological functioning, as it worsened depression [7]. Pregnant women may experience a multidimensional negative impact on QoL, such as social-emotional relationships, physical activities, employment issues, limitation to travel, sleep disruption, and obstacles in performing their prayers [8]. Pregnant women who experience UI during pregnancy reported having 5 times the risk of having UI during their postpartum period [9]. Therefore, there is an urgent need to screen and treat them as early as possible by performing pelvic floor muscle training (PFMT).

PFMT or Kegel exercise is an essential exercise among pregnant women. Strengthening their pelvic floor muscles is recommended as it is minimally invasive and does not involve any complications [1,10]. The advantages of the exercise are to shorten the duration of the second stage of labor, reduce severe perineal lacerations [11], and shorten the painful experience of the postpartum period. Pregnant women should be aware of the benefit and be able to perform the correct techniques of the exercise.

Regretfully, not all pregnant women are aware of PFMT despite UI affecting their daily activities because PFMT uptake may be constrained by the antenatal service provision or challenges in accessing services at a primary care clinic [12]. Pregnant women experience challenges related to PFMT education as it involves an individualized approach, for example, the specific language, different levels of health literacy, and cultural variations [13]. In some cultures, the anatomical involvement may be sensitive to certain women [14], leading to necessary adjustments in disseminating the correct information of PFMT according to those cultures.

Additionally, there may be limited attention toward pelvic floor health during pregnancy provided by health care providers (HCPs) as it may be unclear to whom these professional responsibilities belong [13]. PFMT is actually under the responsibility of a physiotherapist; however, pregnant women are under antenatal care follow-up, which is conducted mainly in primary care clinics where there is limited access to physiotherapists [13]. Moreover, the lack of standardized guidelines, inadequate information, and a lack of continuity of care may result in organizational variation in antenatal care services, which worsens the accessibility and acceptability of PFMT services [13].

Regarding individual factors causing training barriers, only one-tenth of pregnant women seek help due to the misperception that UI will resolve by itself [15], assuming that it is “normal” to have UI, and having a misconception that pelvic training will lead to miscarriage [16]. Hence, women face difficulties in achieving the necessary knowledge and skills, resulting in poor attitude and adherence towards pelvic training [13]. Adherence to daily training is one of the most important prognostic factors for PFMT effectiveness in both the short term and longer term [17]. A new method to disseminate PFMT education is necessary to manage both (individual and HCPs) barriers.

Mobile health (mHealth) is defined as “a medical and public health practice supported by mobile devices, such as mobile phones, patient monitoring devices, personal digital assistants, and other wireless devices” [18]. The National Institute for Health and Care Excellence Guideline has categorized the mHealth apps into three-tier functional classifications:

With regards to the tier C classification, the mHealth app intervention consists of six items which are (1) addressing preventative behavior change, such as smoking; (2) addressing self-management specific behaviors using behavior change techniques; (3) guiding the treatment; (4) providing active monitoring, for example, tracking patients’ location; (5) providing diagnosis, care, or calculated treatment; and (6) providing or guiding the diagnoses [19].

What makes mHealth a powerful tool for behavior modification are its strengths, which include its ease of access and user-friendliness, resulting in its widespread adoption worldwide [20,21]. mHealth can be used by an individual anywhere, at any time, and patients can communicate with health care providers or even a chatbot on specific issues related to the app they are using [22,23]. The strength of a well-designed app is the ability to be well-accepted by the users, change their attitudes, and reduce the acceptability barriers in receiving health care services.

However, qualitative reviews on midwives reported unfavorable findings which did not fully support the apps for several reasons [24]. Midwives were concerned about the accuracy of the apps [25] and their negative impacts on the patient-professional relationship, such as shifting the patient’s trust from trusted to untrusted sources [24]. Hence, there is a need to design a validated pregnancy mHealth app that has undergone the necessary steps based on a framework to improve its effectiveness.

An mHealth app needs to be carefully designed to ensure its impact on the users and its effectiveness. Despite mHealth becoming popular, there is still limited evidence of its effectiveness [26], which is most probably due to the unmet need of incorporating users in the design process [27]. mHealth apps designed for the users but not with the users will lead to high rates of technology rejection [28,29]. Hence, mHealth apps should be designed to fulfill the user’s requirement using a user-centered design (UCD) framework.

Designing and developing an mHealth app using the user-centered element by involving the users in all stages has its robustness and is a gold standard approach in accomplishing mHealth apps that are useful, easy to use, and satisfying to the users [30]. For example, an mHealth app on the diet and oral health for parents of preschool children, which was developed based on their needs, included information on how to prevent oral disease in their children and has been scored with good usability [31]. UCD requires an iterative design process to understand and internalize the users’ needs, goals, strengths, limitations, contexts, and intuitive processes [32]. Additionally, the iterative process includes observing the users’ interaction with the app during the development process [32]. After understanding the importance and iterative UCD element, the app is then developed and incorporated within the software development life cycle (SDLC), which has been reported as the most liked technique to develop a high-quality software product [33].

The mHealth development and evaluation framework was developed using the iteration SDLC version, and it includes six stages: (1) conceptualization, (2) formative research, (3) pretesting, (4) pilot testing, (5) randomized controlled trial, and (6) qualitative research for further refinement before moving to a more scaled-up intervention [34].

Conceptualization involved experts’ decisions regarding the theoretical basis, reviewing the evidence, and planning the development process via several brainstorming sessions. The research team includes a persuasion element to improve the user’s engagement with the app.

Persuasive systems may be defined as “computerized software or information systems designed to reinforce, change, or shape attitudes or behaviors or both without using coercion or deception” [35]. the persuasive system design (PSD) categorized the 28 strategies into four main categories: primary task support, dialogue support, credibility support, and social support [35]. Among the most PSD used was the primary task support using the self-monitoring tracking [36] in physical activity apps.

Formative research, which is the next stage, involved focus group discussions (FGDs) with pregnant women to determine how they used the app and which design they preferred to use on their mobile phones. The pretesting stage, which stressed the importance of the message context (PFMT adherence), was set and strengthened the collection of responses from pregnant women. The reason for the differences was that not all participants would be able to grasp every message, and key messages (PFMT) could be repeated in different contexts to reach more pregnant women.

A pilot study stage was conducted to obtain further feedback from pregnant women after they had used it over several weeks. The next stage is the randomized control trial stage aimed at obtaining rigorous evidence and, finally, a qualitative study to explore the use of the app in depth.

This study aims to document the validation process of a newly designed mHealth app called the Kegel Exercise Pregnancy Training (KEPT) app running on the Android platform. The KEPT app is intended to deliver training sessions, send reminders, and chart PFMT and UI symptoms. The expectation of the app is to deliver the correct method of PFMT efficiently and conveniently according to the pregnant women’s time and place, without the need to be in the clinic or consult with physiotherapists or doctors. The KEPT app is expected to fill the information gap between clinical visits and has undergone its usability evaluation by the experts [37].

We incorporated an intervention mapping (IM) approach with a UCD SDLC framework called the mHealth development and evaluation framework. This study focused on conceptualizing the app and using the UCD-11 checklist to evaluate the user-centeredness properties of the app.

IM is a framework that was designed to plan an effective intervention from the needs assessment up to its evaluation. It comprises stepwise decision-making in developing, implementing, and evaluating interventions using community-based research methods [38,39]. According to Fernandez et al, the participation from the community is to ensure that the intervention matches priority population needs and intervention contexts [39]. The characteristics of this approach involve three aspects that are applied during the intervention planning process: (1) participatory planning, (2) the comprehensive theory used, and (3) an ecological and systems approach for understanding health problems and intervening to address them [40].

Intervention development, according to IM, comprises six steps: (1) demonstrate a comprehensive understanding of the health problem; (2) outline the behavioral and environmental outcomes; (3) identify theory-based and evidence-based behavior change methods that affect the determinants and translate these into practical applications that fit the intervention context; (4) combine the intervention components into a coherent program that uses delivery channels that fit the context; (5) develop implementation strategies to facilitate adoption, implementation, and maintenance of the program; and (6) plan both process and outcome evaluations to assess program implementation, and efficacy or effectiveness [39].

Accordingly, the intervention mapping of this mHealth app involved the stepwise approach from the needs assessment followed with other steps as illustrated in Figure 1.

The behavior matrix of this intervention can be divided into knowledge (14 questions), attitude (8 questions), and practices (4 questions) with regards to pelvic floor muscle exercise (PFME). The outcomes of the intention are self-efficacy (17 questions) [41] and adherence (6 questions) [42]. A few examples are listed in Textbox 1. Additional information about the questionnaire has been presented in the results section and published elsewhere [43].

This study involved documenting the project identification stage followed by the user experience design (Figure 2).

In the project identification phase (conceptualization), we determined pregnant women’s (users) understanding of pelvic exercise and the severity of this problem (UI) affecting their QoL. From this foundation, we produced a low-fidelity design, and we conducted FGDs to find out the users’ experience using the design.

The team assessed the latest evidence regarding the users' UI and pelvic floor training, finalizing the relevant theoretical behavioral change to be used and planning the development process according to the IM.

A detailed understanding of at-risk pregnant women was established by conducting a cross-sectional study at a primary care clinic [43]. Within this study, we also determined the prevalence and severity of UI and its impact on participants’ QoL at a primary care clinic (user’s study setting) [43,44]. The findings from this study provided input for the content of their educational videos and short notes on PFMT [45,46], which were captured as frequently asked questions (FAQ).

The team brainstormed and decided to identify theory-based and evidence-based behavior change methods concerning PFMT and UI. The interventions were found to be effective when they were developed based on behavioral changed theories, for example, the social cognitive theory and the health belief model [47]. Another overarching framework of behavior used to identify appropriate targets for enhancing adherence in clinical practice is the capability, opportunity, and motivation-behavior (COM-B) model [48,49].

The COM-B model states that motivation is a crucial source of strength to perform a certain behavior, with the assistance from the capability (physical and psychological) and opportunity (physical and social) to engage in the behavior [50], and the strength of motivation to engage in the behavior must be greater than for any other competing behavior.

A focus group discussion (FGD) study was conducted to understand which design was preferred by the end users. A low-fidelity prototype was given to 5 groups of end users as they chose their preferred design, including their reasoning.

This study was conducted according to the guidelines of the Declaration of Helsinki. The study approvals were obtained in August 2019 from the Ethics Committee for Research Involving Human Subjects, Universiti Putra Malaysia (JKEUPM-2019-368) and the Medical Research and Ethics Committee (MREC), Ministry of Health Malaysia (NMRR-19-412-45606).

The theoretical framework of choice was the COM-B model with a combination of the PSD. The PSD is used with the COM-B is to reinforce the behavior voluntarily and to shape the attitude towards PFMT.

The COM-B model that was built into this app was expected to motivate pregnant women to perform PFMT regularly. Capability was divided into two subcategories: (1) physical capability, whereby pregnant women were fit to contract the affected muscles to perform PFMT, and (2) psychological capability signifying pregnant women understood the correct method of performing PFMT.

Opportunity also had two categories: (1) physical opportunity, such as the KEPT-app itself, and (2) social opportunity in which pregnant women were able to understand further and perform PFMT at their own time. Meanwhile, motivation has relationship with cognitive ability, which boosted women’s confidence to perform PFMT. There were two types of motivation: (1) reflective motivation in which pregnant women incorporated their thought processes to arrange PFMT to be done three times daily, and (2) automatic motivation in which the pregnant women adopted PFMT as part of their routine.

Additionally, the COM-B was integrated with the persuasiveness of the app (Table 5). The KEPT app should be tailored (primary task support) based on the intensity of the exercise, and users have the opportunity to self-monitor (primary task support). The app should be able to send reminders (dialogue support) to remind the user to perform the exercise at a certain time. The expertise and authority (system-credibility support) involved in developing the app are available in the video to convince the app user. Finally, an app should be designed and developed from credible and trustworthy sources to bolster users’ confidence and trust.

FGDs were conducted among 24 pregnant women at two primary care clinics to understand the desirability of the app design. The participants were invited via purposive sampling while waiting for their modified oral glucose tolerance test. A total of 5 sessions were conducted as listed in Table 6.

These interviews were conducted from September to November 2019. The discussion was conducted in a deductive manner to understand the most preferred user interfaces. The questions aimed to make the app as simple as possible since the participants will be using the app 3 times every day. The low-fidelity app designs were provided, and participants could select either Figure 3 (with 6 user interfaces) or Figure 4 (with 4 user interfaces).

Study participants were being asked about their experiences adhering to the exercise:

The statement suggested that pregnant women with or without urinary symptoms were motivated to perform and adhere to the exercise even after delivery. A correct understanding of PFMT importance assisted in compliance with daily exercise.

Following this, they were asked to share their opinion regarding adding PFMT notes into the KEPT app. The majority of study participants (19/24, 79.1%) preferred the apps without notes due to the time factor. However, one participant disagreed and mentioned:

The response suggested mixed opinions on whether or not to include additional notes on the exercise.

Finally, regarding the design selection of the app, all the study participants from the FGD preferred design 2 (a more straightforward and minimalist concept). A high-fidelity prototype design was developed based on all the above findings as illustrated in Figure 5.

The KEPT app prototype was proposed and discussed with the developer. The developer has designed the prototype with a blue color background [37] as illustrated in Figure 6. It consisted of both languages: Malay and English, whereby common English words were used. A minimum number of words were used to ensure the minimalist concept as target users were busy taking care of their families and/or working. There were 2 timer sounds: a high-pitch sound for exercise and a low-pitch sound during 6 seconds of rest. The reminder will be delivered if the PFMT is not completed at 6 pm on the same day.

The KEPT app interface did not require participants to log out due to the need to train three times daily. The UCD-11 checklist has been used as a guide for developing the KEPT app according to its user-centeredness properties as listed in Table 7.

The KEPT app was developed from a UCD-based behavioral change theory and accompanied by the PSD to improve users’ engagement. The integration of the UCD-11 checklist with COM-B and PSD has prevailed to benefit the target user effectively. Therefore, it is crucial that the targeted users evaluate the usability and user acceptance of the final prototype in our next study.