Anxiety disorders such as generalized anxiety disorder, obsessive-compulsive disorder, panic disorder, and phobias in the first year after birth (postpartum) are common with prevalence rates ranging between 9.9% and 20% [1-3]. Anxieties in the postpartum period are often life-stage specific, for example, worries about baby’s care and health and fear of criticism and inadequacy as a mother [4]. Postpartum anxiety disorders can either be a recurrence of a previous disorder or develop as a first episode. Symptom intensity and associated degree of impairment of these anxiety disorders can vary over the course of the postpartum period [5]. Despite available effective treatments [6-10], postpartum mental health problems often go undetected or untreated [11,12]. Low screening and diagnosis rates play a role, but some women with emotional difficulties postpartum are often more reluctant to disclose and seek help [13-15]. Possible reasons for this include being too busy to get around to seeking help and feeling too embarrassed or having no-one they felt comfortable talking to [14], as well as child care concerns [16].

The importance of having efficient and timely treatments is highlighted by the adverse effects of untreated mental health problems on the physical and psychological health of the mother, child, and family [17,18], as well as potential costs to society [11]. For example, it has been shown that maternal anxiety can affect infant bonding and feeding [19,20] and may negatively affect the child’s cognitive and social development [17]. Considering the importance of treatment and unique postpartum barriers to accessing treatment, providing convenient and anonymous treatment seems essential.

One approach of offering anonymous and convenient treatment is Web-based self-help treatments, which run on computers, tablets, or smartphones and allow individuals to work through written therapy material without or with minimal therapist or mental health professionals’ assistance. Many of today’s parents search for information and support on the Web [21]. In addition, postpartum women who feel isolated or restricted by their baby’s schedule experience Web-based resources as useful [22]. In a thematic analysis of motivators and barriers to a Web-based postpartum treatment, it was found that the offered flexibility and anonymity fitted women’s postpartum circumstances [23]. This suggests that Web-based treatments may be an appropriate alternative or supplement to conventional face-to-face therapy for postpartum women.

Systematic reviews and one meta-analysis focusing on the perinatal period suggest that Web-based treatments can help improve postpartum depressive symptoms [24-26], but so far none are specifically developed for postpartum anxiety [24]. A Web-based survey demonstrated that women with postpartum anxiety are interested in Web-based treatments [27], and a qualitative study of postpartum health care professionals (health visitors) in the United Kingdom reported a need for more treatment options for postpartum anxiety and that Web-based treatments could be useful to address this issue [28].

On the basis of the identified interest and need, a self-help treatment for women experiencing moderate or severe symptoms of postpartum generalized anxiety disorder called What Am I Worried About (WaWa) developed in Australia [4] was transformed into a Web-based version called Internet-based WaWa (iWaWa). This study aimed to evaluate the feasibility and acceptability of iWaWa for women with postpartum anxiety problems in England. On the basis of the stage model of behavioral therapy research [29,30], the primary study objectives were to determine study feasibility by examining recruitment and attrition, examine iWaWa’s feasibility in terms of engagement and usability, and examine user’s acceptability of iWaWa in terms of usefulness, helpfulness, and satisfaction. The secondary objective was to examine potential changes in anxiety over the course of the treatment and compare with a wait-list control group.

Table 1 presents detailed information about participant characteristics of all randomized participants, those lost to follow-up, and those completing the 8-week follow-up assessment.

Participants were predominantly “Caucasian” (84/89, 94%), married (62/89, 69%), living with their partner/husband (79/89, 89%), and aged between 22 and 43 years (mean 32.02 years [SD 4.15 years]). Over half of the women had a bachelor’s degrees or higher and were on maternity leave. Slightly less than half of the women (38/89, 43%) had an income below £50,000 and about half (44/89, 49%) an income equal to or above £50,000. Women were between 1 and 52 weeks postpartum (mean 28.58 [SD 13.76]) and 72% (64/89) were first-time mothers (range 1-5 children [mean 1.36 (SD 0.68)]). The majority of participants reported having access to two or more technological devices (84/89, 94%). The two most commonly accessible devices were smartphone (88/89, 99%) and laptop (74/89, 83%).

There were no significant differences between the treatment and wait-list control groups in demographic characteristics, except for relationship status (P=.03). No differences were found between participants of the treatment and wait-list group who completed the 8-week follow-up assessment, and no differences were found between participants of the treatment and wait-list group who did not complete the 8-week follow-up assessment. There were also no differences between treatment group participants who started the iWaWa treatment and those who did not.

Qualitative data comprised 31 comments from iWaWa modules and follow-up questionnaires and five interviews (13-18 min). Data saturation was assumed when no new themes emerged from the interviews and comments. Of the interviewees, one completed the first module and the remaining four between 4 to 9 modules.

Table 2 contains the mental health outcomes for the treatment and wait-list control groups at baseline, the 8-week follow-up, and 12-week follow-up. A total of 28% (25/89) women scored in the mild anxiety range, 42% (37/89) in the moderate anxiety range, and 33% (27/89) in the severe anxiety range on the GAD-7 at the baseline assessment. There were no significant differences between the treatment and wait-list control groups on all mental health measures at baseline. At the 8-week assessment, no significant group differences were found for anxiety (GAD-7 and DASS-21 anxiety). For both groups, anxiety scores significantly reduced from baseline to the 8-week follow-up. There was no significant difference between the 8-week and 12-week follow-up in anxiety scores for the treatment group. Multimedia Appendix 3 illustrates the GAD-7 and DASS-21 anxiety scores for the 3 participants who started the iWaWa program and completed both follow-up assessments. There was no significant correlation between any of the anxiety scores and time since birth and number of children.

This study aimed to assess the trial’s feasibility, iWaWa’s treatment feasibility and acceptability, and explore changes in postpartum anxiety in association with iWaWa. Regarding the study’s feasibility, the minimum sample size required was exceeded within the relatively short recruitment period (8 weeks). Facebook proved most successful for recruitment. However, there was a high dropout attrition from the study and also a high treatment nonusage attrition. Among women who accessed iWaWa, the treatment was experienced as generally useful and helpful but not user-friendly enough in terms of treatment format and content. Participants felt that iWaWa could be improved by having it as a smartphone app and by making the content more concise and inclusive of different parenting styles. Anxiety levels decreased significantly for both groups from baseline to the 8-week follow-up assessment, but there were no differences between the treatment and wait-list control group. However, due to high drop-out and nonusage attrition, especially in the treatment group, the results for mental health cannot be interpreted reliably and will therefore not be further discussed. The results regarding recruitment and attrition, usability, and acceptability will be discussed in more detail below, including a comparison with previous literature and potential program improvements.

Within a relatively short recruitment period, the iWaWa trial successfully recruited more than the minimum calculated sample size through Facebook. The response rate is comparable to similar studies. A study recruiting women with postpartum anxiety for a Web-based online survey through Facebook had a similar response rate (220 respondents over 4 months) [44]. Two studies recruiting women in the United Kingdom with postpartum depression through website advertisement banners for a Web-based treatment had similar (249 respondents over 5 months [33]) and higher response rates (1403 respondents over two waves of 2-week recruitment periods [34]). The successful recruitment might indicate that there is an interest and potential need for a treatment such as iWaWa among postpartum Facebook users. Due to the inability to recruit participants in health care settings, it remains to be investigated whether recommendation or endorsement from health care professionals might have increased iWaWa’s recruitment rates.

Dropout attrition in the iWaWa trial was high (49% to 91%). In comparison, in the open pilot study evaluating WaWa, dropout attrition ranged from 39% to 61%. However, it has been pointed out that many trials testing Web-based interventions often suffer from high attrition rates [45]. Due to the lack of other trials of Web-based postpartum anxiety treatments, attrition can only be compared with trials of postpartum depression. Regarding postpartum Web-based treatments, one study evaluating the feasibility for depression with minimal support (Netmums program) also reported a high attrition rate (76%) [34]. The study found that, like this trial, there was an initial high access with a small user subgroup continuing to access and use the program. The authors discussed the role of curiosity, and it has also been suggested that one session may satisfy the user’s need [46]. The Netmums program’s attrition reduced to approximately 20% at the posttreatment assessment when telephone support was added [33]. Similarly, the Web-based MomMoodBooster program for postpartum depression with telephone guidance also reported comparably very low attrition rates [31,46]. When comparing iWaWa’s attrition to postpartum depression trials, it has to be noted that a systematic review found higher Web-based intervention adherence rates for depression than anxiety [47].

There are a variety of factors that could have caused iWaWa’s high attrition. Commonly identified reasons for dropout include the burden of the program, program information structure, content relevance, level of support, technical access issues, time constraints, lack of motivation, and improvement of the group [34,47,48]. The potential role of these factors regarding iWaWa’s high attrition will be briefly discussed.

This study suggests that some women indeed experienced the access, length, and exercises of iWaWa as a burden. With the infant’s demanding schedule, it is possible that iWaWa participants felt a lack of motivation or experienced time constraints. For Web-based postpartum depression treatments, adherence was high when scheduled support was offered [31,33,46]. Scheduled support was part of the original WaWa treatment and valued as strength of the treatment [4]. On the basis of a survey which found that telephone support was not the preferred mean of support among women with postpartum anxiety [44], it was decided for iWaWa’s telephone support to be optional. Scheduled support might improve adherence of women wanting this type of support, but it might also put off women who do not feel the need or find it unsuitable. Offering to opt in for scheduled support at the treatment start might be an alternative for future iWaWa versions. Regarding content relevance, a qualitative study suggested that being able to identify with a an online intervention program helps with adherence in Web-based psychological interventions [48]. As some of iWaWa’s topics were not experienced as relevant, it could therefore be that some participants logged into iWaWa but dropped out because they experienced the topics as irrelevant. Concerning symptom improvement, a recent systematic review and meta-analysis found that the prevalence of postpartum anxiety decreases from 1-4 weeks to 5-12 weeks [1], and it could therefore be that the participant’s symptoms improved and they no longer felt the need for iWaWa. In addition, almost half of the participants scored in the severe to extremely severe range of stress. It might be that participants were in need for stress management techniques, which could be incorporated in future iWaWa versions. The iWaWa program used an organic information structure design that allowed users to freely explore the modules. The use of a tunnel design, in which users navigate through content in a sequential order, has been identified as less likely to overwhelm users with options [49] and may increase use [50].

Participants enjoyed iWaWa’s accessibility, anonymity, and weekly reminders, as well as the introduction to the principles of CBT and mindfulness. This is in line with previous qualitative research regarding Web-based treatments for postpartum depression [23,32,51], as well as postpartum UK health care professionals (health visitors) [52].

iWaWa was rated and experienced as generally useful and helpful but not user-friendly enough. The iWaWa content was experienced as too long. A preference for brief modules was also identified by Web-basedsurveys among a sample of adults [53] and among women with postpartum anxiety [44]. Participants also felt that iWaWa’s content was not inclusive of different parenting styles. WaWa’s content was found to be acceptable in a small sample (n=7, in the posttreatment evaluation interview) pilot study in Australia [4]. It might therefore be important to investigate in more depth the needs of a larger and more diverse sample of women with postpartum anxiety so that iWaWa’s content and format can be adapted to better meet their needs.

Women in this study also felt that not all topics were relevant. The need for Web-based treatments to be relevant to their own needs and circumstances was also discovered among women with postpartum depression [23,32]. Future iWaWa versions might benefit from presenting users with content most relevant to them (eg, anxiety topic needs and relevance assessment before the treatment starts).

Interactive components with space for responses were previously found to be valued among women with postpartum depression [23]. Even though most iWaWa users engaged with the exercises, many experienced them as difficult. The ease of usage of the interactive components seems important among iWaWa users, and therefore exercises of future iWaWa versions should be tested by potential users before implementation (eg, think-out aloud technique).

Regarding the format, iWaWa participants experienced the log-in process as difficult. The log-in process was also identified as difficult by women testing a Web-based postpartum depression treatment [32]. Participants also expressed that iWaWa was not smartphone-friendly, which was also reported by women testing a Web-based treatment for postpartum depression [32]. The preference for a smartphone-compatible treatment was also found in a survey among women with postpartum anxiety [44]. iWaWa was smartphone compatible, but iWaWa as a smartphone app might improve the ease of access and presentation on smartphones.

The generalizability of the results of this study is compromised due the use of convenience sampling and a homogenous sample. Findings may therefore not be representative of women from different cultural or ethnic backgrounds, lower socioeconomic status, and more severe anxiety. In addition, most women were recruited through Facebook, so findings are limited to this self-selected pool of women. However, the homogeneous nature of this sample can also be a strength, as results can be generalized for this specific group. Furthermore, no incentives were offered during the enrollment, allowing for a sample with a genuine and intrinsic need. The generalizability may also be affected by the fact that the anxiety status was established by a screening instrument and not by a diagnostic tool. However, the GAD-7 is a frequently used anxiety screening tool and has been suggested as suitable for postpartum anxiety [38]. In addition, diagnostic interviews are less anonymous and iWaWa’s anonymity was highlighted as an important strength by the participants.

High dropout and nonusage attrition may have also biased the results. Therefore, findings regarding iWaWa’s usability, acceptability, and mental health changes are limited to the experiences and views of a small sample. None of the women who did not start iWaWa took part in the follow-up interviews; so, no knowledge could be gained regarding what caused the early dropout.

The interviewer of this study was MA, who is a doctoral student investigating Web-based treatments for postpartum anxiety with personal experience related to this subject. These experiences may affect the collection, analysis, and interpretation of the data. Therefore, several actions were taken to minimize the likelihood of this risk. Participants were aware that the interviewer was the main study lead, but reassured of the importance of any feedback (positive and negative) for improving iWaWa. The qualitative analysis also included comments from the iWaWa modules and the follow-up assessment in which women might have felt more comfortable providing negative feedback. MA kept a reflective journal to consider how the treatment and interviewee responses affected her own views on iWaWa throughout data collection, analysis, and reporting. An effort was made to consider this when analyzing the data. Interviewees were offered monetary compensation for their time, which might have confounded their responses.

This study demonstrated that there is an interest in a postpartum anxiety Web-based treatment like iWaWa. However, the iWaWa study and program in the current format is not yet feasible and acceptable, and due to high dropout rates, the results on the impact of iWaWa on anxiety cannot be interpreted reliably. Nonetheless, the study revealed strengths and weaknesses of the iWaWa content and format, as well as highlighted potential areas of improvements. As a first study investigating the usability and acceptability of Web-based treatment specifically targeted at postpartum anxiety, these results provide useful information about Web-based treatment preferences that can help improve iWaWa and inform research and development to optimize usability, acceptability, and Web-based treatment adherence in this population. This study also contributes to filling the gap in evidence-based self-help for mild-to-moderate postpartum anxiety symptoms.