key: cord-1051996-wz57hgn0
authors: Dol, J.; Tomblin Murphy, G.; McMillan, D.; Aston, M.; Campbell-Yeo, M.
title: Process evaluation of the development and remote recruitment for Essential Coaching for Every Mother during COVID-19
date: 2021-03-12
journal: nan
DOI: 10.1101/2021.03.09.21253071
sha: 7b1f125ded8f11ecfe298f9716ffdf24053db67d
doc_id: 1051996
cord_uid: wz57hgn0

Background: With the sudden decrease in in-person support and increase in perinatal mental health concerns during the coronavirus pandemic, innovative strategies, such as mHealth, are more important than ever. This study has two objectives: (1) to describe the modification of Essential Coaching for Every Mother during the coronavirus pandemic, and (2) to describe the process evaluation of recruitment and retention of pregnant and postpartum women for a pre-post intervention study. Methods: For objective 1, modified messages were piloted with mothers and postpartum healthcare providers simultaneously. Semi-structured interviews were conducted with a subset of 10 participants from the original development study. For objective 2, three methods were used for recruitment: social media, posters in hospital, and media outreach. First time mothers were eligible for enrollment antenatally (37+ weeks) and postnatally (<3 weeks). Eligibility screening occurred remotely via text message with participants initiating contact. Data were collected via TextIt and REDCap. Outcomes were days to recruit 75 participants, eligibility vs. ineligibility rates, dropout and exclusion reasons, survey completion rates, perinatal timing of enrollment, and recruitment sources. Results: For objective 1, three mothers (M age=30.67 years) and seven healthcare providers (M age = 46.0 years) participated in the modification of the messages. Participants felt the messages were appropriate and relevant related to changes in postpartum care during the coronavirus pandemic. Nine messages were modified related to coronavirus and five messages were added to the program. For objective 2, recruitment ran July 15th-September 19th (67 days) with 200 screened and 88 enrolled, 70% antenatally. It took 50 days to enroll 75 participants. Mothers recruited antenatally (n=53) were more likely to receive all intervention message (68% vs. 19%). Mothers recruited postnatally (n=35) missed more messages on average (13.8 vs. 6.4). Participants heard about the study through family/friends (31%), news (20%), Facebook groups (16%), Facebook ads (14%), posters (12%), or other ways (7%). Conclusion: Antenatal recruitment resulted in participants enrolling earlier and receiving more of the study messages. Word of mouth and media outreach were successful, followed by advertisement on Facebook. Remote recruitment was a feasible way to recruit for Essential Coaching for Every Mother.

Introduction 56

Irrespective of a pandemic, mothers living in Nova Scotia and beyond face gaps in access to 57 information and often struggle to find adequate support during the postpartum period, defined as 58 the first six weeks after birth. [1] [2] [3] These gaps may be magnified during the coronavirus pandemic 59 and may significantly impact the transition for new mothers. 4 Compliance with physical 60 distancing recommendations contribute to mothers isolating at home, being physically isolated 61

from not only health providers, but also from their extended family and support systems. 4 In 62

Nova Scotia, all public health drop-ins were closed indefinitely, there was a reduction in in-63 person healthcare support, and midwifery-led home births and home visits were temporarily 64 deferred during the coronavirus peak from March to May 2020. 5 This significantly differed from 65

pre-coronavirus procedures, where mothers were recommended to have a postnatal contact 66

shortly after birth by a public health nurse 6 and mothers frequently engaged in visits with family, 67 friends, or new parent groups. 7,8 Emerging evidence shows that the pandemic has resulted in 37-68 54% of mothers experiencing perinatal depression and 57-72% experiencing symptoms of 69 perinatal anxiety. 9,10 With the sudden decrease in in-person support and the increase in perinatal 70 mental health concerns, innovative strategies, such as mHealth, are more important than ever as a 71

means to offer information and support during the postpartum period. 72

Prior to the coronavirus outbreak, the Essential Coaching for Every Mother program was 73 developed to send daily text messages to mothers during the immediate six-week postpartum 74

period. 11 As a result of changes in care and postpartum recommendations after the coronavirus 75 outbreak, changes were necessary in some of the messages. Additionally, given the requirement 76

of physical distancing and limitations on the number of visitors in hospital, exploration was 77 needed on the ability to recruitment remotely rather than the traditional, in-person approach for a 78 planned randomized control trial. Therefore, this study has two aims: (1) to describe the 79 modification of Essential Coaching for Every Mother to be applicable during the coronavirus 80 pandemic, and (2) to describe the process evaluation of remote recruitment of pregnant and 81 postpartum women for the Essential Coaching for Every Mother pre-post intervention study. 82

This study focuses on describing the methodological processes, opportunities, and challenges for 83 recruiting participation of women in a postpartum text message program during pandemic 84 restrictions. 85 2 Methods 86

Essential Coaching for Every Mother is a six-week postpartum text message program that was 88 previously developing in consultation with postpartum mothers and healthcare providers with the 89 goal of improving women's psychosocial outcomes. 11 Details of the original development has 90 been previously published. 11 With the outbreak of COVID-19 in early 2020 and the readiness of 91

the Essential Coaching for Every Mother program to fill the sudden gap in postpartum support, a 92 decision was made to modify the program to be offered immediately. Given that Essential 93

Coaching for Every Mother was developed prior to the coronavirus outbreak but not previously 94

implemented, some modifications were necessary of existing messages and for the inclusion of 95 coronavirus related content. To ensure that the revised content of Essential Coaching for Every 96

Mother was appropriate and acceptable, the modified messages were piloted with mothers and 97 postpartum healthcare providers simultaneously. Messages were updated using the Government 98 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) would deliver within the three-month recruitment period. Women were eligible to enrol 150 postnatally up to 21 days following the birth of their child. The postpartum limit was set to 151 ensure there was a least a three-week gap between baseline and 6-week follow-up surveys and to 152 ensure participants received enough of the message to provide evaluative feedback. 153 154

Three primary methods of recruitment were used: social media, posters in the hospital, and 156 media outreach. First, social media advertisements were used to recruit mothers via Facebook 157

and Instagram advertisements and tweets on Twitter. Social media outreach and paid 158 advertisements started on July 15 th , 2020 and ran until August 16 th , 2020. Second, postpartum 159 mothers were targeted for recruitment via posters at the IWK Perinatal Clinic at IWK Health and 160

in each room on the Family Newborn Unit. Posters were placed on August 5 th , 2020 and taken 161 down on September 15 th . Finally, media interviews also occurred with the first author, with one 162 televised news interview, two radio interviews, and several written media pieces occurring after 163 a media release was published by IWK Health on August 5 th , 2020. This targeted both prenatal 164 and postpartum women. 165

All eligibility screening occurred remotely via text message through the TextIt platform 16 166 with interested participants initiating contact. Pregnant women started the recruitment process by 167

texting 'pregnant' to the study number and proceeded through the antenatal eligibility screening 168 process. Eligible mothers were instructed to text 'delivered' within 48 hours of giving birth to be 169 enrolled in the study. During the antenatal screening process, participants were not truly enrolled 170

in the study but were monitored for enrollment post-birth. During the antenatal recruitment flow, 171

only the mother's phone number and due date was collected via TextIt once she was deemed 172 eligible. Women received reminder messages to text 'delivered' at 39 weeks, 40 weeks, 41 173 weeks and 42 weeks if they had not yet enrolled or withdrawn. Mothers who were deemed 174 ineligible as part of the antenatal screening due to being less than 37 weeks were sent a message 175

to remind them to text 'pregnant' if they were still interested. This occurred until August 27 th 176 when the number of interest and enrolled participants was beyond the desired 75 participants. 177

Postpartum women who initiated contact using 'birth' proceeded through the postpartum 178 eligibility flow. Once deemed eligible, postpartum participants and antenatal women who texted 179 'delivered' completed the same flow to be enrolled in the study and start receiving messages 180

based on their delivery date. During this phase, additional details about newborn's name, 181 preferred gender pronoun, date of birth, mother's name, and preference for breastfeeding or 182 formula messages were collected through TextIt. This was used to personalize the messages with 183 names and ensure messages were sent based on child's age and preference for breastfeeding or 184 formula messages. 185

Participants were asked to complete a consent form and survey at baseline (survey #1) 186

and once the messages ended at 6-weeks (survey #2). Participants were reminded about the 187 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) TextIt which was kept separate from survey data collected via REDCap. Survey data included 207 questionnaires about psychosocial outcomes and evaluated the impact of the program. This data 208 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 12, 2021. ; https://doi.org/10.1101/2021.03.09.21253071 doi: medRxiv preprint will be reported elsewhere and is not the focus of this manuscript. Ethics approval was obtained 209 by the IWK Health Centre (#1024984). 210

The first message of Essential Coaching for Every Mother is designed to start the evening 211 of the second day after giving birth. This could be as early as 17 hours after birth (if a mom 212 delivered at 11:59 the night before) or as late as 41 hours (if a mom delivered at midnight). If a 213 participant signed up beyond this time frame, they started the messages based on when they 214 delivered. 215

To determine feasibility, data on implementation extent was collected via output data 216

available through the TextIt platform as well as REDCap. For the outcomes identified above, the 217 following information was used: 218

• Days required to recruit participants = number of days from start of study to enrollment 219 of 75 participants, total number of participants enrolled, and time required for recruitment 220

• Eligibility vs. ineligibility rates = number of individuals who were eligible and enrolled 221

in the study vs. the number of individuals who contacted but were not eligible 222 (comparing both antenatal and postpartum ineligibility and reasons) 223

• The healthcare providers ranged in type of postpartum healthcare provider role, with one 244 each of: postpartum unit registered nurse, public health nurse, family doctor, reproductive care 245 program representative, midwife, family practice nurse, and a physiotherapist. Health care 246 providers mean age was 46.0 years old (SD =9.87) and had a mean of 17.71 years (SD = 11.37) 247

working with mothers. All postpartum healthcare providers were white women. 248 249 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 12, 2021. ; https://doi.org/10.1101/2021.03.09.21253071 doi: medRxiv preprint

Overall, the mothers and healthcare providers felt the messages were appropriate and relevant 251 related to changes in postpartum care during the coronavirus pandemic. General feedback was 252 related to ensuring that the messages shared information about what mothers could do (e.g., go 253

for walks when safely physically distancing, going to their family doctors for vaccinations and 254 follow-ups) and who to contact if they had questions (i.e., call 811 if you show COVID-19 255 symptoms). There were also recommendations to provide links to current guidelines in case 256

recommendations changed, which happened as Nova Scotia shifted from household isolation, to 257 family bubbles, to the Atlantic province bubble. 258

Nine messages were modified from the original program to include information related to 259 the coronavirus, of which four messages were collapsed into two, and five messages were added 260 to the program. Three of the five messages were provided during the first two weeks of the 261 program in which participants received on three days an additional message (three messages total 262 per day) versus the usual two messages. The other two messages were sent in place of one of the 263 previously collapsed message. In weeks three through six, usual messaging was reduced to one 264 message a day. Figure 2 provides an example of two messages included -the first is a message 265 that was revised to reflect the current standard of care in postnatal follow-up by a public health 266 nurse which increased from within 48 hours to 3-5 days and may occur over the phone or online 267 rather than in person. The second message is an example of a message that was added related 268 specifically to postnatal care during COVID-19. 

Timing required to recruit participants 276

The study started recruitment on July 15, 2020 and stopped all enrollment on September 19, 277 2020. This is a period of 67 days during which, 96 participants were enrolled in the program and 278

were assigned a study identification number. Timing to enroll 75 participants (our initial target) 279 took 50 days (July 15, 2020 to September 2, 2020). 280 281 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 12, 2021. ; https://doi.org/10.1101/2021.03.09.21253071 doi: medRxiv preprint

A total of 200 initial contact messages were sent to the study contact number by potential 283 participants during the recruitment period. Figure 3 outlines the perinatal timing and reasons for 284 ineligibility. A total of 140 participants initiated contact antenatally and 60 participants had 285 initiated contact in the postpartum period. For the antenatal participants, 30 were not eligible, 286 seven were not interested, and 45 were excluded not being based in Nova Scotia. For the 287 postpartum period, 20 were not eligible and two were not interested. 288 We had to exclude 45 participants who were not based in Nova Scotia and were using a 289

United States (US) number. We initially thought this could be individuals who are temporarily 290 residing in Canada since they provided valid Nova Scotia postal codes in the demographic 291 questionnaire. However, upon further analysis, we believe these were not actual mothers from 292 the US wanting to participate due to a discrepancy in the standardized questionnaire responses 293 with extremely high scores on these measures (well beyond the standard mean) and quick 294 completion time of their REDCap surveys (immediately after they enrolled and on their delivery 295 date), suggesting these were not actual participants. When analyzing location through Twilio, the 296 US had a predominant send/receive rate, providing further evidence that these respondents were 297 not in Canada. Therefore, with the triangulation of these findings, it was deemed that these 298 responses are not actual potential participants and if they were, they were not residing in Nova 299

Scotia as required by the study protocol. 300 301

Dropout & survey completion rates 302

Of the 96 enrolled participants, four withdrew from the program after receiving 0, 5, 6 and 303 9 messages respectively (Mean [M] = 4, Standard Deviation [SD] = 3.7). Three of the 304 participants who withdrew enrolled in the postpartum period and one enrolled antenatally. As 305 none of the participants who withdrew completed the baseline survey, we were unable to 306 determine if these participants were different from those who completed the program. Four 307 participants did not complete any aspect of the baseline survey, thus were excluded from the 308 analysis. 309 Therefore, the study had a total of 88 participants who did not opt out and who completed 310 at least some of the baseline survey. All participants received a reminder text messages to 311 complete the survey every other day from enrollment until two weeks postpartum or survey 312 completion, whichever came first, as well as a final email or text reminder to complete the 313 survey. Ninety percent (90.1%) of participants completed the full survey, taking on average 5.0 314 days to complete the baseline survey from enrollment (Median = 3 days, SD= 5.3 days, range 0-315 19 days). Nine percent (n=8) did not complete baseline survey in full -on average, participants 316 completed 56.25% of the survey (range: 25%-75%). 317 318

Timing of recruitment 319

Of the 88 participants who were enrolled, 42 (47.7%) received full messages. Of these 42, 320 36 were antenatally recruited and six were recruited postnatally. There were differences in the 321 number of messages missed if participants were recruited antenatally or postnatally. Late 322 enrollment during the antenatal period resulted in missing on average 6.4 messages (SD=6.2) 323

whereas late enrollment during the postpartum period resulted in missing on average 13.8 324 messages (SD=10.6). Similarly, when women enrolled antenatally, their infants were a mean of 325 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 12, 2021. ; https://doi.org/10.1101/2021.03.09.21253071 doi: medRxiv preprint 1.4 days old (SD=2.0) whereas when they enrolled postnatally, their infants were a mean of 6.0 326 days old (SD=5.3). 327

Most (40.4%) enrollment messages (text of 'birth' or 'delivery') were received in the 328 morning hours (between 0600-1200), followed by 24.7% in the afternoon (1200-1800), 20.7% in 329 the evening (1800-2400) and 14.6% overnight (0000-0600 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) Mother is not a static intervention, such as print media, which is harder to update and ensure 367 everyone has the updated version, suggesting that mHealth could be a viable option to keep 368 content up to date, especially during emerging health concerns. 369

Additionally, the online and remote recruitment of pregnant and postpartum women for a 370 pre-post intervention study for Essential Coaching for Every Mother was a success as we were 371 able to recruit over our target of 75 participants within 50 days, with recruitment suspended 372 within 67 days due to significant interest. This suggests that mothers were interested in receiving 373 information during the postpartum period, which may have been enhanced due to the pandemic. 374

Emerging evidence shows that the pandemic has resulted in 37-54% of mothers experiencing 375 perinatal depression and 57-72% experiencing symptoms of perinatal anxiety, 9,10 suggesting that 376 a preventative mHealth program for mothers could have a positive effect on mothers postpartum 377 adjustment and experience. Given the growing evidence of the mental health consequences of 378 physical distancing recommendations, 20 particularly during an intensely vulnerable period as is 379 the postpartum period, having evidence-based information provided via text message may help 380 cover this gap. Digital health during COVID-19 has the potential to bridge the healthcare service 381 gap while maintaining physical distancing recommenations. 21,22 382

We found that mothers who were recruited antenatally received more of the study 383 messages than participants who were recruited postnatally, with the latter missing on average 7.4 384 messages more. As participants who initiated contact during the antenatal period were sent 385 reminder messages starting at 39 weeks, they were more likely to enroll earlier than mothers who 386 had already delivered. No mother who expressed interest during the antenatal period and was 387 deemed eligible failed to enroll. Thus, antenatal recruitment may be a more efficient way to 388 target recruitment for the larger clinical trial to ensure mothers receive as much of the program as 389 possible. Additionally, given the delay in baseline survey completion after delivery, shifting to 390 have participants complete the baseline survey upon enrollment and prior to delivery may result 391 in more timely completion than during the postpartum period. 392

Looking at the direct recruitment methods, the most successful approach was promotion 393 through Facebook. Both advertisement through mother-focused Facebook groups and paid 394 advertisements were similarly effective. This finding is supported by previous systematic 395 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 12, 2021. ; https://doi.org/10.1101/2021.03.09.21253071 doi: medRxiv preprint reviews which found that Facebook recruitment was an effective way to reach participants for 396 health research. 23,24 Within our study, we also found that sharing the study in the media and news 397 reached 17% of participants and posters in the hospital reached 13.6% of participants, which 398

suggests that using a multi-pronged approach to recruitment is more efficient than solely using 399 social media. 400

Despite the successes, there were some challenges in recruitment. First, most participants 402 heard about the study through family and friends, but it is unclear how these family and friends 403 heard about it. Additionally, we were unable to gather how mothers who contacted us but did not 404 enroll in the study heard about it. Both of these factors limit the interpretation of recruitment 405 source analysis. A second challenge was the high potential for people to misuse the self-406 identification of eligibility screening which occurred exclusively via text message. This occurred 407

in relation to the large number of non-Nova Scotia based on phones. We hypothesize that 408 someone(s) had been completing the eligibility screening and baseline questionnaire to gain 409 access to the honorarium. While TextIt cannot limit to provincial locations, we continued to 410 monitor recruitment closely to ensure we identified any issues related to this through regularly 411 monitoring of area codes. This may have potentially excluded individuals who were residing in 412

Nova Scotia but had US numbers, this was required to ensure safety and adherence to study 413 protocol inclusion criteria. 414

Another limitation was that limited sample size and diversity representation in the 415 modification testing of Essential Coaching for Every Mother. However, as participants were 416 recruited from the original development to ensure they were familiar with the program, having 417 10 participants represents over a third of the original sample (10/28 = 35.7%). In terms of the 418 sample being primarily white and women, this was recognized as a limitation of the original 419 development and thus was not able to be controlled for this purpose. Further work should have a 420 more direct focus on collecting a diverse sample. 421

Despite these challenges, this study found that Essential Coaching for Every Mother was 423 able to be successfully modified for implementation during the COVID-19 pandemic and remote 424 recruitment of pregnant and postpartum women for a pre-post intervention study was possible 425 using a variety of recruitment sources. Findings from this study will be applied in the 426 development of a clinical trial to determine effectiveness in a real-world setting. 427 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted March 12, 2021. ;

Transition to parenthood: The needs of parents in pregnancy 441 and early parenthood

Views of women and clinicians on postpartum 443 preparation and recovery

The Canadian Maternity Experiences 445 Survey: An Overview of Findings

Impact of 447 COVID-19 restrictions on the postpartum experience of women living in Eastern Canada: A 448 mixed method cohort study. medRxiv

Nova Scotia Health Authority. Temporary service and facility closure notices from Nova 451

Reproductive Care Program of Nova Scotia. Healthy Babies, Healthy Families: Postpartum & 455 Postnatal Guidelines

Navigating and negotiating 458 information and support: Experiences of first-time mothers

Intimate mothering publics': comparing face-to-face support groups and Internet 460 use for women seeking information and advice in the transition to first-time motherhood. Cult 461 Health Sex

Moms are not OK: COVID-19 463 and maternal mental health. Front Glob Women's Heal

Elevated depression 465 and anxiety among pregnant individuals during the COVID-19 pandemic

Design, development and 468 usability testing of Essential Coaching for Every Mother: A postnatal text message educational 469 intervention

World Health Organization. Pregnancy, Childbirth, breastfeeding and COVID-19

Government of Canada. COVID-19: Pregnancy, childbirth and caring for newborns: Advice 475 for mothers

Government of Nova Scotia

Reproductive Care Program of Nova Scotia. The Nova Scotia Atlee Perinatal Database. 481 Halifax

Research electronic data 485 capture (REDCap) -A metadata-driven methodology and workflow process for providing 486 translational research informatics support

The Pace of Technologic 488 Change: Implications for Digital Health Behavior Intervention Research

The Mental Health Consequences of COVID-19 and Physical 492 Distancing: The Need for Prevention and Early Intervention

Digital mental health and COVID-19: 495 Using technology today to accelerate the curve on access and quality tomorrow

Digital health strategies to 498 fight COVID-19 worldwide: Challenges, recommendations, and a call for papers

The use of Facebook in recruiting participants for health 501 research purposes: A systematic review

Recruiting for 503 health, medical or psychosocial research using Facebook: Systematic review