Enhancing Periconceptional Health by Targeting Postpartum Mothers at Rural WIC Clinics

Abstract

The overall goal of this pilot quality improvement (QI) intervention was to (1) assess the feasibility of making a WIC (Women, Infants, and Children) systems-level change that added measurement of maternal weight and discussion of maternal health habits into each postpartum maternal and offspring visit in rural clinics in Colorado and (2) assess the impacts of the intervention on maternal diet, physical activity, and weight status. A mixed-method evaluation approach was used involving the collection of quantitative data (HeartSmartMoms usage reports, manual WIC chart reviews [to calculate screening rates], pre-/postsurveys, and weight status [body mass index]) and qualitative data (focus groups and project team meeting minutes). It was determined it is feasible to make a short-term systems-level change; however, many barriers were encountered in doing so, and the results were not sustained. The QI intervention did decrease participants’ daily consumption of sugar-sweetened beverages and maternal weight status (controlling for maternal age and language), but did not improve any other eating/physical activity behaviors. Lessons learned and recommendations to improve the implementation of health promotion interventions aimed at improving postpartum maternal health, which can increase health during the periconceptional phase, and in turn, improve the health outcomes for a child, are discussed.

INTRODUCTION

The prevalence of obesity continues to rise among Americans and continues to be one of the top public health crises of our time. More than one third of the adults in the United States are obese (Ogden, Carroll, Kit, & Flegal, 2014). Obesity disproportionately affects people who live in poverty and ethnic minority people (Wang & Beydoun, 2007).

Although there are no gender differences in the annual rate of increase in the prevalence of obesity (Wang & Beydoun, 2007), among women, the age-group with the fastest increase in obesity prevalence is 20 to 34 years (for men the rate of increase was similar across age-groups; Wang & Beydoun, 2007). This age range also coincides with childbearing years for women. Pregnancy is a risk factor for excessive weight gain (Siega-Riz, Evenson, & Dole, 2004), and this is of concern because maternal obesity can result not only in pregnancy and delivery complications, such as pregnancy induced-hypertension, gestational diabetes, and higher rates of cesarean delivery, but also in fetal and newborn complications, such as congenital abnormalities, large-for-gestational-age infants, and still births (Aviram, Hod, & Yogev, 2011). Additionally, excessive weight gain during pregnancy often leads to weight retention postpartum (Siega-Riz et al., 2004), and this, in turn, has the potential to compromise the well-being of future pregnancies and deliveries, and the health of subsequent infants born to obese mothers. This evidence highlights the need for an increased focus on the nutritional health of women before pregnancy and on weight loss interventions in overweight and obese women in the periconceptional period (just prior to pregnancy) to improve optimal pregnancy and infant outcomes (Zhang et al., 2011).

BACKGROUND

Increasing evidence supports the importance of the periconceptional phase, the time period just weeks before conception to early pregnancy, in predicting later health outcomes of the child. Studies indicate that the intrauterine environment (after conception) affects the incidence of obesity in the child and that as a result of maternal obesity, the child faces a lifelong risk of obesity (Aviram et al., 2011; Lawlor, Lichtenstein, Fraser, & Langstrom, 2011; Salsberry & Reagan, 2005; Zhang et al., 2011) and health-related problems.

Animal studies also suggests that changes in maternal diet even prior to conception result in major changes in the fetal epigenome and corresponding changes in the phenotype (Aagaard-Tillery et al., 2008; Howie, Sloboda, Reynolds, & Vickers, 2013; Nivoit et al., 2009), which can result in metabolic damage (Heerwagen, Miller, Barbour, & Friedman, 2010; Howie et al., 2013). Given that the foundation for lifelong risk and susceptibility to numerous diseases, including obesity, begins in the periconceptional phase, prevention should encompass a life course perspective, which commences prior to conception (Nader, Huang, & Gahagan, 2012).

A challenge to intervening during the periconceptional phase is that many pregnancies are unplanned and, thus, the timing of this phase is often unknown. The Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) program is a federally supported program through the U.S. Department of Agriculture, which provides supplemental foods and nutrition education for low-income pregnant, breastfeeding, and non–breastfeeding postpartum women, and their infants and children up to age 5 (U.S. Department of Agriculture, Food and Nutrition Services, 2016). WIC offers an opportunity to intervene in this periconception period because it has regular contact with women who already have one infant or child and who are likely to become pregnant again. Additionally, standard WIC visits focus on nutrition and well-being, thereby making it an ideal location for the integration of evidence-based interventions that address the behaviors leading to the intergenerational transmission of obesity.

The overall goal of this pilot quality improvement (QI) intervention between an academic and community partner was to augment the postpartum maternal care currently provided by the WIC program with two innovative, cost-effective strategies that were incorporated into postpartum maternal and infant/toddler (0–2 years) visits. Although primiparous women do not enter the WIC program prior to conception under present guidelines, the WIC program does provide an opportunity to maintain contact with women (through child visits up to age 5) who are likely to conceive again. WIC is positioned to promote counseling on the role of diet and physical activity in improving the woman’s health and the health of any future children. A large proportion of women retain excess weight after the delivery of a child (Melzer & Schutz, 2010), which can be detrimental to their health and the health of any future children. As such, the long-term goal of this project was to reestablish prepregnancy weight status. More immediate QI intervention objectives were the following:

1. Assess the feasibility, including the barriers and facilitators, to making a WIC systems-level change that added measurement of maternal weight and discussion of maternal health habits into each postpartum maternal and offspring visit in rural clinics in Colorado. The discussion would incorporate modifications of the HeartSmartKids™ (HSK) program (Gance-Cleveland, Gilbert, Kopanos, & Gilbert, 2010) to be appropriate for women (HeartSmartMoms™ [HSM]) and motivational interviewing (MI) techniques.

2. Assess the impacts of the intervention on maternal diet, physical activity, and weight status (body mass index [BMI]).

This QI intervention was developed with an emphasis on the individual level of the socioecological model (McLeroy, Bibeau, Steckler, & Glanz, 1988). A combination of the health belief model (Janz & Becker, 1984) and the transtheoretical (stages of change) model (Prochaska & DiClemente, 1983) were used to develop this intervention. The health belief model focuses on people’s beliefs about whether or not a negative health outcome will affect them, their perceptions of the benefits of trying to avoid the health outcome, the barriers to taking action, and how all these factors influence their readiness to act. The transtheoretical model focuses on a person’s attempt to change behavior as he or she progresses through five stages of change: precontemplation, contemplation, preparation, action, and maintenance. The QI intervention, described in more depth below, used questions on an interactive electronic system and MI counseling techniques to assess how important, how confident, and how ready participants were to making change. Barriers to making change and goals to achieve the desired change were discussed and every effort was made throughout the intervention process to “meet participants where they were at” in terms of making healthy eating and physical activity changes.

METHOD

Setting

This study was conducted in the Valley-Wide WIC program located in the San Luis Valley region of south-central Colorado. The San Luis Valley is composed of six counties (three frontier with less than six people per square mile and three rural) with a total population of 46,449; 47% of residents are of Hispanic origin (Department of Local Affairs, State of Colorado, n.d.). Approximately the size of New Jersey, the estimated median household income is $28,138 compared to the Colorado median of $47,203 (San Luis Valley Development Resources Group, 2012). Approximately 20% of residents have household incomes below the poverty level compared to 12.5% in Colorado (Department of Local Affairs, State of Colorado, n.d.), 10.5% of residents were uninsured in 2015 (Colorado Health Institute, 2015), and many rely on Medicaid for access to health care.

There are three WIC offices in the San Luis Valley; the Alamosa WIC office is the largest of the three. The Monte Vista WIC office is located approximately 35 miles from Alamosa, and the smallest WIC office is located in Center, nearly 60 miles from Alamosa. The three WIC sites serve an average monthly caseload of nearly 1,450 participants as follows: Alamosa = 900, Monte Vista = 350, and Center = 200.

Intervention

The QI intervention included two interconnected strategies: (1) implementation of the interactive HSM electronic program consisting of a bilingual kiosk and decision support system that provided feedback to the mother and provider regarding weight status/BMI; weight and BMI trends, along with relevant health risks and tailored recommendations, are highlighted in the HSM summary and (2) training of WIC educators and other staff in MI to support maternal health behavior change. MI is an empathic counseling style that manifests through specific techniques and strategies, such as reflective listening, shared decision making, and agenda setting (Resnicow, Davis & Rollnick, 2006). One of the goals of MI is to assist individuals in working through their ambivalence about behavior change. A style of MI termed Brief Negotiation allows clinicians to incorporate these techniques in a targeted clinical encounter such as a WIC visit (Rollnick, Heather, & Bell, 1992).

Prior to the project’s inception, institutional review board approval was sought and received. Starting in January 2013, the pilot QI intervention was implemented. Just before the implementation of the intervention, WIC educators underwent a full-day (8-hour) training lead by an academic researcher and clinical specialist in weight management. Training included orientation to the HSM interactive kiosk and summary, as well as interactive training on MI principles and skills. MI training methods included didactic instruction, demonstration of techniques, five case-based practice sessions using mock HSM summary with standardized scenarios of WIC participant mothers portrayed by study team members, audit, and feedback from peers and the expert trainer on the use of MI counseling skills during the timed practice scenarios. Throughout the day of training, steps of the Brief Negotiation MI process were added iteratively to the patient scenarios, building on the skills practiced during previous practice interviews. An additional 1-day training session in MI was conducted 15 months after initial implementation to offer a booster to staff previously trained and to train newly hired staff.

At the end of the day-long training, another academic researcher facilitated the discussion with all WIC staff about how best to integrate the new intervention into standard WIC visits, so that it did not feel like something “on top of” their current work load. Feedback from this discussion was used to inform the script to invite postpartum women to do the intervention, as well as the office flow as it pertained to the intervention, including where to place the HSM kiosks and the scales to weigh the mothers. The day after the WIC staff training session, one of the academic researchers conducted cognitive interviews (using a standardized protocol) with WIC mothers after they completed the HSM interview. The cognitive interviews sought to understand how the mothers interpreted each of the HSM questions in an effort to ensure the content validity of the HSM interview. A few questions were then modified based on feedback from the cognitive interviews.

Throughout the implementation of the pilot intervention, the Plan-Do-Study-Act (PDSA) process was used to assess the implementation of the intervention and to improve the quality of it. The PDSA cycle is shorthand for testing a change by developing a plan to test the change (Plan), carrying out the plan (Do), observing and learning from the consequences (Study), and determining what modifications should be made to the test (Act; Taylor et al., 2014). Table 1 describes the PDSA cycles and QI intervention activities (Acts) conducted as part of this pilot project.

TABLE 1.

PDSA Cycles and Activities

PDSA Cycle	Quality Improvement Intervention Activity (Act)
1 (Jan–April 2013)	April 2013: Revised enrollment script, asked WIC educators to record two of their maternal counseling sessions for feedback, held a conference call with WIC team and university researchers
2 (Jan–May 2014)a	January 2014: University researchers went to the SLV to try to raise morale April 2014: University staff conducted a “booster” training; converted “control” sites to intervention sites, changed enrollment practices in the waiting room, set goal of 20 visits per month
3 (July–October 2014)	July 2014: Gave farmers’ market incentives to WIC educators for meeting goal October 2014: Provided a group lunch incentive for the WIC educators

NOTE: PDSA = Plan-Do-Study-Act; WIC = Women, Infants, and Children; SLV = San Luis Valley.

^aCycle 2 occurred so much later after Cycle 1 because of all the staff turnover, the government “shut-down,” and the holidays.

Participants

Women were invited to participate in the QI intervention if they were (1) postpartum (up to 2 years) and (2) visiting the WIC clinic for either a maternal and/or infant/toddler visit. A total of 292 women who met these criteria participated in the QI intervention. Of these, 59 received the intervention a second time and 12 received the intervention a third time. The results included in this article will report only on the subset of women who received the intervention at least two times and, thus, have pre- and postmeasurements. The average amount of time between these women’s visits was 193 days (approximately 6½ months).

Data Collection and Analysis

To address the first objective related to assessing the feasibility of making a WIC systems-level change that incorporated HSM and MI techniques into each postpartum maternal and offspring visit, both quantitative and qualitative data were collected. Quantitative data primarily included the intervention reach collected via monthly HSM usage reports and manual WIC chart reviews (to calculate screening rates) done by the WIC Director (KB; for the months of January 2014 and January 2015). Qualitative data were collected via focus group sessions with the WIC educator team (January 2013) and (April 2014). For each focus group session the purpose of the focus group session was explained. The focus group protocols consisted of only five questions each, with follow-up probes (0–6 probes per question). At the end of each focus group session, the collective answers to each question were reviewed with the respondents, and they were given the opportunity to correct any misunderstandings or add any additional information that may have been left out of the initial responses (member checking) in an effort to ensure the accuracy and completeness of the data collected. The focus group sessions were recorded and transcribed. A thematic analysis involving mining the data for categories and themes related to the feasibility of making a systems-level WIC change was conducted. Additionally, monthly project team¹ meeting minutes, which captured the WIC Director’s update on how the QI intervention implementation was going, as well as project team members’ thoughts about the actions taken in the PDSA cycle were also analyzed and used to triangulate the themes that emerged from the focus group sessions.

The HSM monthly usage data were compiled into a graph for analysis (these data informed the PDSA cycles). The screening rates were calculated from the WIC chart reviews (number of intervention participants/number of eligible participants [up to 2-years postpartum mothers]).

To address the second objective related to the impacts of the intervention on the participating mothers’ behaviors and weight status, a subset of mothers who had a pre- and postassessment were included in the analyses. Their survey responses to seven eating/drinking behaviors and three physical activity behaviors from the HSM survey (adapted from the Youth Risk Behavior Surveillance System; Centers for Disease Control and Prevention [CDC], 2009) were analyzed using the nonparametric Friedman test (Friedman, 1937) since all variables were ordinal in nature. Change in their BMIs over time was also assessed using a repeated measures analysis of covariance. Maternal age and language at baseline were entered into the models as covariates.

RESULTS

Descriptive Statistics

The demographics characteristics for this subset of women are included in Table 2. The mean age is 26 years at Time 1 (T1) and 28 years at Time 2 (T2). The majority of parents completed the interview in English (89% at T1 and 85% at T2). At T1, just over three fourths of the participants had infants less than 1 year old (77%); they were, on average, 4 months postpartum at T1. For these women, the majority were formula feeding (63%), followed by breastfeeding (23%), and doing both breast and formula feeding (14%).

TABLE 2.

Demographic Descriptive Statistics for SLV Postpartum WIC Sample Over Time

Demographics	Time 1a (n = 292)	Time 2b (n = 59)
Age (years), M (SD)	26.1 (5.4)	27.9 (5.8)
Language, %
English	89.0	84.7
Spanish	11.0	15.3
Pregnant (yes), %	—	13.6
Has infant <12 months old (yes), %	77.7	62.7
If has infant, feeding method
Breast	23.1	33.3
Formula	63.1	55.6
Both	13.8	11.1

NOTE: SLV = San Luis Valley; WIC = Women, Infants, and Children. The average time between Time 1 and 2 visits was 193 days (SD = 158 days).

^aTime 1 included visits between January 2013 and October 2014.

^bTime 2 included visits between February 2013 and February 2015.

Objective 1: Feasibility

The HSM usage reach data showed that there was high initial “uptake” (46 postpartum mothers were enrolled in project in Feb 2013), but there was a sharp decline over subsequent months (see Figure 1). Three PDSA cycles were completed in an effort to improve the number of eligible participants receiving the intervention. However, due to many program barriers, including staff turnover and the temporary cessation of program funding due to the October 2013 government “shut-down,” the success with the initial uptake was never repeated, although there were times that the enrollment “bounced back” and was higher because of the PDSA efforts.

FIGURE 1. HSM Usage Data for 2 Years (January 2013–January 2015).

NOTE: HSM = HeartSmartMoms; PDSA = Plan-Do-Study-Act; WIC = Women, Infants, and Children; MI = motivational interviewing.

The HSM usage data presented in Figure 1 provide the number of participating mothers in the intervention only. To understand systems change feasibility, it was critical to know how the HSM data mapped to the total number of mothers eligible to participate in the intervention. As such, the WIC Director conducted manual chart reviews for the months of January 2014 and January 2015 to determine this. The results of the chart reviews are included in Table 3. Only one in every six eligible postpartum mother received the QI intervention; this was fairly consistent from the middle of the intervention to the end of the intervention.

TABLE 3.

Eligible Postpartum Visits for the Month of January 2014 and 2015

Month	Total Eligible Postpartum Visits (Across All sites)	No. (%) of HSM Interviews Completed	No. (%) of HSM Interviews Completed + Counseling
January 2014	95	21 (22)	15 (16)
January 2015	116	16 (14)	14 (12)

NOTE: HSM = HeartSmartMoms.

Perceived barriers to implementing the intervention were assessed via the focus groups with the WIC educators and the Director and from the project team meeting minutes. They were categorized into three main categories: perceived barriers for the WIC educators, perceived project barriers, and perceived barriers for the WIC participants. They were then organized even further in each of these categories using a socioecological framework (see Figure 2). For barriers pertaining to WIC educators, subcategories included “Individual Barriers” (lack of knowledge/job inexperience, lack of self-efficacy to effectively counsel), “Structural Barriers” (staff turnover, limited visit time, unbalanced workloads), and “Economic Barriers” (cut in funding). For project-related barriers, the subcategories were similar: “Attitudinal Barriers” (lack of staff buy-in), “Structural Barriers” (not integrated into Compass computer system, limited wi-fi connectivity), and “Economic Barriers” (limited budget for project). Last, the barriers pertaining to WIC participants were also classified similarly: “Individual Barriers” (lack of knowledge about connection between their health and diet), “Attitudinal Barriers” (resistant to change, weight is a sensitive subject, many competing priorities), and “Visit Barriers” (infants and young children present, receipt of too much information at visit).

FIGURE 2. Perceived Barriers to Implementing the HSM/MI Intervention in WIC Clinics.

NOTE: HSM = HeartSmartMoms; WIC = Women, Infants, and Children; MI = motivational interviewing.

To give examples of excerpts that illustrated some of these barriers, one WIC educator started one of the focus groups by saying, “It’s [the intervention] ok … It’s hard. I find it harder than I thought to ask permission (one of the components of MI),” which supports the barrier of the lack of self-efficacy to effectively counsel. Similarly, another said later in the focus group,

I think with the Motivational Interviewing, it’s hard because again the time, I don’t always have time to say “Well, how motivated are you to do this.” I just say you know “This is what they recommended and this is what you can take home because it is the last thing that I had to do to get them out the door.”

The barrier of time came up over and over again. Another such exchange that illustrates the structural barrier of time for WIC educators and the structural barrier of wi-fi connectivity for the project: “Our clients walk in late, it is kind of hard to get them to do the kiosk … and sometimes the kiosk isn’t working the way it should …” A different WIC educator followed and said, “You have to reboot it and restart it cus it is just blank white, blank … several times.” A third participant contributed, “We’re connected to the internet because our Compass works off the Internet and we’re all working fine, but the kiosk will say, ‘Lost internet connection, hit retry,’ but we’re all fine.” These are just a few examples of the barriers that were conveyed by WIC educators.

Objective 2: Impact on Maternal Eating and Physical Activity Behaviors and on Weight Status

The quantitative results for the impacts on maternal behaviors are included in Table 4. Postpartum women who participated in the intervention significantly decreased their sugar-sweetened beverage consumption (mean rank T1 = 1.61, mean rank T2 = 1.55; χ² = 5.54; p < .05). Mean rank reflects the number of times participants drink a can, bottle, or glass of a sweetened beverage, on a typical day. There were no other significant improvements in their eating/drinking or physical activity behaviors. They did, however, improve their weight status, F(3, 56) = 4.44, p < .05, controlling for maternal age and language at baseline. There was a significant interaction between time and language, F(3, 56) = 4.50, p < .05. It appeared that the intervention had a much greater impact on Spanish-speaking mothers, although the results should be interpreted with caution because of the low number of Spanish-speaking mothers (n = 9) and the unequal sample sizes in each language group (English-speaking = 50 vs. Spanish-speaking = 9).

TABLE 4.

Impacts on Maternal Behaviors and Weight Status for Postpartum Women

Variable (n = 55)	Mean Rank T1	Mean Rank T2	Friedman Test (χ2)
Eating and drinking behaviors
Daily fruit consumption	1.52	1.48	0.13
Daily vegetable consumption	1.53	1.47	0.36
Daily “junk food” consumption	1.48	1.52	0.12
Weekly eating out	1.46	1.54	0.80
Weekly breakfast consumption	1.57	1.43	1.09
Daily sugar-sweetened beverage consumption	1.61	1.39	5.54*
Daily milk consumption	1.45	1.55	0.70
Physical activity behaviors
Daily physical activity	1.49	1.51	0.21
Daily screen time	1.58	1.42	2.46
Intent to be more physical active	1.47	1.53	0.43
Weight Status (n = 59)	Adjusted M, T1	Adjusted M, T1	F Statistic
Time BMIa	28.56	26.96	4.45*
Language × Time			4.50*
Maternal age × Time			2.67

NOTE: BMI = body mass index.

^aChange over time in BMI was tested controlling for maternal age and language at baseline.

^*p < .05.

DISCUSSION

This QI project set out to determine if making a systems-level change, to implement a maternal weight component, supported by HSM technology and MI at postpartum visits, in an effort to improve maternal periconceptional health, in three rural WIC clinics, was feasible and whether the intervention improved postpartum women’s eating and physical activity behaviors, as well as their weight status. Via HSM usage data, manual WIC chart reviews, focus groups, and project team meeting transcripts, it was determined that, in fact, it is feasible to make a short-term systems-level change. However, many barriers were encountered in doing so, and the results were not sustained. The current project was unable to overcome all the barriers, resulting in only partial implementation of the systems-level change.

The second objective was assessed by examining pre-/post-HSM survey data for those participants who had at least one repeat visit. The results suggest that the intervention did improve participants’ daily consumption of sugar-sweetened beverages and maternal weight status (controlling for maternal age and language), but did not improve any other eating/drinking or physical activity behaviors. These results need to be interpreted with caution, however, because there were limitations to this QI project evaluation, namely, the sample size is very small and there is no control group. Additionally, it might be expected that BMI will change over time for postpartum women even without an intervention.

CONCLUSIONS

There were several lessons learned from this academic/community pilot QI project, which have implications for practice. At the end of the project, each project team member had the opportunity to reflect on “lessons learned” from the project. They sent their lists of lessons learned to the lead author, and she compiled them and then circulated them to the team and they were discussed and agreed on at a final team meeting. General lessons learned and recommendations are as follows:

• A long-distance University–WIC partnership is feasible (225 miles), but frequent communications (both in-person and via phone, Skype, etc.) are necessary.

• WIC is an appropriate place to address maternal– offspring overweight/obesity; however, the link between the two need to be elucidated for both WIC educators and WIC participants.

• Low-cost adjustments to WIC counseling can be made to address maternal–offspring overweight/obesity.

• It should not be assumed that WIC educators have bought into a healthy lifestyle for themselves. Their knowledge, beliefs, and motivations differ greatly from mid-level health practitioners and doctors. This difference should be considered carefully when they are participate in a health promotion intervention.

Intervention-specific lessons learned and recommendations are as follows:

• WIC Director buy-in and support are necessary, but not sufficient; the buy-in of the WIC educators is also essential for implementation.

• There is a need to better integrate the intervention into the normal work load of the WIC educators.

  • ○
    This can be accomplished by taking a community-based participatory research-approach at the beginning of the project and engaging staff in the development of the intervention and better integrating the intervention into the agencies computer software program (for CO WIC it is Compass).

• WIC staff respond well to incentives used by the team to increase their participation in the intervention.

• Integration of MI skills can be challenging:

  • ○
    The initial all-day intensive training was reported as beneficial, and booster training resulted in increased uptake of the intervention; however, longitudinal periodic training with continued audit and feedback, or step-based training over time may have further enhanced uptake.
  • ○
    There is likely benefit from using the train-the-trainer model so that new staff could be trained internally when they begin their jobs. This could take the form of an “intervention champion” on staff to aid with the successful uptake and maintenance of the intervention on staff.

In sum, this pilot QI project revealed great potential for making a systems-level change involving maternal health promotion in rural WIC clinics, but many barriers need to be overcome in order to successfully implement a WIC intervention. It is hoped that this article sheds light on some of those barriers and that the information presented here can be used to contribute to the successful implementation of health promotion interventions aimed at improving postpartum maternal health, which can increase health during the periconceptional phase and, in turn, improve the health outcomes for a child.