Abstract
Introduction:
Group prenatal care (GPNC) represents a promising alternative to individual prenatal care (IPNC) for better perinatal outcomes. Some studies suggest better perinatal outcomes with more GPNC sessions attended. The present study sought to examine patient characteristics associated with higher or lower GPNC session attendance.
Methods:
We conducted a secondary analysis of data collected in a single-site randomized controlled trial comparing GPNC to IPNC, focusing solely on patients assigned to GPNC. The outcome of interest was GPNC session attendance. Predictors evaluated included patient sociodemographic, psychosocial, health, and health behavior characteristics. The association of patient characteristics with session attendance was assessed using Zero-Inflated Poisson regression.
Results:
The sample (n = 1,068) was racially diverse (40.7% Black, 35.8% white, 22.1% Hispanic) and largely Medicaid eligible (92.3%). The characteristics of older age (adjusted relative risk [aRR] = 1.01; 95% confidence interval [CI] [1.00, 1.01]; p = .012), foreign nativity (aRR = 1.14; 95% CI [1.06, 1.22]; p < .001), and prenatal distress (aRR = 1.09; 95% CI [1.01, 1.19]; p = .030) were associated with higher session attendance. Being in a committed relationship but unmarried (aRR = .91; 95% CI [.85, .97]; p = .007), less than high school education (aRR = .89; 95% CI [.82, .97]; p = .007), depressive symptoms (aRR = .93; 95% CI [.87, .99]; p = .027), housing instability (aRR = .92; 95% CI [.84, .99]; p = .049), housing issues (aRR = .88; 95% CI [.80, .97]; p = .008), life stressors (aRR = .92; 95% CI [.86, .98]; p = .010), gestational diabetes (aRR = .81; 95% CI [.71, .91]; p < .001), and smoking (aRR = .92; 95% CI [.85, .99]; p = .040) were associated with lower session attendance.
Discussion:
Patient characteristics were differentially associated with GPNC session attendance. Given that patients with less education, more depressive symptoms, housing instability, stressors, gestational diabetes, and prenatal cigarette use attended fewer sessions, adaptations to the GPNC model to reduce attendance barriers for these groups should be considered.
Group prenatal care (GPNC) is an innovative model of prenatal care that integrates clinical assessment with extensive health education and socialization within a group of pregnant persons of similar weeks’ gestation (Rising, 1998; Schindler Rising et al., 2004). Shared medical appointments such as GPNC are theorized to benefit patients through increased time spent with providers, decreased social hierarchy in the patient-provider relationship, enhanced education, peer support, and the formation of healthy group norms (Cunningham et al., 2021). GPNC in particular has been suggested to improve perinatal health through bolstered patient-provider communication, social support, stress reduction, health education (Sheeder et al., 2012), and less practitioner bias (Carter & Mazzoni, 2021).
Improving patient engagement with GPNC is essential, not only because GPNC may improve perinatal health outcomes compared with traditional individual prenatal care (IPNC) (Liu et al., 2021; Tucker et al., 2021), but also because ensuring that patients recruited into groups will attend assigned sessions contributes to the success of GPNC programs within obstetric practices (Pekkala et al., 2020). Yet, evidence regarding the association of individual-level characteristics with GPNC session attendance is limited. Although some studies suggest associations between patient characteristics and interest in GPNC participation (Berman et al., 2020; McDonald et al., 2016; Weber Yorga & Sheeder, 2015), factors associated with interest may not necessarily translate into actual GPNC session attendance.
Previous studies of GPNC session attendance have considered attendance using attendance thresholds, as well as the number and percentage of sessions attended. In an examination of the percentage of GPNC sessions attended of those participants who were eligible to attend, foreign nativity was found to be associated with a higher rate of session attendance (Cunningham et al., 2017). Further, when expanding from individual characteristics to examine the influence of group composition, being in a group of patients with a more diverse range of ages was associated with higher session attendance through patient satisfaction (Earnshaw et al., 2016).
When characteristics were compared between those with low (fewer than five sessions) versus high (five or more sessions) GPNC session attendance in a subsample of the current study, patients with lower perceived family support were more likely attend fewer than five sessions (Francis et al., 2019). Wagijo et al. (2022) found initiation of GPNC (one or more sessions) was more common among patients who were 22–26 years old, nulliparous, cohabitating but unmarried, with average or above average stress, an average lifestyle and pregnancy knowledge, and nonsmoking; however, no difference was seen based on subgroups of attendance (low: fewer than four sessions, medium: four to six sessions, high: seven or more sessions). Despite having been used in previous studies (Crockett et al., 2019; Ickovics et al., 2007), these attendance thresholds may not represent meaningful cut points, and analyzing attendance as a continuous variable may therefore add value.
Drawing from a large sample of pregnant persons randomized to attend GPNC at a health facility with an established GPNC program, the present study aimed to determine whether pregnant persons’ sociodemographic, psychosocial, health, and health behavior characteristics are associated with the number of sessions an individual attends. Identification of patient characteristics associated with higher or lower attendance can inform targeted recruitment and model adaptation.
Methods
Setting and Design
This study is a secondary analysis of the Centering and Racial Disparities (CRADLE) randomized controlled trail (RCT), which aimed to assess whether participation in GPNC reduced rates of preterm birth and low birth weight when compared with IPNC, and whether GPNC reduced racial disparities in these rates. Findings from the primary analysis have been published previously (Crockett et al., 2022; Smith et al., 2023) and a detailed study protocol is available (Chen et al., 2017). The CRADLE study received institutional review board approval from the Prisma Health Office of Human Research Protection (Pro00043994) and is registered on clinicaltrials.gov (identifier: NCT02640638).
In brief, CRADLE study participants were recruited at a single hospital-affiliated obstetrics practice in South Carolina that serves a largely medically underserved population. Enrollment took place between February 2016 and March 2020. Patients were eligible for inclusion if they were between the ages of 14 and 45 and entered care prior to 20 6/7 weeks' gestation. Patients with medical, pregnancy, social, or behavioral complications that precluded prenatal care delivery by a nurse practitioner and those with planned preterm birth were excluded. Eligible patients provided informed written consent on study enrollment. After enrollment, patients were randomly assigned (1:1), stratified by race and ethnicity, to attend either CenteringPregnancy GPNC or IPNC. Patients were recruited on entry to prenatal care and followed through the postpartum period, with data collected at three time points: baseline, the third trimester, and 8 weeks postpartum. Upon enrollment and after 30 weeks’ gestation, participants completed surveys composed of validated sociodemographic, psychosocial, and behavioral measures. Small financial incentives were provided with completion of the first survey at enrollment ($25) and after attendance of five visits in the assigned study arm ($50), which coincided with completion of the second survey. Following 8 weeks postpartum, maternal health history, service use, childbirth, and birth outcomes were abstracted from electronic health records (EHRs).
Participants
The current analysis was limited to participants who were assigned to the GPNC arm (n = 1,175). Participants with pregnancies resulting in miscarriage (n = 47) or missing attendance information (n = 60) were excluded from analysis due to ineligibility to attend sessions after 21 weeks’ gestation or missingness on the outcome variables of interest, respectively. This resulted in an analytic sample of 1,068 pregnant persons.
CenteringPregnancy GPNC
In CenteringPregnancy GPNC, groups of 8–12 patients receive prenatal care together over a series of 10 2-hour sessions, following the same visit schedule as IPNC (Rising, 1998; Schindler Rising et al., 2004). Sessions are facilitated by a certified health provider (midwife, obstetrician, or nurse practitioner) and co-facilitator (registered nurse, social worker, or similar). During the first 30 minutes, the clinician conducts brief physical assessments and routine ultrasounds in the group space. Concurrently, patients socialize, complete worksheets connected to that day's discussion, and perform self-assessments including measurement of weight and blood pressure. The remaining 90 minutes are spent in facilitated discussion centered around topics of pregnancy, childbirth, and parenting, matched to group gestational age. The consistency of facilitators and group members over the 6 months’ course is considered an essential element of the GPNC model, to encourage the formation of relationships and trust.
Measures
Selection of patient characteristics was informed by previous studies of GPNC. In addition to sociodemographic, psychosocial, and health behavior variables introduced in previous literature, common health conditions in pregnancy were also considered, given the interest in GPNC in these populations (Carter et al., 2024; Chen et al., 2023). Measures of interest were largely collected at Survey 1 or through EHR abstraction. In cases in which measures were collected at both surveys, only initial survey reports were included to avoid capturing potential intervention effects. However, some characteristics were only captured in the second survey; where this is the case, it has been noted.
Sociodemographic Characteristics
Sociodemographic variables included maternal age (calculated based on date of birth), parity (nulliparous; primiparous/multiparous), and race and ethnicity. Participants self-reported race and ethnicity in survey questions from the U.S. Federal Government 1997 Office of Management and Budget standards on race and ethnicity used by the U.S. Census Bureau, which allowed selection of multiple categories (Wallman & Hodgdon, 1977). Selections were then collapsed into four categories of race and ethnicity to provide greater cell size for subgroup comparison: Black; Hispanic, race other than Black; nonHispanic white; and multiracial or of other racial groups (i.e., Native American or Alaska Native, Asian, Middle Eastern or North African, Native Hawaiian or Other Pacific Islander, or Other). Participants who identified as Black and other races or ethnicities were categorized with those who identified as Black only. Participants who identified as Hispanic alone or with a race other than Black were categorized as Hispanic (Chen et al., 2017). Race and ethnicity categories are hereafter referred to as “Black,” “Hispanic,” “white,” and “other race and ethnicity.” We also collected nativity (born in the United States; born outside the United States), relationship with baby's father (married; committed dating relationship or engaged; single or other), educational attainment (less than high school; high school; above high school), Medicaid eligibility (eligible; ineligible), and work status (full time; part time; unemployed).
Psychosocial Characteristics
Several items assessed participants' perceived stress and experience of stressors in pregnancy. Participants’ initial feeling about pregnancy was measured using an item adapted from the Pregnancy Risk Assessment Monitoring System (PRAMS) questionnaire, with responses classified as happy, unhappy, or unsure (Centers for Disease Control and Prevention, 2014). Symptoms of depression were measured using the Center for Epidemiological Studies Depression Scale (CES-D) (Radloff, 1977), adapted for use in pregnant populations with five of the 20 questions that overlapped with common pregnancy symptoms excluded (Ickovics et al., 2011). CES-D scores were classified as reaching clinical concern (≥12) or not reaching clinical concern (<12). General stress, perceived control, and confidence related to one's life circumstances was measured using four items from the Perceived Stress Scale (PSS), with responses on a 5-point scale from never (0) to very often (4) (Cohen & Hoberman, 1983). To assess anxiety related to pregnancy, participants completed the 13-item Pregnancy-Specific Anxiety Scale (PSAS), with responses on a 5-point scale from never (0) to always (4) (Roesch et al., 2004). Participants reported how bothered they felt by 17 common stressors in pregnancy on the Revised Prenatal Distress Questionnaire (NUPDQ), using a 3-point scale ranging from not at all (0) to very much (2) (Lobel et al., 2008). PSS, PSAS, and NUPDQ item responses were summed, and summed totals were then classified as below average (a score of −1 standard deviation [SD] below the mean), average (a score within 1SD above and below the mean), and above average (a score of +1SD above the mean).
Perceived neighborhood safety was measured using four items adapted from the Los Angeles Mommy and Baby survey, on which participants rated police protection, property protection, safety from violence, and friendliness in their neighborhood (County of Los Angeles, 2016). Mean scores were classified as average, above average, and below average. Housing instability was measured as two or more moves in the past year, as has been done in previous research (Carrion et al., 2015). Participants also reported other issues of housing instability in the previous year (e.g., eviction or foreclosure, seeking temporary shelter, and being behind on mortgage or rent payments) and indicated whether they experienced any of seven life stressors during pregnancy (e.g., job loss, death of a loved one). Similar to housing instability, housing issues and life stressors during pregnancy were categorized as fewer than two or two or more. Food insecurity in the past month was measured using five items from the U.S. Household Food Security Survey moduleshort form, with scores categorized as secure (0), marginally secure (1), and insecure (2 or more) (Economic Research Service, 2012). Perceived experience of discrimination was measured using the Everyday Discrimination Scale (EDS) and mean scores calculated (Williams et al., 1997). Perceived family support was measured using seven items adapted from the Provisions of Social Relations Scale, with responses on a 4-point scale ranging from strongly disagree (0) to strongly agree (3) (Turner et al., 1990). Mean scores were classified as lower (<4) or higher family support (≥4). Perceived partner support was assessed using seven items derived from previous measures (Ghosh et al., 2010; Hahn-Holbrook et al., 2013; Lobel et al., 1992; Turner et al., 1990), with responses on a 4-point scale ranging from not at all (1) to very much (4). Mean scores were classified as lower (<3) and higher partner support (≥3). Housing instability, housing issues, life stressors in pregnancy, and perceived partner support were measured at Survey 2 in the third trimester.
Health and Health Behavior Characteristics
Diagnoses of common maternal health conditions (gestational diabetes, chronic hypertension, and gestational hypertension) were identified via medical chart review. Participants self-reported use of cigarettes, alcohol, and marijuana in pregnancy at Survey 1 using questions adapted from the PRAMS questionnaire (Centers for Disease Control and Prevention, 2014). Responses were classified as none, past use, or any cigarette use, and none or any alcohol and marijuana use.
Session Attendance
Participants’ prenatal care attendance was abstracted from EHRs as a count variable (number of GPNC sessions; number of IPNC visits; total prenatal care [TPNC] visits/sessions). The primary outcome of interest in the present study was total GPNC session attendance. Number of GPNC sessions attended was deemed to be more representative of the intervention “dose” received, as compared with alternatives such as the threshold of attendance (e.g., five or more sessions) or the percentage of eligible sessions attended. We calculated participant eligibility to attend GPNC sessions from the month of study enrollment and date of childbirth with a maximum of 10 sessions.
Statistical Analysis
Data were analyzed using SAS (Version 9.4). Descriptive statistics were calculated to characterize the study sample: mean and SD for normally distributed continuous variables; median, first quartile, and third quartile for continuous variables with a non-normal distribution; and counts with percentages for categorical variables. As session attendance was a count variable with a high prevalence of zero values and overdispersion, a Zero-Inflated Poisson (ZIP) regression was used to assess the association between patient characteristics and session attendance, each individually. The model consisted of two parts, a Poisson count model with log link for not-always zero counts and a binomial model with logit link for always zero probability. The count regression identified factors associated with an increase in the total number of GPNC sessions attended, whereas the zero-inflated regression identified factors contributing to participants attending zero GPNC sessions. Adjusted models controlled for gestational age at study enrollment and gestational age at birth. Missing data were coded as its own “missing” category for data analysis and complete case analysis employed. Significance was determined at a p < .05.
A post hoc analysis of TPNC visit/session count was conducted for variables found to have a significant association in ZIP models, utilizing Poisson regression.
Results
Participants’ median age was 24.5 years (Q1 = 20.97, Q3 = 28.45). Forty-one percent of participants were in the Black, 22% the Hispanic, 36% the white and 1% the “other race and ethnicity” groups. Most participants were born in the United States (82%), had a high school education (53%), were Medicaid eligible (92%), and were multiparous (56%) (Table 1). On average, participants attended 4.81 GPNC sessions (SD = 3.54, median = 6, Q1 = 1, Q3 = 8). Based on timing of enrollment and date of childbirth, participants were eligible to attend an average of 9.52 sessions (SD = 1.17, median = 10, Q1 = 10, Q3 = 10) and attended 50.27% of the sessions they were eligible to attend (SD = 36.47, median = 60%, Q1 = 10%, Q3 = 80%) (Figure 1). Despite assignment to the GPNC arm, participants attended an average of 5.69 IPNC visits (SD = 3.83, median = 5, Q1 = 3, Q3 = 8), contributing to an overall average of 10.50 TPNC visits/sessions (SD = 3.45, median = 11, Q1 = 8, Q3 = 13).
Table 1.
Characteristics of Patients Assigned to Group Prenatal Care
| Participant Characteristics (n = 1,068) | n (%) |
|---|---|
| Sociodemographic characteristics | |
| Age, median [Q1, Q3] | 24.50 [20.97, 28.45] |
| Race and ethnicity | |
| Black | 435 (40.73) |
| Hispanic | 236 (22.10) |
| White | 382 (35.77) |
| Other race | 15 (1.40) |
| Nativity | |
| Born in the United States | 878 (82.21) |
| Born outside the United States | 181 (16.95) |
| Missing | 9 (0.84) |
| Relationship status | |
| Single or other | 225 (21.07) |
| Committed dating relationship or engaged | 448 (41.95) |
| Married | 212 (19.85) |
| Missing | 183 (17.13) |
| Educational attainment | |
| Less than high school | 255 (23.88) |
| High school | 566 (53.00) |
| Above high school | 207 (19.38) |
| Missing | 40 (3.75) |
| Medicaid eligibility | |
| Eligible | 986 (92.32) |
| Ineligible | 40 (3.75) |
| Missing | 42 (3.93) |
| Employment | |
| Unemployed | 459 (42.98) |
| Working part time | 189 (17.70) |
| Working full time | 345 (32.30) |
| Missing | 75 (7.02) |
| Parity | |
| Nulliparous | 475 (44.48) |
| Primiparous/Multiparous | 593 (55.52) |
| Psychosocial characteristics | |
| Initial feelings about pregnancy | |
| Happy | 630 (58.99) |
| Unhappy | 103 (9.64) |
| Unsure | 323 (30.24) |
| Missing | 12 (1.12) |
| Symptoms of depression | |
| Not reaching clinical concern | 683 (63.95) |
| Of clinical concern | 294 (27.53) |
| Missing | 91 (8.52) |
| Perceived stress | |
| Below average | 172 (16.10) |
| Average | 728 (68.16) |
| Above average | 125 (11.70) |
| Missing | 43 (4.03) |
| Pregnancy-specific anxiety | |
| Below average | 195 (18.26) |
| Average | 703 (65.82) |
| Above average | 148 (13.86) |
| Missing | 22 (2.06) |
| Prenatal distress | |
| Below average | 161 (15.07) |
| Average | 673 (63.01) |
| Above average | 163 (15.26) |
| Missing | 71 (6.65) |
| Perceived neighborhood safety | |
| Below average | 166 (15.54) |
| Average | 529 (49.53) |
| Above average | 245 (22.94) |
| Missing | 128 (11.99) |
| Housing instability* | |
| <2 moves | 770 (72.10) |
| ≥2 moves | 118 (11.05) |
| Missing | 180 (16.85) |
| Housing issues* | |
| <2 housing issues | 810 (75.84) |
| ≥2 housing issues | 109 (10.21) |
| Missing | 149 (13.95) |
| Food insecurity | |
| Secure | 588 (55.06) |
| Marginally secure | 152 (14.23) |
| Insecure | 237 (22.19) |
| Missing | 91 (8.52) |
| Number of life stressors in pregnancy* | |
| <2 life stressors | 674 (63.11) |
| ≥2 life stressors | 244 (22.85) |
| Missing | 150 (14.04) |
| Everyday discrimination, median [Q1, Q3] | 1.10 [1.0, 1.6] |
| Missing | 28 (2.62) |
| Family support | |
| Lower support | 173 (16.20) |
| Higher support | 699 (65.45) |
| Missing | 196 (18.35) |
| Partner support* | |
| Lower support | 321 (30.06) |
| Higher support | 565 (52.90) |
| Missing | 182 (17.04) |
| Health and health behavior | |
| Gestational diabetes | |
| No | 990 (92.70) |
| Yes | 78 (7.30) |
| Chronic hypertension | |
| No | 953 (89.23) |
| Yes | 115 (10.77) |
| Gestational hypertension | |
| No | 893 (83.61) |
| Yes | 175 (16.39) |
| Cigarette smoking | |
| None | 691 (64.70) |
| Smoked in the 3 months prior | 174 (16.29) |
| Any smoking during pregnancy | 168 (15.73) |
| Missing | 35 (3.28) |
| Alcohol use during pregnancy | |
| None | 991 (92.79) |
| Any | 45 (4.21) |
| Missing | 32 (3.00) |
| Marijuana use during pregnancy | |
| None | 984 (92.13) |
| Any | 50 (4.68) |
| Missing | 34 (3.18) |
Measured at Survey 2 in the third trimester.
Figure 1.
(A) The number of group prenatal care sessions participants attended. Notably, 23.5% of participants attended zero GPNC sessions, whereas 4.3% attended 10 sessions. (B) The proportion of eligible sessions participants attended, calculated as the number of participants who attended/the number of participants eligible to attend based on gestational age at enrollment and birth.
Characteristics Associated With Session Attendance
Tables 2 and 3 summarize the results of ZIP regression. In the adjusted Poisson model, sociodemographic characteristics of older age (adjusted risk ratio [aRR] = 1.01; 95% confidence interval [CI] [1.00, 1.01]; p = .012), Hispanic race and ethnicity group (aRR = 1.08; 95% CI [1.08, 1.16]; p = .045), and foreign nativity (aRR = 1.14; 95% CI [1.06, 1.22]; p < .001) were positively associated with GPNC session attendance. Being in a committed dating relationship or engaged rather than married (aRR = .91; 95% CI [.85, .97]; p = .007) and having less than a high school education (aRR = .89; 95% CI [.82, .97]; p = .002) were inversely associated with session attendance. Of the psychosocial characteristics examined, depressive symptoms were inversely associated with session attendance (aRR = .93; 95% CI [.87, .99]). Average relative to below average prenatal distress was positively associated with session attendance (aRR = 1.09; 95% CI [1.01, 1.19]; p = .030). By contrast, housing instability (aRR = .92; 95% CI [.84, .99]; p = .049), housing issues (aRR = .88; 95% CI [.80, .97]; p = .008), and life stressors in pregnancy (aRR = .92; 95% CI [.86, .98]; p = .010) were inversely associated with session attendance. The health characteristics of gestational diabetes (aRR = .81; 95% CI [.71, .91]; p < .001) and smoking during pregnancy (aRR = .92; 95% CI [.85, .99]; p = .040) were inversely associated with session attendance.
Table 2.
Likelihood of Group Prenatal Care Session Attendance (Count Model) and of Attending Zero Sessions (Zero-Inflated Model)
| Participant Characteristics | Count |
Zero Inflated |
||
|---|---|---|---|---|
| RR [95% CI] | p Value | OR [95% CI] | p Value | |
| Sociodemographic characteristics | ||||
| Age (n = 1,068) | 1.01 [.99, 1.01] | .0632 | .97 [.94, .99] | .0312 |
| Race and ethnicity (n = 1,068) | ||||
| Black | 1.04 [.97, 1.11] | .2584 | .99 [.72, 1.36] | .9296 |
| Hispanic | 1.06 [.98, 1.14] | .1464 | .66 [.44, 0.99] | .0481 |
| White | Ref | Ref | ||
| Other race | 1.21 [.97, 1.51] | .0940 | 1.09 [.34, 3.51] | .8844 |
| Nativity (n = 1,059) | ||||
| Born in the United States | Ref | Ref | ||
| Born outside the United States | 1.11 [1.04, 1.19] | .0029 | .68 [.45, 1.03] | .0659 |
| Relationship status (n = 993) | ||||
| Single or other | .95 [.88, 1.03] | .2337 | 1.26 [.77, 2.04] | .3586 |
| Committed dating relationship or engaged | .90 [.84, .97] | .0041 | 1.14 [.74, 1.76] | .5512 |
| Married | Ref | Ref | ||
| Educational attainment (n = 1,028) | ||||
| Less than high school | .88 [.81, 0.96] | .0027 | 1.42 [.91, 2.23] | .1248 |
| High school | .98 [.91, 1.05] | .5077 | 1.35 [.91, 2.00] | .1416 |
| Above high school | Ref | Ref | ||
| Medicaid eligibility (n = 1,026) | ||||
| Eligible | Ref | Ref | ||
| Ineligible | 1.07 [.93, 1.23] | .3589 | 1.17 [.56, 2.44] | .6724 |
| Employment (n = 993) | ||||
| Unemployed | .93 [.87, .99] | .0265 | 1.07 [.78, 1.48] | .6755 |
| Working part time | 1.00 [.93, 1.09] | .9322 | .68 [.44, 1.07] | .0918 |
| Working full time | Ref | Ref | ||
| Parity (n = 1,068) | ||||
| Nulliparous | Ref | Ref | ||
| Primiparous/Multiparous | 1.04 [.98, 1.09] | .2267 | .82 [.62, 1.09] | .1780 |
| Psychosocial characteristics | ||||
| Feelings about pregnancy (n = 1,056) | ||||
| Happy | Ref | Ref | ||
| Unhappy | 1.02 [.93, 1.13] | .6473 | .72 [.45, 1.16] | .1756 |
| Unsure | 1.09 [.98, 1.21] | .1145 | .86 [.52, 1.41] | .5445 |
| Symptoms of depression (n = 997) | ||||
| Not reaching clinical concern | Ref | Ref | ||
| Of clinical concern | .93 [.87, .99] | .0303 | 1.34 [.97, 1.84] | .0745 |
| Perceived stress (n = 1,025) | ||||
| Below average | Ref | Ref | ||
| Average | .99 [.92, 1.07] | .8277 | 1.67 [1.07, 2.59] | .0231 |
| Above average | .95 [.85, 1.05] | .2969 | 2.00 [1.14, 3.53] | .0163 |
| Pregnancy-specific anxiety (n = 1,046) | ||||
| Below average | Ref | Ref | ||
| Average | 1.02 [.95, 1.09] | .5918 | 1.06 [.72, 1.56] | .7660 |
| Above average | 1.02 [.93, 1.13] | .6423 | 1.41 [.86, 2.31] | .1783 |
| Prenatal distress (n = 997) | ||||
| Below average | ref | Ref | ||
| Average | 1.09 [1.01, 1.18] | .0374 | .98 [.65, 1.48] | .9286 |
| Above average | 1.05 [.94, 1.16] | .4004 | 1.09 [.65, 1.82] | .7372 |
| Perceived neighborhood safety (n = 940) | ||||
| Below average | 1.04 [.95, 1.14] | .3995 | 1.36 [.87, 2.13] | .1717 |
| Average | 1.04 [.97, 1.12] | .2814 | .89 [.62, 1.27] | .5128 |
| Above average | Ref | Ref | ||
| Housing instability (n = 888)* | ||||
| <2 moves | Ref | Ref | ||
| ≥2 moves | .89 [.82, 0.98] | .0177 | 1.08 [.66, 1.78] | .7517 |
| Housing issues (n = 919)* | ||||
| <2 housing issues | Ref | Ref | ||
| ≥2 housing issues | .87 [.79, 0.95] | .0027 | 1.32 [.81, 2.15] | .2590 |
| Food security (n = 997) | ||||
| Secure | Ref | Ref | ||
| Marginally secure | 1.05 [.97, 1.13] | .2390 | .89 [.58, 1.39] | .6238 |
| Insecure | .95 [.88, 1.02] | .1631 | 1.40 [.99, 1.98] | .0539 |
| Number of life stressors in pregnancy* (n = 918) | ||||
| <2 life stressors | Ref | Ref | ||
| ≥2 life stressors | .92 [.86, .98] | .0090 | 1.34 [.93, 1.92] | .1194 |
| Everyday discrimination (n = 1,040) | .98 [.93, 1.04] | .5122 | 1.20 [.92, 1.58] | .1744 |
| Family support (n = 872) | ||||
| Lower support | .97 [.89, 1.05] | .4325 | 1.33 [.91, 1.94] | .1452 |
| Higher support | Ref | Ref | ||
| Partner support *(n = 887) | ||||
| Lower support | .99 [.94, 1.06] | .8436 | .79 [.56, 1.15] | .2236 |
| Higher support | Ref | Ref | ||
| Health and health behavior | ||||
| Gestational diabetes (n = 1,068) | ||||
| No | Ref | Ref | ||
| Yes | .78 [.69, .88] | <.00001 | 1.28 [.76, 2.17] | .3515 |
| Chronic hypertension (n = 1,068) | ||||
| No | Ref | Ref | ||
| Yes | .88 [.81, .98] | .0171 | 1.49 [.97, 2.28] | .0698 |
| Gestational hypertension (n = 1,068) | ||||
| No | Ref | Ref | ||
| Yes | 1.04 [.97, 1.12] | .2585 | .85 [.57, 1.27] | .4271 |
| Cigarette smoking (n = 1,033) | ||||
| None | Ref | Ref | ||
| Smoked in the 3 months prior | .98 [.91, 1.06] | .5972 | 1.38 [.94, 2.01] | .1009 |
| Any smoking during pregnancy | 0.90 [0.83, 0.98] | .0154 | 1.39 [.95, 2.05] | .0896 |
| Alcohol use during pregnancy (n = 1,036) | ||||
| None | Ref | Ref | ||
| Any | −.94 [.81, 1.08] | .3742 | 1.06 [.53, 2.14] | .8688 |
| Marijuana use during pregnancy (n = 984) | ||||
| None | Ref | Ref | ||
| Any | .90 [.79, 1.03] | .1387 | 1.02 [.53, 2.02] | .9344 |
Abbreviations: CI, confidence interval; OR, odds ratio; RR, relative risk.
Note: Bolding indicates statistically significant p values < .05.
Measured at Survey 2 in the third trimester.
Table 3.
Adjusted Likelihood of Group Prenatal Care Session Attendance (Count Model) and of Attending Zero Sessions (Zero-Inflated Model)
| Patient Characteristics | Count* |
Zero Inflated* |
||
|---|---|---|---|---|
| aRR [95% CI] | p Value | aOR [95% CI] | p Value | |
| Sociodemographic characteristics | ||||
| Age (n = 1,068) | 1.01 [1.00, 1.01] | .012 | .97 [.94, .99] | .016 |
| Race and ethnicity (n = 1,068) | ||||
| Black | 1.05 [.98, 1.12] | .209 | .97 [.70, 1.34] | .864 |
| Hispanic | 1.08 [1.00, 1.16] | .045 | .64 [.42, .96] | .030 |
| White | Ref | Ref | ||
| Other race | 1.23 [.99, 1.54] | .068 | 1.02 [.32, 3.31] | .971 |
| Nativity (n = 1,059) | ||||
| Born in the United States | Ref | Ref | ||
| Born outside the United States | 1.14 [1.06, 1.22] | <.001 | .64 [.42, .97] | .035 |
| Relationship status (n = 993) | ||||
| Single or other | .96 [.88, 1.04] | .273 | 1.21 [.74, 1.97] | .451 |
| Committed dating relationship or engaged | .91 [.85, .97] | .007 | 1.13 [.73, 1.73] | .601 |
| Married | Ref | Ref | ||
| Educational attainment (n = 1,028) | ||||
| Less than high school | .89 [.82, .97] | .007 | 1.38 [.87, 2.16] | .169 |
| High school | .97 [.91, 1.04] | .426 | 1.38 [.92, 2.06] | .119 |
| Above high school | Ref | Ref | ||
| Medicaid eligibility (n = 1,026) | ||||
| Eligible | Ref | Ref | ||
| Ineligible | 1.07 [.93, 1.23] | .349 | 1.19 [.57, 2.51] | .638 |
| Employment (n = 993) | ||||
| Unemployed | .95 [.89, 1.01] | .096 | 1.00 [.72, 1.39] | .990 |
| Working part time | 1.01 [.93, 1.09] | .896 | .67 [.42, 1.05] | .078 |
| Working full time | Ref | Ref | ||
| Parity (n = 1,068) | ||||
| Nulliparous | Ref | Ref | ||
| Primiparous/Multiparous | 1.04 [.97, 1.10] | .136 | .78 [.58, 1.05] | .101 |
| Psychosocial characteristics | ||||
| Initial feelings about pregnancy (n = 1,056) | ||||
| Happy | Ref | Ref | ||
| Unhappy | 1.01 [.92, 1.12] | .819 | .78 [.48, 1.26] | .303 |
| Unsure | 1.07 [.96, 1.19] | .214 | .90 [.54, 1.49] | .689 |
| Symptoms of depression (n = 997) | ||||
| Not reaching clinical concern | Ref | Ref | ||
| Of clinical concern | .93 [.87, .99] | .027 | 1.33 [.95, 1.84] | .084 |
| Perceived stress (n = 1,025) | ||||
| Below average | Ref | Ref | ||
| Average | .99 [.93, 1.07] | .943 | 1.72 [1.09, 2.69] | .019 |
| Above average | .92 [.83, 1.03] | .147 | 2.05 [1.15, 3.64] | .015 |
| Pregnancy-specific anxiety (n = 1,046) | ||||
| Below average | Ref | Ref | ||
| Average | 1.02 [.95, 1.09] | .577 | 1.07 [.73, 1.58] | .725 |
| Above average | 1.04 [.93, 1.13] | .479 | 1.46 [.89, 2.41] | .138 |
| Prenatal distress (n = 997) | ||||
| Below average | Ref | Ref | ||
| Average | 1.09 [1.01, 1.19] | .030 | .97 [.64, 1.47] | .887 |
| Above average | 1.05 [.94, 1.16] | .389 | 1.10 [.66, 1.85] | .708 |
| Perceived neighborhood safety (n = 940) | ||||
| Below average | 1.04 [.95, 1.14] | .387 | 1.32 [.84, 2.08] | .222 |
| Average | 1.06 [.99, 1.14] | .106 | .83 [.57, 1.19] | .321 |
| Above average | Ref | Ref | ||
| Housing instability† (n = 888) | ||||
| <2 moves | Ref | Ref | ||
| ≥2 moves | .92 [.84, .99] | .049 | 1.01 [.61, 1.68] | .958 |
| Housing issues† (n = 919) | ||||
| <2 housing issues | Ref | Ref | ||
| ≥2 housing issues | .88 [.80, .97] | .008 | 1.26 [.77, 2.06] | .351 |
| Food security (n = 997) | ||||
| Secure | Ref | Ref | ||
| Marginally secure | 1.04 [.97, 1.13] | .282 | .92 [.59, 1.43] | .706 |
| Insecure | .96 [.89, 1.04] | .331 | 1.38 [.98, 1.96] | .068 |
| Number of life stressors in pregnancy† (n = 918) | ||||
| <2 life stressors | Ref | Ref | ||
| ≥2 life stressors | .92 [.86, .98] | .010 | 1.33 [.92, 1.92] | .130 |
| Everyday discrimination (n = 1,040) | .98 [.93, 1.04] | .586 | 1.16 [.89, 1.53] | .274 |
| Family support (n = 872) | ||||
| Lower support | .98 [.91, 1.06] | .612 | 1.29 [.89, 1.89] | .194 |
| Higher support | Ref | Ref | ||
| Partner support (n = 887)† | ||||
| Lower support | .99 [.94, 1.05] | .835 | .79 [.55, 1.13] | .199 |
| Higher support | Ref | Ref | ||
| Health and health behavior | ||||
| Gestational diabetes (n = 1,068) | ||||
| No | Ref | Ref | ||
| Yes | .81 [.71, .91] | <.001 | 1.33 [.78, 2.25] | .293 |
| Chronic hypertension (n = 1,068) | ||||
| No | Ref | Ref | ||
| Yes | .94 [.85, 1.03] | .187 | 1.50 [.97, 2.34] | .069 |
| Gestational hypertension (n = 1,068) | ||||
| No | Ref | Ref | ||
| Yes | 1.04 [.97, 1.12] | .267 | .88 [.59, 1.31] | .537 |
| Cigarette smoking (n = 1,033) | ||||
| None | Ref | Ref | ||
| Smoked in the 3 months prior | .96 [.88, 1.04] | .295 | 1.41 [.96, 2.07] | .080 |
| Any smoking during pregnancy | .92 [.85, .99] | .040 | 1.34 [.91, 1.99] | .144 |
| Alcohol use during pregnancy (n = 1,036) | ||||
| None | Ref | Ref | ||
| Any | .88 [.76, 1.02] | .082 | 1.16 [.57, 2.36] | .680 |
| Marijuana use during pregnancy (n = 984) | ||||
| None | Ref | Ref | ||
| Any | .90 [.79, 1.03] | .139 | 1.01 [.51, 1.99] | .974 |
Abbreviations: aOR, adjusted odds ratio; aRR, adjusted relative risk; CI, confidence interval.
Note: Bolding indicates statistically significant p values < .05.
Adjusted for gestational age at study enrollment and gestational age at birth.
Measured at Survey 2 in the third trimester.
Characteristics Associated With Attending Zero Sessions
In the adjusted logit model, older age (adjusted odds ratio [aOR] = .97; 95% CI [.94, .99]; p = .025), Hispanic race and ethnicity group (aOR = .64; 95% CI [.42, .96]; p = .30), and foreign nativity (aOR = .64; 95% CI [.42, .97]; p = .04) were associated with lower odds of attending zero sessions. The psychosocial characteristic of average and above average perceived stress were associated with higher odds of attending zero sessions relative to below average stress (aOR = 1.69; 95% CI [1.08, 2.64]; p = .022; aOR = 1.98; 95% CI [1.11, 3.51]; p = .020).
Post Hoc Analysis
Post hoc analysis of TPNC visit/session attendance among participants assigned to the GPNC arm is presented in Table 4. Older age (aRR = 1.01; 95% CI [1.00, 1.01]; p < .001) and a diagnosis of gestational diabetes (aRR = 1.25; 95% CI [1.17, 1.33]; p < .001) were positively associated with TPNC visits/session attendance. Less than high school and high school education (aRR = .93; 95% CI [.88, .98]; p = .010; aRR = .94; 95% CI [.89, .98]; p = .009), depressive symptoms (aRR = .96; 95% CI [.96, 1.00]; p = .048), and above average perceived stress (aRR = .92; 95% CI [.86, .99]; p = .026) were inversely associated with TPNC visit/session attendance.
Table 4.
Post Hoc Poisson Regression Analysis of Total Prenatal Care Visit/Session Count
| Patient Characteristics | RR [95% CI] | p Value | aRR [95% CI]* | p Value |
|---|---|---|---|---|
| Age (n = 1,068) | .01 [.00, .01] | .008 | 1.01 [1.00, 1.01] | <.001 |
| Nativity (n = 1,059) | ||||
| Born in the United States | Ref | Ref | ||
| Born outside the United States | .99 [.94, 1.04] | .540 | 1.02 [.97, 1.07] | .509 |
| Race and ethnicity (n = 1,068) | ||||
| Black | .98 [.94, 1.02] | .337 | .99 [.95, 1.03] | .6874 |
| Hispanic | .97 [.92, 1.02] | .242 | .99 [.95, 1.05] | .8723 |
| White | Ref | Ref | ||
| Other race | .98 [.83, 1.15] | .760 | 1.02 [.87, 1.19] | .8521 |
| Relationship status (n = 993) | ||||
| Single or Other | .95 [.89, 1.01] | .081 | .97 [.91, 1.02] | .226 |
| Committed dating relationship or engaged | .95 [.90, .99] | .039 | .96 [.91, 1.01] | .082 |
| Married | Ref | Ref | ||
| Educational attainment (n = 1,028) | ||||
| Less than high school | .90 [.85, .96] | <.001 | .93 [.88, .98] | .010 |
| High school | 094 [.89, .99] | .015 | .94 [.89, .98] | .009 |
| Above high school | Ref | Ref | ||
| Symptoms of depression (n = 997) | ||||
| Not reaching clinical concern | Ref | Ref | ||
| Of clinical concern | .96 [.92, .99] | .033 | .96 [.92, 1.00] | .048 |
| Perceived stress (n = 1,025) | ||||
| Below average | Ref | Ref | ||
| Average | .98 [.93, 1.03] | .398 | .98 [.93, 1.03] | .322 |
| Above average | .93 [.87, 1.00] | .053 | .92 [.86, .99] | .026 |
| Prenatal distress (n = 997) | ||||
| Below average | Ref | Ref | ||
| Average | 1.03 [.97, 1.09] | .281 | 1.04 [.98, 1.09] | .197 |
| Above average | 1.03 [.96, 1.10] | .379 | 1.03 [.96, 1.09] | .458 |
| Housing instability † (n = 888) | ||||
| <2 moves | Ref | Ref | ||
| ≥2 moves | .96 [.92, 1.02] | .149 | .98 [.94, 1.06] | .909 |
| Housing issues† (n = 919) | ||||
| <2 housing issues | Ref | Ref | ||
| ≥2 housing issues | .92 [.87, .98] | .007 | .94 [.88, 1.00] | .059 |
| Number of life stressors in pregnancy† (n = 918) | ||||
| <2 life stressors | Ref | Ref | ||
| ≥2 life stressors | .99 [.95, 1.03] | .633 | .99 [.95, 1.04] | .799 |
| Gestational diabetes (n = 1,068) | ||||
| No 0 | Ref | Ref | ||
| Yes 1 | 1.23 [1.16, 1.32] | <.001 | 1.25 [1.17, 1.33] | <.001 |
| Cigarette smoking (n = 1,033) | ||||
| None 0 | Ref | Ref | ||
| Smoked in the 3 months prior 1 | .98 [.93, 1.03] | .476 | .96 [.92, 1.01] | .154 |
| Any smoking during pregnancy 2 | .93 [.88, .98] | .004 | .95 [.90, 1.00] | .059 |
Abbreviations: CI, confidence interval; RR, relative risk; aRR, adjusted relative risk.
Note: Bolding indicates statistically significant p values < .05.
Adjusted for gestational age at study enrollment and gestational age at birth.
Measured at Survey 2 in the third trimester.
Discussion
GPNC represents an alternative to IPNC that shows promise for reducing adverse perinatal health outcomes, particularly among disadvantaged groups. We sought to identify patient characteristics associated with GPNC session attendance among pregnant persons randomly assigned to attend GPNC. Several patient characteristics (older age, Hispanic race and ethnicity group, foreign nativity, and prenatal distress) were positively associated with session attendance. Other patient characteristics (being in a committed relationship but unmarried, less than high school education, depressive symptoms, housing instability, housing issues, life stressors in pregnancy, gestational diabetes, and smoking during pregnancy) were inversely associated with session attendance. Further, younger participants and those with greater perceived stress were more likely to attend zero sessions (i.e., not initiate GPNC), and participants in the Hispanic race and ethnicity group and those with foreign nativity were more likely to initiate GPNC.
In the present study, nearly a quarter of participants assigned to attend GPNC did not attend a single session. Participants attended about half of the sessions they were eligible to attend (an average of 4.8 sessions); those who attended at least one GPNC session attended an average of 6.7 sessions. Another RCT reported similarly high rates of no attendance (22%) and comparable rates of session attendance (5.3 among participants who attended one or more session) (Ickovics et al., 2016). Attendance rates were also largely in keeping with previous studies of patient characteristics and attendance in which participants attended 58% of sessions eligible (an average of 4.3 sessions) (Cunningham et al., 2017), or an average of 6.5 sessions among participants who attended one or more sessions (Earnshaw et al., 2016).
Like our study, which found foreign nativity to be positively associated with session attendance, a study of pregnant adolescents found being born outside the United States was associated with higher rates of GPNC session attendance (Cunningham et al., 2017). One possible explanation for this finding is that the social support offered in GPNC is particularly attractive to pregnant persons who have moved away from larger social support systems in their home countries. For some, GPNC may be consistent with the more collectivist culture in which they were born. We also found an association between being in the Hispanic race and ethnicity group and GPNC initiation and session attendance. Groups were offered in both English and Spanish, possibly attracting Spanish-speaking patients of Hispanic ethnicity. By contrast, patients of younger age, those with less than a high school education, and those who were in a committed relationship but unmarried attended fewer GPNC sessions. Young pregnant persons have been targeted for GPNC, given the psychosocial challenges of pregnancy in adolescence (Mazzoni & Carter, 2017); however, it could be that younger patients still in high school encounter scheduling conflicts to attending GPNC due to the fixed session schedule or length of sessions, which could cause longer school absences. Offering sessions in the evening might better meet the needs of these individuals (Pekkala et al., 2020).
Differences in attendance were also seen based on participant stress and depressive symptoms. In keeping with previous findings for GPNC initiation, average prenatal distress was positively associated with session attendance relative to below average prenatal distress (Wagijo et al., 2022). By contrast, participants with average or above average versus below average perceived stress were more likely to attend zero GPNC sessions. Further, depressive symptoms were inversely associated with session attendance. Although commonly co-occurring, previous studies suggest pregnancy-specific distress to be distinct from general distress (Alderdice et al., 2012). Although GPNC has been suggested to reduce stress and depression (Benediktsson et al., 2013; Buultjens et al., 2021), the enhanced education and peer support offered may be more attractive to pregnant persons experiencing pregnancy-related distress rather than general overwhelm and depression.
Unstable housing, housing issues, and multiple life stressors have not been previously identified as factors associated with GPNC attendance. These hardships are likely linked to barriers to GPNC attendance. Indeed, housing instability has been linked to poor prenatal care utilization more generally (Bellerose et al., 2022; DiTosto et al., 2021); however, an association was not found with TPNC visits/sessions in the current sample. Previous studies, including among a subset of the current sample, have suggested scheduling, transportation, and childcare are commonly reported barriers to GPNC attendance (Berman et al., 2020; Francis et al., 2019; Phillippi & Myers, 2013). Although IPNC is not dissimilar, unique challenges may be posed by the structure of GPNC (e.g., longer session length, fixed schedule, and limited chances for makeup sessions). It may be that although GPNC can attenuate some stressors, sufficient resources are necessary for regular attendance (Novick et al., 2012).
Literature suggests higher medical risk, such as having gestational diabetes, is associated with receiving greater than adequate IPNC (Krans et al., 2013). Similarly, our post hoc analysis found a gestational diabetes diagnosis to be positively associated with TPNC visits/sessions; however, diagnosis was inversely associated with GPNC session attendance. This discrepancy suggests pregnant persons with gestational diabetes may experience unique barriers to attending GPNC, such as interference of diabetes-related services (e.g., testing or education) with sessions or a lack of needed gestational diabetes-specific content shared in groups. Some of these barriers might be addressed in a tailored diabetes-specific GPNC model. A recent study piloted GPNC that incorporates diabetes management interventions and demonstrated some improvement when compared with IPNC for diabetes self-care activities (Carter et al., 2022). Smoking during pregnancy was also found to be associated with lower GPNC attendance. It is possible pregnant persons who smoke fear stigma and judgment in a group environment.
Strengths and Limitations
The present study has several strengths. First, the study sample was a relatively large and racially diverse subset of RCT participants from a practice with an established GPNC program. This program has continually been a certified CenteringPregnancy provider since 2009, indicating high fidelity. By examining data collected in an RCT, the present study benefits from race and ethnicity-stratified random assignment. Although we solely examined those assigned to GPNC, participants had an equal chance of being in the IPNC arm. Second, a broad range of sociodemographic, psychosocial, and health factors were measured using validated instruments, and participant attendance was abstracted via medical chart review by trained study staff.
Our study also had limitations. First, our sample was largely Medicaid-eligible pregnant persons from a single practice enrolled in an RCT. Consistent with GPNC eligibility, inclusion was limited to pregnant persons of low medical risk. Fewer than half of eligible participants enrolled in the trial. Therefore, findings may not necessarily be generalizable to other populations. Second, although response rates were relatively high, 78.7% in the GPNC arm (Smith et al., 2023), there are missing data on several survey measures. As participants were approached to complete the second survey at subsequent prenatal appointments, missingness was likely associated with poor prenatal care attendance or transfer of care. Further, incentive at the second time point was tied to the completion of five visits in their assigned model of care. Some participants may have attended five sessions to receive this incentive, when they might not have otherwise done so. Finally, analysis was exploratory in nature and thus should be interpreted cautiously, as the use of multiple independent tests of association could inflate the chance of false positive findings.
Implications for Practice
The current study offers insight for GPNC recruitment, retention, and model adaptation. Findings of lower GPNC attendance among some patient groups—including those with more depressive symptoms, stressors, and diagnosis of gestational diabetes—indicate a need to explore whether some of these groups may prefer IPNC, or whether modified versions of GPNC might better meet their needs. Modifications to the GPNC model that could further facilitate session attendance should be considered, such as coordination of additional medical monitoring with group sessions for persons with gestational diabetes or the availability of makeup groups for missed sessions. Qualitative interviews with groups identified in association with lower session attendance should be conducted (e.g., patients with depressive symptoms, those who are housing insecure, and those who have gestational diabetes or smoke in pregnancy) to identify care preferences and supports that might help them attend more GPNC sessions. Application of adaptive trial designs in the assessment of whether model modifications keep participants engaged in the group may yield further actionable insights.
Aknowledgments
We would like to thank the pregnant persons who participated, as well as research staff involved in data collection. Emily Doherty had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Research reported in this publication was supported by the National Institutes of Health, Eunice Kennedy Shriver National Institute of Child Health and Human Development (R01HD082311). The content is solely the responsibility of the authors.
Biographies
Emily A. Doherty, PhD, is a Postdoctoral Research Fellow in the Department of Rural Health at Oklahoma State University Center of Health Sciences. Her research interests include maternal child health, women’s health, health disparities, and substance use.
Sarah Griffin, PhD, MPH, is a Professor and Graduate Programs Coordinator in the Department of Public Health Sciences at Clemson University. Her research interests include obesity prevention, physical activity, school health, health disparities, community-based intervention effectiveness and population health.
Kathleen B. Cartmell, PhD, MPH, is an Associate Research Professor in the Department of Public Health Sciences at Clemson University. Her research interests include HPV vaccination, COVID vaccination, tobacco cessation, palliative care, mixed methods, dissemination, and implementation science.
Moonseong Heo, PhD, is a Professor in the Department of Public Health Sciences at Clemson University. His research interests include clinical trials, substance use disorder, hepatitis C virus, medication and care adherence pattern analysis, and obesity epidemiology.
Liwei Chen, PhD, MD, MS, MHS, is a Professor in the Department of Epidemiology at University of California Los Angeles Fielding School of Public Health. Her research interests include cardiometabolic disease epidemiology, maternal and child health, and health disparities in cardiometabolic diseases.
Jessica L. Britt, PhD, is the Maternal-Child Health Projects and Grants Coordinator in the OB/GYN Research Division at Prisma Health. Her research interests include maternal child health, placenta, maternal stress, pregnancy, fetal development, and fetal programming.
Amy H. Crockett, MD, MPH, is a specialist in Maternal Fetal Medicine and professor of Obstetrics and Gynecology at Prisma Health/University of South Carolina Greenville. Her research interests include outpatient obstetrics, prevention of preterm birth, and racial equity in birth outcomes.
Footnotes
Emily A. Doherty is now in the Department of Rural Health, Oklahoma State University Center for Health Sciences, Tulsa, OK 74107, USA.
CRediT authorship contribution statement
Emily A. Doherty: Writing – review & editing, Writing – original draft, Formal analysis, Conceptualization. Sarah Griffin: Writing – review & editing, Supervision. Kathleen B. Cartmell: Writing – review & editing, Supervision. Moonseong Heo: Writing – review & editing, Supervision, Funding acquisition. Liwei Chen: Writing – review & editing, Supervision, Funding acquisition. Jessica L. Britt: Writing – review & editing, Data curation. Amy H. Crockett: Writing – review & editing, Supervision, Funding acquisition.
The authors report no conflicts of interest.
References
- Alderdice F, Lynn F, & Lobel M (2012). A review and psychometric evaluation of pregnancy-specific stress measures. Journal of Psychosomatic Obstetrics & Gynecology, 33(2), 62–77. 10.3109/0167482X.2012.673040 [DOI] [PubMed] [Google Scholar]
- Bellerose M, Rodriguez M, & Vivier PM (2022). A systematic review of the qualitative literature on barriers to high-quality prenatal and postpartum care among low-income women. Health Services Research, 57(4), 775–785. 10.1111/1475-6773.14008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Benediktsson I, McDonald SW, Vekved M, McNeil DA, Dolan SM, & Tough SC (2013). Comparing CenteringPregnancy® to standard prenatal care plus prenatal education. BMC Pregnancy and Childbirth, 13(Suppl 1), S5. 10.1186/1471-2393-13-S1-S5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Berman R, Weber Yorga K, & Sheeder J (2020). Intention to participate in group prenatal care: Moving beyond yes or no. Health Promotion Practice, 21 (1), 123–132. 10.1177/1524839918784943 [DOI] [PubMed] [Google Scholar]
- Buultjens M, Farouque A, Karimi L, Whitby L, Milgrom J, & Erbas B (2021). The contribution of group prenatal care to maternal psychological health outcomes: A systematic review. Women and Birth, 34(6), e631–e642. 10.1016/j.wombi.2020.12.004 [DOI] [PubMed] [Google Scholar]
- Carrion BV, Earnshaw VA, Kershaw T, Lewis JB, Stasko EC, Tobin JN, & Ickovics JR (2015). Housing instability and birth weight among young urban mothers. Journal of Urban Health, 92(1), 1–9. 10.1007/s11524-014-9913-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Carter EB, Barbier K, Hill PK, Cahill AG, Colditz GA, Macones GA, Tuuli MG, & Mazzoni SE (2022). Pilot randomized controlled trial of diabetes group prenatal care. American Journal of Perinatology, 39(01), 045–053. 10.1055/s-0040-1714209 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Carter EB, & Mazzoni SE (2021). A paradigm shift to address racial inequities in perinatal healthcare. American Journal of Obstetrics and Gynecology, 224 (4), 359–361. 10.1016/j.ajog.2020.11.040 [DOI] [PubMed] [Google Scholar]
- Carter EB, Thayer SM, Paul R, Barry VG, Iqbal SN, Ehrenberg S, Doering M, Mazzoni SE, Frolova AI, Kelly JC, Raghuraman N, & Debbink MP (2024). Diabetes group prenatal care: A systematic review and meta-analysis. Obstetrics & Gynecology, 144(5), 621–632. 10.1097/AOG.0000000000005442 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Centers for Disease Control and Prevention. (2014). PRAMS Questionnaires. https://www.cdc.gov/prams/Questionnaire.htm. [Google Scholar]
- Chen L, Crockett AH, Covington-Kolb S, Heberlein E, Zhang L, & Sun X (2017). Centering and Racial Disparities (CRADLE study): Rationale and design of a randomized controlled trial of centering pregnancy and birth outcomes. BMC Pregnancy and Childbirth, 17(1), 118. 10.1186/s12884-017-1295-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chen Y, Crockett AH, Britt JL, Zhang L, Nianogo RA, Qian T, Nan B, & Chen L (2023). Group vs individual prenatal care and gestational diabetes outcomes: A secondary analysis of a randomized clinical trial. JAMA Network Open, 6(8): Article e2330763. 10.1001/jamanetworkopen.2023.30763 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cohen S, & Hoberman HM (1983). Positive events and social supports as buffers of life change stress. Journal of Applied Social Psychology, 13(2), 99–125. 10.1111/j.1559-1816.1983.tb02325.x [DOI] [Google Scholar]
- County of Los Angeles. (2016). The Los Angeles Mommy & Baby Project: Healthy mommies and healthy babies. County of Los Angeles Public Health. https://publichealth.lacounty.gov/mch/lamb/lambsurvey.html. [Google Scholar]
- Crockett AH, Chen L, Heberlein EC, Britt JL, Covington-Kolb S, Witrick B, Doherty E, Zhang L, Borders A, Keenan-Devlin L, Smart B, & Heo M (2022). Group vs traditional prenatal care for improving racial equity in preterm birth and low birthweight: The Centering and Racial Disparities randomized clinical trial study. American Journal of Obstetrics and Gynecology, 227(6), 893.e1–893.e15. 10.1016/j.ajog.2022.06.066 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Crockett AH, Heberlein EC, Smith JC, Ozluk P, Covington-Kolb S, & Willis C (2019). Effects of a multi-site expansion of group prenatal care on birth outcomes. Maternal and Child Health Journal, 23(10), 1424–1433. 10.1007/s10995-019-02795-4 [DOI] [PubMed] [Google Scholar]
- Cunningham SD, Grilo S, Lewis JB, Novick G, Rising SS, Tobin JN, & Ickovics JR (2017). Group prenatal care attendance: Determinants and relationship with care satisfaction. Maternal and Child Health Journal, 21(4), 770–776. 10.1007/s10995-016-2161-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cunningham SD, Sutherland RA, Yee CW, Thomas JL, Monin JK, Ickovics JR, & Lewis JB (2021). Group medical care: A systematic review of health service performance. International Journal of Environmental Research and Public Health, 18(23): Article 12726. 10.3390/ijerph182312726 [DOI] [PMC free article] [PubMed] [Google Scholar]
- DiTosto JD, Holder K, Soyemi E, Beestrum M, & Yee LM (2021). Housing instability and adverse perinatal outcomes: A systematic review. American Journal of Obstetrics & Gynecology MFM, 3(6): Article 100477. 10.1016/j.ajogmf.2021.100477 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Earnshaw VA, Rosenthal L, Cunningham SD, Kershaw T, Lewis J, Rising SS, Stasko E, Tobin J, & Ickovics JR (2016). Exploring group composition among young, urban women of color in prenatal care: Implications for satisfaction, engagement, and group attendance. Women’s Health Issues, 26(1), 110–115. 10.1016/j.whi.2015.09.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Economic Research Service. (2012). Survey tools. https://www.ers.usda.gov/topics/food-nutrition-assistance/food-security-in-the-u-s/survey-tools/.
- Francis E, Johnstone MB, Convington-Kolb S, Witrick B, Griffin SF, Sun X, Crockett A, & Chen L (2019). Group prenatal care attendance and women’s characteristics associated with low attendance: Results from Centering and Racial Disparities (CRADLE Study). Maternal and Child Health Journal, 23(10), 1371–1381. 10.1007/s10995-019-02784-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ghosh JKC, Wilhelm MH, Dunkel-Schetter C, Lombardi CA, & Ritz BR (2010). Paternal support and preterm birth, and the moderation of effects of chronic stress: A study in Los Angeles County mothers. Archives of Women’s Mental Health, 13(4), 327–338. 10.1007/s00737-009-0135-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hahn-Holbrook J, Schetter CD, Arora C, & Hobel CJ (2013). Placental corticotropin-releasing hormone mediates the association between prenatal social support and postpartum depression. Clinical Psychological Science: A Journal of the Association for Psychological Science, 1(3), 253–264. 10.1177/2167702612470646 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ickovics JR, Earnshaw V, Lewis JB, Kershaw TS, Magriples U, Stasko E, Rising SS, Cassells A, Cunningham S, Bernstein P, & Tobin JN (2016). Cluster randomized controlled trial of group prenatal care: Perinatal outcomes among adolescents in New York City Health Centers. American Journal of Public Health, 106(2), 359–365. 10.2105/AJPH.2015.302960 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ickovics JR, Kershaw TS, Westdahl C, Magriples U, Massey Z, Reynolds H, & Rising SS (2007). Group prenatal care and perinatal outcomes: A randomized controlled trial. Obstetrics & Gynecology, 110(2), 330–339. 10.1097/01.AOG.0000275284.24298.23 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ickovics JR, Reed E, Magriples U, Westdahl C, Schindler Rising S, & Kershaw TS (2011). Effects of group prenatal care on psychosocial risk in pregnancy: Results from a randomised controlled trial. Psychology & Health, 26(2), 235–250. 10.1080/08870446.2011.531577 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Krans EE, Davis MM, & Palladino CL (2013). Disparate patterns of prenatal care utilization stratified by medical and psychosocial risk. Maternal and Child Health Journal, 17(4), 639–645. 10.1007/s10995-012-1040-9 [DOI] [PubMed] [Google Scholar]
- Liu Y, Wang Y, Wu Y, Chen X, & Bai J (2021). Effectiveness of the Centering Pregnancy program on maternal and birth outcomes: A systematic review and meta-analysis. International Journal of Nursing Studies, 120: Article 103981. 10.1016/j.ijnurstu.2021.103981 [DOI] [PubMed] [Google Scholar]
- Lobel M, Cannella DL, Graham JE, DeVincent C, Schneider J, & Meyer BA (2008). Pregnancy-specific stress, prenatal health behaviors, and birth outcomes. Health Psychology, 27(5), 604–615. 10.1037/a0013242 [DOI] [PubMed] [Google Scholar]
- Lobel M, Dunkel-Schetter C, & Scrimshaw SC (1992). Prenatal maternal stress and prematurity: A prospective study of socioeconomically disadvantaged women. Health Psychology, 11(1), 32–40. 10.1037/0278-6133.11.1.32 [DOI] [PubMed] [Google Scholar]
- Mazzoni SE, & Carter EB (2017). Group prenatal care. American Journal of Obstetrics and Gynecology, 216(6), 552–556. 10.1016/j.ajog.2017.02.006 [DOI] [PubMed] [Google Scholar]
- McDonald SD, Sword W, Eryuzlu LN, Neupane B, Beyene J, & Biringer AB (2016). Why are half of women interested in participating in group prenatal care? Maternal and Child Health Journal, 20(1), 97–105. 10.1007/s10995-015-1807-x [DOI] [PubMed] [Google Scholar]
- Novick G, Sadler LS, Knafl KA, Ellen Groce N, & Kennedy HP (2012). The intersection of everyday life and group prenatal care for women in two urban clinics. Journal of Health Care for the Poor and Underserved, 23(2), 589–603. 10.1353/hpu.2012.0060 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Pekkala J, Cross-Barnet C, Kirkegaard M, Silow-Carroll S, Courtot B, & Hill I (2020). Key considerations for implementing group prenatal care: Lessons from 60 practices. Journal of Midwifery & Women’s Health, 65(2), 208–215. 10.1111/jmwh.13047 [DOI] [PubMed] [Google Scholar]
- Phillippi JC, & Myers CR (2013). Reasons women in appalachia decline centering pregnancy care. Journal of Midwifery & Women’s Health, 58(5), 516–522. 10.1111/jmwh.12033 [DOI] [PubMed] [Google Scholar]
- Radloff LS (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1(3), 385–401. 10.1177/014662167700100306 [DOI] [Google Scholar]
- Rising S. (1998). Centering pregnancy: An interdisciplinary model of empowerment. Journal of Nurse-Midwifery, 43(1), 46–54. 10.1016/S0091-2182(97)00117-1 [DOI] [PubMed] [Google Scholar]
- Roesch SC, Schetter CD, Woo G, & Hobel CJ (2004). Modeling the types and timing of stress in pregnancy. Anxiety, Stress & Coping, 17(1), 87–102. 10.1080/1061580031000123667 [DOI] [Google Scholar]
- Schindler Rising S, Kennedy HP, & Klima CS (2004). Redesigning prenatal care through centering pregnancy. Journal of Midwifery & Women’s Health, 49(5), 398–404. 10.1016/j.jmwh.2004.04.018 [DOI] [PubMed] [Google Scholar]
- Sheeder J, Weber Yorga K, & Kabir-Greher K (2012). A review of prenatal group care literature: The need for a structured theoretical framework and systematic evaluation. Maternal and Child Health Journal, 16(1), 177–187. 10.1007/s10995-010-0709-1 [DOI] [PubMed] [Google Scholar]
- Smith JC, Heberlein EC, Domingue A, LaBoy A, Britt J, & Crockett AH (2023). Randomized controlled trial on the effect of group versus individual prenatal care on psychosocial outcomes. Journal of Obstetric, Gynecologic & Neonatal Nursing, 52(6), 467–480. 10.1016/j.jogn.2023.07.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tucker CM, Felder TM, Dail RB, Lyndon A, & Allen K-C (2021). Group prenatal care and maternal outcomes: A scoping review. MCN: The American Journal of Maternal/Child Nursing, 46(6), 314–322. 10.1097/NMC.0000000000000766 [DOI] [PubMed] [Google Scholar]
- Turner RJ, Grindstaff CF, & Phillips N (1990). Social support and outcome in teenage pregnancy. Journal of Health and Social Behavior, 31(1), 43–57. 10.2307/2137044 [DOI] [PubMed] [Google Scholar]
- Wagijo MR, Crone MR, Van Zwicht BS, Van Lith JMM, Schindler Rising S, & Rijnders MEB (2022). Centering pregnancy in the Netherlands: Who engages, who doesn’t, and why. Birth, 49(2), 329–340. 10.1111/birt.12610 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wallman KK, & Hodgdon J (1977). Race and ethnic standards for federal statistics and administrative reporting. Statistical Reporter/Office of Federal Statistical Policy and Standards, (77–110), 450–454. [PubMed] [Google Scholar]
- Weber Yorga KD, & Sheeder JL (2015). Which pregnant adolescents would be interested in group-based care, and why? Journal of Pediatric and Adolescent Gynecology, 28(6), 508–515. 10.1016/j.jpag.2015.03.006 [DOI] [PubMed] [Google Scholar]
- Williams DR, Yan Yu, Jackson JS, & Anderson NB (1997). Racial differences in physical and mental health: Socio-economic status, stress and discrimination. Journal of Health Psychology, 2(3), 335–351. 10.1177/135910539700200305 [DOI] [PubMed] [Google Scholar]