Abstract
In 1992, the American Academy of Pediatrics (AAP) recommended that infants be placed in the supine position to sleep. The Safe to Sleep campaign (formerly known as Back to Sleep) began in 1994 to promote risk reduction strategies for sleep-related infant death. In subsequent years, a 53% decline in deaths attributed to sudden infant death syndrome (SIDS) was reported. 1 The prevalence of prone positioning in the United States, as assessed with ongoing national sampling, decreased from 70% in 1992 to 11% in 2002 and increased slightly to 13% in 2005. 2 Despite these improvements, SIDS remains the third leading causes of infant death. The Safe to Sleep campaign alone could not address racial disparities in reducing SIDS.The incidence of sudden infant death is 2 to 3 times higher in the black population when compared with the US population as a whole.1-3
Health care and community social service settings have the opportunity to be a primary source of information for pregnant women of all races as previous research reported low-income women rely heavily on interpersonal sources for health care information. 4 Safe sleep attitudes and knowledge should be assessed initially to facilitate infant safe sleep communication and identify potential discrepancies. The objective of this study was to use data collected for larger projects in a secondary data analysis to compare self-reported safe sleep data from a community-based social service agency setting and a primary care setting. We hypothesized that respondents from the primary care setting would be more likely to report safe sleep practices when compared with the community-based cohort. Analyzing the community wide assessment of safe sleep practices will help enable focused future interventions to maximize impact on a community scale.5,6
Methods
Community Agency Survey
The community social service agency facilitated a county-based nutrition education and supplemental food program, including nutritional education/referral services and financial support for healthy foods. This component of the study used a nonrandomized cross-sectional survey. Participants shared a couple of high-risk characteristics, including unemployment and low-income. Community safe sleep data were collected using a Pregnancy Risk Assessment Monitoring System (PRAMS)–based questionnaire. Safe sleep questions included the following: “In which one position do you most often lay your baby down to sleep?” and “How often does your new baby sleep in the same bed with you or anyone else?” Surveys were mailed to all clients who were new mothers, prior to first appointment, for a period of 6 months, and were available in English and Spanish. Implied consent was obtained on surveys. The surveys were returned via mail or in-person by the client. The survey results were not reviewed by community social service agency staff members and to ensure client anonymity, the surveys never included identifying information. Survey data were presented descriptively.
Primary Care Center Survey
The primary care centers were 2 academic family medicine residency clinics which offered typical primary care services for patients of all ages, specifically serving low-income, at-risk families in the community. This component of the study also used a nonrandomized cross-sectional survey. Safe sleep checklists were given to mothers of patients under the age of 6 months as part of a health care provider safe sleep pilot study in both primary care centers. Checklists were given by the front desk when mothers checked in for their infants’ appointments. Implied consent was obtained on the survey. Mothers completed the checklist and it was collected and reviewed by the nurse or physician during the appointment to inform providers about patients’ intentions and behaviors regarding safe sleep as well as to act as a catalyst for discussion of infant safe sleep. Providers then delivered safe sleep counseling, at minimum, for those responses not following the AAP’s recommendations. Checklists were collected weekly from the clinics by a research assistant.
Statistical Analysis
The Statistical Package for the Social Sciences (IBM, Armonk, NY) was used for analysis. All variables were presented using descriptive statistics. The proportion of safe sleep environments was compared using a chi-square with Yate’s correction and Fisher’s exact test. Two binary logistic regressions were performed to examine the association between demographic differences and safe sleep position/bed sharing. Covariates were entered into the model if they had P value of less than .2. For parsimony, indicator variables were created for race (white as reference group) and education (college educated as reference group).The study protocols were approved by 2 academic institutional review boards and the clinical sites’ institutional review boards.
Results
Community Agency Survey
There were 166 PRAMS survey respondents and age ranged from 15 to 43 years. The majority of women had an income less than $26 000 (82%). Average family size, including parents, was reported as 3 (SD = 1.6) and most women reported living in an urban setting (n = 141, 85%; Table 1).
Frequency Distribution of Demographic Variables.
Missing data in the community group for race (n = 16), ethnicity (n = 8), education (n = 8), and age (n = 7); missing data in the primary care center group for race (n = 28), ethnicity (n = 25), education (n = 25), and age (n = 24).
Eighty-nine percent (n = 148) of women responded to survey questions regarding safe sleep. Of these, most of the women (n = 126, 85%) reported they lay their baby on his or her back while sleeping, 10% (n = 14) on the side and 5% (n = 8) on the stomach. Sixty-four percent (n = 94) of women reported their baby never or rarely sleeps with them/anyone else in the same bed, 30% (n = 45) reported often or sometimes bed sharing, and 6% (n = 9) reported always bed sharing. One third of women (33%, n = 49) reported never bed sharing and placing their baby on his or her back while sleeping, consistent with the AAP Safe Sleep recommendations 1 (Table 2).
Safe Sleep Surveys.
χ2(1) = 0.79, P = .372.
Fisher’s exact test, P< .001.
Overall safe sleep was defined as never bed sharing and placing their baby on his or her back while sleeping (community agency) and safe position, safe location, both recommended items (firm mattress and fitted sheet), and no unsafe items (loose blankets, bumper, pillows, stuffed toys, etc) (primary care center).
Primary Care Center Survey
There were 79 safe sleep checklists completed by mothers. Age ranged from 16 to 41 years. Number of children ranged from 1 to 5, with the majority reporting 3 or less children (n = 48, 61%).
Assessment of Maternal Responses to Checklist Items (Prior to Appointment)
As shown in Table 2, 85% of participants were aware of safe location; 79% were aware of placing infant supine, 44% reported no unsafe items in the bed, and 63% identified having a firm mattress and fitted sheet in the crib. Overall safe sleep, reporting compliance with all the above recommendations, was reported by 20% of participants.
Comparison of Community Agency and Primary Care Center Results
Demographic Differences
Participants from both locations were similar in Hispanic ethnicity (n = 36, 23% community, n = 15, 28% primary care center)—χ2(1) = 0.31, P = .578—and age (25 years, SD = 5.4, community; 24 years, SD = 5.4, primary care center)—t(212) = 0.95, P = .346. A larger proportion of community agency survey participants were white (n = 101, 67%) compared with 49% (n = 25) of primary care center participants, though the difference was determined not to be statistically significant—χ2(2) = 5.46, P = .065. Similarly, education exhibited a nonsignificant difference with primary care center participants reporting greater proportions of high school graduates (n = 49, 91%) compared with community agency survey participants (n = 126, 80%)—χ2(1) = 2.66, P = .103.
Sleep Items Responses
Safe sleep position responses (supine) did not differ significantly between the community (n = 126, 85%) and primary care center (n = 62, 79%) participants—χ2(1) = 0.79, P = .372. However, reported bed sharing was significantly higher in the community sample (n = 54, 36%) than the primary care center sample (n = 4, 5%; Fisher’s exact test, P< .001; Table 2).
Adjusting for Demographic Differences
Black and other race were identified as covariate predictors of unsafe sleep positioning. The model correctly classified 83.3% of cases with a c-index of 0.633. Adjusting for covariates, there was still no significant difference between community and primary care centers in sleep positioning. Black race was determined to be significantly associated with increased unsafe sleep positioning (odds ratio [OR] = 2.86, P = .022; Table 3).
Logistic Regression Models.
The logistic regression model for bed sharing was adjusted for less than high school education in addition to study cohort. The model correctly classified 71.8% of cases with a c-index of 0.688. The community center cohort continued to be the only significant predictor of bed sharing (OR = 25.40, P = .002; Table 3).
Discussion
Our findings suggest most parents, regardless of source of information, use supine positioning for their infant. This supports previous findings and is not surprising as the Back to Sleep/Safe to Sleep campaign is celebrating its 20th year. 7 However, when comparing the community agency and primary care center participants, 30% of respondents in the community setting (compared with 5% in primary care location) reported bed sharing. There are several possible reasons for this. First, mothers may report better compliance with safe sleep guidelines than actual behaviors implemented at home. Chung-Park 8 determined 85% of patients reported that laying infants on their back was the safest sleep position, but only 69% practiced this. Second, safe sleep habits are highly influenced by cultural factors, which could be changed through a comprehensive community-based effort. 9 For example, bed sharing is common practice among Black mothers, though race was only associated with unsafe sleep positioning in this study.10-12 Reasons for bed sharing include convenient nighttime feedings and parental vigilance toward SIDS prevention. 12 Furthermore, barriers to placing an infant in the supine position for sleep include a lack of trust with physicians, fear of infant choking in supine position and lack of appropriate advice. 13 Findings from the logistic analyses support previous research and further highlight the need for targeted interventions toward black and less educated populations. The continued importance of infant sleep location was indicated in the 2014 Kansas Child Death Review Board report, as 64% (n = 16/25) of SIDS deaths occurred in the adult bed (based on 2012 data). 14
Parents may choose to follow instructions from non–health care professionals because they find physicians’ messages to be conflicting, confusing, or lacking expertise.15,16 Research suggests that consistent communication on safe sleep conditions might help in building the trust between patients and community service/primary care providers. Smith et al. 17 concluded that high maternal ratings of physician qualifications on safe sleep advice were associated with increased safe sleep practices by mothers. One study used a trained health educator to deliver a 15-minute educational intervention to groups (n = 3-10) of parents in Washington. The in-person intervention resulted in 75% of infants placed on their back to sleep. 18 Direct, interactive educational interventions may be more effective in reducing SIDS. However, before comprehensive community interventions are implemented, additional baseline data are needed. 19
Collaborative advice by community social service agencies and staff at primary care centers supportive of exclusively supine position could increase the chance that parents and caregivers will place the infant supine for sleep. There has been a dose-dependent effect of advice for exclusively supine sleep reported, which was determined by the amount and source of advice from external resources. 5 Differences in reported safe sleep environments may be due to differences in knowledge of safe sleep guidelines between community- and clinic-based respondents, or clinic-based respondents may have been more likely to provide socially desirable responses due to the setting.
A proactive collaborative community-based safe sleep counseling approach plays an important role in improving safe sleep practices. Previous studies reported, in addition to counseling patients during clinical encounters, evaluating with educational questionnaires proved to be consistent in reducing SIDS. 18 One interventional study with evaluation of mothers by telephone after educative questionnaire during the stay in hospital showed increased awareness and compliance in the follow up of required recommendations. 6 Another reported improvement in physician-patient safe sleep discussions during clinical encounters using a health care educational checklist intervention. 20 These preliminary studies suggest that such interventions may help improve knowledge and behaviors, potentially affecting the risk of SIDS.
Limitations
This study was not without limitations. Results were based on self-reported data. The data were collected for independent projects and therefore data collection instruments, including question wording and variables assessed were not the same. Respondents were not equally matched by race or educational level limiting the strength of the comparative analysis. Finally, the 2 study locations were not in close proximity but were located within the same urban community. Overlap of participants was a possibility not controlled for during data collection of original studies limiting this secondary analysis.
Conclusions
Safe sleep education should begin during pregnancy and be consistently reinforced throughout pregnancy and child’s first year. Consistent verbal counseling and safe sleep messaging with infant caregivers through the safe sleep campaign have had significant impacts in safe sleep behaviors. 7 However, before comprehensive collaborative community action is launched within a community, baseline data are needed. The comparison data from this study further highlighted the need for collaborative effort to improve safe sleep behaviors to reduce SIDS beginning with a systematic assessment of safe sleep practices.
Footnotes
Acknowledgements
The authors would like to thank Matt Engel, MPH, Department of Pediatrics, University of Kansas School of Medicine–Wichita; Christy Schunn, LSCSW, KIDS Network of Kansas; Sugantha Yuvaraj, MBBS, Department of Public Health Science, Wichita State University; Medical Society of Sedgwick County Safe Sleep Taskforce; and Sedgwick County Women, Infants and Children Program.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: A grant from Wichita State University’s Office of Research and Technology Transfer was used to support part of this research.