Implementing the Eat, Sleep, Console Model of Care: A Scoping Review Free

Neonatal abstinence syndrome (NAS) or neonatal opioid withdrawal syndrome (NOWS) is a unique and complex clinical diagnosis that has seen significant growth over the last 2 decades across North America.^1–3 The global opioid epidemic has led to the development of a more specified term, NOWS, which describes infants experiencing withdrawal specifically from opioid exposure, prescription or not, during pregnancy.^4,5 Infants with NAS/NOWS experience a wide variety of withdrawal signs, including irritability, interruptions in sleep, tremors, loose or watery stools, sweating, increased respiratory rate, and a high-pitched cry.^2,4,5 

Increased length of hospital stay and the use of pharmacological treatment are among the most common challenges within the care for infants with NOWS.^1,2,6 The length of hospital stay for this population, as compared with those without NAS/NOWS, is staggering. In Canada, the mean length of hospital stay between 2010 and 2020 averaged between 14.6 and 16.4 days for infants with NAS/NOWS. This average length of stay is 8 times higher compared with infants without NAS/NOWS who averaged a length of stay around 2.09 days between 2008 and 2019.⁷ Filteau et al¹ further characterize the financial impact of length of stay in their 2018 report, demonstrating total hospital costs to have doubled between 2010 and 2014, rising from $15.7 to $26.9 million.

Historically, infants with NAS/NOWS were cared for using the Finnegan Neonatal Abstinence Scoring Tool. The negative patient and health system outcomes associated with the use of the Finnegan Neonatal Abstinence Scoring Tool, paired with the exponential growth of the NAS/NOWS population, has encouraged health care providers to seek out novel, evidence-based models of care such as the Eat, Sleep, Console (ESC) model.^8–10 

The ESC model of care focuses on the functional abilities of the infant rather than the symptoms of withdrawal.¹⁰ The ESC model of care has been successfully implemented in various settings, including large tertiary facilities and smaller community hospitals.^11,12 Moreover, implementation of this model has demonstrated decreases in length of hospital stay and pharmacological treatment for infants, along with decreasing total hospital costs.^9,12–16 Despite the promising outcomes demonstrated with implementing the ESC model of care into clinical settings, little is known about sustainably implementing the model into practice.

The process of implementing evidence-based interventions, such as the ESC model, into multifaceted health systems is complex.¹⁷ Context plays an essential role in implementation, determining whether an intervention is a success or a failure.¹⁸ Understanding the implementation process, including essential components of the ESC model, implementation strategies used, and potential barriers and facilitators, can help to facilitate the spread and scale of the ESC model into practice. Exploring the unique contextual elements of the implementation process, such as strategies, barriers, and facilitators, is critical in developing a theory-informed and systematic approach to sustainable implementation of the model into practice.^19,20 To our knowledge, no high-quality scoping reviews have explored the implementation of the ESC model into clinical practice.

This scoping review aimed to map out available evidence of implementing the ESC model of care into clinical practice. We sought to describe the implementation process, the essential components of the ESC model, and the strategies used. We explored barriers and facilitators to implementation, along with reported impact, implementation, and effectiveness outcomes described in the literature. Finally, we used this evidence synthesis to highlight areas for future research and the next steps in ensuring sustainable implementation of this model into practice to ultimately improve patient and health system outcomes.

Our published protocol²¹ outlining our methodological processes has been previously published (Author 1 Protocol Publication). Below is an overview of our methods and any changes made to the published protocol.

The overarching objective of this scoping review was to explore and characterize the landscape of implementation and evaluation research completed on the ESC model of care in clinical practice. We conducted our scoping review with the guidance of the JBI Methodology for Scoping Reviews²² and Arksey and O’Malley’s Scoping Review Framework.²³ Our reporting confirms to the Preferred Reporting Items for Systematic Reviews and Meta-analyses Extension for Scoping Reviews (PRISMA-ScR), which can be viewed in Supplemental Material 2.²⁴ The research questions were the following:

1. How has the ESC model been implemented and evaluated in practice?

   • 1.1 What strategies have been used to implement the ESC model of care into practice?

   • 1.2 What are the reported barriers and facilitators to implementing the ESC model of care?

   • 1.3 What are the reported measures and outcomes?

We completed our searches in collaboration with a health science librarian (MR). The search strategy was peer reviewed by a second librarian using the peer review of electronic search strategies template.²⁵ We completed systematic searches of the following databases in September 2023: MEDLINE (Ovid), Embase (Elsevier), CINAHL (EBSCO), PsycINFO (EBSCO), and Google Scholar (see Supplemental Material 3 for all search strategies). We completed a gray literature search in February 2024 following Godin’s methodology for gray literature searching.²⁶ Sources included ProQuest dissertations and theses and open access theses and dissertations. Additionally, we consulted the Canadian Agency for Drugs and Technologies in Health Checklist Grey Matters: A Practical Tool for Searching Health-Related Grey Literature for relevant sources. We searched 56 of the suggested resources with only 1 result found, which was a duplicate from the database searches.²⁷ 

The detailed inclusion and exclusion are in Supplemental Material 1, reflective of our published protocol.²¹ We made the following alterations to the study protocol due to emerging reflections throughout the search and screening processes. First, we altered the exclusion criteria to include infants aged over and including 35 weeks’ gestation because many studies included well infants aged under 37 weeks (initial exclusion). Content experts determined that the confounding element of prematurity would be more prevalent with newborns aged less than 35 weeks. Second, we included commentaries if the article spoke to implementation or evaluation of the model in clinical practice because we noted a trend in commentaries encompassing key implementation reflections and considerations.

We collated all identified citations into Covidence, where 200 duplicates were automatically removed. Initial database searches, including gray literature databases, yielded 853 articles (MEDLINE: 96; Embase: 25; CINAHL: 51; PsychINFO: 13; Google Scholar: 126; Scopus: 688; ProQuest theses and dissertations: 51; open access theses and dissertations: 1; citation searching: 2). Two independent reviewers screened title and abstracts according to inclusion criteria. One hundred seventeen studies met initial inclusion criteria and were included for full-text review. Two independent reviewers reviewed the full text of articles/documents against detailed inclusion and exclusion criteria (found in Supplemental Material 1). At final assessment, 29 studies remained for data charting. Google searches yielded 5 articles/documents for data charting. Figure 1 demonstrates our PRISMA-ScR flow diagram created from Covidence software (adapted to include results from gray literature).

Two independent reviewers extracted data from all articles based on a predetermined data charting tool. We extracted data on basic article characteristics, including article citation, research or reporting design, implementation setting, and geographical location. Next, we extracted data describing the intervention and the implementation or evaluation processes, including intervention description, implementation strategies used, barriers and facilitators to implementation, and reported outcomes (impact, implementation and effectiveness). Finally, we extracted evidence of equity, collaboration, or intersectional considerations.

We used the Consolidated Framework for Implementation Research (CFIR),²⁸ Expert Recommendations for Implementing Change (ERIC) taxonomy (clustered categories),²⁹ Implementation Outcome,¹⁹ and Bick and Graham’s Evidence-Based Practice Measures³⁰ frameworks to code the data through a directed content analysis and to thematically organize findings. The CFIR and implementation outcome frameworks were added after the protocol for their ability to explore and characterize foundational elements of the implementation process.^19,28 We presented data in diagram and tabular form with narrative summaries to describe the results in the lens of our overarching review questions and objectives.

We identified 34 articles and documents in the final review. Supplemental material highlights key article characteristics and implementation processes (Supplemental Material 1). The publication dates ranged from 2017 to 2023. Most studies were qualitative improvement or retrospective pre/postimplementation publications (n = 22; 63%), followed by commentaries (n = 6; 17%), health system reports (n = 2; 6%), qualitative descriptive studies (n = 2; 6%), program evaluation (n = 1; 3%), secondary analysis (n = 1: 3%), and randomized control trial (n = 1: 3%). The United States was the primary country of origin (97%) with only 1 included document from Canada (3%). The settings for the included literature were largely underreported (40%), with the majority of those reported from urban settings (26%), followed by rural (20%), then mixed urban and rural settings (14%). The results are organized in the following sections according to scoping review objectives.

The most reported implementation strategy category was “train and educate stakeholders” (31.6%), followed by “develop stakeholder relationships” (22.4%). The “train and educate stakeholders” category included strategies such as building an education curriculum,¹² in-person education sessions such as grand rounds,^12,31 online education through modules, webinars and video training,^11,32–34 and interprofessional education (including families).^34–36 The “develop stakeholder relationships” category included strategies such as a multidisciplinary task force or network established^{10,13,38–40} and community outreach and engagement.^14,41,Supplemental Material 1 includes a detailed list of all categories, strategies used, and associated citations. Figure 2 demonstrates the reported strategies mapped onto Waltz et al²⁹ categorization of the ERIC taxonomy. The only strategies not reported were found in the “provide interactive assistance” category. Most studies used a variety of strategies, with 29.4% reporting the use of 1 to 2 strategies, 41.2% reporting 3 to 4 strategies, and 8.8% reporting 5 or more strategies; however, 20.6% of studies did not report specific implementation strategies.

The implementation process was largely underreported with minimal details provided for most studies. Supplemental Material 1 provides narrative summaries of the implementation processes used for each study. Most studies reported using a multidisciplinary task force to guide the implementation process (n = 12), followed by studies using a combination of a plan-do-study-act methodology and multidisciplinary teams (n = 5). Only 6 studies explicitly named a framework that guided their implementation process, including the use of the following: (1) IOWA Model,⁴⁰ (2) W.K. Kellogg Foundation Logic Model,⁴² (3) Roger’s Diffusion,³⁵ (4) Quality Enhance Research Initiative Framework and Lewin’s Change Theory,⁴³ (5) Break-Through Series Framework,³³ and (6) IOWA Model and Define, Measure Analyze, Improve, and Control Processes⁴⁴; however, these studies reported limited details on how the framework theoretically guided the implementation process.

Barriers and facilitators were poorly reported by many studies, with only 57% of studies exploring barriers and facilitators to implementation. These 20 studies explored barriers and facilitators using various methods, including qualitative interviews, quantitative surveys, informal surveys, and inductive discussion among experts. We initially mapped barriers and facilitators to the CFIR domains, then further categorized them by health care provider reported and patient reported.

The most reported barriers to the ESC model implementation for health care providers (n = 6 [30%]) were in the inner setting, which highlight the negative impacts of resource limitations such as limited private rooms, high turnover rates, availability for training, limitations in diverse language abilities, and staff time management.^{31,38,40,41,43,45} This was followed by the reported barriers of systemic oppression (n = 5 [25%]) such as racism, stigma, and bias.^{15,31,38,41,45} 

Determinants within the implementation setting were most reported as facilitators, with education (n = 6 [30%]) being the most common. This was followed by systematic and collaborative approaches to implementation (n = 5 [25%]), such as using a phased approach, quality improvement approach with feedback or pre/postreview, and using champions.^{12,31,40,41,46} 

The most reported barriers were found in the outer and inner settings and implementation process domains. First, in the outer setting (N = 3 [15%]), competing priorities (ie, caregiving needs of other children or jobs) was the most commonly cited barrier.^42,47,48 Second, in the inner setting (N = 3 [15%]), limited hospital supports, such as availability of rooms, breastfeeding support, and childcare, were commonly cited.^42,48,49 Finally, in the implementation domain, lack of education/preparedness (N = 3 [15%]) was noted as a common barrier reported by patients.^42,47,49 Table 1 describes the health care provider and patient-reported barriers and facilitators for each CFIR domain.

Table 2 outlines the frequency of outcomes reported, along with thematic categorization of effectiveness outcomes.

The most reported impact outcomes were patient health outcomes (n = 24), followed by health system outcomes (n = 23). The subcategories of patient health outcomes are discussed in the section titled “Effectiveness outcomes”; however, they are largely related to the need for pharmacological treatment (n = 23). The categories least reported were healthcare provider outcomes and patient experience, with only 5 studies exploring each category.

Implementation outcomes were largely underreported and explored, with only 8 studies exploring specific implementation outcomes. Of those studies, only 4 used consistent terminology as described by Proctor et al’s¹⁹ Implementation Outcome Framework. The most reported implementation outcome was acceptability (n = 7), followed by appropriateness (n = 5). Implementation cost was the only outcome not explored in any study.

Effectiveness outcomes were the most reported type of outcomes across all studies. The effectiveness outcomes were thematically organized into 11 categories, including: length of stay (n = 21), pharmacological treatment of any kind (n = 23), hospital cost (n = 3), readmission rates (n = 14), nonpharmacological interventions (n = 4), transfer to higher acuity (n = 8), breastfeeding rates (n = 10), weight changes (n = 6), adverse maternal/neonatal outcomes (n = 11), interrater reliability of the tool (n = 1), and medical readiness for discharge (n = 1). The most reported effectiveness outcome was pharmacological treatment of any kind (n = 23). Most studies reported substantial decreases in pharmacological treatment (n = 22 [of 23 studies exploring pharmacological treatment]), with only 1 study reporting no change in the pharmacological treatment.⁴⁸ Length of stay was the second most reported effectiveness outcome (n = 21). Most studies reported decreases in their length of stay (n = 20 [out of 21 studies reporting length of stay]). Amin et al⁴⁸ noted the lack of changes in pharmacological treatment and length of stay were likely attributed to the limitations of rooming-in and lack of parental presence due to the rural setting and increased geographical distances of families from hospitals. Hospital costs were noted to decrease with implementing the ESC model in all 3 reporting studies. No studies reported significant changes in hospital readmissions. All studies (n = 4) reported improvements in nonpharmacological interventions with the implementation of the ESC model. All studies (n = 8) reported no NICU admissions or decreases in transfers to higher acuity care, such as a postpartum unit to a neonatal intensive care unit. Most studies reported an increase in breastfeeding rates, with only 1 study out of 10 reporting no change in rates. Studies reported mixed impacts on weight changes (n = 6), with 4 studies reporting an increase in weight loss and 2 studies reported no difference in weight loss. Minimal adverse maternal and neonatal outcomes were reported across all 11 studies. However, Disilvo⁴² reported 1 maternal death that was deemed a suspected overdose. Finally, high interrater reliability was noted in the 1 study that explored this outcome, along with a decrease in the mean readiness for discharge (comparable to length of hospital stay) for all infants cared for using the ESC model.^33,34 

Our scoping review aimed to map the landscape of how the ESC model of care has been implemented and/or evaluated in health care practice. We examined 34 studies and used the CFIR,²⁸ ERIC taxonomy (cluster categorization),²⁹ and Implementation Outcome Framework,¹⁹ along with narrative descriptions to categorize and characterize findings across studies. The ESC model is shown to have positive patient and health system outcomes, such as decreasing the need for pharmacological treatment in infants with NAS/NOWS and decreasing the length of hospital stay. We know the ESC model works, given the breadth of literature exploring effectiveness outcomes of the ESC model; however, our review highlighted significant gaps in understanding the “how” and “why” of successful implementation.

Our scoping review explored “how” the ESC has been implemented into practice. Minimal description of the implementation processes, such as guiding frameworks, was noted in the literature. The most reported approach to implementation was using a multidisciplinary task force; however, this approach is more of an implementation strategy than a systematic process. The implementation science literature has shown that interventions are more likely to succeed when a theory-informed, systematic approach is used to design, implement, and evaluate interventions.²⁰ The use of frameworks, such as the CFIR (Consolidated Framework for Implementation Research), ensures thoughtful and intentional implementation of an evidence-based intervention into practice, contributing to the advancement of implementation knowledge.⁵⁸ Specifically, using implementation frameworks supports the development of a common language, which allows for the comparison of findings across contexts.^58–62 Although there was limited transparency on how the frameworks guided implementation, only 6 studies explicitly shared the use of frameworks to guide the implementation process. We recommend health system leaders use implementation frameworks to ensure systematic implementation of the ESC model into practice. Moullin et al⁵⁸ provide an excellent article outlining the top 10 recommendations for using implementation frameworks in practice.

Following the systematic implementation, guided by implementation frameworks, researchers must examine these processes and their impact on implementation.¹⁹ Contextual factors that affect implementation can only be explored by examining implementation outcomes, developing knowledge on how to implement an intervention effectively, and ultimately creating the desired impacts for the population of interest.¹⁹ The ESC model is a multilayered intervention encompassing multiple components, factors, and equity concerns, given its development for an equity-deserving population. Unique intervention factors complicate the way we examine implementation outcomes.^63,64 We cannot understand the complex implementation process without examining implementation outcomes. Our review revealed a limited exploration of implementation outcomes across studies, with only 7 studies exploring implementation outcomes, and only 4 of those studies using common implementation language. This knowledge gap ultimately limits the implementation and sustainment of the intervention across contexts over time⁶⁵; therefore, we recommend future research should be completed exploring implementation outcomes specific to the ESC model of care.

Common reported barriers for health care providers and patients across the CFIR contexts were resource limitations (found in the inner setting). Resource limitations, including high turnover rates and provider shortages, are known challenges within today’s health care system landscape that further compound the identified barrier of a lack of education for both patients and health care providers.⁶⁶ High turnover rates and shortages make it difficult to maintain staff competencies in caring for equity-deserving populations, such as birth parents with opioid use disorder (OUD). As such, tailoring implementation strategies to address these challenges can help to mitigate the impacts of turnover and improve sustainability of interventions into clinical practice.⁶⁷ Strategies such as continuous education, along with using low-cost education interventions, such as online education modules, are ways to mitigate the impacts of high turnover rates and improve the sustainability of evidence-based intervention implementation.⁶⁷ 

Systemic oppression was another barrier reported by health care providers, and it permeated all settings. This finding was echoed in the patient-reported barriers; however, it was articulated in social factors including experiences of discrimination, stigma, and oppression. This finding is not surprising given the historical and longstanding stigma and discrimination experienced by birth parents with OUD. Stigma contributes to health inequities, vulnerability, and dehumanization of birth parents with OUD.^68,69 Stigma within this population has been coined as a multileveled, toxic cycle within the health care system that presents externally (health care provider–created stigma), internally (internalized judgment and self-blame), and associatively (infants receiving poor care due to their association to substance use of the birth parent).⁶⁸ Tailoring implementation strategies to mitigate the negative impacts of stigma can help to improve successful implementation of evidence-based interventions in equity-deserving populations. For example, collaborating with persons who have experience in implementation can support mitigation of barriers created by stigma.⁷⁰ 

During our thematic analysis, 3 studies identified an important consideration of ethnic disparities in implementation of the ESC model.^38,45,54 Studies highlight a delay in outcomes seen in equity-deserving populations due to a predominantly English implementation of the ESC model, affecting families whose first language is not English or who experience systemic oppression.^38,45,54 These studies highlight the important impact that inconsistent application of an equity lens has on implementation success. For example, in Weikel et al’s⁵³ study, authors highlight numerous inequities experienced by families of Hispanic ethnicity. These inequities include systemic racism and discrimination and low income, which, when paired with barriers of language comprehension, are theorized to have led to gaps identified between Hispanic and non-Hispanic families. Authors encourage the importance of understanding unique barriers experienced by equity-deserving populations in designing implementation for evidence-based interventions.⁵⁴ 

Implementation strategies must be responsive to the inequities experienced within the population. Future research must consider the impact of equity and intersectionality on implementation, specifically exploring the impact of language and experiences of systemic oppression on implementation outcomes. Including descriptive characteristics such as race, ethnicity, and language will help researchers to better explore this gap in knowledge.⁴⁵ Incorporating frameworks, such as the intersectionality-supplemented CFIR, can support the intentional and systematic incorporation of equity considerations in implementation science research.⁷¹ Moreover, using a partnership approach to implementation by including persons experiencing disparities in health will support implementation and sustainment of evidence-based interventions in practice.⁷² 

The top strategies used were “train and educate stakeholders and develop stakeholder relationships.” These strategies parallel with the reported facilitators highlighted in Table 1, which demonstrates the importance of engaging in educational opportunities for health care providers and patients, along with empowerment within patients. Literature demonstrates that to support successful implementation, strategies used in implementation must address contextual determinants that affect implementation.^73,74 Our review highlighted a gap in the barriers identified and the strategies selected for implementation because no studies reported systematic processes used in selecting and tailoring implementation strategies. For example, resource limitations are noted as a top barrier in implementation in our review; however, strategies to address this barrier, such as “utilizing financial strategies or use evaluative and iterative strategies,” were not frequently reported.^29,75 The literature echoes this finding, highlighting a frequent disconnect between implementation strategies selected and the implementation barriers encountered.^76,77 An implementation strategy will only be successful if it has the capacity to overcome local and existing barriers to implementation.⁷⁸ Waltz et al⁷³ developed an easy-to-use mapping tool to facilitate this intentional consideration and adaptation of strategies based on identified local contextual barriers.

A few limitations should be considered when reviewing this study. First, although our search was comprehensive and completed in collaboration with a health science librarian, we cannot be certain that we have identified all sources on this subject. We added the gray literature search to expand our reach beyond academic databases because this is a relatively new topic in the literature. Second, results were limited to English due to limitations in the authors’ abilities to examine non-English articles critically.

The ESC model of care demonstrates positive patient and health system outcomes when implemented in various settings. This scoping review found 34 articles and documents that explore the implementation of the model into clinical practice. We highlight key barriers and facilitators identified across studies, including barriers of systemic oppression, resource limitations at the health care provider level, and competing priorities and lack of education identified at the patient level. Key facilitators identified include health care provider education, systematic collaboration, and patient empowerment and engagement in the model.

Given the mismatch of selected strategies used to identify contextual barriers reported, a systematic approach is recommended in selecting and tailoring implementation strategies to address locally-identified barriers. Implementation processes, including the use of guiding frameworks, are poorly reported across studies. A significant gap is noted in implementation outcomes research, which is critical to explore to support sustainable and effective implementation of the model. This scoping review provides essential insights into the current landscape of implementation science research available on the ESC model of care. Future research is needed to explore implementation outcomes specific to implementing the ESC model of care that incorporate equity-intentional frameworks and considerations.

CFIR

Consolidated Framework for Implementation Research

ERIC

Expert Recommendations for Implementing Change

ESC

Eat, Sleep, Console (model of care)

NAS

neonatal abstinence syndrome

NOWS

neonatal opioid withdrawal syndrome

PRISMA-ScR

Preferred Reporting Items for Systematic Reviews and Meta-analyses Extension for Scoping Reviews

OUD

opioid use disorder