Health care organizations are leveraging machine-learning techniques, such as artificial neural networks (ANN), to improve delivery of care at a reduced cost. Applications of ANN to diagnosis are well-known; however, ANN are increasingly used to inform health care management decisions. We provide a seminal review of the applications of ANN to health care organizational decision-making. We screened 3,397 articles from six databases with coverage of Health Administration, Computer Science and Business Administration. We extracted study characteristics, aim, methodology and context (including level of analysis) from 80 articles meeting inclusion criteria. Articles were published from 1997–2018 and originated from 24 countries, with a plurality of papers (26 articles) published by authors from the United States. Types of ANN used included ANN (36 articles), feed-forward networks (25 articles), or hybrid models (23 articles); reported accuracy varied from 50% to 100%. The majority of ANN informed decision-making at the micro level (61 articles), between patients and health care providers. Fewer ANN were deployed for intra-organizational (meso- level, 29 articles) and system, policy or inter-organizational (macro- level, 10 articles) decision-making. Our review identifies key characteristics and drivers for market uptake of ANN for health care organizational decision-making to guide further adoption of this technique.

BackgroundImplementation scientists and practitioners, alike, recognize the importance of sustaining practice change, however post-implementation studies of interventions are rare. This is a protocol for the Sustainment, Sustainability and Spread Study (SSaSSy). The purpose of this study is to contribute to knowledge on the sustainment (sustained use), sustainability (sustained benefits), and spread of evidence-based practice innovations in health care. Specifically, this is a post-implementation study of an evidence-informed, Care Aide-led, facilitation-based quality-improvement intervention called SCOPE (Safer Care for Older Persons (in long-term care) Environments). SCOPE has been implemented in nursing homes in the Canadian Provinces of Manitoba (MB), Alberta (AB) and British Columbia (BC). Our study has three aims: (i) to determine the role that adaptation/contextualization plays in sustainment, sustainability and spread of the SCOPE intervention; (ii) to study the relative effects on sustainment, sustainability and intra-organizational spread of high-intensity and low-intensity post-implementation “boosters”, and a “no booster” condition, and (iii) to compare the relative costs and impacts of each booster condition.Methods/designSSaSSy is a two-phase mixed methods study. The overarching design is convergent, with qualitative and quantitative data collected over a similar timeframe in each of the two phases, analyzed independently, then merged for analysis and interpretation. Phase 1 is a pilot involving up to 7 units in 7 MB nursing homes in which SCOPE was piloted in 2016 to 2017, in preparation for phase 2. Phase 2 will comprise a quasi-experiment with two treatment groups of low- and high-intensity post-implementation “boosters”, and an untreated control group (no booster), using pretests and post-tests of the dependent variables relating to sustained care and management practices, and resident outcomes. Phase 2 will involve 31 trial sites in BC (17 units) and AB (14 units) nursing homes, where the SCOPE trial concluded in May 2019.DiscussionThis project stands to advance understanding of the factors that influence the sustainment of practice changes introduced through evidence-informed practice change interventions, and their associated sustainability. Findings will inform our understanding of the nature of the relationship of fidelity and adaptation to sustainment and sustainability, and afford insights into factors that influence the intra-organizational spread of practice changes introduced through complex interventions.

Background There is recognition that the overuse of procedures, testing, and medications constitutes low-value care which strains the healthcare system and, in some circumstances, can cause unnecessary stress and harm for patients. Initiatives across dozens of countries have raised awareness about the harms of low-value care but have had mixed success and the levels of reductions realized have been modest. Similar to the complex drivers of implementation processes, there is a limited understanding of the individual and social behavioral aspects of de-implementation. While researchers have begun to use theory to elucidate the dynamics of de-implementation, the research remains largely atheoretical. The use of theory supports the understanding of how and why interventions succeed or fail and what key factors predict success. The purpose of this scoping review was to identify and characterize the use of theoretical approaches used to understand and/or explain what influences efforts to reduce low-value care. Methods We conducted a review of MEDLINE, EMBASE, CINAHL, and Scopus databases from inception to June 2021. Building on previous research, 43 key terms were used to search the literature. The database searches identified 1998 unique articles for which titles and abstracts were screened for inclusion; 232 items were selected for full-text review. Results Forty-eight studies met the inclusion criteria. Over half of the included articles were published in the last 2 years. The Theoretical Domains Framework (TDF) was the most commonly used determinant framework (n = 22). Of studies that used classic theories, the majority used the Theory of Planned Behavior (n = 6). For implementation theories, Normalization Process Theory and COM-B were used (n = 7). Theories or frameworks were used primarily to identify determinants (n = 37) and inform data analysis (n = 31). Eleven types of low-value care were examined in the included studies, with prescribing practices (e.g., overuse, polypharmacy, and appropriate prescribing) targeted most frequently. Conclusions This scoping review provides a rigorous, comprehensive, and extensive synthesis of theoretical approaches used to understand and/or explain what factors influence efforts to reduce low-value care. The results of this review can provide direction and insight for future primary research to support de-implementation and the reduction of low-value care.

Background Complex interventions are increasingly applied to healthcare problems. Understanding of post-implementation sustainment, sustainability, and spread of interventions is limited. We examine these phenomena for a complex quality improvement initiative led by care aides in 7 care homes (long-term care homes) in Manitoba, Canada. We report on factors influencing these phenomena two years after implementation. Methods Data were collected in 2019 via small group interviews with unit- and care home-level managers (n = 11) from 6 of the 7 homes using the intervention. Interview participants discussed post-implementation factors that influenced continuing or abandoning core intervention elements (processes, behaviors) and key intervention benefits (outcomes, impact). Interviews were audio-recorded, transcribed verbatim, and analyzed with thematic analysis. Results Sustainment of core elements and sustainability of key benefits were observed in 5 of the 6 participating care homes. Intra-unit intervention spread occurred in 3 of 6 homes. Factors influencing sustainment, sustainability, and spread related to intervention teams, unit and care home, and the long-term care system. Conclusions Our findings contribute understanding on the importance of micro-, meso-, and macro-level factors to sustainability of key benefits and sustainment of some core processes. Inter-unit spread relates exclusively to meso-level factors of observability and practice change institutionalization. Interventions should be developed with post-implementation sustainability in mind and measures taken to protect against influences such as workforce instability and competing internal and external demands. Design should anticipate need to adapt interventions to strengthen post-implementation traction.