Tackling the wicked problem of health networks: the design of an evaluation framework

Learning to make a difference for chILD: Value creation through network collaboration and team science

Addressing the recognized challenges and inequalities in providing high quality healthcare for rare diseases such as children's interstitial lung disease (chILD) requires collaboration across institutional, geographical, discipline, and system boundaries. The Children's Interstitial Lung Disease Respiratory Network of Australia and New Zealand (chILDRANZ) is an example of a clinical network that brings together multidisciplinary health professionals for collaboration, peer learning, and advocacy with the goal of improving the diagnosis and management of this group of rare and ultra-rare conditions. This narrative review explores the multifaceted benefits arising from social learning spaces within rare disease clinical networks by applying the value creation framework. The operation of the chILDRANZ network is used as an example across the framework to highlight how value is generated, realized, and transferred within such collaborative clinical and research networks. The community of practice formed in the chILDRANZ multidisciplinary meetings provides a strong example of social learning that engages with the uncertainty inherent in rare disease diagnosis and management and pays attention to generate new knowledge and best practice to make a difference for children and families living with chILD. This review underscores international calls for further investment in, and support of, collaborative clinical networks and virtual centers of excellence for rare disease.

Applying ordered network analysis to video-recorded physician-nurse interactions to examine communication patterns associated with shared understanding in inpatient oncology care settings

OBJECTIVES

The main aim of this study was to demonstrate how ordered network analysis of video-recorded interactions combined with verbal response mode (VRM) coding (eg, edification, disclosure, reflection and interpretation) can uncover specific communication patterns that contribute to the development of shared understanding between physicians and nurses. The major hypothesis was that dyads that reached shared understanding would exhibit different sequential relationships between VRM codes compared with dyads that did not reach shared understanding.

DESIGN

Observational study design with the secondary analysis of video-recorded interactions.

SETTING

The study was conducted on two oncology units at a large Midwestern academic health care system in the USA.

PARTICIPANTS

A total of 33 unique physician-nurse dyadic interactions were included in the analysis. Participants were the physicians and nurses involved in these interactions during patient care rounds.

PRIMARY AND SECONDARY OUTCOME MEASURES

The primary outcome measure was the development of shared understanding between physicians and nurses, as determined by prior qualitative analysis. Secondary measures included the frequencies, orders and co-occurrences of VRM codes in the interactions.

RESULTS

A Mann-Whitney U test showed that dyads that reached shared understanding (N=6) were statistically significantly different (U=148, p=0.00, r=0.93) from dyads that did not reach shared understanding (N=25) in terms of the sequential relationships between edification and disclosure, edification and advisement, as well as edification and questioning. Dyads that reached shared understanding engaged in more edification followed by disclosure, suggesting the importance of this communication pattern for reaching shared understanding.

CONCLUSIONS

This novel methodology demonstrates a robust approach to inform interventions that enhance physician-nurse communication. Further research could explore applying this approach in other healthcare settings and contexts.

New Zealand Pae Ora Healthcare Reforms 2022: Viable by Design? A Qualitative Study Using the Viable System Model

BACKGROUND

The New Zealand (NZ) Pae Ora (Healthy Futures) health reforms came into effect in July 2022 with the establishment of Health New Zealand (HNZ) (Te Whatu Ora) and the Māori Health Authority (MHA) (Te Aka Whai Ora) - the organisations charged for healthcare provision and delivery. Given these changes represent major health system reform, we aimed to conduct an early evaluation of the design of the reforms to determine if they can deliver a viable and sustainable NZ health system going forward.

METHODS

The evaluation was informed by Beer's viable system model (VSM). A qualitative exploratory design with semi-structured interviews and documents analysis using thematic analysis was used. We conducted 28 interviews with senior healthcare managers and reviewed over 300 official documents and news analyses.

RESULTS

The VSM posits that for a system to be viable, all its five sub-systems (operations; co-ordination; operational control; development and governance) need to be strong. Our analysis suggests that the health reforms, despite their strengths, do not satisfy this requirement. The reforms do appreciate the complexity of the healthcare environment: multiple stakeholders, social inequalities, interdependencies. However, our analysis suggests a severe lack of detail regarding the implementation and operationalisation of the reforms. Furthermore, resourcing and coordination within the reformed system is also unclear.

CONCLUSION

The health system reforms may not lead to a viable future NZ health system. Poor communication of the reform implementation and operationalisation will likely result in system failure and inhibit the ability of frontline health organisations to deliver care.

Primary care networks as a means of supporting primary care: findings from qualitative case study-based evaluation in the English NHS

OBJECTIVES

This study aimed to evaluate primary care networks (PCNs) in the English National Health Service. We ask: How are PCNs constituted to meet their defined goals? What factors can be discerned as affecting their ability to deliver benefits to the community, the network as a whole and individual members? What outcomes or outputs are associated with PCNs so far? We draw policy lessons for PCN design and oversight, and consider the utility of the chosen evaluative framework.

DESIGN AND SETTING

Qualitative case studies in seven PCN in England, chosen for maximum variety around geography, rurality and population deprivation. Study took place between May 2019 and December 2022.

PARTICIPANTS

PCN members, staff employed in additional roles and local managers. Ninety-one semistructured interviews and approximately 87 hours of observations were undertaken remotely. Interview transcripts and observational field notes were analysed together using a framework approach. Initial codes were derived from our evaluation framework, with inductive coding of new concepts during the analysis.

RESULTS

PCNs have been successfully established across England, with considerable variation in structure and operation. Progress is variable, with a number of factors affecting this. Good managerial support was helpful for PCN development. The requirement to work together to meet the specific threat of the global pandemic did, in many cases, generate a virtuous cycle by which the experience of working together built trust and legitimacy. The internal dynamics of networks require attention. Pre-existing strong relationships provided a significant advantage. While policy cannot legislate to create such relationships, awareness of their presence/absence is important.

CONCLUSIONS

Networked approaches to service delivery are popular in many health systems. Our use of an explicit evaluation framework supports the extrapolation of our findings to networks elsewhere. We found the framework to be useful in structuring our study but suggest some modifications for future use.

Evaluating primary care networks in low-income and lower middle-income countries: a scoping review

INTRODUCTION

Primary care networks (PCNs) are claimed to be an effective model to organise and deliver primary healthcare through collaborative relationships and effective coordination of primary care activities. Though increasingly implemented in different contexts, there is limited evidence on the effectiveness of PCNs in low-income and lower middle-income countries (LLMICs).

OBJECTIVE

Our scoping review aims to understand how PCNs in LLMICs have been conceptualised, implemented and analysed in the literature and further explores the evidence of the effectiveness of these networks.

METHODS

We structured our review using Arksey and O'Malley's framework for scoping reviews and recommendations by Levac et al. We also used the population, concept and context (PCC) guide of the Joanna Briggs Institute (JBI) methodology for scoping reviews to define the search strategy. The identified documents were then mapped, using Cunningham's evaluation framework for health networks, to understand how PCNs are conceived in LLMIC settings.

RESULTS

We identified 20 documents describing PCNs in five LLMICs. The selected documents showed differing forms and complexities of networks, with a majority resourced by government, non-governmental and donor entities. Most networks were mandated, and established with defined goals, although these were not always understood by stakeholders. Unlike PCNs in developed settings, the scoping review did not identify integration of care as a major goal for the establishment of PCNs in LLMICs. Network evaluation relationships, outputs and outcomes also varied across the five networks in the identified documents, and perceptions of effectiveness differed across stakeholder groups.

CONCLUSION

PCNs in LLMICs benefit from clearly stated goals and measurable outcomes, which facilitates evaluation. In order to maximise the benefits, careful attention to the aspects of network design and operation is required. Future research work could shed light on some of the missing pieces of evidence on their effectiveness by, for example, considering differential consequences of modes of network establishment and operation, including unintended consequences in the systems within which they reside, and evaluating long-term implications.

Describing the state of a research network: A mixed methods approach to network evaluation

Diabetes Action Canada Strategy for Patient-Oriented Research (SPOR) Network in Chronic Disease was formed in 2016 and is funded primarily through the Canadian Institutes of Health Research (CIHR). We propose a novel mixed-methods approach to a network evaluation integrating the State of Network Evaluation framework and the Canadian Academy of Health Sciences (CAHS) preferred framework and indicators. We measure key network themes of connectivity, health and results, and impact and return on investment associated with health research networks. Our methods consist of a longitudinal cross-sectional network survey of members and social network analysis to examine Network Connectivity and assess the frequency of interactions, the topics discussed during them, and how networking effectively facilitates interactions and collaboration among members. Network Health will be evaluated through semistructured interviews, a membership survey inquiring about satisfaction and experience with the Network, and a review of documentary sources related to funding and infrastructure to evaluate Network Sustainability. Finally, we will examine Network Results and Impact using the CAHS preferred framework and indicators to measure returns on investment in health research across the five domains of the CAHS framework, which include: advancing knowledge, capacity building, informing decision making, health impact, and economic and social impact. Indicators will be assessed with various methods, including bibliometric analyses, review of relevant documentary sources (annual reports), member activities informing health and research policy, and Patient Partner involvement. The Network Evaluation will provide members and stakeholders with information for planning, improvements, and funding future Network endeavors.

Using social networks to scale up and sustain community-based programmes to improve physical activity and diet in low-income and middle-income countries: a scoping review

BACKGROUND

Community-based programmes [CBPs], targeting increased physical activity and/or healthier eating, have been used in the prevention and management of non-communicable diseases. However, CBPs are only useful, insofar as they can be scaled up and sustained in some meaningful way. Social networks-defined as "social structures that exists between actors, individuals or organizations"-may serve as an important tool to identify underlying mechanisms that contribute to this process. This scoping review aimed to map and collate literature on the role of social network research in scaling-up and sustaining physical activity and/or diet CBPs in low-and middle-income countries [LMICs].

METHODS

Arksey and O'Malley's framework and its enhancement were followed. Inclusion criteria were peer-reviewed articles exploring the role of social networks in scaled-up and/or sustained physical activity and/or diet CBPs in adult populations, published in English since 2000, and based in a LMIC. Databases searched were PubMed, Cochrane, Scopus, Web of Science, CINAHL, SocIndex, International Bibliography of the Social Sciences, and Google Scholar. Books, conference abstracts, and programmes focused on children were excluded. Two reviewers independently selected and extracted eligible studies. Included publications were thematically analysed using the Framework Approach.

RESULTS

Authors identified 12 articles for inclusion, covering 13 CBPs. Most were based in Latin America, with others in the Caribbean, the Pacific Islands, Iran, and India. All articles were published since 2009. Only three used social network analysis methods (with others using qualitative methods). Five main social network themes were identified: centralisation, cliques, leaders, quality over quantity, and shared goals. Contextual factors to be considered when scaling-up programmes in LMICs were also identified.

CONCLUSIONS

This review has shown that the evidence of the use of social network research in programme scale-up has not yet caught up to its theoretical possibilities. Programmes aiming to scale should consider conducting social network research with identified network themes in mind to help improve the evidence-base of what network mechanisms, in what contexts, might best support the strengthening of networks in physical activity and diet programmes. Importantly, the voice of individuals and communities in these networks should not be forgotten.

Using Network and Complexity Theories to Understand the Functionality of Referral Systems for Surgical Patients in Resource-Limited Settings, the Case of Malawi

BACKGROUND

A functionally effective referral system that links district level hospitals (DLHs) with referral hospitals (RHs) facilitates surgical patients getting timely access to specialist surgical expertise not available locally. Most published studies from low- and middle-income countries (LMICs) have examined only selected aspects of such referral systems, which are often fragmented. Inadequate understanding of their functionality leads to missed opportunities for improvements. This research aimed to investigate the functionality of the referral system for surgical patients in Malawi, a low-income country.

METHODS

This study, conducted in 2017-2019, integrated principles from two theories. We used network theory to explore interprofessional relationships between DLHs and RHs at referral network, member (hospital) and community levels; and used principles from complex adaptive systems (CAS) theory to unpack the mechanisms of network dynamics. The study employed mixed-methods, specifically surveys (n=22 DLHs), interviews with clinicians (n=20), and a database of incoming referrals at two sentinel RHs over a six-month period.

RESULTS

Obstacles to referral system functionality in Malawi included weaknesses in formal coordination structures, notably: unclear scope of practice of district surgical teams; lack of referral protocols; lack of referral communication standards; and misaligned organisational practices. Deficiencies in informal relationships included mistrust and uncollaborative operating environments, undermining coordination between DLHs and RHs. Poor system functionality adversely impacted the quality, efficiency and safety of patient referral-related care. Respondents identified aspects of the district-RH relationships, which could be leveraged to build more collaborative and productive inter-professional relationships in the future.

CONCLUSION

Multi-level interventions are needed to address failures at both ends of the referral pathway. This study captured new insights into longstanding problems in referral systems in resource-limited settings, contributing to a better understanding of how to build more functional systems to optimise the continuum and quality of surgical care for rural populations in similar settings.

Too much theory and not enough practice? The challenge of implementation science application in healthcare practice

BACKGROUND

Implementation science (IS) should contribute to maintaining high standards of care across healthcare systems and enhancing care practices. However, despite the evident need for greater and more rapid uptake and integration of evidence in practice, IS design and methodology fall short of the needs of effective translation.

AIM

In this paper we examine what it is about IS that makes it so appealing for effective uptake of interventions in routine practice, and yet so difficult to achieve. We propose a number of ways that implementation scientists could build mutual relationships with healthcare practitioners and other stakeholders including public members to ensure greater shared care practices, and highlight the value of IS training, collaborative educational events, and co-designed research.

DISCUSSION

More consideration should be given to IS applications in healthcare contexts. Implementation scientists can make a valuable contribution by mobilizing theory and improving practice. However, goals for an evidence-based system may be more appropriately achieved through greater outreach and collaboration, with methods that are flexible to support rapid implementation in complex adaptive systems. Collective learning and mutual trust can be cultivated by embedding researchers into healthcare services while offering greater opportunities for practitioners to learn about, and engage in, implementation research.

CONCLUSION

To bridge the worlds of healthcare practice and IS, researchers could be more consistent in the relationships they build with professionals and the public, communicating through a shared language and co-joining practical approaches to effective implementation. This will build capacity for improved collaboration and foster respectful, interdisciplinary relationships.

Network evaluation of an innovation platform in continuous quality improvement in Australian Indigenous primary healthcare

Background

From 2014 to 2019, the Centre for Research Excellence in Integrated Quality Improvement (CRE-IQI) was evaluated as an innovation platform focusing on continuous quality improvement in Indigenous Australian primary healthcare. Although social network analysis (SNA) is a recognized method for evaluating the functioning, collaboration and effectiveness of innovation platforms, applied research is limited. This study applies SNA to evaluate the CRE-IQI’s functioning as an innovation platform.

Methods

Two surveys (2017, 2019) were conducted using social survey and network methods. Survey items covered respondent characteristics, their perceptions of the CRE-IQI’s performance, and its impact and sociometric relationships. Members’ relationship information was captured for the CRE-IQI at three time points, namely start (retrospectively), midpoint and final year, on three network types (knew, shared information, collaborated). SNA software was used to compute standard network metrics including diameter, density and centrality, and to develop visualizations. Survey and network results were addressed in a workshop held by members to develop improvement strategies.

Results

The response rate was 80% in 2017 and 65% in 2019 (n = 49 and 47, respectively). Between 2017 and 2019, respondents’ mean ratings of the CRE-IQI’s functioning and achievements in meeting its goals were sustained. They perceived the CRE-IQI as multidisciplinary, having effective management and governance, and incorporating Indigenous research leadership, representation and ways of working. Respondents recognized high levels of trust amongst members, rated “good communication and coordination with participants” highly, and “facilitating collaboration” as the CRE’s most strongly recognized achievement. In collaboration and information-sharing networks, average path length remained low in 2017 and 2019, indicating good small-world network properties for relaying information. On average, respondents shared information and collaborated with more CRE members in 2017 than 2019. However, in both 2017 and 2019 there were new collaborations and information-sharing outside of direct collaborations. CRE-IQI outcomes included: evidence generation; knowledge transfer and skills development in quality improvement; research capacity-building, career development; mentoring; grant support; development of new projects; health service support; and policy impact.

Conclusions

This study shows the utility of network analysis in evaluating the functioning, and collaboration, at the individual, organizational and health system levels, of an innovation platform, and adds to our understanding of factors enabling successful innovation platforms.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12961-022-00909-z.

How can the healthcare system deliver sustainable performance? A scoping review

BACKGROUND

Increasing health costs, demand and patient multimorbidity challenge the sustainability of healthcare systems. These challenges persist and have been amplified by the global pandemic.

OBJECTIVES

We aimed to develop an understanding of how the sustainable performance of healthcare systems (SPHS) has been conceptualised, defined and measured.

DESIGN

Scoping review of peer-reviewed articles and editorials published from database inception to February 2021.

DATA SOURCES

PubMed and Ovid Medline, and snowballing techniques.

ELIGIBILITY CRITERIA

We included articles that discussed key focus concepts of SPHS: (1) definitions, (2) measurement, (3) identified challenges, (4) identified solutions for improvement and (5) scaling successful solutions to maintain SPHS.

DATA EXTRACTION AND SYNTHESIS

After title/abstract screening, full-text articles were reviewed, and relevant information extracted and synthesised under the five focus concepts.

RESULTS

Of 142 included articles, 38 (27%) provided a definition of SPHS. Definitions were based mainly on financial sustainability, however, SPHS was also more broadly conceptualised and included acceptability to patients and workforce, resilience through adaptation, and rapid absorption of evidence and innovations. Measures of SPHS were also predominantly financial, but recent articles proposed composite measures that accounted for financial, social and health outcomes. Challenges to achieving SPHS included the increasingly complex patient populations, limited integration because of entrenched fragmented systems and siloed professional groups, and the ongoing translational gaps in evidence-to-practice and policy-to-practice. Improvement strategies for SPHS included developing appropriate workplace cultures, direct community and consumer involvement, and adoption of evidence-based practice and technologies. There was also a strong identified need for long-term monitoring and evaluations to support adaptation of healthcare systems and to anticipate changing needs where possible.

CONCLUSIONS

To implement lasting change and to respond to new challenges, we need context-relevant definitions and frameworks, and robust, flexible, and feasible measures to support the long-term sustainability and performance of healthcare systems.

How does a Network Platform Work for Participating Actors Towards Integrated Care Governance? A Case Study of a Dutch Hospital Region

Introduction

Network platforms are interesting for integrated care governance as they seek solutions for the problem of competition and tensions between networks. In this paper, we analyse how a network platform functions for the actors involved, and how it is used in their work.

Methods

We employed a case study in a Dutch urbanised hospital region, and conducted 17 interviews with hospital physicians, directors, and supporting staff who are involved in a network platform called "BeterKeten" (BK).

Results

Actors assign different functions and purposes to BK: facilitating and legitimising professional (learning) communities; adapting to a changing policy context; enlarging professionals' and the networks' circle of influence; and extending governing possibilities. Network platform' dynamics and frictions entail changing professional and managerial practices; embedding a BK network in a partner network; and alignment of (conflicting) network platforms.

Discussion

Network platforms are a promising strategy to govern, facilitate, and nurture network-building actions to enhance integrated care, offering new ways of working to cope with its multi-level nature.

Conclusion

BK is a dynamic actor with steering capacities that enables the co-existence of multiple purposes. Further research could pay attention to how network platforms are able to develop modalities of integrated care governance that suit healthcare system's networked character.

Utility of the AHRQ Learning Collaboratives Taxonomy for Analyzing Innovations from an Australian Collaborative

BACKGROUND

Despite the proliferation of learning collaborations such as innovation platforms, the factors contributing to their success or failure are rarely documented. The Agency for Healthcare Research and Quality learning collaboratives taxonomy provides a framework for understanding how such collaborations work in different settings according to four primary elements: innovation, communication, time, and social systems. This study applied the taxonomy to assess an innovation platform and the utility of applying the taxonomy.

METHODS

The study focus was a five-year national research collaboration operating as an innovation platform to strengthen primary health care quality improvement efforts for Indigenous Australians. The study team analyzed project records, reports and publications, and interviews that were conducted with 35 stakeholders. Data were mapped retrospectively against the taxonomy domains and thematically analyzed.

RESULTS

The taxonomy proved useful in understanding how and why the innovation platform generated innovations. It revealed that time was particularly important, both to see innovations through and to establish a social system that enabled interconnectivity between members. However, the taxonomy did not provide useful guidance on identifying the types of innovations from the collaboration or the importance of a culture of continuous adaptation and learning. The study also found that the primary and secondary elements of the taxonomy were not discrete, which meant that it was difficult to align themes with only one element.

CONCLUSION

To improve the utility of the taxonomy, several elaborations are proposed, including reconfiguring it to a more dynamic form that recognizes the interconnections and links between the elements.

What Works in Community-Led Suicide Prevention: Perspectives of Wesley LifeForce Network Coordinators

Community coalitions have been recognised as an important vehicle to advance health promotion and address relevant local health issues in communities, yet little is known about their effectiveness in the field of suicide prevention. The Wesley Lifeforce Suicide Prevention Networks program consists of a national cohort of local community-led suicide prevention networks. This study drew on a nationally representative survey and the perspectives of coordinators of these networks to identify the key factors underpinning positive perceived network member and community outcomes. Survey data were analysed through descriptive statistics and linear regression analyses. Networks typically reported better outcomes for network members and communities if they had been in existence for longer, had a focus on the general community, and had conducted more network meetings and internal processes, as well as specific community-focused activities. Study findings strengthen the evidence base for effective network operations and lend further support to the merit of community coalitions in the field of suicide prevention, with implications for similar initiatives, policymakers, and wider sector stakeholders seeking to address suicide prevention issues at a local community level.

Towards developing a comprehensive conceptual understanding of positive hospital culture and approaches to healthcare organisational culture change in Australia

PURPOSE

The key aim of this narrative literature review, therefore, is to identify the key conceptual categories that inform the construction of positive person-centred culture within hospitals, and how these frameworks are brought to bear upon organisational culture within healthcare systems in Australia.

DESIGN/METHODOLOGY/APPROACH

This narrative review presents a thematic synthesis of literature identified through a systematic search protocol undertaken across 19 academic databases and Google Scholar as an additional search tool. Thematic qualitative analysis was performed on the research results to determine the common themes within the diverse literature presented within this study.

FINDINGS

Culture change interventions in hospitals attempt to address the problem of widespread unprofessional behaviour within healthcare systems. However, diverse definitions and seemingly fragmented approaches to understanding and enacting organisational culture change present a significant hurdle in achieving cohesive and sustainable healthcare reform. This narrative literature review offers a comprehensive conceptual view of the key approaches that inform positive person-centred culture within hospital settings. In total, three primary dimensions, belonging, behaving and being, aligned against organisational goals, individual behaviours and worker as well as organisational identity were identified. Other individual and group interactional dynamics that give rise to negative organisational culture are further analysed to understand the fault lines along which existing culture change interventions are typically operationalised.

RESEARCH LIMITATIONS/IMPLICATIONS

This review is not exhaustive and is limited in its methodological scope. The central values and themes identified within the literature are integral to designing humanised healthcare systems. However, owing to the qualitative nature and contextual variability of these factors, these themes do not lend themselves to replicable quantification.

SOCIAL IMPLICATIONS

This analysis contributes to foundational research efforts towards transforming healthcare practice to be more aligned with humanised and equitable values within increasingly complex healthcare organisational settings. Designing culture change interventions that align more suitably with the values-driven categories identified in this literature review may increase the effectiveness and sustainability of these interventions and reform efforts at organisational and systemic levels.

ORIGINALITY/VALUE

This article presents a comprehensive framework to approach healthcare organisational reform through shared and equitable models of operation, management and governance rather than continuing to promote narrowly defined outcomes derived from commodified models of healthcare practice.

The three numbers you need to know about healthcare: the 60-30-10 Challenge

BACKGROUND

Healthcare represents a paradox. While change is everywhere, performance has flatlined: 60% of care on average is in line with evidence- or consensus-based guidelines, 30% is some form of waste or of low value, and 10% is harm. The 60-30-10 Challenge has persisted for three decades.

MAIN BODY

Current top-down or chain-logic strategies to address this problem, based essentially on linear models of change and relying on policies, hierarchies, and standardisation, have proven insufficient. Instead, we need to marry ideas drawn from complexity science and continuous improvement with proposals for creating a deep learning health system. This dynamic learning model has the potential to assemble relevant information including patients' histories, and clinical, patient, laboratory, and cost data for improved decision-making in real time, or close to real time. If we get it right, the learning health system will contribute to care being more evidence-based and less wasteful and harmful. It will need a purpose-designed digital backbone and infrastructure, apply artificial intelligence to support diagnosis and treatment options, harness genomic and other new data types, and create informed discussions of options between patients, families, and clinicians. While there will be many variants of the model, learning health systems will need to spread, and be encouraged to do so, principally through diffusion of innovation models and local adaptations.

CONCLUSION

Deep learning systems can enable us to better exploit expanding health datasets including traditional and newer forms of big and smaller-scale data, e.g. genomics and cost information, and incorporate patient preferences into decision-making. As we envisage it, a deep learning system will support healthcare's desire to continually improve, and make gains on the 60-30-10 dimensions. All modern health systems are awash with data, but it is only recently that we have been able to bring this together, operationalised, and turned into useful information by which to make more intelligent, timely decisions than in the past.