Investigating the use of data-driven artificial intelligence in computerised decision support systems for health and social care: A systematic review.

There is growing interest in the potential of artificial intelligence to support decision-making in health and social care settings. There is, however, currently limited evidence of the effectiveness of these systems. The aim of this study was to investigate the effectiveness of artificial intelligence-based computerised decision support systems in health and social care settings. We conducted a systematic literature review to identify relevant randomised controlled trials conducted between 2013 and 2018. We searched the following databases: MEDLINE, EMBASE, CINAHL, PsycINFO, Web of Science, Cochrane Library, ASSIA, Emerald, Health Business Fulltext Elite, ProQuest Public Health, Social Care Online, and grey literature sources. Search terms were conceptualised into three groups: artificial intelligence-related terms, computerised decision support -related terms, and terms relating to health and social care. Terms within groups were combined using the Boolean operator OR, and groups were combined using the Boolean operator AND. Two reviewers independently screened studies against the eligibility criteria and two independent reviewers extracted data on eligible studies onto a customised sheet. We assessed the quality of studies through the Critical Appraisal Skills Programme checklist for randomised controlled trials. We then conducted a narrative synthesis. We identified 68 hits of which five studies satisfied the inclusion criteria. These studies varied substantially in relation to quality, settings, outcomes, and technologies. None of the studies was conducted in social care settings, and three randomised controlled trials showed no difference in patient outcomes. Of these, one investigated the use of Bayesian triage algorithms on forced expiratory volume in 1 second (FEV1) and health-related quality of life in lung transplant patients. Another investigated the effect of image pattern recognition on neonatal development outcomes in pregnant women, and another investigated the effect of the Kalman filter technique for warfarin dosing suggestions on time in therapeutic range. The remaining two randomised controlled trials, investigating computer vision and neural networks on medication adherence and the impact of learning algorithms on assessment time of patients with gestational diabetes, showed statistically significant and clinically important differences to the control groups receiving standard care. However, these studies tended to be of low quality lacking detailed descriptions of methods and only one study used a double-blind design. Although the evidence of effectiveness of data-driven artificial intelligence to support decision-making in health and social care settings is limited, this work provides important insights on how a meaningful evidence base in this emerging field needs to be developed going forward. It is unlikely that any single overall message surrounding effectiveness will emerge - rather effectiveness of interventions is likely to be context-specific and calls for inclusion of a range of study designs to investigate mechanisms of action.

NHS Scotland's Decision Support Platform: a formative qualitative evaluation.

CONTEXT: The Scottish Government has identified computerised decision support as a strategic priority in order to improve knowledge management in health and social care settings. A national programme to build a pilot Decision Support Platform was funded in 2015. AIMS: We undertook a formative evaluation of the Decision Support Platform to inform plans for its national roll-out in primary care. METHODS: We conducted a series of in-depth semistructured interviews and non-participant observations of workshops demonstrating decision support systems. Participants were policymakers and clinical opinion leaders from primary care. As the Platform was in its early stages of development at the time of data collection, we focused on exploring expectations and drivers of the pilot decision support system tested in primary care. Our methodological approach had to be tailored to changing circumstances and offered important opportunities for realising impact through ongoing formative feedback to policymakers and active engagement of key clinical stakeholders. We drew on sociotechnical principles to inform data analysis and coded qualitative data with the help of NVivo software. FINDINGS: We conducted 30 interviews and non-participant ethnographic observations of eight stakeholder engagement workshops. We observed a strong sense of support from all stakeholders for the Platform and associated plans to roll it out across NHS Scotland. Strategic drivers included the potential to facilitate integration of care, preventive care, patient self-management, shared decision-making and patient engagement through the ready availability of clinically important information. However, in order to realise these benefits, participants highlighted the need for strong national eHealth leadership to drive a coherent strategy and ensure sustained funding, system usability (which stakeholders perceived to be negatively affected by alert fatigue and integration with existing systems) and ongoing monitoring of potential unintended consequences emerging from implementations (eg, increasing clinical workloads). CONCLUSIONS AND IMPLICATIONS: In order to address potential tensions between national leadership and local usability as well as unintended consequences, there is a need to have overall national ownership to support the implementation of the Platform. Potential local tensions could be addressed through allowing a degree of local customisation of systems and tailoring of alerts, and investing in a limited number of pilots that are carefully evaluated to mitigate emerging risks early.

Improving public health evaluation: a qualitative investigation of practitioners' needs.

BACKGROUND: In 2011, the House of Lords published a report on Behaviour Change, in which they report that "a lot more could, and should, be done to improve the evaluation of interventions." This study aimed to undertake a needs assessment of what kind of evaluation training and materials would be of most use to UK public health practitioners by conducting interviews with practitioners about everyday evaluation practice and needed guidance and materials. METHODS: Semi-structured interviews were conducted with 32 public health practitioners in two UK regions, Cambridgeshire and the South West. Participants included directors of public health, consultants in public health, health improvement advisors, public health intelligence, and public health research officers. A topic guide included questions designed to explore participants existing evaluation practice and their needs for further training and guidance. Data were analysed using thematic analyses. RESULTS: Practitioners highlighted the need for evaluation to defend the effectiveness of existing programs and protect funding provisions. However, practitioners often lacked training in evaluation, and felt unqualified to perform such a task. The majority of practitioners did not use, or were not aware of many existing evaluation guidance documents. They wanted quality-assured, practical guidance that relate to the real world settings in which they operate. Practitioners also mentioned the need for better links and support from academics in public health. CONCLUSION: Whilst numerous guidance documents supporting public health evaluation exist, these documents are currently underused by practitioners - either because they are not considered useful, or because practitioners are not aware of them. Integrating existing guides into a catalogue of guidance documents, and developing a new-quality assured, practical and useful document may support the evaluation of public health programs. This in turn has the potential to identify those programs that are effective; thus improving public health and reducing financial waste.

A review of Grey and academic literature of evaluation guidance relevant to public health interventions.

BACKGROUND: Public Health evaluation is essential to understanding what does and does not work, and robust demonstration of effectiveness may be crucial to securing future funding. Despite this, programs are often implemented with poor, incomplete or no evaluation. Public health practitioners are frequently required to provide evidence for the effectiveness of their services; thus, there is a growing need for evaluation guidance on how to evaluate public health programs. The aim of this study is to identify accessible high-quality, evaluation guidance, available to researchers and practitioners and to catalogue, summarise and categorise the content of a subset of accessible, quality guides to evaluation. METHODS: We systematically reviewed grey and academic literature for documents providing support for evaluation of complex health interventions. Searches were conducted January to March 2015, and included academic databases, internet search engines, and consultations with academic and practicing public health experts. Data were extracted by two authors and sent to the authors of the guidance documents for comments. RESULTS: Our initial search identified 402 unique documents that were screened to identify those that were (1) developed by or for a national or international organization (2) freely available to all (3) published during or after 2000 (4) specific to public health. This yielded 98 documents from 43 organisations. Of these, 48 were reviewed in detail. This generated a detailed catalogue of quality evaluation guidance. The content included in documents covers 37 facets of evaluation. CONCLUSIONS: A wide range of guidance on evaluation of public health initiatives is available. Time and knowledge constraints may mean that busy practitioners find it challenging to access the most, up-to-date, relevant and useful guidance. This review presents links to and reviews of 48 quality guides to evaluation as well as categorising their content. This facilitates quick and each access to multiple selected sources of specific guidance.

Creating positive experiences for people living with dementia in care homes.

The global incidence of dementia is set to increase. It is estimated that two thirds of people with dementia live in care homes. To ensure good care experiences for these people, it is essential that the factors which create positive care home environments are understood. This article presents findings from a literature review of factors influencing care home experiences for people with dementia. The main findings were that organisational culture and workforce development are important factors in creating positive care home experiences for people with dementia. Although the evidence is limited, a person-centred approach based on research evidence is essential if the workforce is to gain the necessary knowledge and skills to provide positive experiences of care for residents with dementia.

What do patients and the public know about clinical practice guidelines and what do they want from them? A qualitative study.

BACKGROUND: Guideline producers are increasingly producing versions of guidelines for the public. The aim of this study was to explore what patients and the public understand about the purpose and production of clinical guidelines, and what they want from clinical guidelines to support their healthcare decisions. METHODS: Participants were purposively selected to represent a range of the likely users of patient versions of guidelines, including individuals with health conditions (diabetes and depression), general members of the public, health communication professionals and a group of young people. Participants were asked about their awareness and understanding of clinical guidelines and presented with scenario recommendations, or draft materials from patient guidelines to prompt discussion. Each discussion was facilitated by one or two researchers. All focus groups were recorded and transcribed prior to analysis. Data were analysed using framework analysis. RESULTS: We ran nine focus groups involving 62 individuals, supplemented by four interviews with people experiencing homelessness. Eight groups were held in Scotland, one in England. The four interviews were held in Scotland. The framework analysis yielded five themes: access and awareness; what patients want to know; properties of guidelines; presenting evidence; and format. Awareness of guidelines was low. Participants emphasised the need for information that enables them to choose between treatment options, including harms. They would like help with this from healthcare professionals, especially general practitioners. Participants differed in their support for the inclusion of numerical information and graphs. CONCLUSIONS: Members of the public want information to help them choose between treatments, including information on harms, particularly to support shared decisions with health professionals. Presenting numerical information is a challenge and layered approaches that present information in stages may be helpful. Ignoring the themes identified in this study is likely to lead to materials that fail to support public and patient healthcare decision making.

Patients, health information, and guidelines: A focus-group study.

BACKGROUND: Evidence-based clinical guidelines could support shared decision-making and help patients to participate actively in their care. However, it is not well known how patients view guidelines as a source of health information. This qualitative study aimed to assess what patients know about guidelines, and what they think of their presentation formats. RESEARCH QUESTION: What is the role of guidelines as health information for patients and how could the implementation of evidence-based information for patients be improved? METHODS: A qualitative study with focus groups that were built around a semi-structured topic guide. Focus groups were audiotaped and transcribed and analysed using a phenomenographic approach. RESULTS: Five focus groups were carried out in 2012 with a total of 23 participants. Patients searched for health information from the Internet or consulted health professionals or their personal networks. The concepts of guidelines included instructions or standards for health professionals, information given by a health professional to the patient, and material to protect and promote the interests of patients. Some patients did not have a concept for guidelines. Patients felt that health information was abundant and its quality sometimes difficult to assess. They respected conciseness, clarity, clear structure, and specialists or well-known organizations as authors of health information. Patients would like health professionals to deliver and clarify written materials to them or point out to them the relevant Internet sites. CONCLUSIONS: The concept of guidelines was not well known among our interviewees; however, they expressed an interest in having more communication on health information, both written information and clarifications with their health professionals.

Patient and public attitudes to and awareness of clinical practice guidelines: a systematic review with thematic and narrative syntheses.

BACKGROUND: Clinical practice guidelines are typically written for healthcare providers but there is increasing interest in producing versions for the public, patients and carers. The main objective of this review is to identify and synthesise evidence of the public's attitudes towards clinical practice guidelines and evidence-based recommendations written for providers or the public, together with their awareness of guidelines. METHODS: We included quantitative and qualitative studies of any design reporting on public, patient (and their carers) attitudes and awareness of guidelines written for providers or patients/public. We searched electronic databases including MEDLINE, PSYCHINFO, ERIC, ASSIA and the Cochrane Library from 2000 to 2012. We also searched relevant websites, reviewed citations and contacted experts in the field. At least two authors independently screened, abstracted data and assessed the quality of studies. We conducted a thematic analysis of first and second order themes and performed a separate narrative synthesis of patient and public awareness of guidelines. RESULTS: We reviewed 5415 records and included 26 studies (10 qualitative studies, 13 cross sectional and 3 randomised controlled trials) involving 24 887 individuals. Studies were mostly good to fair quality. The thematic analysis resulted in four overarching themes: Applicability of guidelines; Purpose of guidelines for patient; Purpose of guidelines for health care system and physician; and Properties of guidelines. Overall, participants had mixed attitudes towards guidelines; some participants found them empowering but many saw them as a way of rationing care. Patients were also concerned that the information may not apply to their own health care situations. Awareness of guidelines ranged from 0-79%, with greater awareness in participants surveyed on national guideline websites. CONCLUSION: There are many factors, not only formatting, that may affect the uptake and use of guideline-derived material by the public. Producers need to make clear how the information is relevant to the reader and how it can be used to make healthcare improvements although there were problems with data quality. Awareness of guidelines is generally low and guideline producers cannot assume that the public has a more positive perception of their material than of alternative sources of health information.

Nursing morale: what is it like and why?

BACKGROUND: There has been increased concern in the United Kingdom about the problems of recruitment and retention in the nursing profession, and how this is influenced by low morale. Despite this, however, there are relatively few studies, particularly of a qualitative nature, where nurses are asked about their morale and what factors affect it. AIMS: The present study aimed to explore nursing morale and determine the factors which nurses believed to influence it. By doing so, it was hoped that factors which affect nurse recruitment and retention could be identified. METHODS: Fifty-eight nurses (28 males, 30 females) working in the National Health Service in Scotland were interviewed in depth about their morale and their concerns about their career. Thematic analysis was carried out and a number of main issues emerged. FINDINGS: Morale in this group was very low. A large number of nurses were considering leaving the profession and the majority would discourage others from becoming a nurse. The themes that emerged, which related to their disillusionment, included low pay, lack of support for education, limited opportunity for promotion, lack of resources and job insecurity. CONCLUSIONS: The findings suggest that while recent salary increases may have helped to improve morale, other factors must also be addressed if further decline in morale and a subsequent nursing shortage is to be avoided.