Implementing AI in Hospitals to Achieve a Learning Health System: Systematic Review of Current Enablers and Barriers.

BACKGROUND: Efforts are underway to capitalize on the computational power of the data collected in electronic medical records (EMRs) to achieve a learning health system (LHS). Artificial intelligence (AI) in health care has promised to improve clinical outcomes, and many researchers are developing AI algorithms on retrospective data sets. Integrating these algorithms with real-time EMR data is rare. There is a poor understanding of the current enablers and barriers to empower this shift from data set-based use to real-time implementation of AI in health systems. Exploring these factors holds promise for uncovering actionable insights toward the successful integration of AI into clinical workflows. OBJECTIVE: The first objective was to conduct a systematic literature review to identify the evidence of enablers and barriers regarding the real-world implementation of AI in hospital settings. The second objective was to map the identified enablers and barriers to a 3-horizon framework to enable the successful digital health transformation of hospitals to achieve an LHS. METHODS: The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines were adhered to. PubMed, Scopus, Web of Science, and IEEE Xplore were searched for studies published between January 2010 and January 2022. Articles with case studies and guidelines on the implementation of AI analytics in hospital settings using EMR data were included. We excluded studies conducted in primary and community care settings. Quality assessment of the identified papers was conducted using the Mixed Methods Appraisal Tool and ADAPTE frameworks. We coded evidence from the included studies that related to enablers of and barriers to AI implementation. The findings were mapped to the 3-horizon framework to provide a road map for hospitals to integrate AI analytics. RESULTS: Of the 1247 studies screened, 26 (2.09%) met the inclusion criteria. In total, 65% (17/26) of the studies implemented AI analytics for enhancing the care of hospitalized patients, whereas the remaining 35% (9/26) provided implementation guidelines. Of the final 26 papers, the quality of 21 (81%) was assessed as poor. A total of 28 enablers was identified; 8 (29%) were new in this study. A total of 18 barriers was identified; 5 (28%) were newly found. Most of these newly identified factors were related to information and technology. Actionable recommendations for the implementation of AI toward achieving an LHS were provided by mapping the findings to a 3-horizon framework. CONCLUSIONS: Significant issues exist in implementing AI in health care. Shifting from validating data sets to working with live data is challenging. This review incorporated the identified enablers and barriers into a 3-horizon framework, offering actionable recommendations for implementing AI analytics to achieve an LHS. The findings of this study can assist hospitals in steering their strategic planning toward successful adoption of AI.

From many voices, one question: Community co-design of a population-based qualitative cancer research study.

PURPOSE: This study formed the development stage of a population-based survey aiming to: (i) understand the needs and experiences of people affected by cancer in Queensland, Australia and (ii) recruit a pool of participants for ongoing cancer survivorship research. The current study aimed to co-design and test a single qualitative survey question and study invitation materials to maximise acceptability of, and participation in, the survey and future research. METHODS: Fifty-two community members, including cancer survivors and caregivers, participated across 15 co-design workshops and 20 pretest interviews. During workshops, participants generated and refined ideas for an open-ended survey question and provided feedback on a study invitation letter. The use of a single, open-ended question aims to minimise participant burden while collecting rich information about needs and experiences. The research team then shortlisted the question ideas and revised study invitation materials based on workshop feedback. Next, using interviews, community members were asked to respond to a shortlisted question to test its interpretability and relevance and to review revised invitation materials. Content analysis of participant feedback was used to identify principles for designing study materials. RESULTS: Principles for designing qualitative survey questions were identified from participant feedback, including define the question timeframe and scope; provide reassurance that responses are valid and valued; and use simple wording. Principles for designing study invitation materials were also identified, including communicate empathy and sensitivity; facilitate reciprocal benefit; and include a 'human element'. The qualitative survey question and study invitation materials created using these principles were considered relevant and acceptable for use in a population-based survey. CONCLUSIONS: Through community consultation and co-design, this study identified principles for designing qualitative data collection and invitation materials for use in cancer survivorship research. These principles can be applied by other researchers to develop study materials that are sensitive to the needs and preferences of community members.

Examining emotional and behavioural trajectories in siblings of children with life-limiting conditions.

BACKGROUND: Healthy siblings of children with life-limiting conditions often experience emotional and behavioural struggles over the course of the ill child's condition(s). Resources to support these siblings are limited due to a lack of understanding about their needs. Therefore, this study was designed to characterize the emotional and behavioural trajectories among siblings of children with progressive, life-limiting genetic, metabolic, or neurological conditions over a 12-month observation period. METHODS: Seventy siblings were recruited from a large-survey based study (Charting the Territory) that examined the bio-psychosocial health outcomes of parents and siblings. Linear mixed effect models were used to assess the association between siblings' emotions and behaviour trajectories and selected demographic variables. Siblings' emotions and behaviour were measured with Child Behaviour Checklist (CBCL). RESULTS: Siblings' mean age was 11.2 years at baseline and Internalizing, Externalizing, and Total Behaviour Problems mean scores were within normal ranges across time. However, 7-25% of siblings had scores within the clinical range. Brothers had higher levels of Internalizing Problems than sisters, whereas sisters had higher levels of Externalizing Problems than brothers. When treatment was first sought for the ill child less than a year prior to study participation, siblings had higher levels of Internalizing and Externalizing Problems compared with siblings who participated more than one year after treatment was sought. CONCLUSION: Healthy siblings experience emotional and behavioural problems early in the child's disease trajectory. Although these problems improve with time, our findings show that brothers and sisters experience different types of challenges. Therefore, timely support for siblings is important as they navigate through the uncertainties and challenges.

What is measured matters: A scoping review of analysis methods used for qualitative patient reported experience measure data.

INTRODUCTION: Hospitals are increasingly turning to patients for valuable feedback regarding their care experience. A common method to collect this information is patient reported experience measures (PREMs) surveys. Health care workers report qualitative PREMs as more interesting, relevant, and informative than quantitative survey responses. However, a major barrier to utilising qualitative PREMs data to drive quality improvements is a lack of resources to analyse the data. This scoping review aimed to review the methods used to analyse qualitative PREMs survey data from routine hospital care. METHODS: We utilised the JBI scoping review methodology, and searched four databases for articles from 2013 to 2023 which analysed qualitative PREMs survey data from routine care in hospitals. Study characteristics were extracted, as well as the analysis method - specifically, whether the study used traditional manual analysis methods in which the researcher reads the text and categorise the data, or automated methods utilising computers and algorithms to read and categorise the data. RESULTS: From 960 unique articles, 123 went through full-text review and 54 were deemed eligible. 75.9 % used only manual content analysis methods to analyse the qualitative responses, 16.7 % of studies used a combination of manual and automated methods, and only 7.4 % used exclusively automated methods. Automated methods were used in 27.5 % of studies published 2019-2023, compared to 14.3 % of studies published 2013-2018. All bar one study using automated methods focused on investigating the validity of the automated methodology or used it to complement manual content analysis. CONCLUSION: The studies included in this review show a transition from traditional time-consuming manual analyses to computerised methods enabling analysis at a larger scale. As the volume of PREMs data collected grows, efficient and effective ways to analyse qualitative PREMs data at scale are required to enable health services to capture the patient voice and drive consumer-centred improvements in care.

Population-level 5-year event-free survival for children with cancer in Australia.

BACKGROUND: Event-free survival (EFS) considers other adverse events in addition to mortality. It therefore provides a more complete understanding of the effectiveness and consequences of treatment than standard survival measures, but is rarely reported at the population level for childhood cancer. PROCEDURE: Our study cohort (n = 7067) was obtained from the Australian Childhood Cancer Registry, including children aged under 15 diagnosed with cancer between 2006 and 2015, with follow-up potentially available to 31 December 2020. The events of interest were relapse following remission, progressive disease, diagnosis of a second primary cancer or death from any cause. Five-year EFS and all-cause observed survival were both calculated, stratified by type of childhood cancer, remoteness of residence and stage at diagnosis. Differences in EFS were assessed using multivariable flexible parametric models. RESULTS: Approximately one quarter of patients (n = 1605 of 7067, 23%) experienced at least one of the events of interest within 5 years of diagnosis. Relapse was twice as common for children with metastatic/advanced disease (22%) versus children with localised/limited cancers (11%). Overall 5-year EFS was 75.0% (95% confidence interval [CI]: 73.9%-76.0%), compared to 85.8% observed survival (95% CI: 85.0%-86.6%). Patients with other gliomas had the lowest EFS (35.4%, 95% CI: 27.8%-43.1%). EFS was significantly lower among children with acute myeloid leukaemia in outer regional/remote areas compared to major cities (adjusted hazard ratio [HR] = 1.90, 95% CI: 1.20-3.00). CONCLUSIONS: Reporting EFS at a population level provides further insight on a wider range of impacts apart from mortality alone, contributing towards efforts to improve the management and outcomes of childhood cancer.

Preterm Birth, Family Income, and Intergenerational Income Mobility.

Importance: Preterm birth (PTB) has been associated with lower income in adulthood, but associations with intergenerational income mobility and the role of family socioeconomic status (SES) as modifying factor are unclear. Objectives: To assess whether the association between PTB and income differs according to family SES at birth and to assess the association between PTB and intergenerational income mobility. Design, Setting, and Participants: This study comprised a matched cohort of live births in Canada between January 1, 1990, and December 31, 1996, with follow-up until December 31, 2018. Statistical analysis was performed between May 2023 and March 2024. Exposure: Preterm birth, defined as birth between 24 and 37 weeks' gestational age (with gestational age subcategories of 34-36, 32-33, 28-31, and 24-27 weeks) vs early and full term births (gestational age, 37-41 weeks). Main Outcomes and Measures: Associations between PTB and annual adulthood income in 2018 Canadian dollars were assessed overall (current exhange rate: $1 = CAD $1.37) and stratified by family income quintiles, using generalized estimating equation regression models. Associations between PTB and percentile rank change (ie, difference between the rank of individuals and their parents in the income distribution within their respective generations) and upward or downward mobility (based on income quintile) were assessed using linear and multinomial logistic regressions, respectively. Results: Of 1.6 million included births (51.1% boys and 48.9% girls), 6.9% infants were born preterm (5.4% born at 34-36 weeks, 0.7% born at 32-33 weeks, 0.5% born at 28-31 weeks, and 0.2% born at 24-27 weeks). After matching on baseline characteristics (eg, sex, province of birth, and parental demographics) and adjusting for age and period effects, PTB was associated with lower annual income (mean difference, CAD -$687 [95% CI, -$788 to -$586]; 3% lower per year), and the differences were greater among those belonging to families in the lowest family SES quintile (mean difference, CAD -$807 [95% CI, -$998 to -$617]; 5% lower per year). Preterm birth was also associated with lower upward mobility and higher downward mobility, particularly for those born earlier than 31 weeks' gestational age (24-27 weeks: mean difference in percentile rank change, -8.7 percentile points [95% CI, -10.5 to -6.8 percentile points]). Conclusions and Relevance: In this population-based matched cohort study, PTB was associated with lower adulthood income, lower upward social mobility, and higher downward mobility, with greater differences among those belonging to economically disadvantaged families. Interventions to optimize socioeconomic outcomes of preterm-born individuals would need to define target population considering SES.

Corrigendum: Second malignant neoplasms within 5 years from first primary diagnosis in pediatric oncology patients in Canada: a population-based retrospective cohort study.

[This corrects the article DOI: 10.3389/fonc.2024.1376652.].

Childhood cancer models of survivorship care: a scoping review of elements of care and reported outcomes.

PURPOSE: This study aimed to systematically map elements of care and respective outcomes described in the literature for different models of post-treatment care for survivors of childhood cancer. METHODS: MEDLINE, CINAHL, and Embase were searched with combinations of free text terms, synonyms, and MeSH terms using Boolean operators and are current to January 2024. We included studies that described post-treatment cancer survivorship models of care and reported patient or service level elements of care or outcomes, which we mapped to the Quality of Cancer Survivorship Care Framework domains. RESULTS: Thirty-eight studies with diverse designs were included representing 6101 childhood cancer survivors (or their parent/caregiver) and 14 healthcare professionals. A diverse range of models of care were reported, including paediatric oncologist-led long-term follow-up, multi-disciplinary survivorship clinics, shared-care, and primary care-led follow-up. Elements of care at the individual level most commonly included surveillance for cancer recurrence as well as assessment of physical and psychological effects. At the service level, satisfaction with care was frequently reported but few studies reported how treatment-related-late effects were managed. The evidence does not support one model of care over another. CONCLUSIONS: Gaps in evidence exist regarding distal outcomes such as costs, health care utilization, and mortality, as well as understanding outcomes of managing chronic disease and physical or psychological effects. The findings synthesized in this review provide a valuable reference point for future service planning and evaluation. IMPLICATIONS FOR CANCER SURVIVORS: Decades of research highlight the importance of survivorship care for childhood cancer survivors who are at risk of serious treatment-related late effects. This review emphasizes there is no single, 'one-size fits all' approach for delivering such care to this vulnerable population.

Second malignant neoplasms within 5 years from first primary diagnosis in pediatric oncology patients in Canada: a population-based retrospective cohort study.

Introduction: From the advancement of treatment of pediatric cancer diagnosis, the five-year survival rate has increased significantly. However, the adverse consequence of improved survival rate is the second malignant neoplasm. Although previous studies provided information on the incidence and risk of SMN in long term survivors of childhood cancer, there is still scarce information known for short term (< 5 years) prognosis. This study aims to assess the incidence, characteristics, management, and outcome of children who develop SMN malignancies within 5 years of diagnosis of their initial cancer. Method: This is a retrospective cohort study of early Second Malignant Neoplasms (SMN) in pediatric oncology patients. The Cancer in Young People - Canada (CYP-C) national pediatric cancer registry was used and reviewed pediatric patients diagnosed with their first cancer from 2000-2015. Results: A total of 20,272 pediatric patients with a diagnosis of a first malignancy were analyzed. Of them, 0.7% were diagnosed with a SMN within the first 5 years following their first cancer diagnosis. Development of a SMN impacted survival, shown by an inferior survival rate in the SMN cohort (79.1%) after three years compared to that of the non-SMN cohort (89.7%). Several possible risk factors have been identified in the study including the use of epipodophyllotoxins, exposure to radiation, and hematopoietic stem cell 169 transplant. Discussion: This is the first national study assessing the incidence, 170 characteristics, risk factors and outcome of early SMN in Canadian children 171 from age 0-15 from 2000-2015.