Combining Machine Learning, Patient-Reported Outcomes, and Value-Based Health Care: Protocol for Scoping Reviews.

BACKGROUND: Patient-reported outcome measures (PROMs) and patient-reported experience measures (PREMs) are self-reporting tools that can measure important information about patients, such as health priorities, experience, and perception of outcome. The use of traditional objective measures such as vital signs and lab values can be supplemented with these self-reported patient measures to provide a more complete picture of a patient's health status. Machine learning, the use of computer algorithms that improve automatically through experience, is a powerful tool in health care that often does not use subjective information shared by patients. However, machine learning has largely been based on objective measures and has been developed without patient or public input. Algorithms often do not have access to critical information from patients and may be missing priorities and measures that matter to patients. Combining objective measures with patient-reported measures can improve the ability of machine learning algorithms to assess patients' health status and improve the delivery of health care. OBJECTIVE: The objective of this scoping review is to identify gaps and benefits in the way machine learning is integrated with patient-reported outcomes for the development of improved public and patient partnerships in research and health care. METHODS: We reviewed the following 3 questions to learn from existing literature about the reported gaps and best methods for combining machine learning and patient-reported outcomes: (1) How are the public engaged as involved partners in the development of artificial intelligence in medicine? (2) What examples of good practice can we identify for the integration of PROMs into machine learning algorithms? (3) How has value-based health care influenced the development of artificial intelligence in health care? We searched Ovid MEDLINE(R), Embase, PsycINFO, Science Citation Index, Cochrane Library, and Database of Abstracts of Reviews of Effects in addition to PROSPERO and the ClinicalTrials website. The authors will use Covidence to screen titles and abstracts and to conduct the review. We will include systematic reviews and overviews published in any language and may explore additional study types. Quantitative, qualitative, and mixed methods studies are included in the reviews. RESULTS: The search is completed, and Covidence software will be used to work collaboratively. We will report the review using the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines and Critical Appraisal Skills Programme for systematic reviews. CONCLUSIONS: Findings from our review will help us identify examples of good practice for how to involve the public in the development of machine learning systems as well as interventions and outcomes that have used PROMs and PREMs. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): DERR1-10.2196/36395.

Examining food insecurity and areas with unmet food needs during COVID-19: A geospatial, community-specific approach.

Food insecurity is a public health issue that has increased in the U.S. since the 2020 COVID-19 pandemic. Understanding how this increase occurs locally is crucial in informing appropriate food insecurity-related responses. Analyzing 2-1-1 call data is one way to examine food insecurity-related needs at a zip code level. The purpose of this work was to: (1) examine overall call trend data to 2-1-1 from March through July 2019 and March through July 2020, (2) examine changes in food need call volume to 2-1-1 during COVID-19 by zip code, and (3) identify areas with unmet food needs during COVID-19 in central Texas. Data for 2-1-1 calls from Travis County zip codes for March through July 2020 were compared to calls for March through July 2019 and categorized by reason for calling. Descriptive statistics and paired t-tests were used to analyze food need calls by zip code and mapped using ArcGIS. Communities with high food call volume and no emergency food assets located within the zip code were categorized as areas with unmet food needs. Results indicated there were more overall calls to 2-1-1 in 2020 (N=37,572) than in 2019 (N=28,623), and significantly more food need calls in 2020 than in 2019 (p<0.01). Eastern Travis County, a racially and ethnically diverse and lower-income area, had the largest increase in food need calls. Two zip codes were identified as having unmet food needs, which informed the strategic placement of emergency food assets. This study illustrates how 2-1-1 data can result in rapid translation of research to policy and program implementation.

Mapping Food Insecurity-Related 2-1-1 Calls in a 10-County Area of Central Texas by Zip Code: Exploring the Role of Geographic Food Access, Urbanicity and Demographic Indicators.

Food insecurity is a public health issue that affects 12% of Americans. Individuals living in food insecure households are more likely to suffer from conditions such as undernutrition, obesity and chronic diseases. Food insecurity has been linked to limited geographic access to food; however, past studies have used limited measures of access which do not fully capture the nuances of community context. The purpose of this study was to explore the association between food insecurity and geographic food access by level of urbanicity. 2-1-1 calls made in 2018 in Central Texas were classified as food needs versus non-food needs. Supermarket and convenience stores were mapped using ArcGIS. Geographic food access was operationalized as the presence of supermarkets and convenience stores: within the zip code; only in neighboring zip codes; and not located within or in neighboring zip codes. Descriptive statistics and binomial logistic regression were used to examine associations between geographic access and 2-1-1 food calls, stratified by level of urbanicity. 11% of the 2-1-1 calls made in 2018 (N = 55,405) were regarding food needs. Results showed that peri-urban and rural callers living in zip codes that only had supermarkets in neighboring zip codes had greater odds of calling about food needs compared to those that had supermarkets within the zip code. These findings indicate that geographic food access is associated with food insecurity, but this relationship varies by urbanicity. Thus, the development of food insecurity mitigation programs in peri-urban and rural areas is needed.

Managing ICU surge during the COVID-19 crisis: rapid guidelines.

Given the rapidly changing nature of COVID-19, clinicians and policy makers require urgent review and summary of the literature, and synthesis of evidence-based guidelines to inform practice. The WHO advocates for rapid reviews in these circumstances. The purpose of this rapid guideline is to provide recommendations on the organizational management of intensive care units caring for patients with COVID-19 including: planning a crisis surge response; crisis surge response strategies; triage, supporting families, and staff.

Middle-aged man who could not afford an angioplasty.

Coronary artery disease managed by percutaneous coronary intervention (PCI) has been noted for profit-driven overuse medicine. Concerns mount over inappropriate use of PCI for patients in India. We describe the case of a 55-year-old Indian man who presented for a second opinion following an urgent recommendation for PCI by two cardiologists following a recent acute myocardial infarction even though the patient was symptom-free and out of the window period for primary PCI. The proposed intervention placed the patient at financial risk for insolvency. This case report highlights the challenges and consequences of inappropriate overuse of PCI. Also, we outline the current lack of shared decision-making among patients and physicians for the PCI procedure. The challenges, inherent in the assumptions that overuse of PCI is evidence-based, are discussed including recommendations for the practice of evidence based medicine for this intervention.

What utility scores do mental health service users, healthcare professionals and members of the general public attribute to different health states? A co-produced mixed methods online survey.

BACKGROUND: Utility scores are integral to health economics decision-making. Typically, utility scores have not been scored or developed with mental health service users. The aims of this study were to i) collaborate with service users to develop descriptions of five mental health states (psychosis, depression, eating disorder, medication side effects and self-harm); ii) explore feasibility and acceptability of using scenario-based health states in an e-survey; iii) evaluate which utility measures (standard gamble (SG), time trade off (TTO) and rating scale (RS)) are preferred; and iv) determine how different participant groups discriminate between the health scenarios and rank them. DESIGN AND METHODS: This was a co-produced mixed methods cross-sectional online survey. Utility scores were generated using the SG, TTO and RS methods; difficulty of the completing each method, markers of acceptability and participants' preference were also assessed. RESULTS: A total of 119 participants (58%) fully completed the survey. For any given health state, SG consistently generated higher utility scores compared to RS and for some health states higher also than TTO (i.e. SG produces inflated utility scores relative to RS and TTO). Results suggest that different utility measures produce different evaluations of described health states. The TTO was preferred by all participant groups over the SG. The three participant groups scored four (of five) health scenarios comparably. Psychosis scored as the worst health state to live with while medication side-effects were viewed more positively than other scenarios (depression, eating disorders, self-harm) by all participant groups. However, there was a difference in how the depression scenario was scored, with service users giving depression a lower utility score compared to other groups. CONCLUSION: Mental health state scenarios used to generate utility scores can be co-produced and are well received by a broad range of participants. Utility valuations using SG, TTO and RS were feasible for use with service users, carers, healthcare professionals and members of the general public. Future studies of utility scores in psychiatry should aim to include mental health service users as both co-investigators and respondents.

Perspectives on involvement in the peer-review process: surveys of patient and public reviewers at two journals.

OBJECTIVE: In 2014/2015, The BMJ and Research Involvement and Engagement (RIE) became the first journals to routinely include patients and the public in the peer review process of journal articles. This survey explores the perspectives and early experiences of these reviewers. DESIGN: A cross-sectional survey. SETTING AND PARTICIPANTS: Patient and public reviewers for The BMJ and RIE who have been invited to review. RESULTS: The response rate was 69% (157/227) for those who had previously reviewed and 31% (67/217) for those who had not yet reviewed. Reviewers described being motivated to review by the opportunity to include the patient voice in the research process, influence the quality of the biomedical literature and ensure it meets the needs of patients. Of the 157 who had reviewed, 127 (81%) would recommend being a reviewer to other patients and carers. 144 (92%) thought more journals should adopt patient and public review. Few reviewers (16/224, 7%) reported concerns about doing open review. Annual acknowledgement on the journals' websites was welcomed as was free access to journal information. Participants were keen to have access to more online resources and training to improve their reviewing skills. Suggestions on how to improve the reviewing experience included: allowing more time to review; better and more frequent communication; a more user-friendly process; improving guidance on how to review including videos; improving the matching of papers to reviewers' experience; providing more varied sample reviews and brief feedback on the usefulness of reviews; developing a sense of community among reviewers; and publicising of the contribution that patient and public review brings. CONCLUSIONS: Patient and public reviewers shared practical ideas to improve the reviewing experience and these will be reviewed to enhance the guidance and support given to them.

Frequency of reporting on patient and public involvement (PPI) in research studies published in a general medical journal: a descriptive study.

OBJECTIVES: While documented plans for patient and public involvement (PPI) in research are required in many grant applications, little is known about how frequently PPI occurs in practice. Low levels of reported PPI may mask actual activity due to limited PPI reporting requirements. This research analysed the frequency and types of reported PPI in the presence and absence of a journal requirement to include this information. DESIGN AND SETTING: A before and after comparison of PPI reported in research papers published in The BMJ before and 1 year after the introduction of a journal policy requiring authors to report if and how they involved patients and the public within their papers. RESULTS: Between 1 June 2013 and 31 May 2014, The BMJ published 189 research papers and 1 (0.5%) reported PPI activity. From 1 June 2015 to 31 May 2016, following the introduction of the policy, The BMJ published 152 research papers of which 16 (11%) reported PPI activity. Patients contributed to grant applications in addition to designing studies through to coauthorship and participation in study dissemination. Patient contributors were often not fully acknowledged; 6 of 17 (35%) papers acknowledged their contributions and 2 (12%) included them as coauthors. CONCLUSIONS: Infrequent reporting of PPI activity does not appear to be purely due to a failure of documentation. Reporting of PPI activity increased after the introduction of The BMJ's policy, but activity both before and after was low and reporting was inconsistent in quality. Journals, funders and research institutions should collaborate to move us from the current situation where PPI is an optional extra to one where PPI is fully embedded in practice throughout the research process.

Minimal Clinical Impact of Intraoperative Examination of Sentinel Lymph Nodes in Patients With Ductal Carcinoma In Situ: An Opportunity for Improved Resource Utilization.

OBJECTIVES: There is little information regarding sentinel lymph node (SLN) frozen-section examination in patients with a history of ductal carcinoma in situ (DCIS). We evaluated the usage, clinical impact, and pathology resources used for SLN cryosectioning in mastectomy cases with a DCIS history. METHODS: Mastectomies with SLNs submitted from 2012 to 2013 at a tertiary care center were analyzed. Medicare reimbursement was used to estimate pathology health care expenditures of intraoperative frozen sections. RESULTS: There was no difference in the rate of SLN frozen-section examination or parts submitted, total blocks frozen, total blocks submitted, or total SLNs identified per case between the DCIS (n = 139) and invasive (n = 369) groups. Nine patients with DCIS had SLN metastases (three macrometastases, two micrometastases, and four isolated tumor cells), all of which were examined by frozen section. Only the macrometastases were identified by cryosectioning, which led to two synchronous axillary lymph node dissections that did not yield any additional positive nodes. A total of $19,313 was spent for pathology per DCIS patient with surgical management affected, whereas only $1,019 was spent per invasive carcinoma patient affected. CONCLUSIONS: Decreasing SLN frozen-section use in patients with a history of DCIS represents an opportunity for pathology cost containment.

Mind the gap in clinical trials: A participatory action analysis with citizen collaborators.

What are the strengths, gaps, expectations, and barriers to research engagement in clinical trials as communicated through social media? Clinical trials test treatments to provide reliable information for safety and effectiveness. Trials are building blocks in which what is learned in earlier research can be used to improve treatments, compare alternatives, and improve quality of life. For 20 years, the percentages of clinical trials volunteers have decreased whereas the costs of running clinical trials have multiplied. Participants enroll in trials to access latest treatments, to help others, and to advance science, but there is growing unrest. The priorities of those running the trials differ from those of the participants, and the roles for public research involvement lack clarity. Changes to bridge these gaps in the research culture are proposed through the use of participatory action research (PAR) in which stakeholders collaborate to improve research methodology, galvanize citizen participation, multiply health knowledge, problem-solve barriers to access, and explore the value of research volunteers as collaborators. PAR enabled the inclusion of citizens as full collaborators. Social media data were gathered for 120 days until saturation was reached. De-identified data were organized into a Strengths Weaknesses, Opportunities and Threats framework and coded into themes for analysis. After the analysis, the authors prioritized potential solutions for improving research engagement. Strengths and opportunities remained constant through trial phases, disease burdens, and interventions. Threats included alienation, litigation, disparity, and shaming. Poor management and barriers to inclusion were identified as weaknesses. Opportunities included improving resource management and information quality. Barriers were minimized when relationships between staff and participants were inclusive, respectful, tolerant, and open to change. Participants' communications ranged from fulfillment through trial involvement to disparities and rights violations. PAR provides a safe space without power imbalances in which researchers and citizen worked as equals rather than as researchers and objects of research.

Are health education interventions effective for the control and prevention of urogenital schistosomiasis in sub-Saharan Africa? A systematic review.

An estimated 129 million people are infected with urogenital schistosomiasis in sub-Saharan Africa. Current control recommendations endorse mass administration of praziquantel. Health education is an important component of effective schistosomiasis prevention and control, but there has been limited research on its effectiveness. This paper reviews the effectiveness of health education as an intervention in the prevention and control of urogenital schistosomiasis in sub-Saharan Africa. The outcomes of interest were prevalence, incidence or transmission of schistosomiasis, behaviour change associated with infection, or changes in knowledge of the disease. The findings from this review suggest that health education has a beneficial impact on knowledge and understanding of schistosomiasis within the target groups. However, further research is needed due to the poor quality of the included studies.