Using neural networks to support high-quality evidence mapping.

BACKGROUND: The Living Evidence Map Project at the Norwegian Institute of Public Health (NIPH) gives an updated overview of research results and publications. As part of NIPH's mandate to inform evidence-based infection prevention, control and treatment, a large group of experts are continously monitoring, assessing, coding and summarising new COVID-19 publications. Screening tools, coding practice and workflow are incrementally improved, but remain largely manual. RESULTS: This paper describes how deep learning methods have been employed to learn classification and coding from the steadily growing NIPH COVID-19 dashboard data, so as to aid manual classification, screening and preprocessing of the rapidly growing influx of new papers on the subject. Our main objective is to make manual screening scalable through semi-automation, while ensuring high-quality Evidence Map content. CONCLUSIONS: We report early results on classifying publication topic and type from titles and abstracts, showing that even simple neural network architectures and text representations can yield acceptable performance.

Femoral nerve blocks for the treatment of acute pre-hospital pain: A systematic review with meta-analysis.

BACKGROUND: Pain management is one of the most important interventions in the emergency medical services. The femoral nerve block (FNB) is, among other things, indicated for pre- and post-operative pain management for patients with femoral fractures but its role in the pre-hospital setting has not been determined. The aim of this review was to assess the effect and safety of the FNB in comparison to other forms of analgesia (or no treatment) for managing acute lower extremity pain in adult patients in the pre-hospital setting. METHODS: A systematic review (PROSPERO registration (CRD42018114399)) was conducted. The Cochrane and GRADE methods were used to assess outcomes. Two authors independently reviewed each study for eligibility, extracted the data and performed risk of bias assessments. RESULTS: Four studies with a total of 252 patients were included. Two RCTs (114 patients) showed that FNB may reduce pain more effectively than metamizole (mean difference 32 mm on a 100 mm VAS (95% CI 24 to 40)). One RCT (48 patients) compared the FNB with lidocaine and magnesium sulphate to FNB with lidocaine alone and was only included here for information regarding adverse effects. One case series included 90 patients. Few adverse events were reported in the included studies. The certainty of evidence was very low. We found no studies comparing FNB to inhaled analgesics, opioids or ketamine. CONCLUSIONS: Evidence regarding the effectiveness and adverse effects of pre-hospital FNB is limited. Studies comparing pre-hospital FNB to inhaled analgesics, opioids or ketamine are lacking.

Targeted mass media interventions promoting healthy behaviours to reduce risk of non-communicable diseases in adult, ethnic minorities.

BACKGROUND: Physical activity, a balanced diet, avoidance of tobacco exposure, and limited alcohol consumption may reduce morbidity and mortality from non-communicable diseases (NCDs). Mass media interventions are commonly used to encourage healthier behaviours in population groups. It is unclear whether targeted mass media interventions for ethnic minority groups are more or less effective in changing behaviours than those developed for the general population. OBJECTIVES: To determine the effects of mass media interventions targeting adult ethnic minorities with messages about physical activity, dietary patterns, tobacco use or alcohol consumption to reduce the risk of NCDs. SEARCH METHODS: We searched CENTRAL, MEDLINE, Embase, PsycINFO, CINAHL, ERIC, SweMed+, and ISI Web of Science until August 2016. We also searched for grey literature in OpenGrey, Grey Literature Report, Eldis, and two relevant websites until October 2016. The searches were not restricted by language. SELECTION CRITERIA: We searched for individual and cluster-randomised controlled trials, controlled before-and-after studies (CBA) and interrupted time series studies (ITS). Relevant interventions promoted healthier behaviours related to physical activity, dietary patterns, tobacco use or alcohol consumption; were disseminated via mass media channels; and targeted ethnic minority groups. The population of interest comprised adults (≥ 18 years) from ethnic minority groups in the focal countries. Primary outcomes included indicators of behavioural change, self-reported behavioural change and knowledge and attitudes towards change. Secondary outcomes were the use of health promotion services and costs related to the project. DATA COLLECTION AND ANALYSIS: Two authors independently reviewed the references to identify studies for inclusion. We extracted data and assessed the risk of bias in all included studies. We did not pool the results due to heterogeneity in comparisons made, outcomes, and study designs. We describe the results narratively and present them in 'Summary of findings' tables. We judged the quality of the evidence using the GRADE (Grading of Recommendations Assessment, Development, and Evaluation) methodology. MAIN RESULTS: Six studies met the inclusion criteria, including three RCTs, two cluster-RCTs and one ITS. All were conducted in the USA and comprised targeted mass media interventions for people of African descent (four studies), Spanish-language dominant Latino immigrants (one study), and Chinese immigrants (one study). The two latter studies offered the intervention in the participants' first language (Spanish, Cantonese, or Mandarin). Three interventions targeted towards women only, one pregnant women specifically. We judged all studies as being at unclear risk of bias in at least one domain and three studies as being at high risk of bias in at least one domain.We categorised the findings into three comparisons. The first comparison examined mass media interventions targeted at ethnic minorities versus an equivalent mass media intervention intended for the general population. The one study in this category (255 participants of African decent) found little or no difference in effect on self-reported behavioural change for smoking and only small differences in attitudes to change between participants who were given a culturally specific smoking cessation booklet versus a booklet intended for the general population. We are uncertain about the effect estimates, as assessed by the GRADE methodology (very low quality evidence of effect). No study provided data for indicators of behavioural change or adverse effects.The second comparison assessed targeted mass media interventions versus no intervention. One study (154 participants of African decent) reported effects for our primary outcomes. Participants in the intervention group had access to 12 one-hour live programmes on cable TV and received print material over three months regarding nutrition and physical activity to improve health and weight control. Change in body mass index (BMI) was comparable between groups 12 months after the baseline (low quality evidence). Scores on a food habits (fat behaviours) and total leisure activity scores changed favourably for the intervention group (very low quality evidence). Two other studies exposed entire populations in geographical areas to radio advertisements targeted towards African American communities. Authors presented effects on two of our secondary outcomes, use of health promotion services and project costs. The campaign message was to call smoking quit lines. The outcome was the number of calls received. After one year, one study reported 18 calls per estimated 10,000 targeted smokers from the intervention communities (estimated target population 310,500 persons), compared to 0.2 calls per estimated 10,000 targeted smokers from the control communities (estimated target population 331,400 persons) (moderate quality evidence). The ITS study also reported an increase in the number of calls from the target population during campaigns (low quality evidence). The proportion of African American callers increased in both studies (low to very low quality evidence). No study provided data on knowledge and attitudes for change and adverse effects. Information on costs were sparse.The third comparison assessed targeted mass media interventions versus a mass media intervention plus personalised content. Findings are based on three studies (1361 participants). Participants in these comparison groups received personal feedback. Two of the studies recorded weight changes over time. Neither found significant differences between the groups (low quality evidence). Evidence on behavioural changes, and knowledge and attitudes typically found some effects in favour of receiving personalised content or no significant differences between groups (very low quality evidence). No study provided data on adverse effects. Information on costs were sparse. AUTHORS' CONCLUSIONS: The available evidence is inadequate for understanding whether mass media interventions targeted toward ethnic minority populations are more effective in changing health behaviours than mass media interventions intended for the population at large. When compared to no intervention, a targeted mass media intervention may increase the number of calls to smoking quit line, but the effect on health behaviours is unclear. These studies could not distinguish the impact of different components, for instance the effect of hearing a message regarding behavioural change, the cultural adaptation to the ethnic minority group, or increase reach to the target group through more appropriate mass media channels. New studies should explore targeted interventions for ethnic minorities with a first language other than the dominant language in their resident country, as well as directly compare targeted versus general population mass media interventions.

A systematic review of hospital accreditation: the challenges of measuring complex intervention effects.

BACKGROUND: The increased international focus on improving patient outcomes, safety and quality of care has led stakeholders, policy makers and healthcare provider organizations to adopt standardized processes for evaluating healthcare organizations. Accreditation and certification have been proposed as interventions to support patient safety and high quality healthcare. Guidelines recommend accreditation but are cautious about the evidence, judged as inconclusive. The push for accreditation continues despite sparse evidence to support its efficiency or effectiveness. METHODS: We searched MEDLINE, EMBASE and The Cochrane Library using Medical Subject Headings (MeSH) indexes and keyword searches in any language. Studies were assessed using the Cochrane Risk of Bias Tool and AMSTAR framework. 915 abstracts were screened and 20 papers were reviewed in full in January 2013. Inclusion criteria included studies addressing the effect of hospital accreditation and certification using systematic reviews, randomized controlled trials, observational studies with a control group, or interrupted time series. Outcomes included both clinical outcomes and process measures. An updated literature search in July 2014 identified no new studies. RESULTS: The literature review uncovered three systematic reviews and one randomized controlled trial. The lone study assessed the effects of accreditation on hospital outcomes and reported inconsistent results. Excluded studies were reviewed and their findings summarized. CONCLUSION: Accreditation continues to grow internationally but due to scant evidence, no conclusions could be reached to support its effectiveness. Our review did not find evidence to support accreditation and certification of hospitals being linked to measurable changes in quality of care as measured by quality metrics and standards. Most studies did not report intervention context, implementation, or cost. This might reflect the challenges in assessing complex, heterogeneous interventions such as accreditation and certification. It is also may be magnified by the impact of how accreditation is managed and executed, and the varied financial and organizational healthcare constraints. The strategies hospitals should impelment to improve patient safety and organizational outcomes related to accreditation and certification components remains unclear.

A case study of the organisation of research knowledge by systematic reviews within the mental health care field.

BACKGROUND: Our aim was to explore how systematic reviews of the effects of complex interventions within the field of organisation of mental health care organised research knowledge. METHODS: We looked up references included in 14 mental health care reviews in the database Epistemonikos to examine overlap between reviews. We classified topic concordance between study reports and reviews as similar, narrower, broader or other topic to indicate how well the research knowledge was organised. FINDINGS: We examined 182 comparisons between the theme of study reports and reviews. In 100 (55%) of the 182 comparisons, the review had a broader theme, 18 (10%) narrower, 34 (19%) similar and in 30 (16%), the review's theme was classified as 'other'. The content analysis indicated that there existed (inconsistent) overlapping of study reports between different topics. CONCLUSIONS: Within the field organisation of mental health care, there exists some unclear and inconsistent organisation of current knowledge. This may diminish the validity and reliability of systematic reviews. It is important that review authors take care in defining the review question precisely, conduct thorough literature searches, consider each study's hypothesis, disciplinary tradition and context, and if necessary, examine which other reviews have included a study in question.

Interventions aimed at increasing the level of physical activity by including organised follow-up: a systematic review of effect.

BACKGROUND: Organised follow-up is a common feature of several strategies at the primary health care level to promote health behaviour change, e.g. to increase physical activity. In Norway, municipal 'healthy living' centres run by health care personnel are established to offer counselling and organised follow-up of health behaviour change during a 12-week programme. We report the results of a systematic review commissioned by the Norwegian Directorate of Health concerning organised follow-up to improve physical activity. METHODS: We searched ten electronic databases up to June 2012, reference lists of included publications, and relevant journals. Study selection and quality risk of bias assessment were carried out independently. Data were synthesised narratively due to heterogeneity of measurements of physical activity. The GRADE approach was used to assess our confidence in the effect estimate for each outcome in each comparison. RESULTS: Fourteen randomised controlled trials from seven countries and with a total of 5,002 participants were included in the systematic review. All studies were carried out in primary care or community settings. The interventions comprised referral to supervised group physical activity (2 studies), referral to local resources with follow-up (6 studies), and self-organised physical activity with follow-up (6 studies). The narrative synthesis, comprising a total of 39 comparisons, indicated effects of self-organised physical activity with follow-up (compared to both advice and no treatment) and referral to local resources with follow-up (compared to advice) in some of the comparisons where we rated our confidence in the effect estimates as moderate. However, the results indicated no difference between intervention and control groups for the majority of comparisons. Follow-up in the studies was mainly short-term with the longest follow-up 9 months post-treatment. We rated our confidence in the effect estimates as low or very low in most comparisons, both for positive and neutral results. CONCLUSIONS: The results of this systematic review indicate considerable uncertainty concerning effects of organised follow-up during 10-14 weeks on physical activity. Major methodological problems concerning the measurement of physical activity are discussed. TRIAL REGISTRATION: Systematic review registration: PROSPERO CRD42011001598.

Accessible methods and tools to estimate chemical exposure in humans to support risk assessment: A systematic scoping review.

Exposure assessment is a crucial component of environmental health research, providing essential information on the potential risks associated with various chemicals. A systematic scoping review was conducted to acquire an overview of accessible human exposure assessment methods and computational tools to support and ultimately improve risk assessment. The systematic scoping review was performed in Sysrev, a web platform that introduces machine learning techniques into the review process aiming for increased accuracy and efficiency. Included publications were restricted to a publication date after the year 2000, where exposure methods were properly described. Exposure assessments methods were found to be used for a broad range of environmental chemicals including pesticides, metals, persistent chemicals, volatile organic compounds, and other chemical classes. Our results show that after the year 2000, for all the types of exposure routes, probabilistic analysis, and computational methods to calculate human exposure have increased. Sixty-three mathematical models and toolboxes were identified that have been developed in Europe, North America, and globally. However, only twelve occur frequently and their usefulness were associated with exposure route, chemical classes and input parameters used to estimate exposure. The outcome of the combined associations can function as a basis and/or guide for decision making for the selection of most appropriate method and tool to be used for environmental chemical human exposure assessments in Ontology-driven and artificial intelligence-based repeated dose toxicity testing of chemicals for next generation risk assessment (ONTOX) project and elsewhere. Finally, the choice of input parameters used in each mathematical model and toolbox shown by our analysis can contribute to the harmonization process of the exposure models and tools increasing the prospect for comparison between studies and consistency in the regulatory process in the future.

Cross-mapping of terms used in chemical risk assessment with those used in systematic review: research protocol.

The focus on implementation of systematic review (SR) principles in chemical risk assessments (CRAs) is growing as it has the potential to advance the rigour and transparency of the CRAs. However, the SR and CRA communities use their own specific terminologies. Understanding the meaning of core SR and CRA terms and where they overlap is critical for application of SR methods and principles in CRAs. Moreover, it will increase the possibility for cross-sectorial collaboration, avoid misunderstandings, and improve communication among risk assessors, researchers, and policy makers. We present a process for the cross-mapping of core CRA terms and core SR terms. Core terms for study appraisal, evidence synthesis and integration used in the SR and CRA communities will be included. The outcome will be an overview of how core SR terms map onto core CRA terms and vice versa, and a description of the relationship and conceptual overlap between the terms. The cross-mapping is divided in three phases, where in the first phase the core SR and CRA terms will be identified. In the second phase, existing SR and CRA definitions will be mapped. In the third phase, descriptions of the relationship and conceptual overlap between the terms will be derived. The third phase will include weekly one-hour online meetings for SR and CRA experts.

Use of Swedish smokeless tobacco during pregnancy: A systematic review of pregnancy and early life health risk.

BACKGROUND AND AIMS: Smokeless tobacco is a heterogeneous product group with diverse composition and prevalence globally. Tobacco use during pregnancy is concerning due to the risk of adverse pregnancy outcomes and effects on child health. Nicotine may mediate several of these effects. This systematic review measured health outcomes from Swedish smokeless tobacco (snus) use during pregnancy. METHOD: Literature search was conducted by an information specialist in May 2022. We included human studies of snus use during pregnancy compared with no tobacco use, assessed risk of bias, conducted a meta-analysis and assessed confidence in effect-estimates using Grading of Recommendations, Assessment, Development and Evaluations (GRADE). RESULTS: We included 18 cohort studies (42 to 1 006 398 participants). Snus use during pregnancy probably (moderate confidence in risk estimates) increase the risk of neonatal apnea, adjusted odds ratio 95% confidence interval [aOR (95% CI)] 1.96 (1.30 to 2.96). Snus use during pregnancy possibly (low confidence in risk estimates) increase the risk of stillbirths aOR 1.43 (1.02 to 1.99), extremely premature births aOR 1.69 (1.17 to 2.45), moderately premature birth aOR 1.26 (1.15 to 1.38), SGA aOR 1.26 (1.09 to 1.46), reduced birth weight mean difference of 72.47 g (110.58 g to 34.35 g reduction) and oral cleft malformations aOR 1.48 (1.00 to 2.21). It is uncertain (low confidence in risk estimates, CI crossing 1) whether snus use during pregnancy affects risk of preeclampsia aOR 1.11 (0.97 to 1.28), antenatal bleeding aOR 1.15 (0.92 to 1.44) and very premature birth aOR 1.26 (0.95 to 1.66). Risk of early neonatal mortality and altered heart rate variability is uncertain, very low confidence. Snus using mothers had increased prevalence of caesarean sections, low confidence. CONCLUSIONS: This systematic review reveals that use of smokeless tobacco (snus) during pregnancy may adversely impact the developing child.