Fatigue and vigilance in medical experts detecting breast cancer.

An abundance of laboratory-based experiments has described a vigilance decrement of reducing accuracy to detect targets with time on task, but there are few real-world studies, none of which have previously controlled the environment to control for bias. We describe accuracy in clinical practice for 360 experts who examined >1 million women's mammograms for signs of cancer, whilst controlling for potential biases. The vigilance decrement pattern was not observed. Instead, test accuracy improved over time, through a reduction in false alarms and an increase in speed, with no significant change in sensitivity. The multiple-decision model explains why experts miss targets in low prevalence settings through a change in decision threshold and search quit threshold and propose it should be adapted to explain these observed patterns of accuracy with time on task. What is typically thought of as standard and robust research findings in controlled laboratory settings may not directly apply to real-world environments and instead large, controlled studies in relevant environments are needed.

Software using artificial intelligence for nodule and cancer detection in CT lung cancer screening: systematic review of test accuracy studies.

OBJECTIVES: To examine the accuracy and impact of artificial intelligence (AI) software assistance in lung cancer screening using CT. METHODS: A systematic review of CE-marked, AI-based software for automated detection and analysis of nodules in CT lung cancer screening was conducted. Multiple databases including Medline, Embase and Cochrane CENTRAL were searched from 2012 to March 2023. Primary research reporting test accuracy or impact on reading time or clinical management was included. QUADAS-2 and QUADAS-C were used to assess risk of bias. We undertook narrative synthesis. RESULTS: Eleven studies evaluating six different AI-based software and reporting on 19 770 patients were eligible. All were at high risk of bias with multiple applicability concerns. Compared with unaided reading, AI-assisted reading was faster and generally improved sensitivity (+5% to +20% for detecting/categorising actionable nodules; +3% to +15% for detecting/categorising malignant nodules), with lower specificity (-7% to -3% for correctly detecting/categorising people without actionable nodules; -8% to -6% for correctly detecting/categorising people without malignant nodules). AI assistance tended to increase the proportion of nodules allocated to higher risk categories. Assuming 0.5% cancer prevalence, these results would translate into additional 150-750 cancers detected per million people attending screening but lead to an additional 59 700 to 79 600 people attending screening without cancer receiving unnecessary CT surveillance. CONCLUSIONS: AI assistance in lung cancer screening may improve sensitivity but increases the number of false-positive results and unnecessary surveillance. Future research needs to increase the specificity of AI-assisted reading and minimise risk of bias and applicability concerns through improved study design. PROSPERO REGISTRATION NUMBER: CRD42021298449.

Optimising breast cancer screening reading: blinding the second reader to the first reader's decisions.

OBJECTIVES: In breast cancer screening, two readers separately examine each woman's mammograms for signs of cancer. We examined whether preventing the two readers from seeing each other's decisions (blinding) affects behaviour and outcomes. METHODS: This cohort study used data from the CO-OPS breast-screening trial (1,119,191 women from 43 screening centres in England) where all discrepant readings were arbitrated. Multilevel models were fitted using Markov chain Monte Carlo to measure whether reader 2 conformed to the decisions of reader 1 when they were not blinded, and the effect of blinding on overall rates of recall for further tests and cancer detection. Differences in positive predictive value (PPV) were assessed using Pearson's chi-squared test. RESULTS: When reader 1 recalls, the probability of reader 2 also recalling was higher when not blinded than when blinded, suggesting readers may be influenced by the other's decision. Overall, women were less likely to be recalled when reader 2 was blinded (OR 0.923; 95% credible interval 0.864, 0.986), with no clear pattern in cancer detection rate (OR 1.029; 95% credible interval 0.970, 1.089; Bayesian p value 0.832). PPV was 22.1% for blinded versus 20.6% for not blinded (p < 0.001). CONCLUSIONS: Our results suggest that when not blinded, reader 2 is influenced by reader 1's decisions to recall (alliterative bias) which would result in bypassing arbitration and negate some of the benefits of double-reading. We found a relationship between blinding the second reader and slightly higher PPV of breast cancer screening, although this analysis may be confounded by other centre characteristics. KEY POINTS: • In Europe, it is recommended that breast screening mammograms are analysed by two readers but there is little evidence on the effect of 'blinding' the readers so they cannot see each other's decisions. • We found evidence that when the second reader is not blinded, they are more likely to agree with a recall decision from the first reader and less likely to make an independent judgement (alliterative error). This may reduce overall accuracy through bypassing arbitration. • This observational study suggests an association between blinding the second reader and higher positive predictive value of screening, but this may be confounded by centre characteristics.

Testing for lynch syndrome in people with endometrial cancer using immunohistochemistry and microsatellite instability-based testing strategies - A systematic review of test accuracy.

BACKGROUND: Lynch syndrome is an inherited genetic condition that is associated with an increased risk of cancer, including endometrial and colorectal cancer. We assessed the test accuracy of immunohistochemistry and microsatellite instability-based testing (with or without MLH1 promoter methylation testing) for Lynch syndrome in women with endometrial cancer. METHODS: We conducted a systematic review of literature published up to August 2019. We searched bibliographic databases, contacted experts and checked reference lists of relevant studies. Two reviewers conducted each stage of the review. RESULTS: Thirteen studies were identified that included approximately 3500 participants. None of the studies was at low risk of bias in all domains. Data could not be pooled due to the small number of heterogeneous studies. Sensitivity ranged from 60.7-100% for immunohistochemistry, 41.7-100% for microsatellite instability-based testing, and 90.5-100% for studies combining immunohistochemistry, microsatellite instability-based testing, and MLH1 promoter methylation testing. Specificity ranged from 60.9-83.3% (excluding 1 study with highly selective inclusion criteria) for immunohistochemistry, 69.2-89.9% for microsatellite instability-based testing, and 72.4-92.3% (excluding 1 study with highly selective inclusion criteria) for testing strategies that included immunohistochemistry, microsatellite instability-based testing, and MLH1 promoter methylation. We found no statistically significant differences in test accuracy estimates (sensitivity, specificity) in head-to-head studies of immunohistochemistry versus microsatellite instability-based testing. Reported test failures were rare. CONCLUSIONS: Sensitivity of the index tests were generally high, though most studies had much lower specificity. We found no evidence that test accuracy differed between IHC and MSI based strategies. The evidence base is currently small and at high risk of bias.

The use of electronic healthcare records for colorectal cancer screening referral decisions and risk prediction model development.

BACKGROUND: The database used for the NHS Bowel Cancer Screening Programme (BCSP) derives participant information from primary care records. Combining predictors with FOBTs has shown to improve referral decisions and accuracy. The richer data available from GP databases could be used to complement screening referral decisions by identifying those at greatest risk of colorectal cancer. We determined the availability of data for key predictors and whether this information could be used to inform more accurate screening referral decisions. METHODS: An English BCSP cohort was derived using the electronic notifications received from the BCSP database to GP records. The cohort covered a period between 13th May 2009 to 17th January 2017. Completeness of variables and univariable associations were assessed. Risk prediction models were developed using Cox regression and multivariable fractional polynomials with backwards elimination. Optimism adjusted performance metrics were reported. The sensitivity and specificity of a combined approach using the negative FOBT model plus FOBT positive patients was determined using a probability equivalent to a 3% PPV NICE guidelines level. RESULTS: 292,059 participants aged 60-74 were derived for the BCSP screening cohort. A model including the screening test result had a C-statistic of 0.860, c-slope of 0.997, and R2 of 0.597. A model developed for negative screening results only had a C-statistic of 0.597, c-slope of 0.940, and R2 of 0.062. Risk predictors included in the models included; age, sex, alcohol consumption, IBS diagnosis, family history of gastrointestinal cancer, smoking status, previous negatives and whether a GP had ordered a blood test. For the combined screening approach, sensitivity increased slightly from 53.90% (FOBT only) to 58.82% but at the expense of an increased referral rate. CONCLUSIONS: This research has identified several potential predictors for CRC in a BCSP population. A risk prediction model developed for BCSP FOBT negative patients was not clinically useful due to a low sensitivity and increased referral rate. The predictors identified in this study should be investigated in a refined algorithm combining the quantitative FIT result. Combining data from multiple sources enables fuller patient profiles using the primary care and screening database interface.

Antenatal screening for fetal trisomies using microarray-based cell-free DNA testing: A systematic review and meta-analysis.

OBJECTIVE: To evaluate the test accuracy of non-invasive prenatal testing (NIPT) for fetal trisomy 21, 18, and 13 using cell-free (cf) DNA analysis in maternal plasma with microarray quantitation. METHOD: Systematic review and meta-analysis. Searches in MEDLINE, Pre-MEDLINE, EMBASE, Web of Science, and the Cochrane Library to 09.07.2018. RESULTS: Five studies analyzing 3074 samples, including 187 trisomy 21, 43 trisomy 18, and 19 trisomy 13 cases, were identified. Risk of bias was high in all studies, introduced particularly by exclusions from analysis and by the role of the sponsor. Sensitivity of microarray-based cfDNA testing was 99.5% (95%CI 96.3%-99.9%) for trisomy 21, 97.7% (95%CI 87.9%-99.6%) for trisomy 18, and 100% (95%CI 83.2%-100%) for trisomy 13. Specificity was 100% (95% CI 99.87%-100%) for trisomy 21, 99.97% (95%CI 99.81%-99.99%) for trisomy 18, and 99.97% (95%CI 99.81%-99.99%) for trisomy 13. Pooled test failure rate was 1.1%. A direct comparison of microarray- and sequencing-based cfDNA found equivalent test accuracy. CONCLUSION: Included studies suggest that NIPT using microarray-based cfDNA testing has high sensitivity and specificity for detecting fetal trisomy 21, 18, and 13. However, the evidence base is small and at high risk of bias.

Risk-adjusted colorectal cancer screening using the FIT and routine screening data: development of a risk prediction model.

BACKGROUND: The faecal immunochemical test (FIT) is replacing the guaiac faecal occult blood test in colorectal cancer screening. Increased uptake and FIT positivity will challenge colonoscopy services. We developed a risk prediction model combining routine screening data with FIT concentration to improve the accuracy of screening referrals. METHODS: Multivariate analysis used complete cases of those with a positive FIT (⩾20 µg g-1) and diagnostic outcome (n=1810; 549 cancers and advanced adenomas). Logistic regression was used to develop a risk prediction model using the FIT result and screening data: age, sex and previous screening history. The model was developed further using a feedforward neural network. Model performance was assessed by discrimination and calibration, and test accuracy was investigated using clinical sensitivity, specificity and receiver operating characteristic curves. RESULTS: Discrimination improved from 0.628 with just FIT to 0.659 with the risk-adjusted model (P=0.01). Calibration using the Hosmer-Lemeshow test was 0.90 for the risk-adjusted model. The sensitivity improved from 30.78% to 33.15% at similar specificity (FIT threshold of 160 µg g-1). The neural network further improved model performance and test accuracy. CONCLUSIONS: Combining routinely available risk predictors with the FIT improves the clinical sensitivity of the FIT with an increase in the diagnostic yield of high-risk adenomas.

Double Reading in Breast Cancer Screening: Cohort Evaluation in the CO-OPS Trial.

Purpose To investigate the effect of double readings by a second radiologist on recall rates, cancer detection, and characteristics of cancers detected in the National Health Service Breast Screening Program in England. Materials and Methods In this retrospective analysis, 805 206 women were evaluated through screening and diagnostic test results by extracting 1 year of routine data from 33 English breast screening centers. Centers used double reading of digital mammograms, with arbitration if there were discrepant interpretations. Information on reader decisions, with results of follow-up tests, were used to explore the effect of the second reader. The statistical tests used were the test for equality of proportions, the χ2 test for independence, and the t test. Results The first reader recalled 4.76% of women (38 295 of 805 206 women; 95% confidence interval [CI]: 4.71%, 4.80%). Two readers recalled 6.19% of women in total (49 857 of 805 206 women; 95% CI: 6.14%, 6.24%), but arbitration of discordant readings reduced the recall rate to 4.08% (32 863 of 805 206 women; 95% CI: 4.04%, 4.12%; P < .001). A total of 7055 cancers were detected, of which 627 (8.89%; 95% CI: 8.22%, 9.55%; P < .001) were detected by the second reader only. These additional cancers were more likely to be ductal carcinoma in situ (30.5% [183 of 600] vs 22.0% [1344 of 6114]; P < .001), and additional invasive cancers were smaller (mean size, 14.2 vs 16.7 mm; P < .001), had fewer involved nodes, and were likely to be lower grade. Conclusion Double reading with arbitration reduces recall and increases cancer detection compared with single reading. Cancers detected only by the second reader were smaller, of lower grade, and had less nodal involvement. © RSNA, 2018.

Adverse events in women and children who have received intrapartum antibiotic prophylaxis treatment: a systematic review.

BACKGROUND: Adverse events from intrapartum antibiotic prophylaxis (IAP) are poorly documented yet essential to inform clinical practice for neonatal group B Streptococcus (GBS) disease prevention. In this systematic review, we appraised and synthesised the evidence on the adverse events of IAP in the mother and/or her child. METHODS: We searched MEDLINE, MEDLINE In-Process & Other Non-Indexed Citations, EMBASE, Cochrane, and Science Citation Index from date of inception until October 16th 2016. Reference lists of included studies and relevant systematic reviews were hand-searched. We included primary studies in English that reported any adverse events from intrapartum antibiotics for any prophylactic purpose compared to controls. The search was not restricted to prophylaxis for GBS but excluded women with symptoms of infection or undergoing caesarean section. Two reviewers assessed the methodological quality of studies, using the Cochrane Risk of Bias tool, and the Risk of Bias Assessment Tool for Nonrandomised Studies. Results were synthesised narratively and displayed in text and tables. RESULTS: From 2364 unique records, 30 studies were included. Despite a wide range of adverse events reported in 17 observational studies and 13 randomised controlled trials (RCTs), the evidence was inconsistent and at high risk of bias. Only one RCT investigated the long-term effects of IAP reporting potentially serious outcomes such as cerebral palsy; however, it had limited applicability and unclear biological plausibility. Seven observational studies showed that IAP for maternal GBS colonisation alters the infant microbiome. However, study populations were not followed through to clinical outcomes, therefore clinical significance is unknown. There was also observational evidence for increased antimicrobial resistance, however studies were at high or unclear risk of bias. CONCLUSIONS: The evidence base to determine the frequency of adverse events from intrapartum antibiotic prophylaxis for neonatal GBS disease prevention is limited. As RCTs may not be possible, large, better quality, and longitudinal observational studies across countries with widespread IAP could fill this gap. TRIAL REGISTRATION: CRD42016037195 .

Associations between behaviours that challenge in adults with intellectual disability, parental perceptions and parental mental health.

OBJECTIVES: This study examined parental perceptions of behaviours that challenge (CB) in their adult children with intellectual disability (ID), and explored whether perceptions mediated associations between CB and parental psychological distress. DESIGN: A within-group correlational design was employed. METHODS: Sixty-five parents reported on individuals with genetic syndromes and ID who had chronic CB. Parents completed the Illness Perception Questionnaire-Revised (IPQ-R) adapted to measure perceptions of self-injury, aggression or property destruction, alongside assessments of parental locus of control, attributions about behaviour, parental psychological distress, and CB. RESULTS: A high proportion of parents evidenced anxiety and depression at clinically significant levels (56.9% and 30.8%, respectively). Contrary to predictions, psychological distress was not significantly associated with CB. The perception that the adult with ID exerted control over the parent's life mediated the association between CB and parental psychological distress. Few parents endorsed operant reinforcement as a cause of CB (< 10%). CONCLUSIONS: The high levels of psychological distress in parents is notable and of concern. Further research should consider the reasons why parents have causal attributions that might be inconsistent with contemporary interventions. PRACTITIONER POINTS: Parents experience high levels of psychological distress while supporting adults with ID who engage in chronic behaviours that challenge. A stronger belief that the adult with ID exerts control over the parent's life may mediate an association between CB exhibited by the individual with ID and parental psychological distress. Few parents endorsed operant reinforcement as a cause of behaviours that challenge.

Where's WALY? : A proof of concept study of the 'wellbeing adjusted life year' using secondary analysis of cross-sectional survey data.

BACKGROUND: The Quality-Adjusted Life Year (QALY) is a measure that combines life extension and health improvement in a single score, reflecting preferences around different types of health gain. It can therefore be used to inform decision-making around allocation of health care resources to mutually exclusive options that would produce qualitatively different health benefits. A number of quality-of-life instruments can be used to calculate QALYs. The EQ-5D is one of the most commonly used, and is the preferred option for submissions to NICE ( https://www.nice.org.uk/process/pmg9/ ). However, it has limitations that might make it unsuitable for use in areas such as public and mental health where interventions may aim to improve well-being. One alternative to the QALY is a Wellbeing-Adjusted Life Year. In this study we explore the need for a Wellbeing-Adjusted Life Year measure by examining the extent to which a measure of wellbeing (the Warwick-Edinburgh Mental Well-being Scale) maps onto the EQ-5D-3L. METHODS: Secondary analyses were conducted on data from the Coventry Household Survey in which 7469 participants completed the EQ-5D-3L, Warwick-Edinburgh Mental Well-being Scale, and a measure of self-rated health. Data were analysed using descriptive statistics, Pearson's and Spearman's correlations, linear regression, and receiver operating characteristic curves. RESULTS: Approximately 75 % of participants scored the maximum on the EQ-5D-3L. Those with maximum EQ-5D-3L scores reported a wide range of levels of mental wellbeing. Both the Warwick-Edinburgh Mental Well-being Scale and the EQ-5D-3L were able to detect differences between those with higher and lower levels of self-reported health. Linear regression indicated that scores on the Warwick-Edinburgh Mental Well-being Scale and the EQ-5D-3L were weakly, positively correlated (with R(2) being 0.104 for the index and 0.141 for the visual analogue scale). CONCLUSION: The Warwick-Edinburgh Mental Well-being Scale maps onto the EQ-5D-3L to only a limited extent. Levels of mental wellbeing varied greatly amongst participants who had the maximum score on the EQ-5D-3L. To evaluate the relative effectiveness of interventions that impact on mental wellbeing, a new measure - a Wellbeing Adjusted Life Year - is needed.

The behavioral characteristics of Sotos syndrome.

In this study we describe the levels of clinically significant behavior in participants with Sotos syndrome relative to three matched contrast groups in which the behavioral phenotype is well documented (Autism Spectrum Disorder, ASD; Prader-Willi, and Down syndromes). Parents and carers of 38 individuals with Sotos syndrome (mean age = 17.3; SD = 9.36), completed questionnaires regarding self-injury, aggression, repetitive behavior, autism spectrum phenomenology, overactivity, impulsivity and mood, interest and pleasure. Individuals with Sotos syndrome showed an increased risk of self-injurious behavior, physical aggression, and destruction of property relative to the Down syndrome group but not a greater risk of stereotyped behavior. Impulsivity and levels of activity were also significantly higher relative to those with Down syndrome and comparable to those with ASD. A large proportion of participants met the cut off score for ASD (70.3%) and Autism (32.4%) on the Social Communication Questionnaire. Social impairments were particularly prominent with repetitive behavior and communication impairments less characteristic of the syndrome. Interestingly, preference for routine and repetitive language were heightened in individuals with Sotos syndrome and the repetitive behavior profile was strikingly similar to that observed in individuals with Prader-Willi syndrome. These findings build upon previous research and provide further evidence of the behavioral phenotype associated with Sotos syndrome.

Artificial intelligence software for analysing chest X-ray images to identify suspected lung cancer: an evidence synthesis early value assessment.

Background: Lung cancer is one of the most common types of cancer in the United Kingdom. It is often diagnosed late. The 5-year survival rate for lung cancer is below 10%. Early diagnosis may improve survival. Software that has an artificial intelligence-developed algorithm might be useful in assisting with the identification of suspected lung cancer. Objectives: This review sought to identify evidence on adjunct artificial intelligence software for analysing chest X-rays for suspected lung cancer, and to develop a conceptual cost-effectiveness model to inform discussion of what would be required to develop a fully executable cost-effectiveness model for future economic evaluation. Data sources: The data sources were MEDLINE All, EMBASE, Cochrane Database of Systematic Reviews, Cochrane CENTRAL, Epistemonikos, ACM Digital Library, World Health Organization International Clinical Trials Registry Platform, clinical experts, Tufts Cost-Effectiveness Analysis Registry, company submissions and clinical experts. Searches were conducted from 25 November 2022 to 18 January 2023. Methods: Rapid evidence synthesis methods were employed. Data from companies were scrutinised. The eligibility criteria were (1) primary care populations referred for chest X-ray due to symptoms suggestive of lung cancer or reasons unrelated to lung cancer; (2) study designs that compared radiology specialist assessing chest X-ray with adjunct artificial intelligence software versus radiology specialists alone and (3) outcomes relating to test accuracy, practical implications of using artificial intelligence software and patient-related outcomes. A conceptual decision-analytic model was developed to inform a potential full cost-effectiveness evaluation of adjunct artificial intelligence software for analysing chest X-ray images to identify suspected lung cancer. Results: None of the studies identified in the searches or submitted by the companies met the inclusion criteria of the review. Contextual information from six studies that did not meet the inclusion criteria provided some evidence that sensitivity for lung cancer detection (but not nodule detection) might be higher when chest X-rays are interpreted by radiology specialists in combination with artificial intelligence software than when they are interpreted by radiology specialists alone. No significant differences were observed for specificity, positive predictive value or number of cancers detected. None of the six studies provided evidence on the clinical effectiveness of adjunct artificial intelligence software. The conceptual model highlighted a paucity of input data along the course of the diagnostic pathway and identified key assumptions required for evidence linkage. Limitations: This review employed rapid evidence synthesis methods. This included only one reviewer conducting all elements of the review, and targeted searches that were conducted in English only. No eligible studies were identified. Conclusions: There is currently no evidence applicable to this review on the use of adjunct artificial intelligence software for the detection of suspected lung cancer on chest X-ray in either people referred from primary care with symptoms of lung cancer or people referred from primary care for other reasons. Future work: Future research is required to understand the accuracy of adjunct artificial intelligence software to detect lung nodules and cancers, as well as its impact on clinical decision-making and patient outcomes. Research generating key input parameters for the conceptual model will enable refinement of the model structure, and conversion to a full working model, to analyse the cost-effectiveness of artificial intelligence software for this indication. Study registration: This study is registered as PROSPERO CRD42023384164. Funding: This award was funded by the National Institute for Health and Care Research (NIHR) Evidence Synthesis programme (NIHR award ref: NIHR135755) and is published in full in Health Technology Assessment; Vol. 28, No. 50. See the NIHR Funding and Awards website for further award information.

Lung cancer is one of the most common types of cancer in the United Kingdom. Early diagnosis may improve survival, as lung cancer is often diagnosed late. Chest X-rays can be used to identify features of lung cancer. There can be delays in getting X-rays, and sometimes features of lung cancer are not seen on them. Artificial intelligence software may help by finding features of cancer on chest X-rays and highlighting them. A radiologist will look at the X-rays and information from the software. There is a lack of information about how lung cancer diagnosis could change if artificial intelligence software is used and what the costs may be to the National Health Service. This project looked at the use of artificial intelligence software in the detection of lung cancer in people referred from primary care. Software companies were invited to provide evidence. There were no studies that looked at this topic among people from primary care. We summarised the closest evidence we could find instead. All of this had flaws, so we could not tell if the results were accurate or helpful to this review. It was not clear if artificial intelligence helped to find cancers or improve people's health. We made a theoretical model to discuss the best way to assess if artificial intelligence software might be cost-effective in detecting lung cancer and what evidence would be needed to do this in a fully working model. Costs and alternative pricing models provided by five companies were used to calculate the cost of adding artificial intelligence software to review chest X-rays in people referred from their general practitioner, for the first 5 years, based on one National Health Service trust. Future studies are needed to identify the impact of adjunct artificial intelligence on test accuracy, clinical decision-making and patient outcomes (e.g. mortality and morbidity).

Use of artificial intelligence for image analysis in breast cancer screening programmes: systematic review of test accuracy.

OBJECTIVE: To examine the accuracy of artificial intelligence (AI) for the detection of breast cancer in mammography screening practice. DESIGN: Systematic review of test accuracy studies. DATA SOURCES: Medline, Embase, Web of Science, and Cochrane Database of Systematic Reviews from 1 January 2010 to 17 May 2021. ELIGIBILITY CRITERIA: Studies reporting test accuracy of AI algorithms, alone or in combination with radiologists, to detect cancer in women's digital mammograms in screening practice, or in test sets. Reference standard was biopsy with histology or follow-up (for screen negative women). Outcomes included test accuracy and cancer type detected. STUDY SELECTION AND SYNTHESIS: Two reviewers independently assessed articles for inclusion and assessed the methodological quality of included studies using the QUality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) tool. A single reviewer extracted data, which were checked by a second reviewer. Narrative data synthesis was performed. RESULTS: Twelve studies totalling 131 822 screened women were included. No prospective studies measuring test accuracy of AI in screening practice were found. Studies were of poor methodological quality. Three retrospective studies compared AI systems with the clinical decisions of the original radiologist, including 79 910 women, of whom 1878 had screen detected cancer or interval cancer within 12 months of screening. Thirty four (94%) of 36 AI systems evaluated in these studies were less accurate than a single radiologist, and all were less accurate than consensus of two or more radiologists. Five smaller studies (1086 women, 520 cancers) at high risk of bias and low generalisability to the clinical context reported that all five evaluated AI systems (as standalone to replace radiologist or as a reader aid) were more accurate than a single radiologist reading a test set in the laboratory. In three studies, AI used for triage screened out 53%, 45%, and 50% of women at low risk but also 10%, 4%, and 0% of cancers detected by radiologists. CONCLUSIONS: Current evidence for AI does not yet allow judgement of its accuracy in breast cancer screening programmes, and it is unclear where on the clinical pathway AI might be of most benefit. AI systems are not sufficiently specific to replace radiologist double reading in screening programmes. Promising results in smaller studies are not replicated in larger studies. Prospective studies are required to measure the effect of AI in clinical practice. Such studies will require clear stopping rules to ensure that AI does not reduce programme specificity. STUDY REGISTRATION: Protocol registered as PROSPERO CRD42020213590.

Testing strategies for Lynch syndrome in people with endometrial cancer: systematic reviews and economic evaluation.

BACKGROUND: Lynch syndrome is an inherited genetic condition that is associated with an increased risk of certain cancers. The National Institute for Health and Care Excellence has recommended that people with colorectal cancer are tested for Lynch syndrome. Routine testing for Lynch syndrome among people with endometrial cancer is not currently conducted. OBJECTIVES: To systematically review the evidence on the test accuracy of immunohistochemistry- and microsatellite instability-based strategies to detect Lynch syndrome among people who have endometrial cancer, and the clinical effectiveness and the cost-effectiveness of testing for Lynch syndrome among people who have been diagnosed with endometrial cancer. DATA SOURCES: Searches were conducted in the following databases, from inception to August 2019 - MEDLINE ALL, EMBASE (both via Ovid), Cochrane Database of Systematic Reviews, Cochrane Central Register of Controlled Trials (both via Wiley Online Library), Database of Abstracts of Reviews of Effects, Health Technology Assessment Database (both via the Centre for Reviews and Dissemination), Science Citation Index, Conference Proceedings Citation Index - Science (both via Web of Science), PROSPERO international prospective register of systematic reviews (via the Centre for Reviews and Dissemination), NHS Economic Evaluation Database, Cost-Effectiveness Analysis Registry, EconPapers (Research Papers in Economics) and School of Health and Related Research Health Utilities Database. The references of included studies and relevant systematic reviews were also checked and experts on the team were consulted. REVIEW METHODS: Eligible studies included people with endometrial cancer who were tested for Lynch syndrome using immunohistochemistry- and/or microsatellite instability-based testing [with or without mutL homologue 1 (MLH1) promoter hypermethylation testing], with Lynch syndrome diagnosis being established though germline testing of normal (non-tumour) tissue for constitutional mutations in mismatch repair. The risk of bias in studies was assessed using the Quality Assessment of Diagnostic Accuracy Studies-2 tool, the Consolidated Health Economic Reporting Standards and the Philips' checklist. Two reviewers independently conducted each stage of the review. A meta-analysis of test accuracy was not possible because of the number and heterogeneity of studies. A narrative summary of test accuracy results was provided, reporting test accuracy estimates and presenting forest plots. The economic model constituted a decision tree followed by Markov models for the impact of colorectal and endometrial surveillance, and aspirin prophylaxis with a lifetime time horizon. RESULTS: The clinical effectiveness search identified 3308 studies; 38 studies of test accuracy were included. (No studies of clinical effectiveness of endometrial cancer surveillance met the inclusion criteria.) Four test accuracy studies compared microsatellite instability with immunohistochemistry. No clear difference in accuracy between immunohistochemistry and microsatellite instability was observed. There was some evidence that specificity of immunohistochemistry could be improved with the addition of methylation testing. There was high concordance between immunohistochemistry and microsatellite instability. The economic model indicated that all testing strategies, compared with no testing, were cost-effective at a willingness-to-pay threshold of £20,000 per quality-adjusted life-year. Immunohistochemistry with MLH1 promoter hypermethylation testing was the most cost-effective strategy, with an incremental cost-effectiveness ratio of £9420 per quality-adjusted life-year. The second most cost-effective strategy was immunohistochemistry testing alone, but incremental analysis produced an incremental cost-effectiveness ratio exceeding £130,000. Results were robust across all scenario analyses. Incremental cost-effectiveness ratios ranged from £5690 to £20,740; only removing the benefits of colorectal cancer surveillance produced an incremental cost-effectiveness ratio in excess of the £20,000 willingness-to-pay threshold. A sensitivity analysis identified the main cost drivers of the incremental cost-effectiveness ratio as percentage of relatives accepting counselling and prevalence of Lynch syndrome in the population. A probabilistic sensitivity analysis showed, at a willingness-to-pay threshold of £20,000 per quality-adjusted life-year, a 0.93 probability that immunohistochemistry with MLH1 promoter hypermethylation testing is cost-effective, compared with no testing. LIMITATIONS: The systematic review excluded grey literature, studies written in non-English languages and studies for which the reference standard could not be established. Studies were included when Lynch syndrome was diagnosed by genetic confirmation of constitutional variants in the four mismatch repair genes (i.e. MLH1, mutS homologue 2, mutS homologue 6 and postmeiotic segregation increased 2). Variants of uncertain significance were reported as per the studies. There were limitations in the economic model around uncertainty in the model parameters and a lack of modelling of the potential harms of gynaecological surveillance and specific pathway modelling of genetic testing for somatic mismatch repair mutations. CONCLUSION: The economic model suggests that testing women with endometrial cancer for Lynch syndrome is cost-effective, but that results should be treated with caution because of uncertain model inputs. FUTURE WORK: Randomised controlled trials could provide evidence on the effect of earlier intervention on outcomes and the balance of benefits and harms of gynaecological cancer surveillance. Follow-up of negative cases through disease registers could be used to determine false negative cases. STUDY REGISTRATION: This study is registered as PROSPERO CRD42019147185. FUNDING: This project was funded by the National Institute for Health Research (NIHR) Evidence Synthesis programme and will be published in full in Health Technology Assessment; Vol. 25, No. 42. See the NIHR Journals Library website for further project information.

Lynch syndrome is an inherited condition that is caused by a problem in the genes. People who have Lynch syndrome have a higher risk of some types of cancer (such as bowel and womb cancers) than people who do not have it. Identifying Lynch syndrome could stop cancers developing, lead to earlier treatment for cancers and help to find other family members who might have it. Currently, the National Institute for Health and Care Excellence guidance recommends testing for Lynch syndrome in people who have bowel cancer. Our aim was to investigate whether or not we should test for Lynch syndrome in women with womb cancer, and their relatives. We investigated two main tests: immunohistochemistry and microsatellite instability. There was no clear evidence that one of these tests is better than the other. There is some evidence that both tests are reasonably accurate. There was no good-quality evidence about whether or not treating women with Lynch syndrome with extra cancer screening and aspirin improves their outcomes. We used the best evidence available in our economic model, but it was at high risk of bias. The economic model suggested that testing women with endometrial cancer for Lynch syndrome is cost-effective. The best test in the model was immunohistochemistry followed by methylation testing. We are unsure of these results because of the low quality of evidence available.

Newborn Screening for Long-Chain 3-Hydroxyacyl-CoA Dehydrogenase and Mitochondrial Trifunctional Protein Deficiencies Using Acylcarnitines Measurement in Dried Blood Spots-A Systematic Review of Test Accuracy.

Background: Long-chain 3-hydroxyacyl-CoA dehydrogenase (LCHAD) and mitochondrial trifunctional protein (MTP) deficiencies are rare autosomal recessive fatty acid ß-oxidation disorders. Their clinical presentations are variable, and premature death is common. They are included in newborn blood spot screening programs in many countries around the world. The current process of screening, through the measurement of acylcarnitines (a metabolic by-product) in dried blood spots with tandem mass spectrometry, is subject to uncertainty regarding test accuracy. Methods: We conducted a systematic review of literature published up to 19th June 2018. We included studies that investigated newborn screening for LCHAD or MTP deficiencies by tandem mass spectrometry of acylcarnitines in dried blood spots. The reference standards were urine organic acids, blood acylcarnitine profiles, enzyme analysis in cultured fibroblasts or lymphocytes, mutation analysis, or at least 10-year follow-up. The outcomes of interest were sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). Assessment of titles, abstracts, and full-text papers and quality appraisal were carried out independently by two reviewers. One reviewer extracted study data. This was checked by a second reviewer. Results: Ten studies provided data on test accuracy. LCHAD or MTP deficiencies were identified in 23 babies. No cases of LCHAD/MTP deficiencies were identified in four studies. PPV ranged from 0% (zero true positives and 28 false positives from 276,565 babies screened) to 100% (13 true positives and zero false positives from 2,037,824 babies screened). Sensitivity, specificity, and NPV could not be calculated as there was no systematic follow-up of babies who screened negative. Conclusions: Test accuracy estimates of screening for LCHAD and MTP deficiencies with tandem mass spectrometry measurement of acylcarnitines in dried blood were variable in terms of PPVs. Screening methods (including markers and thresholds) varied between studies, and sensitivity, specificity, and NPVs are unknown.

Double reading in breast cancer screening: considerations for policy-making.

In this article, we explore the evidence around the relative benefits and harms of breast cancer screening using a single radiologist to examine each female's mammograms for signs of cancer (single reading), or two radiologists (double reading). First, we briefly explore the historical evidence using film-screen mammography, before providing an in-depth description of evidence using digital mammography. We classify studies according to which exact version of double reading they use, because the evidence suggests that effectiveness of double reading is contingent on whether the two radiologists are blinded to one another's decisions, and how the decisions of the two radiologists are integrated. Finally, we explore the implications for future mammography, including using artificial intelligence as the second reader, and applications to more complex three-dimensional imaging techniques such as tomosynthesis.

Rapid antigen detection and molecular tests for group A streptococcal infections for acute sore throat: systematic reviews and economic evaluation.

BACKGROUND: Sore throat is a common condition caused by an infection of the airway. Most cases are of a viral nature; however, a number of these infections may be caused by the group A Streptococcus bacterium. Most viral and bacterial sore throat infections resolve spontaneously within a few weeks. Point-of-care testing in primary care has been recognised as an emerging technology for aiding targeted antibiotic prescribing for sore throat in cases that do not spontaneously resolve. OBJECTIVE: Systematically review the evidence for 21 point-of-care tests for detecting group A Streptococcus bacteria and develop a de novo economic model to compare the cost-effectiveness of point-of-care tests alongside clinical scoring tools with the cost-effectiveness of clinical scoring tools alone for patients managed in primary care and hospital settings. DATA SOURCES: Multiple electronic databases were searched from inception to March 2019. The following databases were searched in November and December 2018 and searches were updated in March 2019: MEDLINE [via OvidSP (Health First, Rockledge, FL, USA)], MEDLINE In-Process & Other Non-Indexed Citations (via OvidSP), MEDLINE Epub Ahead of Print (via OvidSP), MEDLINE Daily Update (via OvidSP), EMBASE (via OvidSP), Cochrane Database of Systematic Reviews [via Wiley Online Library (John Wiley & Sons, Inc., Hoboken, NJ, USA)], Cochrane Central Register of Controlled Trials (CENTRAL) (via Wiley Online Library), Database of Abstracts of Reviews of Effects (DARE) (via Centre for Reviews and Dissemination), Health Technology Assessment database (via the Centre for Reviews and Dissemination), Science Citation Index and Conference Proceedings [via the Web of Science™ (Clarivate Analytics, Philadelphia, PA, USA)] and the PROSPERO International Prospective Register of Systematic Reviews (via the Centre for Reviews and Dissemination). REVIEW METHODS: Eligible studies included those of people aged ≥ 5 years presenting with sore throat symptoms, studies comparing point-of-care testing with antibiotic-prescribing decisions, studies of test accuracy and studies of cost-effectiveness. Quality assessment of eligible studies was undertaken. Meta-analysis of sensitivity and specificity was carried out for tests with sufficient data. A decision tree model estimated costs and quality-adjusted life-years from an NHS and Personal Social Services perspective. RESULTS: The searches identified 38 studies of clinical effectiveness and three studies of cost-effectiveness. Twenty-six full-text articles and abstracts reported on the test accuracy of point-of-care tests and/or clinical scores with biological culture as a reference standard. In the population of interest (patients with Centor/McIsaac scores of ≥ 3 points or FeverPAIN scores of ≥ 4 points), point estimates were 0.829 to 0.946 for sensitivity and 0.849 to 0.991 for specificity. There was considerable heterogeneity, even for studies using the same point-of-care test, suggesting that is unlikely that any single study will have accurately captured a test's true performance. There is some randomised controlled trial evidence to suggest that the use of rapid antigen detection tests may help to reduce antibiotic-prescribing rates. Sensitivity and specificity estimates for each test in each age group and care setting combination were obtained using meta-analyses where appropriate. Any apparent differences in test accuracy may not be attributable to the tests, and may have been caused by known differences in the studies, latent characteristics or chance. Fourteen of the 21 tests reviewed were included in the economic modelling, and these tests were not cost-effective within the current National Institute for Health and Care Excellence's cost-effectiveness thresholds. Uncertainties in the cost-effectiveness estimates included model parameter inputs and assumptions that increase the cost of testing, and the penalty for antibiotic overprescriptions. LIMITATIONS: No information was identified for the elderly population or pharmacy setting. It was not possible to identify which test is the most accurate owing to the paucity of evidence. CONCLUSIONS: The systematic review and the cost-effectiveness models identified uncertainties around the adoption of point-of-care tests in primary and secondary care settings. Although sensitivity and specificity estimates are promising, we have little information to establish the most accurate point-of-care test. Further research is needed to understand the test accuracy of point-of-care tests in the proposed NHS pathway and in comparable settings and patient groups. STUDY REGISTRATION: The protocol of the review is registered as PROSPERO CRD42018118653. FUNDING: This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 24, No. 31. See the NIHR Journals Library website for further project information.

Sore throat is a common condition caused by an infection of the airway. Most cases are viral; however, a small number may be caused by the group A Streptococcus bacterium. Most viral and bacterial sore throat infections resolve spontaneously within a few weeks; however, some may be more serious and require antibiotics. Currently, National Institute for Health and Care Excellence guidance recommends the use of clinical scoring tools to identify patients for whom antibiotic treatment is appropriate. Ideally, a throat swab culture should be obtained to identify the organism causing the infection in cases in which diagnosis is uncertain. However, this takes time, causing potential delays in administering the correct treatment. Point-of-care tests can be administered at or near the site of the patient; therefore, they are much faster. Our review considered evidence for the test accuracy and cost-effectiveness of 21 point-of-care tests for detecting group A Streptococcus bacteria. We built an economic model, predicting costs and benefits for adults and children in a primary care or hospital setting. The findings will support the National Institute for Health and Care Excellence to make recommendations about the use of these point-of-care tests for detecting group A Streptococcus bacteria in the NHS in England and Wales. The clinical effectiveness review found 38 relevant studies; of these, 26 reported on the accuracy of point-of-care tests. These studies found wide variation in the accuracy of the tests. The quality of the evidence was weak and there was little information on some of the 21 tests. As the studies were all so different, it was not possible to identify which test is the most accurate. The economic model found considerable uncertainty about how costs and benefits would change if point-of-care tests were introduced in different care settings. Further research is needed to see whether or not point-of-care testing provides value for money.

Fatigue in radiology: a fertile area for future research.

Fatigue in radiologists may be responsible for a large number of medical errors. This review describes the latest research on fatigue in radiology. This includes measurement methods, and recent evidence on how fatigue affects accuracy in laboratory test conditions and in clinical practice. The extensive opportunities for future research in the area are explored, including testing interventions to reduce fatigue-related error, and further understanding of which fatigue measures correlate with errors. Finally we explore the possibility of answering these questions using large population-based observational studies and pragmatic integrated randomised controlled trials.

Lifespan trajectory of affect in Cornelia de Lange syndrome: towards a neurobiological hypothesis.

BACKGROUND: Depressive symptomology and low affect are comparatively common in individuals with genetic disorders such as Cornelia de Lange syndrome. However, lifespan trajectories and associated person characteristics have not been examined. In this study, the trajectories for affect and associated behavioural characteristics were investigated in individuals with Cornelia de Lange syndrome with individuals with fragile X syndrome (FXS) comparable for chronological age and total number of behavioural indicators of ASD included for the purpose of contrast. METHODS: A 7-year longitudinal study of affect (mood, interest and pleasure) was conducted in individuals with CdLS (n = 44) and FXS (n = 95). The trajectories of low affect were explored, as well as associations between Time 1 behavioural characteristics and affect at Time 1 and Time 3 (7 years later). RESULTS: The CdLS group were lower in mood than the FXS group overall (p < .001). Interest and pleasure scores showed a significant decline over the lifespan for individuals with CdLS (p < .001) but not the FXS group. Lower level of ability at Time 1 was associated with lower mood at Time 1 and Time 3 in the FXS group only. Higher levels of ASD symptomology at Time 1 were associated with low mood and interest and pleasure in both syndrome groups at Time 1 and Time 3. Greater insistence on sameness at Time 1 was associated with lower mood at Time 1 in the FXS group and lower interest and pleasure at Time 1 and Time 3 in the CdLS group. CONCLUSIONS: Low affect in specific genetic syndromes may be associated with differing lifespan trajectories and behavioural profiles. Specifically, individuals with CdLS appear at risk for experiencing declines in levels of interest and pleasure whereas individuals with FXS show no significant change in the level of affect with age.