Interspecialty referral of oesophagogastric and pharyngolaryngeal cancers delays diagnosis and reduces patient survival: A matched case-control study.

OBJECTIVES: Pharyngolaryngeal and oesophagogastric cancers present with swallowing symptoms and as such, their clinical evaluation traverses boundaries between different specialties. We studied the incidence and significance of interspecialty cancer referrals (ICRs), that is, pharyngolaryngeal cancers first evaluated by gastroenterology and oesophagogastric cancers first evaluated by otolaryngology. DESIGN: A subset analysis of our Integrated Aerodigestive Partnership's audit dataset, of all ICR patients, and an equal number of controls matched for age, sex and cancer subsite. MAIN OUTCOME MEASURES: Information about patient age and presenting symptoms was recorded. The relationship between symptoms and ICR risk was examined with binary logistic regression. Referral-to-diagnosis latency was compared between ICR and control patients with unpaired Student's t test. Cox regression was used to identify independent predictors of overall survival. RESULTS: Of 1130 patients with pharyngolaryngeal and oesophagogastric cancers between 2008 and 2018, 60 diagnoses (5.3%) were preceded by an ICR. Referral-to-diagnosis latency increased from 43 ± 50 days for control patients to 115 ± 140 days for ICR patients (P < .0001). Dysphagia significantly increased the risk of an ICR (odds ratio 3.34; 95% CI 1.30-8.56), and presence of classic gastroesophageal reflux symptoms (heartburn or regurgitation; OR 0.25; 95% CI 0.08-0.83) and "distal" symptoms (nausea/vomiting, abdominal pain or dyspepsia; OR 0.23; 95% CI 0.08-068) significantly reduced the risk. Eleven pharyngolaryngeal cancers (of 26; 42%) were missed by gastroenterology, and eight (of 34; 24%) oesophageal cancers were missed by otolaryngology. An ICR was an independent adverse prognostic risk factor on multivariable analysis (hazard ratio 1.76; 95% CI 1.11-2.73; P < .02; log-rank test). Two systemic root causes were poor visualisation of pharynx and larynx by per-oral oesophago-gastro-duodenoscopy (OGD) for pharyngolaryngeal cancers, and poor sensitivity (62.5%) of barium swallow when it was used to 'evaluate' oesophageal mucosa. CONCLUSIONS: An interspecialty cancer referral occurs in a significant proportion of patients with foregut cancers. It almost triples the time to cancer diagnosis and is associated with a high incidence of missed cancers and diminished patient survival. It is a complex phenomenon, and its reduction requires an integrated approach between primary and secondary care, and within secondary care, to optimise referral pathways and ensure appropriate and expeditious specialist evaluation.

An integrated approach to processing WHO-2016 verbal autopsy data: the InterVA-5 model.

BACKGROUND: Verbal autopsy is an increasingly important methodology for assigning causes to otherwise uncertified deaths, which amount to around 50% of global mortality and cause much uncertainty for health planning. The World Health Organization sets international standards for the structure of verbal autopsy interviews and for cause categories that can reasonably be derived from verbal autopsy data. In addition, computer models are needed to efficiently process large quantities of verbal autopsy interviews to assign causes of death in a standardised manner. Here, we present the InterVA-5 model, developed to align with the WHO-2016 verbal autopsy standard. This is a harmonising model that can process input data from WHO-2016, as well as earlier WHO-2012 and Tariff-2 formats, to generate standardised cause-specific mortality profiles for diverse contexts. The software development involved building on the earlier InterVA-4 model, and the expanded knowledge base required for InterVA-5 was informed by analyses from a training dataset drawn from the Population Health Metrics Research Collaboration verbal autopsy reference dataset, as well as expert input. RESULTS: The new model was evaluated against a test dataset of 6130 cases from the Population Health Metrics Research Collaboration and 4009 cases from the Afghanistan National Mortality Survey dataset. Both of these sources contained around three quarters of the input items from the WHO-2016, WHO-2012 and Tariff-2 formats. Cause-specific mortality fractions across all applicable WHO cause categories were compared between causes assigned in participating tertiary hospitals and InterVA-5 in the test dataset, with concordance correlation coefficients of 0.92 for children and 0.86 for adults. The InterVA-5 model's capacity to handle different input formats was evaluated in the Afghanistan dataset, with concordance correlation coefficients of 0.97 and 0.96 between the WHO-2016 and the WHO-2012 format for children and adults respectively, and 0.92 and 0.87 between the WHO-2016 and the Tariff-2 format respectively. CONCLUSIONS: Despite the inherent difficulties of determining "truth" in assigning cause of death, these findings suggest that the InterVA-5 model performs well and succeeds in harmonising across a range of input formats. As more primary data collected under WHO-2016 become available, it is likely that InterVA-5 will undergo minor re-versioning in the light of practical experience. The model is an important resource for measuring and evaluating cause-specific mortality globally.

Collecting and sharing self-generated health and lifestyle data: Understanding barriers for people living with long-term health conditions - a survey study.

Background: The growing popularity of collecting self-generated health and lifestyle data presents a valuable opportunity to develop our understanding of long-term health conditions and improve care. Barriers remain to the effective sharing of health and lifestyle data by those living with long-term health conditions which include beliefs around concepts of Trust, Identity, Privacy and Security, experiences of stigma, perceptions of risk and information sensitivity. Method: We surveyed 250 UK adults who reported living with a range of long-term health conditions. We recorded data to assess self-reported behaviours, experiences, attitudes and motivations relevant to sharing self-generated health and lifestyle data. We also asked participants about their beliefs about Trust, Identity, Privacy and Security, stigma, and perceptions of risk and information sensitivity regarding their health and lifestyle data. Results: Three-quarters of our sample reported recording information about their health and lifestyle on a daily basis. However, two-thirds reported never or rarely sharing this information with others. Trust, Identity, Privacy and Security concerns were considered to be 'very important' by those with long-term health conditions when deciding whether or not to share self-generated health and lifestyle data with others, with security concerns considered most important. Of those living with a long-term health condition, 58% reported experiencing stigma associated with their condition. The greatest perceived risk from sharing with others was the potential for future harm to their social relationships. Conclusions: Our findings suggest that, in order for health professionals and researchers to benefit from the increased prevalence of self-generated health and lifestyle data, more can be done to address security concerns and to understand perceived risks associated with data sharing. Digital platforms aimed at facilitating the sharing of self-generated health and lifestyle data may look to highlight security features, enable users to control the sharing of certain information types, and emphasise the practical benefits to users of sharing health and lifestyle data with others.

Added value of an open narrative in verbal autopsies: a mixed-methods evaluation from Malawi.

Background: The WHO standardised verbal autopsy (VA) instrument includes closed questions, ascertaining signs and symptoms of illness preceding death, and an optional open narrative. As VA analyses increasingly use automated algorithms, inclusion of narratives should be justified. We evaluated the role of open narratives on VA processes, data quality and respondent's emotional stress. Methods: A mixed-methods analysis was conducted using VA data for child deaths (0-59 months), between April 2013 and November 2016 in Mchinji district, Malawi. Deaths were prospectively randomised to receive closed questions only or open narrative followed by closed questions. On concluding the VA, interviewers self-completed questions on respondents' emotional stress. Logistic regression was used to determine associations with visible emotional distress during VAs. A group discussion with interviewers was conducted at the project end, to understand field experiences and explore future recommendations; data were coded using deductive themes. Results: 2509 VAs were included, with 49.8% (n=1341) randomised to open narratives. Narratives lasted a median of 7 minuntes (range: 1-113). Interviewers described improved rapport and felt narratives improved data quality, although there was no difference in the proportion of deaths with an indeterminate cause using an automated algorithm (5.3% vs 6.1%). The majority of respondents did not display visible emotional stress (81%). Those with a narrative had higher, but not statistically significant, odds of emotional distress (adjusted OR: 1.20; 95% CI: 0.98 to 1.47). Factors associated with emotional stress were: infant deaths versus neonates; deaths at a health centre or en-route to hospital versus home; and higher socioeconomic status. Non-parental respondents and increased time between death and interview were associated with lower odds of emotional distress. Conclusion: Conducting an open narrative may help build rapport, something valued by the interviewers. However, additional time and emotional burdens should be further justified, with quality and utility of narratives promoted through standardised recommendations.

Modulation of Medical Condition Likelihood by Patient History Similarity.

INTRODUCTION: We describe an analysis that modulates the simple population prevalence derived likelihood of a particular condition occurring in an individual by matching the individual with other individuals with similar clinical histories and determining the prevalence of the condition within the matched group. METHODS: We have taken clinical event codes and dates from anonymised longitudinal primary care records for 25,979 patients with 749,053 recorded clinical events. Using a nearest neighbour approach, for each patient, the likelihood of a condition occurring was adjusted from the population prevalence to the prevalence of the condition within those patients with the closest matching clinical history. RESULTS: For conditions investigated, the nearest method performed well in comparison with standard logistic regression. CONCLUSIONS: Results indicate that it may be possible to use histories to identify 'similar' patients and thus to modulate future likelihoods of a condition occurring.

A multidimensional approach to assessing intervention fidelity in a process evaluation of audit and feedback interventions to reduce unnecessary blood transfusions: a study protocol.

BACKGROUND: In England, NHS Blood and Transplant conducts national audits of transfusion and provides feedback to hospitals to promote evidence-based practice. Audits demonstrate 20% of transfusions fall outside guidelines. The AFFINITIE programme (Development & Evaluation of Audit and Feedback INterventions to Increase evidence-based Transfusion practIcE) involves two linked, 2×2 factorial, cluster-randomised trials, each evaluating two theoretically-enhanced audit and feedback interventions to reduce unnecessary blood transfusions in UK hospitals. The first intervention concerns the content/format of feedback reports. The second aims to support hospital transfusion staff to plan their response to feedback and includes a web-based toolkit and telephone support. Interpretation of trials is enhanced by comprehensively assessing intervention fidelity. However, reviews demonstrate fidelity evaluations are often limited, typically only assessing whether interventions were delivered as intended. This protocol presents methods for assessing fidelity across five dimensions proposed by the Behaviour Change Consortium fidelity framework, including intervention designer-, provider- and recipient-levels. METHODS: (1) Design: Intervention content will be specified in intervention manuals in terms of component behaviour change techniques (BCTs). Treatment differentiation will be examined by comparing BCTs across intervention/standard practice, noting the proportion of unique/convergent BCTs. (2) Training: draft feedback reports and audio-recorded role-play telephone support scenarios will be content analysed to assess intervention providers' competence to deliver manual-specified BCTs. (3) Delivery: intervention materials (feedback reports, toolkit) and audio-recorded telephone support session transcripts will be content analysed to assess actual delivery of manual-specified BCTs during the intervention period. (4) Receipt and (5) enactment: questionnaires, semi-structured interviews based on the Theoretical Domains Framework, and objective web-analytics data (report downloads, toolkit usage patterns) will be analysed to assess hospital transfusion staff exposure to, understanding and enactment of the interventions, and to identify contextual barriers/enablers to implementation. Associations between observed fidelity and trial outcomes (% unnecessary transfusions) will be examined using mediation analyses. DISCUSSION: If the interventions have acceptable fidelity, then results of the AFFINITIE trials can be attributed to effectiveness, or lack of effectiveness, of the interventions. Hence, this comprehensive assessment of fidelity will be used to interpret trial findings. These methods may inform fidelity assessments in future trials. TRIAL REGISTRATION: ISRCTN 15490813 . Registered 11/03/2015.

Comparing verbal autopsy cause of death findings as determined by physician coding and probabilistic modelling: a public health analysis of 54 000 deaths in Africa and Asia.

BACKGROUND: Coverage of civil registration and vital statistics varies globally, with most deaths in Africa and Asia remaining either unregistered or registered without cause of death. One important constraint has been a lack of fit-for-purpose tools for registering deaths and assigning causes in situations where no doctor is involved. Verbal autopsy (interviewing care-givers and witnesses to deaths and interpreting their information into causes of death) is the only available solution. Automated interpretation of verbal autopsy data into cause of death information is essential for rapid, consistent and affordable processing. METHODS: Verbal autopsy archives covering 54 182 deaths from five African and Asian countries were sourced on the basis of their geographical, epidemiological and methodological diversity, with existing physician-coded causes of death attributed. These data were unified into the WHO 2012 verbal autopsy standard format, and processed using the InterVA-4 model. Cause-specific mortality fractions from InterVA-4 and physician codes were calculated for each of 60 WHO 2012 cause categories, by age group, sex and source. Results from the two approaches were assessed for concordance and ratios of fractions by cause category. As an alternative metric, the Wilcoxon matched-pairs signed ranks test with two one-sided tests for stochastic equivalence was used. FINDINGS: The overall concordance correlation coefficient between InterVA-4 and physician codes was 0.83 (95% CI 0.75 to 0.91) and this increased to 0.97 (95% CI 0.96 to 0.99) when HIV/AIDS and pulmonary TB deaths were combined into a single category. Over half (53%) of the cause category ratios between InterVA-4 and physician codes by source were not significantly different from unity at the 99% level, increasing to 62% by age group. Wilcoxon tests for stochastic equivalence also demonstrated equivalence. CONCLUSIONS: These findings show strong concordance between InterVA-4 and physician-coded findings over this large and diverse data set. Although these analyses cannot prove that either approach constitutes absolute truth, there was high public health equivalence between the findings. Given the urgent need for adequate cause of death data from settings where deaths currently pass unregistered, and since the WHO 2012 verbal autopsy standard and InterVA-4 tools represent relatively simple, cheap and available methods for determining cause of death on a large scale, they should be used as current tools of choice to fill gaps in cause of death data.

Electronic data capture in a rural African setting: evaluating experiences with different systems in Malawi.

BACKGROUND: As hardware for electronic data capture (EDC), such as smartphones or tablets, becomes cheaper and more widely available, the potential for using such hardware as data capture tools in routine healthcare and research is increasing. OBJECTIVE: We aim to highlight the advantages and disadvantages of four EDC systems being used simultaneously in rural Malawi: two for Android devices (CommCare and ODK Collect), one for PALM and Windows OS (Pendragon), and a custom-built application for Android (Mobile InterVA--MIVA). DESIGN: We report on the personal field and development experience of fieldworkers, project managers, and EDC system developers. RESULTS: Fieldworkers preferred using EDC to paper-based systems, although some struggled with the technology at first. Highlighted features include in-built skip patterns for all systems, and specifically the 'case' function that CommCare offers. MIVA as a standalone app required considerably more time and expertise than the other systems to create and could not be customised for our specific research needs; however, it facilitates standardised routine data collection. CommCare and ODK Collect both have user-friendly web-interfaces for form development and good technical support. CommCare requires Internet to build an application and download it to a device, whereas all steps can be done offline with ODK Collect, a desirable feature in low connectivity settings. Pendragon required more complex programming of logic, using a Microsoft Access application, and generally had less technical support. Start-up costs varied between systems, and all were considered more expensive than setting up a paper-based system; however running costs were generally low and therefore thought to be cost-effective over the course of our projects. CONCLUSIONS: EDC offers many opportunities for efficient data collection, but brings some issues requiring consideration when designing a study; the decision of which hardware and software to use should be informed by the aim of data collection, budget, and local circumstances.

Picture this: the state of the art in visualization for complex adaptive systems.

Visualization has an increasingly important role to play in scientific research. Moreover, visualization has a special role to play within artificial life as a result of the informal status of its key explananda: life and complexity. Both are poorly defined but apparently identifiable via raw inspection. Here we concentrate on how visualization techniques might allow us to move beyond this situation by facilitating increased understanding of the relationships between an ALife system's (low-level) composition and organization and its (high-level) behavior. We briefly review the use of visualization within artificial life, and point to some future developments represented by the articles collected within this special issue.