Patterns of general practitioner usage among Pacific people: indicative results from the Waikato Medical Care Survey 1991-2

AIMS: To report patterns of medical contact in a representative sample of Pacific people attending the general practitioner. METHODS: The data were drawn from a survey of general practice in the Waikato region representing a one per cent sample of all weekday encounters. In total, 12,833 patient encounter forms were completed. Just over one per cent of all encounters were recorded for patients of Pacific Islands background. RESULTS: Rates of medical contact for Pacific patients were lower-3.4 visits per year versus 4.5 for the whole sample-fewer follow up visits were requested (71% versus 76.2%), presentation was delayed (4.9 days from onset versus 3.7 for the sample) and there was an apparently lower level of rapport achieved. CONCLUSION: Overall levels of medical contact and return visits among Pacific patients appear to be lower and presentation delayed in this Waikato sample.

Mass spectrometry cancer data classification using wavelets and genetic algorithm

This paper introduces a hybrid feature extraction method applied to mass spectrometry (MS) data for cancer classification. Haar wavelets are employed to transform MS data into orthogonal wavelet coefficients. The most prominent discriminant wavelets are then selected by genetic algorithm (GA) to form feature sets. The combination of wavelets and GA yields highly distinct feature sets that serve as inputs to classification algorithms. Experimental results show the robustness and significant dominance of the wavelet-GA against competitive methods. The proposed method therefore can be applied to cancer classification models that are useful as real clinical decision support systems for medical practitioners.

Autoethnography: introducing 'I' into medical education research

Context: Autoethnography is a methodology that allows clinician-educators to research their own cultures, sharing insights about their own teaching and learning journeys in ways that will resonate with others. There are few examples of autoethnographic research in medical education, and many areas would benefit from this methodology to help improve understanding of, for example, teacher-learner interactions, transitions and interprofessional development. Objectives: We wish to share this methodology so that others may consider it in their own education environments as a viable qualitative research approach to gain new insights and understandings. Methods: This paper introduces autoethnography, discusses important considerations in terms of data collection and analysis, explores ethical aspects of writing about others and considers the benefits and limitations of conducting research that includes self. Results: Autoethnography allows medical educators to increasingly engage in self-reflective narration while analysing their own cultural biographies. It moves beyond simple autobiography through the inclusion of other voices and the analytical examination of the relationships between self and others. Autoethnography has achieved its goal if it results in new insights and improvements in personal teaching practices, and if it promotes broader reflection amongst readers about their own teaching and learning environments. Conclusions: Researchers should consider autoethnography as an important methodology to help advance our understanding of the culture and practices of medical education. Discuss ideas arising from the article at www.mededuc.com discuss. © 2015 John Wiley

Marginalisation of dental students in a shared medical and dental education programme

OBJECTIVES: Internationally, there are a number of universities at which medical and dental education programmes share common elements. There are no studies about the experiences of medical and dental students enrolled in different programmes who share significant amounts of learning and teaching. METHODS: Semi-structured interviews and focus groups were conducted with 36 students and staff in a learning programme shared between separate medical and dental faculties. They were transcribed and an iterative process of interpretation and analysis within the theoretical framework of the contact hypothesis and social identity theory was used to group data into themes and sub-themes. RESULTS: Dental students felt 'marginalised' and felt they were treated as 'second-class citizens' by medical students and medical staff in the shared aspects of their programmes. Contextual factors such as the geographical location of the two schools, a medical : dental student ratio of almost 3 : 1, along with organisational factors such as curriculum overload, propagated negative attitudes towards and professional stereotyping of the dental students. Lack of understanding by medical students and faculty of dental professional roles contributed further. CONCLUSIONS: Recommendations for reducing the marginalisation of dental students in this setting include improving communication between faculties and facilitating experiential contact. This might be achieved through initiating a common orientation session, stronger social networks and integrated learning activities, such as interprofessional problem-based learning and shared clinical experiences.

A predictive framework for modeling healthcare data with evolving clinical interventions

Medical interventions critically determine clinical outcomes. But prediction models either ignore interventions or dilute impact by building a single prediction rule by amalgamating interventions with other features. One rule across all interventions may not capture differential effects. Also, interventions change with time as innovations are made, requiring prediction models to evolve over time. To address these gaps, we propose a prediction framework that explicitly models interventions by extracting a set of latent intervention groups through a Hierarchical Dirichlet Process (HDP) mixture. Data are split in temporal windows and for each window, a separate distribution over the intervention groups is learnt. This ensures that the model evolves with changing interventions. The outcome is modeled as conditional, on both the latent grouping and the patients' condition, through a Bayesian logistic regression. Learning distributions for each time-window result in an over-complex model when interventions do not change in every time-window. We show that by replacing HDP with a dynamic HDP prior, a more compact set of distributions can be learnt. Experiments performed on two hospital datasets demonstrate the superiority of our framework over many existing clinical and traditional prediction frameworks.

Understanding toxicities and complications of cancer treatment: a data mining approach

Cancer remains a major challenge in modern medicine. Increasing prevalence of cancer, particularly in developing countries, demands better understanding of the effectiveness and adverse consequences of different cancer treatment regimes in real patient population. Current understanding of cancer treatment toxicities is often derived from either “clean” patient cohorts or coarse population statistics. It is difficult to get up-to-date and local assessment of treatment toxicities for specific cancer centres. In this paper, we applied an Apriori-based method for discovering toxicity progression patterns in the form of temporal association rules. Our experiments show the effectiveness of the proposed method in discovering major toxicity patterns in comparison with the pairwise association analysis. Our method is applicable for most cancer centres with even rudimentary electronic medical records and has the potential to provide real-time surveillance and quality assurance in cancer care.

A data as a product model for future consumption of big stream data in clouds

Data is becoming the world’s new natural resourceand big data use grows quickly. The trend of computingtechnology is that everything is merged into the Internet and‘big data’ are integrated to comprise completeinformation for collective intelligence. With the increasingsize of big data, refining big data themselves to reduce data sizewhile keeping critical data (or useful information) is a newapproach direction. In this paper, we provide a novel dataconsumption model, which separates the consumption of datafrom the raw data, and thus enable cloud computing for bigdata applications. We define a new Data-as-a-Product (DaaP)concept; a data product is a small sized summary of theoriginal data and can directly answer users’ queries. Thus, weseparate the mining of big data into two classes of processingmodules: the refine modules to change raw big data into smallsizeddata products, and application-oriented mining modulesto discover desired knowledge further for applications fromwell-defined data products. Our practices of mining big streamdata, including medical sensor stream data, streams of textdata and trajectory data, demonstrated the efficiency andprecision of our DaaP model for answering users’ queries

How do postgraduate qualifications in medical education impact on health professionals?

Background and Purpose: The number of degree-awarding programmes in medical education is steadily increasing. Despite the popularity and extensive investment in these courses, there is little research into their impact. This study investigated the perceived impact of an internationally-renowned postgraduate programme in medical education on health professionals’ development as educators.

Methods: An online survey of the 2008–12 graduates from the Centre for Medical Education, University of Dundee was carried out. Their self-reported shifts in various educational competencies and scholarship activities were analysed using non-parametric statistics. Qualitative data were also collected and analysed to add depth to the quantitative findings.

Results: Of the 504 graduates who received the online questionnaire 224 responded. Participants reported that a qualification in medical education had significantly (p < 0.001) improved their professional educational practices and engagement in scholarly activities. Masters graduates reported greater impact compared to Certificate graduates on all items, including ability to facilitate curriculum reforms, and in assessment and feedback practices. Masters graduates also reported more engagement in scholarship activities, with significantly greater contributions to journals. These qualifications equally benefited all participants regardless of age. International graduates reported greater impact of the qualification than their UK counterparts.

Conclusion: A postgraduate medical education programme can significantly impact on the practices and behaviours of health professionals in education, improving self-efficacy and instilling an increased sense of belonging to the educational community.

Selecting, training and assessing new general practice community teachers in UK medical schools

Background: Standards for undergraduate medical education in the UK, published in Tomorrow’s Doctors, include the criterion ‘everyone involved in educating medical students will be appropriately selected, trained, supported and appraised’. Aims: To establish how new general practice (GP) community teachers of medical students are selected, initially trained and assessed by UK medical schools and establish the extent to which Tomorrow’s Doctors standards are being met. Method: A mixed-methods study with questionnaire data collected from 24 lead GPs at UK medical schools, 23 new GP teachers from two medical schools plus a semi-structured telephone interview with two GP leads. Quantitative data were analysed descriptively and qualitative data were analysed informed by framework analysis. Results: GP teachers’ selection is non-standardised. One hundred per cent of GP leads provide initial training courses for new GP teachers; 50% are mandatory. The content and length of courses varies. All GP leads use student feedback to assess teaching, but other required methods (peer review and patient feedback) are not universally used. Conclusions: To meet General Medical Council standards, medical schools need to include equality and diversity in initial training and use more than one method to assess new GP teachers. Wider debate about the selection, training and assessment of new GP teachers is needed to agree minimum standards.

Speed up health research through topic modeling of coded clinical data

Although random control trial is the gold standard in medical research, researchers are increasingly looking to alternative data sources for hypothesis generation and early-stage evidence collection. Coded clinical data are collected routinely in most hospitals. While they contain rich information directly related to the real clinical setting, they are both noisy and semantically diverse, making them difficult to analyze with conventional statistical tools. This paper presents a novel application of Bayesian nonparametric modeling to uncover latent information in coded clinical data. For a patient cohort, a Bayesian nonparametric model is used to reveal the common comorbidity groups shared by the patients and the proportion that each comorbidity group is reflected individual patient. To demonstrate the method, we present a case study based on hospitalization coding from an Australian hospital. The model recovered 15 comorbidity groups among 1012 patients hospitalized during a month. When patients from two areas of unequal socio-economic status were compared, it reveals higher prevalence of diverticular disease in the region of lower socio-economic status. The study builds a convincing case for routine coded data to speed up hypothesis generation.

Learning vector representation of medical objects via EMR-driven nonnegative restricted Boltzmann machines (eNRBM)

Electronic medical record (EMR) offers promises for novel analytics. However, manual feature engineering from EMR is labor intensive because EMR is complex - it contains temporal, mixed-type and multimodal data packed in irregular episodes. We present a computational framework to harness EMR with minimal human supervision via restricted Boltzmann machine (RBM). The framework derives a new representation of medical objects by embedding them in a low-dimensional vector space. This new representation facilitates algebraic and statistical manipulations such as projection onto 2D plane (thereby offering intuitive visualization), object grouping (hence enabling automated phenotyping), and risk stratification. To enhance model interpretability, we introduced two constraints into model parameters: (a) nonnegative coefficients, and (b) structural smoothness. These result in a novel model called eNRBM (EMR-driven nonnegative RBM). We demonstrate the capability of the eNRBM on a cohort of 7578 mental health patients under suicide risk assessment. The derived representation not only shows clinically meaningful feature grouping but also facilitates short-term risk stratification. The F-scores, 0.21 for moderate-risk and 0.36 for high-risk, are significantly higher than those obtained by clinicians and competitive with the results obtained by support vector machines.

Data clustering using side information dependent Chinese restaurant processes

Side information, or auxiliary information associated with documents or image content, provides hints for clustering. We propose a new model, side information dependent Chinese restaurant process, which exploits side information in a Bayesian nonparametric model to improve data clustering. We introduce side information into the framework of distance dependent Chinese restaurant process using a robust decay function to handle noisy side information. The threshold parameter of the decay function is updated automatically in the Gibbs sampling process. A fast inference algorithm is proposed. We evaluate our approach on four datasets: Cora, 20 Newsgroups, NUS-WIDE and one medical dataset. Types of side information explored in this paper include citations, authors, tags, keywords and auxiliary clinical information. The comparison with the state-of-the-art approaches based on standard performance measures (NMI, F1) clearly shows the superiority of our approach.

Physical health comorbidities in women with personality disorder: data from the Geelong osteoporosis study

BACKGROUND: Associations between common psychiatric disorders, psychotic disorders and physical health comorbidities are frequently investigated. The complex relationship between personality disorders (PDs) and physical health is less understood, and findings to date are varied. This study aims to investigate associations between PDs with a number of prevalent physical health conditions. METHODS: This study examined data collected from women (n=765;≥25years) participating in a population-based study located in south-eastern Australia. Lifetime history of psychiatric disorders was assessed using the semi-structured clinical interviews (SCID-I/NP and SCID-II). The presence of physical health conditions (lifetime) were identified via a combination of self-report, medical records, medication use and clinical data. Socioeconomic status, and information regarding medication use, lifestyle behaviors, and sociodemographic information was collected via questionnaires. Logistic regression models were used to investigate associations. RESULTS: After adjustment for sociodemographic variables (age, socioeconomic status) and health-related factors (body mass index, physical activity, smoking, psychotropic medication use), PDs were consistently associated with a range of physical health conditions. Novel associations were observed between Cluster A PDs and gastro-oesophageal reflux disease (GORD); Cluster B PDs with syncope and seizures, as well as arthritis; and Cluster C PDs with GORD and recurrent headaches. CONCLUSIONS: PDs were associated with physical comorbidity. The current data contribute to a growing evidence base demonstrating associations between PDs and a number of physical health conditions independent of psychiatric comorbidity, sociodemographic and lifestyle factors. Longitudinal studies are now required to investigate causal pathways, as are studies determining pathological mechanisms.

Fusing data mining, machine learning and traditional statistics to detect biomarkers associated with depression

BACKGROUND: Atheoretical large-scale data mining techniques using machine learning algorithms have promise in the analysis of large epidemiological datasets. This study illustrates the use of a hybrid methodology for variable selection that took account of missing data and complex survey design to identify key biomarkers associated with depression from a large epidemiological study.

METHODS: The study used a three-step methodology amalgamating multiple imputation, a machine learning boosted regression algorithm and logistic regression, to identify key biomarkers associated with depression in the National Health and Nutrition Examination Study (2009-2010). Depression was measured using the Patient Health Questionnaire-9 and 67 biomarkers were analysed. Covariates in this study included gender, age, race, smoking, food security, Poverty Income Ratio, Body Mass Index, physical activity, alcohol use, medical conditions and medications. The final imputed weighted multiple logistic regression model included possible confounders and moderators.

RESULTS: After the creation of 20 imputation data sets from multiple chained regression sequences, machine learning boosted regression initially identified 21 biomarkers associated with depression. Using traditional logistic regression methods, including controlling for possible confounders and moderators, a final set of three biomarkers were selected. The final three biomarkers from the novel hybrid variable selection methodology were red cell distribution width (OR 1.15; 95% CI 1.01, 1.30), serum glucose (OR 1.01; 95% CI 1.00, 1.01) and total bilirubin (OR 0.12; 95% CI 0.05, 0.28). Significant interactions were found between total bilirubin with Mexican American/Hispanic group (p = 0.016), and current smokers (p<0.001).

CONCLUSION: The systematic use of a hybrid methodology for variable selection, fusing data mining techniques using a machine learning algorithm with traditional statistical modelling, accounted for missing data and complex survey sampling methodology and was demonstrated to be a useful tool for detecting three biomarkers associated with depression for future hypothesis generation: red cell distribution width, serum glucose and total bilirubin.

Enhancement of medical named entity recognition using graph-based features

Named Entity Recognition (NER) is a crucial step in text mining. This paper proposes a new graph-based technique for representing unstructured medical text. The new representation is used to extract discriminative features that are able to enhance the NER performance. To evaluate the usefulness of the proposed graph-based technique, the i2b2 medication challenge data set is used. Specifically, the 'treatment' named entities are extracted for evaluation using six different classifiers. The F-measure results of five classifiers are enhanced, with an average improvement of up to 26% in performance.