Cohort Of Australian Men Research Articles

Predicting the diagnosis of HIV and sexually transmitted infections among men who have sex with men using machine learning approaches

ObjectivesWe aimed to develop machine learning models and evaluate their performance in predicting HIV and sexually transmitted infections (STIs) diagnosis based on a cohort of Australian men who have sex with men (MSM). MethodsWe collected clinical records of 21,273 Australian MSM during 2011–2017. We compared accuracies for predicting HIV and STIs (syphilis, gonorrhoea, chlamydia) diagnosis using four machine learning approaches against a multivariable logistic regression (MLR) model. ResultsMachine learning approaches consistently outperformed MLR. Gradient boosting machine (GBM) achieved the highest area under the receiver operator characteristic curve for HIV (76.3%) and STIs (syphilis, 85.8%; gonorrhoea, 75.5%; chlamydia, 68.0%), followed by extreme gradient boosting (71.1%, 82.2%, 70.3%, 66.4%), random forest (72.0%, 81.9%, 67.2%, 64.3%), deep learning (75.8%, 81.0%, 67.5%, 65.4%) and MLR (69.8%, 80.1%, 67.2%, 63.2%). GBM models demonstrated the ten greatest predictors collectively explained 62.7-73.6% of variations in predicting HIV/STIs. STIs symptoms, past syphilis infection, age, time living in Australia, frequency of condom use with casual male sexual partners during receptive anal sex and the number of casual male sexual partners in the past 12 months were most commonly identified predictors. ConclusionsMachine learning approaches are advantageous over multivariable logistic regression models in predicting HIV/STIs diagnosis.

Management of low-risk prostate cancer in a population-based cohort of Australian men between 2008 and 2016.

13 Background: The aim of this study is to evaluate the practice pattern of management of NCCN low-risk prostate cancer (LRPC) in a population-based cohort of Australian men. Methods: This is prospective cohort of men captured in Prostate Cancer Outcomes Registry Victoria (PCOR-Vic), who were diagnosed with LRPC between Aug 2008 and Dec 2016. Conservative management was defined as no active treatment (surgery, radiotherapy, or other local therapy) within 12-month of LRPC diagnosis. Chi-squared test for trend was used to evaluate change in practice over time. Multivariate logistics regressions were used to patient-, tumour- and institutional factors influencing the likelihood of conservative management for LRPC. Results: A total of 3238 men with LRPC were identified in the PCOR-Vic database. The median age was 62.6 (range:37-94). The median PSA level was 5.1ng/mL (range: 0.01-9.96). Overall, 1934 (60%) had conservative management, of which 1668 (86%) were documented as being on active surveillance. Of the 1304 (40%) men who active treatment within 12-month of diagnosis, 977 (30%) had surgeries, 289 (9%) had radiotherapy, and 38 (1%) had other local treatment. Overall, there is increasing trend in conservative management for LRPC from 52% in 2009 to 73% in 2016 (P<0.001). In multivariate analyses, age, PSA, clinical stage, institutions and year of diagnosis were all independently associated with conservative management. Men diagnosed in private and regional centres were 26% (95%CI=0.63-0.88, P=0.001) and 40% (95% CI=0.51-0.72, P<0.001) less likely to have conservative management for LRPC. Conclusions: This is the largest Australian series on management of LRPC to date. We observe increasing use of conservative management for LRPC over time, however, there is large institutional variations in care with men diagnosed in private and regional centres more likely to have active local management for LRPC.

Why do men go to the doctor? Socio-demographic and lifestyle factors associated with healthcare utilisation among a cohort of Australian men.

BackgroundMen use health services less often than women and frequently delay seeking help even if experiencing serious health problems. This may put men at higher risk for developing serious health problems which, in part, may explain men’s higher rates of some serious illnesses and shorter life span relative to women. This paper identifies factors that contribute to health care utilisation in a cohort of Australian men by exploring associations between socio-economic, health and lifestyle factors and the use of general practitioner (GP) services.MethodsWe used data from Ten to Men, the Australian Longitudinal Study on Male Health. Health care utilisation was defined in two ways: at least one GP visit in the past 12 months and having at least yearly health check-ups with a doctor. Associations between these two measures and a range of contextual socio demographic factors (education, location, marital status, country of birth, employment, financial problems etc.) as well as individual health and lifestyle factors (self-rated health, smoking, drinking, healthy weight, pain medication) were examined using logistic regression analysis. The sample included 13,763 adult men aged 18 to 55 years. Analysis was stratified by age (18 to 34 year versus 35 to 55 years).ResultsOverall, 81 % (95 % CI: 80.3–81.6) of men saw a GP for consultation in the 12 months prior to the study. The odds of visiting a GP increased with increasing age (p < 0.01), but decreased with increasing remoteness of residence (p < 0.01). Older men, smokers and those who rate their health as excellent were less likely to visit a GP in the last 12 months, but those on daily pain medication or with co-morbidities were more likely to have visited a GP. However, these factors were not associated with consulting a GP in the last 12 months among young men.Overall, 39 % (95 % CI: 38.3–39.9) reported having an annual health check. The odds of having an annual health check increased with increasing age (p < 0.01), but showed no association with area of residence (p = 0.60). Across both age groups, the odds of a regular health check increased with obesity and daily pain medication, but decreased with harmful levels of alcohol consumption.ConclusionThe majority of men (61 %) did not engage in regular health check-up visits, representing a missed opportunity for preventative health care discussions. Lower consultation rates may translate into lost opportunities to detect and intervene with problems early and this is where men may be missing out compared to women.

AB0222 Incidence of ar in albania (based on hospital data, 2004-2011)

BackgroundRheumatoid arthritis (RA) is a systemic autoimmune disease that causes chronic inflammation of the joints and surrounding tissues.In this paper we provide the first data on the incidence of AR...

Chronic disease prevalence and associations in a cohort of Australian men: The Florey Adelaide Male Ageing Study (FAMAS)

BackgroundAn increasing proportion of Australia's chronic disease burden is carried by the ageing male. The aim of this study was to determine the prevalence of asthma, cancer, diabetes, angina and musculoskeletal conditions and their relationship to behavioural and socio-demographic factors in a cohort of Australian men.MethodsSelf-reports of disease status were obtained from baseline clinic visits (August 2002 – July 2003 & July 2004 – May 2005) from 1195 randomly selected men, aged 35–80 years and living in the north-west regions of Adelaide. Initially, relative risks were assessed by regression against selected variables for each outcome. Where age-independent associations were observed with the relevant chronic disease, independent variables were fitted to customized multiadjusted models.ResultsThe prevalence of all conditions was moderately higher in comparison to national data for age-matched men. In particular, there was an unusually high rate of men with cancer. Multiadjusted analyses revealed age as a predictor of chronic conditions (type 2 diabetes mellitus, angina, cancer & osteoarthritis). A number of socio-demographic factors, independent of age, were associated with chronic disease, including: low income status (diabetes), separation/divorce (asthma), unemployment (cancer), high waist circumference (diabetes), elevated cholesterol (angina) and a family history of obesity (angina).ConclusionSocio-demographic factors interact to determine disease status in this broadly representative group of Australian men. In addition to obesity and a positive personal and family history of disease, men who are socially disadvantaged (low income, unemployed, separated) should be specifically targeted by public health initiatives.