Machine-learning based identification of undiagnosed dementia in primary care: a feasibility study.

BACKGROUND: Up to half of patients with dementia may not receive a formal diagnosis, limiting access to appropriate services. It is hypothesised that it may be possible to identify undiagnosed dementia from a profile of symptoms recorded in routine clinical practice. AIM: The aim of this study is to develop a machine learning-based model that could be used in general practice to detect dementia from routinely collected NHS data. The model would be a useful tool for identifying people who may be living with dementia but have not been formally diagnosed. DESIGN & SETTING: The study involved a case-control design and analysis of primary care data routinely collected over a 2-year period. Dementia diagnosed during the study period was compared to no diagnosis of dementia during the same period using pseudonymised routinely collected primary care clinical data. METHOD: Routinely collected Read-encoded data were obtained from 18 consenting GP surgeries across Devon, for 26 483 patients aged >65 years. The authors determined Read codes assigned to patients that may contribute to dementia risk. These codes were used as features to train a machine-learning classification model to identify patients that may have underlying dementia. RESULTS: The model obtained sensitivity and specificity values of 84.47% and 86.67%, respectively. CONCLUSION: The results show that routinely collected primary care data may be used to identify undiagnosed dementia. The methodology is promising and, if successfully developed and deployed, may help to increase dementia diagnosis in primary care.

What is the relationship between age and deprivation in influencing emergency hospital admissions? A model using data from a defined, comprehensive, all-age cohort in East Devon, UK.

OBJECTIVES: To clarify the relationship between social deprivation and age as two factors associated with emergency admissions to hospital. DESIGN: Emergency admissions for 12 months were analysed for patients in the NHS NEW Devon CCG. Social deprivation was measured by the Index of Multiple Deprivation (IMD). Logistic regression models estimated the separate and combined effects of social deprivation and age on the risk of emergency admissions for people aged under and over 65. SETTING: East Devon, UK-area of the NEW Devon CCG. POPULATION: 765 861 patients in the CCG database. MAIN OUTCOME MEASURE: Emergency admission to any English hospital. RESULTS: Age (p<0.001) and social deprivation (p<0.001) were significantly associated with emergency admission to hospital, but there was a significant interaction between age and social deprivation (p<0.001). From the third quintile of age upwards, age progressively overtakes deprivation and age has a dominant effect on emergency admissions over the age of 65. The effect of age was J-shaped in all deprivation groups, increasing exponentially after age 40. For patients under 65, age and social deprivation had similar risks for emergency admissions, the differences in risk between the top and bottom quintiles of IMD and age being ∼1.5 and 0.9 percentage points. In patients over 65, age had a much greater effect on the risk of admissions than social deprivation, the differences in risk between the top and bottom quintiles of IMD and age being ∼2.8 and 18.7 percentage points. CONCLUSIONS: Risk curves for all social groups have similar shapes, implying a common biological pattern for ageing in any social group. Over age 65, the biological effects of ageing outweigh the social effects of deprivation. Our model enables CCGs to anticipate and plan for emergency admissions to hospital. These findings provide a new logic for allocating resources to different populations.

Women's preferences for masculinity in male faces are highest during reproductive age range and lower around puberty and post-menopause.

Masculinity in male faces is thought to be a sign of mate quality and is associated with measures of long-term health. Previous studies have demonstrated that women's masculinity preferences change across the menstrual cycle with women preferring more masculine men during phases of the menstrual cycle where fertility is highest (i.e. the late follicular phase). Given the hormonal correlates of such preferences and that these hormones change across the life span, we tested for differences in female masculinity preferences at different ages. We compared the masculinity preferences of peri-pubescent girls and young adult women (Study 1), circum-menopausal women reporting to either be pre- or post-menopause (Study 2), and a large sample of women across a wide range of ages (Study 3). In all three studies, preferences for masculinity in male faces were highest in women who were at a reproductively active age. Preferences for masculinity were lower when females were peri-pubescent, post-menopausal, or at ages corresponding to these groups. These data support the notion that masculinity in male faces is an important trait for reproductively relevant mate choice decisions. These data also highlight a shift in female visual preferences for men that is associated with important stages of the lifespan. Visual preferences appear to track important hormonal changes associated with age; as women pass puberty their preferences shift towards facial traits associated with mate quality and as women undergo menopause their preferences for such facial traits decrease. Overall, these results demonstrate the important role of reproductive status and support the notion that preferences for male faces are tied to reproductively relevant hormones.