Predicting depression in old age: Combining life course data with machine learning.

With ageing populations, understanding life course factors that raise the risk of depression in old age may help anticipate needs and reduce healthcare costs in the long run. We estimate the risk of depression in old age by combining adult life course trajectories and childhood conditions in supervised machine learning algorithms. Using data from the Survey of Health, Ageing and Retirement in Europe (SHARE), we implement and compare the performance of six alternative machine learning algorithms. We analyse the performance of the algorithms using different life-course data configurations. While we obtain similar predictive abilities between algorithms, we achieve the highest predictive performance when employing semi-structured representations of life courses using sequence data. We use the Shapley Additive Explanations method to extract the most decisive predictive patterns. Age, health, childhood conditions, and low education predict most depression risk later in life, but we identify new predictive patterns in indicators of life course instability and low utilization of dental care services.

IDS Transposer: A Users Guide.

The Intermediate Data Structure (IDS) provides a standard format for storing and sharing individual-level longitudinal life-course data (Alter and Mandemakers 2014; Alter, Mandemakers and Gutmann 2009). Once the data are in the IDS format, a standard set of programs can be used to extract data for analysis, facilitating the analysis of data across multiple databases. Currently, life-course databases store information in a variety of formats, and the process of translating data into IDS can be long and tedious. The IDS Transposer is a software tool that automates this process for source data in any format, allowing database administrators to specify how their datasets are to be represented in IDS. This article describes how the IDS Transposer works, first by going through an example step-by-step, and then by discussing each part of the process and potential options and exceptions in detail.

Application of the Real World Data for Pharmacoepidemiology Reserach / 薬剤疫学

Recently, medical real world data involving claims data, drug data, and electrical medical record database have been developed worldwide and can be utilized for the phrmacoepidemiology research.The results of the pharmacoepidemiology research with these database contribute not only to post-marketing safety research but also to the outcomes research and pharmacoeconomic evaluation. On the other hand, collaborating with a number of local governments, we have developed maternal and child health check-up and school health check-up database in Japan. The development of healthcare lifecourse data like this will contribute to the preventive medicine, the understanding of the rare diseases, and drug development.

<b>Application of the Real World Data for Pharmacoepidemiology Reserach </b> / 薬剤疫学

<p>Recently, medical real world data involving claims data, drug data, and electrical medical record database have been developed worldwide and can be utilized for the phrmacoepidemiology research.The results of the pharmacoepidemiology research with these database contribute not only to post-marketing safety research but also to the outcomes research and pharmacoeconomic evaluation. On the other hand, collaborating with a number of local governments, we have developed maternal and child health check-up and school health check-up database in Japan. The development of healthcare lifecourse data like this will contribute to the preventive medicine, the understanding of the rare diseases, and drug development.</p><p></p>