Clinical dataset

Question

Does anyone have any idea where I can get free clinical data (it does not matter what it is) for a data mining project I have? I need to use the data to predict something like mortality or hospitalization or... to do classification basically.

An example might be: I have data on cancer patients containing conditions (stroke, chf, etc.) and demographics. Using these features (some or all) I can predict one year mortality. There are no specific criteria. I just need a relatively big and slightly complex (health-related) data.

I need to do some preprocessing first (so it should not be clean data) and use something like Weka to do the prediction.

The data can be in the following formats csv, tsv, xls

Answer 1

If you want data about hospitals, especially about the quality of caregiving and effectiveness of treatments and operations, check this file, it contains a subset of http://hospitalcompare.hhs.gov/, a site run by the U.S. Department of Health and Human Services. The purpose of the web site is to provide data and information about the quality of care at over 4,000 Medicare-certified hospitals in the U.S.

If you want data from clinical trials, check out clinicaltrials.gov, a huge "database of databases".

If you want the data from Europe, read this news article published in Oct 2016 on nature.com: Europe’s drug regulator opens vaults of clinical-trials data

If you want lab data, go to bioconductor.org and check their "Illustrative experiment data packages".

The R packages linked there are actually small mini-databases of 3-5 "tables", patient-sample-condition-experiment-result. All databases have a similar design.

For Neuroscience data, check http://synapse.org.

There are many other databases. Sorry this is a borad answer but you also asked a broad question.

Answer 2

You should also take a look at MIMIC-III dataset. It has almost 50K patients with many features. This github repository has the preprocessing scripts for some prediction tasks such as mortality. Also, there are standard cookbook of queries published by the owners of the MIMIC-III dataset here.