Upload Files

This web application predicts opioid use disorder from an input CSV file. The CSV file must include the following columns: QUESTID2, EDUD5PNRMIS, SUTRTPY, MEDICARE, ASTHMAEVR, PYUD5HER, SMIPY, PYUD5ALC, CATAG6, DIABETEVR, AMDEYR MHTRTPY, MRJYR, SUICTHNK, IRWRKSTAT, PYUD5COC, PRVHLTIN, IREDUHIGHST2, TOBYR, NEWRACE2, PYUD5MRJ, ALCYR, INCOME IRMARIT, NOBOOKY2, COUTYP4, IRHHSIZ2, HRTCONDYR, IRSEX, BOOKED, and UD5OPIANY. If any of these columns are missing, the application will return an error. A sample input file is provided below. The user manual can be found under the Help menu.

To predict OUD , please upload a CSV file using the data fields listed above. If you wish to add new data to the model (training set), please use the 'Add new opioid use disorder data to training (CSV)' upload box for opioid use disorder and the 'Add new non-opioid use disorder data to training (CSV)' upload box for non-opioid use disorder, respectively.

After uploading the necessary file, click the 'OUD prediction' button. Once the classification results are generated, you will be automatically redirected to the 'Predicted OUD' page. Then, click the 'Probability estimation' button to view the probability scores on the 'Probability Score's page.

The prediction and probability results can be downloaded by clicking the 'Prediction results' and 'Probability results' buttons, respectively. Additionally, the training/testing datasets are available in the 'Data' menu.

An example CSV file can be obtained by clicking on the 'Download Input Samples' button. To predict new data, the example CSV file should be in the form shown below:

If you find our web application useful, please cite the following paper.

Akhter, S. and Miller, J.H., 2025. Identifying Key Predictive Features for Opioid Use Disorder Using Machine Learning. medRxiv, pp.2025-07.