OpenVigil v1.2.6b

This web application permits you to process a query on the FDA Adverse Event Reporting System (AERS) pharmacovigilance data. Click here for a tutorial on how to use it. Further information/specifications can be found in the documentation. Limitations can be found in the OpenVigil Cave-At document.
Press to display more information about the database structure and content.
This installation uses data from 2003-10-06 to 2013-12-31 (according to DEMO.FDA_DT).

Registered tables are:
DEMO with columns ISR CASENO I_F_COD FOLL_SEQ IMAGE EVENT_DT MFR_DT FDA_DT REPT_COD MFR_NUM MFR_SNDR AGE AGE_COD GNDR_COD E_SUB WT WT_COD REPT_DT OCCP_COD DEATH_DT TO_MFR CONFIG REPORTER_COUNTRY DSRC containing 5337037 records.
DMAP with columns ISR DRUG BRAND EXTRA DOSE ROUTE containing 0 records.
DRUG with columns ISR DRUG_SEQ ROLE_COD DRUGNAME VAL_VBM ROUTE DOSE_VBM DECHAL RECHAL LOT_NUM EXP_DT NDA_NUM DSRC containing 19388354 records.
DSRC with columns FNAME DT NERR_PARSER NERR_SQL TERR_PARSER TERR_SQL containing 281 records.
INDI with columns ISR DRUG_SEQ INDI_PT DSRC containing 9441806 records.
OUTC with columns ISR OUTC_CODE DSRC containing 4810836 records.
REAC with columns ISR PT DSRC containing 19239224 records.
RPSR with columns ISR RPSR_COD DSRC containing 1992810 records.
THER with columns ISR DRUG_SEQ START_DT END_DT DUR DUR_COD DSRC containing 7835655 records.
Interesting fields for novice users might be DRUG.DRUGNAME (mostly the USAN (U.S. adopted name, see entry in wikipedia) drug name but other names are used (list), too) and REAC.PT (the adverse reaction coded as MedDRA Preferred Term (click here for a list).

For advanced users: a description of the content of tables and columns in these tables can be found in chapters B. FILE DESCRIPTIONS, C. DATA ELEMENT DESCRIPTIONS, D. DATA ELEMENT CONTENTS AND MAXIMUM LENGTHS of the FDA Asc_nts.doc database description document (created by the FDA and can also be found in the archive of the original data). Note that DEMO.CASE is called DEMO.CASENO in our database. Import failures from the original FDA data into this database can be found in the data import table from aers2sql for this instance of OpenVigil.
Pitfalls and shortcomings of pharmacovigilance data analyses are described in the OpenVigil Cave-At document.


Step 1: Chose how to construct your query. Create query...
Step 2: Construct a SQL query and enter it in the text box below. Here is an example that shows you the MySQL syntax, database table and field adressing and how to link the tables by forcing the ISR numbers to be the same (this example uses some further MySQL magic like AS that is explained in the MySQL manual):
SELECT D.* FROM DRUG AS D, REAC AS R WHERE D.ISR=R.ISR AND (R.PT="DIZZINESS" AND DRUGNAME="HALOPERIDOL");
Type your SQL query to AERS here:



Please enter additionally the SQL user password for this installation:
Step 2: Select either a drugname or an adverse event for which a safety analysis via PRR shall be conducted!
Notez bien that the AERS data is not up-to-date but rather several months old.

Drugname
Adverse event

Step 2: Enter a ISR number to get all records associated with this case!


Step 2: Fill out at least one of the fields below to filter out the cases you are interested in. As result, you will get either a list of case numbers (ISR number) which you can click to get further information about every reported case or statistics on the frequency. Note that this search more is not very powerful; consider watching the tutorial and using the Wizard in professional mode or writing SQL queries yourself! Also note that you have to use the USAN drug name and that both indication and adverse event are named according to the MedDRA terminology. OpenVigil will attempt to find the best match for the drugname and/or the adverse event that you have entered.

Drugname Matching:
Adverse event Matching:
Show results as as raw data (i.e., a list of each single individual safety report)
as statistics (likelihood that a drug is connected to an adverse event*)

*) i.e., if either drugname or adverse event is given, a sorted list of the number of occurences of each adverse event linked to drug or - vice versa - of each drug linked to an adverse event; if both drugname and adverse event are given, a proportional reporting rating for this drug and this putative adverse reaction.
Step 2: Select the filtering conditions! Which records shall be extracted? E.g., focus on a specific drugname (DRUG.DRUGNAME is equal HALOPERIDOL) or adverse reaction (REAC.PT is equal DIZZINESS).

concat.database fieldoperatorvalue


Step 3: Select whether the above filtered records shall be extracted as raw data or whether this data shall be postprocessed. If you want to analyse the data (a horizontalized table) yourself, select raw data, otherwise request a statistical analysis:

Step 4: Select which data fields shall be put into the horizontalized table. E.g., if you filtered for a certain drugname (DRUG.DRUGNAME is equal HALOPERIDOL) you might be interested which adverse reactions were reported (select REAC.PT), or vice versa.


This analysis module is currently very basic. Consider exporting your query results and using a real statistics program like Gnumeric/R.

Step 4a: You can select a data field from which categories/groups shall be made (a so called "factor"). Each group is counted and the number of records per group is presented. If you want this, select a data field here:



Step 4b: The numerical data in a data field can be descriptive-statistically analysed (average, standard deviation, minimum, maximum). If you want this, select a data field here:
If a factor is selected, each group will be analyzed.



Note: You can also use both factor and descriptive statistics simultanously.




Export results as human readable HTML or standard CSV or Microsoft Excel CSV?
Please note that no pictures (e.g., PRR/ChiSquare charts or pie charts) can be exported via CSV!
Only show or count unique ISRs?

.


Security code:
Please enter the characters/numbers that you see in the picture above:

(this protection from automated queries requires the use of cookies)



Depending on query complexity and server load your query might take some seconds to several minutes to complete. There is no progress indicator, so please wait! Once the results are ready, the new window will show the results. You can start a new query in the meantime in this window.