Transcript OPEN_Randonode_Data Validation_20080919

Data Validation
OPEN Development Conference
September 19, 2008
Sushmita De
Systems Analyst
Data Validation Overview….
OPEN application accepts registration data from the
user, and transfers the registration data to the lead
group’s RandoNode for processing. This document
describes the process to verify the data transfer
integrity between OPEN and RandoNode.
09/19/08
Data Validation
2
Purpose of Data Validation….
These validation processes do not handle any
physical transmission error. The primary purpose of
these tests is to uncover any system defects and to
prove system functionality.
09/19/08
Data Validation
3
Scope of Data Validation…
•
Validate the persistence of data from OPEN screen into OPEN
Randohub database.
•
Validates the integrity of the data sent from OPEN to RandoNode.
Scope: OpenRequest , OpenRegistration and ODM (metadata and
clinical data).
•
Validates the integrity of the data sent from RandoNode to OPEN.
Scope: OpenResponse
•
CTSU’s local RandoNode was used for persisting the data to the
database.
09/19/08
Data Validation
4
Steps for Data Validation from OPEN screen to
OPEN database…
•
Enroll a patient using OPEN portal
Enter group information, protocol information, Credit information
and all associated administrators information and check list data.
•
There are two ways to verify Screen data persistence :
1. Compare hard copy data to OPEN screen.
2. Compare hard copy data to Randohub database.
09/19/08
Data Validation
5
Steps for Data Validation from OPEN screen to
OPEN database…(continued)
1. Compare hard copy data to OPEN screen.
 Query for the selected registration data in OPEN
 Manually verify the summary page & eligibility checklist with
hard copy.
2. Compare hard copy data to Randohub database.
 In Randohub database, MYRESP view contains the registration
data
 Query MYRESP view with the registration id (Tracking #)
 Manually compare the query result with hard copy.
09/19/08
Data Validation
6
Steps for Data Validation from OPEN to
Randonode & Vice Versa
•
Open Portal saves the data in Randohub schema for
both request sent and response received.
•
Randonode saves the data in Randonode schema
•
Sql validation scripts compare data between Randohub
(OPEN database) and Randonode.
•
Validation scripts create and store data in temporary
tables in Randonode to compare the data.
•
In all cases if test is passed, resultant temporary table
will not contain any data, otherwise it will contain the
records which are not matching in either schema.
09/19/08
Data Validation
7
Randonode Data Model…
09/19/08
Data Validation
8
Validation Observations
•
Identified few data discrepancy which were difficult to
identify through system integration testing, since in most
of the cases generated final output was correct.
•
Few of these observations are related to persisting data
in Randonode, not affecting group transaction.
•
All of the observations have already been fixed except
two for which the work is under progress.
09/19/08
Data Validation
9
Observations corrected……
•
LF(Line Feed) was getting appended to few of the text
fields during data transmission to RandoNode.
•
CreationDateTime of ODM node in metadata was not
getting saved in database.
•
ODM.Creation_Date_Time, Open_registration.
Randomized_Date was saving only the date part.
•
For all kind of investigators (treating investigator, drug
shipment investigator etc.), “IVR” was sent as
OPENUserType in clinical data. This was creating
confusion for same person playing multiple roles.
09/19/08
Data Validation
10
Observations corrected……(Continued)
•
For Treating & Drug shipment investigators, user type
sent in clinical data supposed to be “Investigator”
instead of “other”.
•
For multiple items having same codeList values,
codeLists were not getting saved properly in database.
•
Randohub.open_responses(in OPEN database) table
was not getting populated with the response detail
received from groups.
09/19/08
Data Validation
11
Observations…Work in progress…..
•
Override requests are associated with only override
codes, not with edit checks.
•
Override requester and reviewer detail are not present
in admin section of clinical data.
•
Both of these depend on procedures that are being
developed.
09/19/08
Data Validation
12
Future Plans for Validation…
•
Validate the integrity of the data sent from OPEN to .Net
RandoNodes.
•
Validate groups’ RandoNode response time during load
testing.
•
Validate meta data files content in the RandoNode
repository against the CaDSR repository based on the
form’s public id.
09/19/08
Data Validation
13