This week we released version 1 of our Automated User Acceptance Testing (UAT) feature to Beta site users and we are starting to work with customers to implement this approach into their processes. Manually testing Edit Checks and other dynamic features of Medidata Rave studies is tedious, error-prone work that takes a long time. Organizations which adopt the Automated UAT approach can test their studies faster, gather better quality testing evidence and free up their skilled staff for higher-value work.

Need convincing? Here's a video which demonstrates in 1 minute 38 seconds:

• Creating a test script for an Edit Check
• Reviewing the test script
• Executing the test and collecting screenshot evidence (also available in PDF)
• Reviewing the results in Medidata Rave

This video shows the testing for a single Edit Check but the system is also able to create and run test cases for hundreds of Edit Checks at a time. It also works for Rave EDC as well as Rave Classic (shown in video).

If you are ready to accelerate your Edit Check testing. Get in touch!

In the last part we introduced the Given..When..Then structure for writing User Acceptance Tests and showed how TrialGrid automates execution of these tests against Medidata Rave and also how the tests can be checked for validity before they are even run. In this final part we'll look at automating the generation of tests for Edit Checks. This blog post is a bit technical but if you are curious (or skeptical) about our claim of generating tests for 1,000 Edit Checks in 60 seconds, this post explains how we do it.

Alternatively, you could sign up for our free webinar https://register.gotowebinar.com/register/2929700804630029324 and watch TrialGrid in action. Seeing is believing!

Solving Constraints

Before we start looking at Edit Checks let's first look at a little puzzle. Maybe you did this kind of puzzle at school.

1
2
3
4
5
6

I have two integer numbers (whole numbers) A and B. 
A is always a positive number 
A is always an even number 
A is always greater than zero
A is always less than B
If B = 10, what are the possible values for A?

The answer is that A can be 2, 4, 6 or 8. You probably solved this puzzle in seconds.

This kind of problem is known as a Constraint Statisfaction Problem and it is well studied in Computer Science circles. My computer solved this puzzle in 0.0007 seconds - though I helped it out a bit by telling it only to look at values for A in the range 1 to 100 since I didn't want to wait while it considered the infinite set of positive even numbers. Computers can be really good at solving these problems if the set of values to search in isn't too large.

Edit Checks are Constraint Problems

Now we've warmed up on a simple constraint problem lets look at another. This is an Edit Check that uses Systolic and Diastolic blood pressure to identify pre-hypertension:

1
2
3
4
5
6
7

 IF:
    field SYSBP in form VS in folder SCREEN is greater than or equal to 120
    AND field SYSBP in form VS in folder SCREEN is less than or equal to 129
    AND field DIABP in form VS in folder SCREEN is less than 80
 THEN:
    Open Query "Subject is pre-hypertensive (SYS 120..129 and DIA < 80). Please confirm." 
    on field SYSBP in form VS in folder SCREEN to Marking group "Site from System"

In TrialGrid CQL the logic test part would look like:

1
2
3

SCREEN.VS.SYSBP[0] >= 120 
AND SCREEN.VS.SYSBP[0] <= 129
AND SCREEN.VS.DIABP[0] < 80

And in Medidata Rave QuickEdit:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11

||StandardValue|VSORRES|SCREEN|VS|SYSBP|0||||||
|120|3|||||||||||
IsGreaterThanOrEqualTo|||||||||||||
||StandardValue|VSORRES|SCREEN|VS|SYSBP|0||||||
|129|3|||||||||||
IsLessThanOrEqualTo|||||||||||||
And|||||||||||||
||StandardValue|VSORRES|SCREEN|VS|DIABP|0||||||
|80|2|||||||||||
IsLessThan|||||||||||||
And|||||||||||||

Now if we wanted to work out what values would fire this Edit Check we can formulate it as a Constraints problem using the variables SYSBP and DIABP and quickly come up with a truth table of possible values and their results:

Okay, the last two are not likely ones you would come up with but the way we have this Edit Check formulated these are valid values. The SYSBP and DIABP Fields are defined in this study with a DataFormat of 3 which means maximum value 999 and minimum value -999. For the computer trying to solve this problem to work out which values will fire the Edit Check the constraints are:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10

From Data Format:
    SYSBP must be >= -999
    SYSBP must be <= 999
    DIABP must be >= -999
    DIABP must be <= 999

From Edit Check:
    SYSBP must be >= 120
    SYSBP must be <= 129
    DIABP must be < 80

That gives us ranges for SYSBP of (120..129) and for DIABP of (-999..79) When values are within those ranges, the Edit Check will fire. The system can also work out what values will cause the Edit Check not to fire.

Solving the Constraint Problem

So if Edit Checks are just a type of Constraint Problem, how does TrialGrid solve them?

We said already that the definition of our example edit check in TrialGrid is:

1
2
3

SCREEN.VS.SYSBP[0] >= 120 
AND SCREEN.VS.SYSBP[0] <= 129
AND SCREEN.VS.DIABP[0] < 80

This format is generated automatically on import of Checks from Medidata Rave but users can also create Edit Checks using this format directly. In order to convert it to QuickEdit or back into the Architect Loader Spreadsheet format we have to parse this syntax. That turns the text into a tree-like structure in-memory:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14

   - LogicalOperator("AND") 
       - LogicalOperator("AND") 
            - Expression
                - DataPoint(folder="SCREEN", form="VS", field="SYSBP" recordposition=0)
                - Comparator(">=")
                - Value("120")
            - Expression     
                - DataPoint(folder="SCREEN", form="VS", field="SYSBP" recordposition=0)
                - Comparator("=<")
                - Value("129")
        - Expression     
            - DataPoint(folder="SCREEN", form="VS", field="DIABP" recordposition=0)
            - Comparator("<")
            - Value("80")

And by examining this tree we can generate a constraint formula by looking at the DataFormats of Fields. For numeric Fields this is simple but for other types of Field this is not too hard. Checkbox fields can only have a value of 1 or 0. Computers already store dates/times as the number of seconds since January 1 1970 and fields with Data Dictionaries are already constrained to a limited set of values (e.g. "YES", "NO", "UNKNOWN")

For any Edit Check there are likely hundreds of value combinations that will fire the check and hundreds more that will not fire the check. For now TrialGrid is set to generate a single positive and a single negative test for each Edit Check using constraints that we feel are reasonable. For example, the range of future dates is infinite but the system won't generate dates out in the year 9800.

Finding invalid Edit Checks

Medidata Rave will allow you to write Edit Checks which are valid but which will never fire. A simple example:

1

SCREEN.VS.SYSBP[0].IsEmpty AND SCREEN.VS.SYSBP[0] > 100 

The SYSBP value can never be both Empty and > 100. Nobody sets out to write a check like this but in a complex Check it can happen. The constraint solver won't be able to find a solution to this problem and will fail with a warning.

Generated test case

The TrialGrid constraint solver can solve many Edit Checks and generate a positive (Check fires) and negative (Check does not fire) testcase for each in the Given..When..Then syntax.

Here is a testcase generated by TrialGrid for our example Edit Check:

TrialGrid has a batch-create option which allows you to select as many Edit Checks as you like and generate tests for them. This generally takes less than a minute for 1,000 Edit Checks.

Running tests can also be done in batch so you can take a study with no tests, generate tests and start looking at the results in about 5 minutes.

Summary

In this mini-series of blog posts we've tried to show the basics of our Automated User Acceptance Testing system. We've described what UAT is, how we format UAT tests so that they are meaningful to humans and can be executed by a computer and finally shown how the system can generate tests to get you started quickly.

That's the high level view, if you want to see this system in action please sign up for our Free Webinar on January 10 2019.

Registration at https://register.gotowebinar.com/register/2929700804630029324

Yesterday I described the cost and effort of creating, executing and maintaining test scripts for User Acceptance Testing. I also made the bold statement that the TrialGrid approach could reduce this effort by a factor of 10. As you might expect, we achieve this through automation. Of the three parts:

1) Create the tests
2) Execute the tests
3) Maintain the tests

The first, Create the tests, is the most technically challenging. Before we can automate creation of tests we first need to understand what tests are going to look like and how they are executed.

If we look at the test script from the first part of this series we can see that it is designed to be read and executed by a human:

Name : Check SCREENING_VISIT_DATE
Version : 1

Once executed the signed and dated test script will be kept as evidence to be reviewed by the Sponsor.

Clearly, if we want to automate the process of executing these scripts we need to keep that readability. We need a format for test scripts that is structured enough for software to execute but also naturally readable for humans.

Executable Specifications

Fortunately, Software Development has had a solution to this problem for nearly a decade. Behaviour Driven Development (BDD) is an approach to writing specifications and acceptance tests which can be read and understood by humans and executed by software. This is exactly what we are looking for. BDD doesn't specify any particular format for these tests but the most widely adopted standard is called "gherkin"[1].

Gherkin uses a simple syntax. Here is a short example:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10

Feature: Buying things from the shop

  If a user has money they can buy things

  Scenario: Buying things
    Given Alice has $1.30 
    When she visits the grocery store
    And she buys 1 banana for $0.25
    Then she will have $1.05    
    And she will have 1 banana

This example starts with a "Feature" declaration. It's documentation telling is about the Scenario tests which follow. At line 3 we have some free text description of the tests. At line 5 we start with a Scenario called "Buying things".

Scenarios follow a format of:

Given some background information that sets up the test conditions
When some action is taken
Then I should see some result

Acceptance tests for Edit Checks

If we convert our original test script to the Gherkin Given..When..Then structure we might get:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13

# Version 1.0

Feature: Testing Edit check SCREENING_VISIT_DATE

  Visit Date should not be before Date of Birth

  Scenario: The Check Fires
    Given I log into EDC using an Investigator role 
    And I navigate to Site 123
    When I create a new Subject
    And I enter "10 JAN 1978" as the Date Of Birth on the Subject Form
    And I enter "13 DEC 1968" as the Visit Date for the Visit Form in the Screening Folder
    Then I will see query text "Visit Date is before subject Date of Birth. Please correct."   

This format is a little wordy, mostly because of the need to specify Fields, Forms and Folders for Data. Gherkin includes a data table structure which can help here and we can combine it with some simple shortcuts for field selection:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18

# Version 1.0

Feature: Testing Edit check SCREENING_VISIT_DATE

  Visit Date should not be before Date of Birth

  Scenario: The Check Fires
    Given I log into EDC using an Investigator role 
    And I navigate to Site 123
    When I create a new Subject
    And I enter data:
    | DataPoint                      | Value       |
    | SUBJECT.SUBJECT.DOB            | 10 JAN 1978 |
    | SCREENING.VISIT.VDATE          | 13 DEC 1968 | 
    Then I will see query text 
       """
       Visit Date is before subject Date of Birth. Please correct.
       """   

That makes the test a bit more concise.

Automated Testing

The Given..When..Then style of test is readable to a human but can also be read by software. The format isn't totally free-form. Each of the Given / When / Then steps must conform to a pattern that the software understands. Currently TrialGrid understands around 50 patterns which can check a wide range of states in Medidata Rave, not just whether a query exists. This means you can write tests which ensure that Forms and Fields are visible (or not visible) to certain EDC user Roles, to check the calculations for Derivations and to verify the results of data integrations such as IxRS feeds which enter data into Medidata Rave forms.

Tests can be read and then executed against a live Rave instance by the TrialGrid UAT module. Provide a Rave URL, study name, environment (e.g. TEST) and credentials to interact with Rave and the TrialGrid system will execute your tests against the Rave instance.

Data is entered via Rave Web Services and results verified automatically. Screenshots of the Rave page showing results of actions such as data entry and queries created can be captured for both Classic Rave and the new Rave EDC (formerly called RaveX). Results are updated in real-time as the system works through each step but you can also leave it to run unattended and view the results when it is done. TrialGrid runs these tasks in the background so you can get on with some other work.

The output is a PDF document that shows the actions taken and the results, comparing expected results against actual and providing screenshots as evidence.

Automated Maintainance

One of the challenges of test scripts is keeping them up-to-date with changes to the study. For example, imagine an Edit Check that ensures that when Race is "Other" then Race Other is specified on the demography form. The check has been programmed with the query text:

"Race is Other and Specify Other Race is missing. Please review and correct."

But in the test we are looking for:

"Race is Other and Specify Other Race is missing."

This could happen if the Specification or the programming of the Edit Check changed. But we want them to match. A human tester might be tempted to pass this test as "close enough" but automated test software looks for an exact match and will fail this test.

The TrialGrid approach can identify these kinds of problems before the test is ever run. In this example we see a warning in the Test editor which identifies that there is an issue:

Here we are giving the system extra hints about what are test relates to through the @EditCheck "tag" (another feature of the gherkin format) and referencing the DM001 Edit Check. This has several benefits:

1. By setting up a link between the Edit Check and the Test we can say whether an Edit Check has been tested or not and calculate what percentage of Edit Checks and other objects are exercised by tests.

2. The system has greater contextual knowledge about what is being tested and can help with warnings like the one shown here.

TrialGrid performs similar validation of data dictionary values, unit dictionary selections, Folder, Field and Form references and more. This capability reduces the effort of maintaining tests and supports risk-based approaches where you don't run tests because "nothing has changed and this test is still valid". This function can tell you something has changed and this test may not be valid any more.

Summary

In this second part of the three part series on Automated User Acceptance Testing we briefly covered the formatting of the tests, how they are executed and how the system helps you ensure that tests stay in synchronization with the Edit Checks, Forms and other objects that they are supposed to be testing.

These features make the execution and maintenance of tests much easier and faster but we are still left with the huge challenge of writing these kinds of tests for hundreds of Edit Checks. In the last part we'll cover how the TrialGrid system can automate that part, creating tests in seconds that would take a human hundreds of hours of effort.

Come back tomorrow. But if you want to see this system in action don't forget our Free Webinar on January 10 2019.

Registration at https://register.gotowebinar.com/register/2929700804630029324

Notes:

[1] Why "gherkin" is a story in itself but in summary, "gherkin" is the format used by a software tool called "cucumber" and it is called cucumber because passing tests are shown in green text so the idea was to get everything to "look as green as a 'cuke". I know, hilarious.