Tips and tricks - Before starting

In Castor, we use specific terms with a certain goal. Using this article, you can learn more about these terms, what they mean, and how you can use them. 

Basic definitions

Form: the eCRF structure that contains all your study phases, reports and surveys, where you will create all the fields that are relevant for your study. The form consists of the following elements:

• Phase: a specific study period which contains one or more steps.
• Step: a sub-section within each phase that contains multiple fields.
• Field: a question that needs to be answered, such as “How old is the subject?”

Record: a “case” - this is usually a participant in your study, but can also be a sample or another instance. A version of the form that you build will be available within each record.

Study vs. Report vs. Survey

While reports and surveys are also part of the study itself, Castor distinguishes these two from the basic study structure. This is an overview of the Castor structure:

Study: the main study structure that you will create, based on your study protocol. Here you should add all of the scheduled study phases in which you will be collecting all the necessary data. For example, a study can consist of the phases: Inclusion, Baseline, Follow-up. In each of these phases, you can create additional steps and fields to represent the data you will be capturing.

Report: part of the eCRF used for unscheduled phases or events, such as adverse events, concomitant medications and repeated measurements. Characteristics of a report:

• Repetitive: multiple report forms can be filled in on multiple occasions for the same subject (has a N:1 relationship with the record). This means that several adverse event reports can be registered for the same subject.
• Unscheduled: it is not part of the planned study, but needs to be recorded in the form.
• Same structure as the study: contains the same step/field structure as the general study form - thus the report will also contain fields representing questions that need to be answered.
• Can be linked to a certain study phase or another report.

Survey: a form used to create a survey that you will send out to participants. Characteristics of a survey:

• Relies on the same step/field structure as the main study: separate steps are the separate sections/pages of a survey and fields are the questions that need to be answered.
• Always belongs to a survey package: surveys are sent in the form of 'packages' and even if you want to send a single survey, it has to be sent in the form of a survey package.
• Contains unique URLs that are sent to the respondents' email addresses and are automatically attached to the records (study participants) they have been sent to. The answers of the respondents are automatically captured in the appropriate record.
  
• Can be linked to a certain study phase or report.
• Possibility to schedule in advance and send repetitively.

Tip: Repeated Measures 

In Castor, a Repeated Measure is a type of a report that can be used for any type of repeated data capture. For example, repeated ECG measurements or a medication diary can be captured using repeated measures. Although considered a report as per the Castor structure, the repeated measures can also be part of your main study structure.

Characteristics of the repeated measure:

• Scheduled or unscheduled: e.g. repetitive blood pressure measurement that can be a scheduled part of the study protocol or a follow-up to an adverse event
• Report structure: considered to be a report, as its structure has to be created under 'Reports'. Can be reflected in the study structure or another report in the form of a table
• Relies on the same step/field structure as other reports: the main difference is that it contains one step and the fields within the step appear as columns in a table, where the rows reflect each separate measurement.

Tip: Create a data dictionary

Before you start building your study in Castor we recommend that you make a data dictionary. A data dictionary is a file containing definitions of every field that you want to use. To help build your data dictionary, use the following 3 steps as a guide:

1. Define your field types:
   • Define what type of data you will be capturing and choose the appropriate field types and variable names. These can be numerical fields, date or date and time fields, dropdown fields etc.
   • Define the lower and upper limits of the fields and think about checking whether the entered value is correct. These checks are called validations.
   • Option labels and values: e.g. Male = 1, Female = 2 or Yes = 1, No = 0 (these definitions depend on your preference).
   • Think about how you are going to analyse your data and if you are going to use these options as stratification factors or for further calculations/dependencies in your CRF.
   • Think about using standards, such as predefined templates.
2. Define what values are “normal”
3. Define option groups

Here is a part of an example data dictionary: 

 

Definitions

• Record: in Castor a record is one "case". This can be a patient, but also an animal, sample etc.
• Form: a list of fields, e.g. a demographic form containing fields with patient information regarding age, gender or country of origin.
• Field: a question, such as: "What is the patient's age?" 
• Phase: a study period which contains forms.

 Adding new data fields

When you create fields always make sure that the question you want to ask is noted in the 'Label' field. Where possible, you should always enter the name of the variable in the ‘Variable name’ field.  Names of the variable will be used in export files and calculation fields. 

note - When using option groups in your study fields (e.g. radio buttons), we strongly advise that you are sure this is the correct method for your research, as it is difficult to change this later.

Copying of fields/forms

When you want to use the same field format multiple times, you can use the copy function. Fields can easily be reproduced by clicking the copy icon (see image). This enables you to quickly compile a list - for example, multiple yes or no questions. You can adjust the labels of the fields by double clicking the label name in the form builder  

You can also copy whole steps in this way. Before copying a step, we advise you to thoroughly test the step to prevent having to amend each copied step. You can test your form by:

• Creating a test record and filling in data for several patients.
• Export your data to see if it fits your expectations.
• Invite colleagues to test the form. 

 For more information on testing your forms, please follow this link. 

Hiding forms

Did you know that phases and steps can be hidden from predetermined users? This ensures that users can only view forms that are relevant to them. You can do this by selecting users that are not allowed to view the step with the ‘assign user roles’ button (see image below). Before doing this, you do need to assign roles to each user. You can find out how to do this here.

User rights

Always check the user rights before starting your data collection. Click here for more information on user rights. 

A few common user rights are:  

Data input	Add, view, edit (For the institute for whom the data is gathered)
Study monitor	View, query (For the institute that needs to be monitored)
Study admin	Depending on the tasks that need to be executed, this user has all management rights
Sending of Surveys	Add, view
Randomization	View, randomize, view randomization (if not blinded)
Principal investigator	View, edit, sign

  

Settings

We advise to always use the ‘clear inapplicable child fields’ function in the 'Settings' tab. This function ensures that fields that are hidden because of dependency are cleared. 

Besides this, consider what kind of record IDs you are planning to use. We advise to use incremental IDs for each institute or to use a unique pattern that the institute uses. For more information on changing record IDs, follow this link. 

Calculations

In Castor you can use calculations to obtain more information from your data. For example, when you wish to calculate a patient's age or calculate the difference between two dates. You can find common calculation templates here.