Randomisation form

You first need to select an existing form to be your randomisation form. When this form is saved, it will trigger randomisation to one of your treatment groups. The randomisation form should contain any prognostic factors you wish to balance on. The dropdown menu will show all the forms you have created in your CRF. Pick the one you want to be the randomisation form from this list.

Choosing the randomisation form

Next, you need to define the treatment group names as you'd like them to appear in the randomisation system. In blinded randomisation systems, these names will not be shown to the user, except when using the unblinding feature. Blinded systems allocate a kit code to the participant when randomised, and the user sees this code instead of the treatment group name. Indicate the allocation ratio, such as 1:1 for equal allocation of two treatments.

Defining treatment groups and allocation ratio

Subsequently, select the API only randomisation option if randomisation is to be performed only via the API by an external EDC system (such as OpenClinica).

Method

Go to the Method tab to select the method for balancing allocations to the treatment groups. Choose from list, dynamic blocks, or minimisation. Comprehensive information and links are provided on this page about the randomisation methods used by Sealed Envelope. If you are unsure which to use, you should consult a statistician.

Additionally, add your prognostic factors by using the drop-down controls, which contain all the category and Yes/No fields on the randomisation form, as well as site. Site is a system variable that always exists and reflects the site the participant is recruited at.

Choosing prognostic factors

Blinding

On this tab, you need to select whether the randomisation system will be blinded or unblinded. Blinded systems return a treatment code from a code list rather than the treatment group. On the other hand, unblinded systems freely show the treatment group to users of the system. You can find more information on blinding here.

If you require a blinded system, you must decide whether administrators are allowed to manage the code list. Choose Yes if your kits will be pre-labelled with kit codes before being shipped to sites or codes are site specific; in this case, an administrator will need to update the code list regularly to reflect the location and availability of kits. If kits are not pre-labelled and, for instance, an unblinded pharmacist makes up the kit on demand by referring to the code list, you do not need to track the location and availability of kits. In this case you can turn code list management off.

Please indicate also whether and how unblinding will take place through the randomisation system (unblinding involves revealing the true treatment of a participant to a designated person when it is felt necessary to do so on clinical grounds). Several options are available: no unblinding, screen, email, screen and email, and third-party unblinding.

Defining unblinding option for your system

Note that a separate unblinding role can be allocated to appropriate individuals who are authorised to unblind. For convenience, you can also allow investigators to use the unblinding facility.

Study size

This section records the following information for your trial:

• Randomisation limit - the number of randomisations will be capped at this limit
• Expected duration of recruitment – this is for information only and does not affect the system
• Expected number of sites – this is for information only and does not affect the system

Randomisation list

If you chose List as your randomisation method, the Randomisation list tab will become active. You will need a randomisation list both for testing and for the live (production) system. You can create and upload your own lists, or they can be generated automatically. Generated lists are stratified by any prognostic factors you've specified. 

If you choose to automatically generate the randomisation list, the seed for both the test and live systems can be chosen, and the list length must be specified. Additionally, you can choose the block sizes, which will apply to both test and live lists. You can have a look at our simulation tool for advice on choice of block sizes.

Code list

For a blinded trial, you can choose whether you want the code lists (one for testing and one for the live system) to be generated by Sealed Envelope. Otherwise, you can create and upload your own lists.

If you choose to automatically generate the code list, the seed for both the test and live systems can be chosen, and the list length needs to be specified. Additionally, you can choose the kit block size, which will apply to both test and live lists. Kits will be balanced within kit blocks and are used by the dispensing policy when choosing a kit. You can also choose the type of kit code: alphanumeric or numeric.