Frequently asked questions

How do I use a blinded code list?

If your trial is blinded a code list will be used to link randomisation (kit) codes to treatments.

When a patient is randomised, a kit code is assigned from this list. Only people who have access to the list know what treatment this code is linked to. Normally the only people who have access to the list are:

  • Sealed Envelope
  • Trial pharmacies or drug packaging company
  • Independent statistician to the data monitoring committee
Kit code Treatment
AF5 Active
GY6 Placebo
YU1 Active
FT2 Placebo

Example code list

Codes are written or printed on the packaging of drugs you give to trial patients. This is usually done either by the trial pharmacies or a specialist drug packaging company. Obviously these parties will need to be unblinded and require a copy of the code list to do this.

If the codes are pre-printed on the kits, you will need to allocate different parts of the code list to different trial sites. You will need to update the randomisation system with this information to ensure the code allocated is available on site when randomising a patient. Conversely, if pharmacies hold unblinded stocks of the trial drugs a pharmacist can write the kit code on the packaging at randomisation. In this case there is no need to allocate different parts of the code list to different trial sites in advance.

Who creates the code list?

The code list can be created by your own statistician or by Sealed Envelope. In the former case please send the code list to us as a spreadsheet or CSV file.

What format should the kit code be?

The kit code can be any format - it can be a simple number (1, 2, 3…), a box/pack number (2000/21) or an alphanumeric code (AF3, TY8) as above. Some issues to bear in mind when devising a format are:

  • Each code must be unique. Do not repeat codes or use the same code at different sites.
  • If using a number format, kits should not be allocated in a way that leaks information inadvertently about the treatment allocation. For instance if you used the list shown in the table and codes were allocated in the order 1, 5, 2, 3 you might guess that the second patient was given a different treatment because codes 2-4 were skipped. To prevent this Sealed Envelope systems randomise the order kit codes are allocated within kit blocks. This means the codes jump around whilst still being allocated in approximately ascending order.
Kit code Treatment
1 Active
2 Active
3 Active
4 Active
5 Placebo

Code list with sequential codes

What is a randomisation list?

A randomisation list determines the order in which treatments will be allocated to subjects. It may be blocked and/or stratified. It does not contain kit codes - these are kept separately in a code list. Randomisation lists are not needed when dynamic methods of allocated are used, such as minimisation.

Sequence Block Block size Treatment
1 1 2 Active
2 1 2 Placebo
3 2 4 Active
4 2 4 Placebo
5 2 4 Placebo
6 2 4 Active

Example randomisation list

What fields should I include on my randomisation form?

If you are using our comprehensive randomisation service, we customise the randomisation form for your trial. The fields to include on this form would normally include:

  • A patient identifier, e.g. pre-existing trial identifier, date of birth and initials or clinic number. This is so that you can later match up randomisations to other patient records.
  • Stratification factors
  • Inclusion and exclusion criteria. Each criteria is a yes/no question. We will set up the form so that the patient must be eligible by these criteria to be randomised.

How do I randomise if I cannot access Sealed Envelope?

There may be rare occasions when you cannot access Sealed Envelope, either because of an issue with your internet access or down-time at Sealed Envelope. The trial coordinating centre should be prepared for this event and have a back-up procedure in place. This will usually be described in a randomisation SOP. Common strategies are:

  • Wait until the service is accessible again
  • If local access is the problem, ask a user with an administrative account (such as the trial manager) to perform the randomisation on behalf of the investigator
  • If randomisation cannot wait, for example because the treatment must be administered very soon after enrolment in the trial, perform a manual randomisation. This may involve phoning the trial coordinating centre to be issued with the next treatment from a predefined emergency randomisation list the centre has drawn up in advance. Alternatively it may involve taking the next available vial/kit/treatment pack from stock in a double blind drug trial.

The appropriate strategies will vary by trial. If a manual randomisation is performed, an administrator account can be used to update the randomisation system when it becomes available again.

What is a forced randomisation?

When a trial uses a code list and a dynamic method of allocation such as minimisation then occasionally the treatment group chosen may not be available on site. This is more likely to happen when the drug supply is constrained so that each site does not hold much trial drug stock.

When this happens one of two things can happen:

  • The randomisation fails
  • A different treatment group is chosen (one which is available on site)

In the latter case this is called a forced randomisation because the first choice treatment allocation was not available.

Sealed Envelope randomisation systems support both outcomes. You can see whether forced allocations are allowed for your trial on the specification page. If an allocation is forced it is recorded in the randomisation database.

What’s wrong with using actual sealed envelopes?

Read about the problems of using real envelopes.

My password doesn’t work

First check that you are trying to log in to the correct system. If you have an email with your login details from Sealed Envelope the URL of the system concerned will be present in the email. In particular be aware that test systems and live system are different and do not share the same login information.

Help, I’ve forgotten my password

Click on the Forgot your password? link on the log in page and enter the email address for your account. If you have an account you will be sent an email with instructions.

If you do not receive the email you either do not have an account or the email has gone to your Spam folder or been blocked by your email provider. Contact your email provider if you are sure you have an account. Contact the trial manager or other holder of an administrator account for the system concerned if you do not have an account.

I’ve been given the first administrator account for a randomisation system - what now?

When a randomisation system is set up, the first administrator account is created by Sealed Envelope and the login details are sent to that person’s email address. The administrator should log in and create the trial sites, unless the sites have been pre-coded by Sealed Envelope.

You do not need to add all your sites at once - you can come back later and add more sites as needed.

Next you should add some investigator accounts for each site so that randomisations can be performed by the sites. You do this through Access.

If your trial has a code list you should update the list to reflect the availability of treatment kits at each site. Randomisation cannot occur if there are no codes available at a site.

Finally check the specification page and randomisation form and report any discrepancies or errors to Sealed Envelope.

How do I test a randomisation system?

When your randomisation system is set up in Sealed Envelope’s evaluation environment, you should test it until you are satisfied it meets your requirements and is randomising as expected.

The test system is an identical copy of the live randomisation system except it will use a dummy randomisation list or code list (where appropriate). To test it, we recommend taking the following steps:

  1. Add at least one site
  2. Check the specification page and make sure it matches your requirements
  3. Check the randomisation form. If detailed inclusion and exclusion criteria are listed make sure these are consistent with the approved protocol. If there are validation checks on fields (e.g. an age range on the date of birth field) make sure these operate correctly.
  4. Perform some randomisations. Check that these are conforming to the randomisation protocol (blocks, minimisation etc). Ask Sealed Envelope for the dummy list or code list if necessary to help you check this. You may find the statistician for your trial can help, particularly if you have strata or are using minimisation.
  5. Create an investigator account using another email account you have. Log in as an investigator and try randomising. This will show you the simplified interface investigators see.
  6. Check you have received the email notifications for the randomisations. Make sure they contain the correct information.
  7. Try marking a randomisation as in error. Check the reports now exclude this record.
  8. If relevant, try unblinding a randomisation. Check the treatment group matches the dummy code list.
  9. Try downloading the randomisations as a CSV file. Check the data is consistent with the randomisations you have performed.

I didn’t receive a notification email

Notification emails are sent out from our servers when randomisations take place or for various other events. Sometimes a recipient doesn’t receive the email. There are several reasons for this which in descending order of likelihood are:

  • Your email server has classified the notification email as spam and either put it in a spam folder or deleted it. If you can’t see the message in your spam folder ask your mail administrator for advice. From our experience this seems to be a particular issue with some addresses.
  • The email has been delayed or queued by your email server due to capacity problems or other issues. Wait a few hours to see if you get the notification.
  • Your inbox was full or the email address held on our system is incorrect. Check that the email address for your account is correct and update it if necessary.
  • There was an issue with Sealed Envelope that prevented the notification being sent out. This is almost never the case but if you have investigated all other sources above we can check our mail logs to see if the notification was sent.

Unfortunately we cannot resend notifications so if your mail administrator cannot find it your only option is to ask someone else who received the same notification to forward a copy to you.

Page updated 6 Jul 2020