Rewritten/Changed/Stopped differences in INPUTEVENTS_MV

[vincentmajor]

Can someone please clarify the roles of CANCELREASON and STATUSDESCRIPTION in the INPUTEVENTS_MV table.
From what I have seen CANCELREASON can be 0, 1 or 2 - what is the difference between 1 and 2?
And STATUSDESCRIPTION can be Rewrtten, Changed or Stopped - what is the difference and why does the AMOUNT field sometimes match the ORIGINALAMOUNT even though the order was rewritten/changed/stopped?

I've also noticed that CANCELREASON and STATUSSDESCRIPTION don't always match. That is rewritten orders can have CANCELREASON of 0, 1 or 2.

[alistairewj]

The documentation is limited on this topic so empirical observations are very welcome here.

I do know that STATUSDESCRIPTION provides some information on cancellations - that is, if an order was "Rewritten" then I don't think it was ever administered to the patient. You can see this for individual drugs: the STARTTIME of the current order will perfectly match the ENDTIME of the previous order after you remove all the rows with STATUSDESCRIPTION = 'Rewritten'.

I really don't know about the rest - do you have any notable observations?

[ilana-intensix]

Does anyone have any further insight on these ambiguities? I'm trying to understand which rows can be safely filtered out and whether a I can assume that a "Changed" order has run from the STARTTIME to the ENDTIME. Alistair, were you able to verify that a status of "Rewritten" indicates that a medicine wasn't administered?

[alistairewj]

For STATUSDESCRIPTION, we have the following possibilities:

• Changed - the dose was modified, as a result the current rate ends now, and a new rate begins.
• Paused - drug delivery has been paused
• Stopped - the drug is no longer being delivered
• Finished Running - the bag/container which the drug was in has been emptied, and so the drug is no longer being delivered
• Rewritten - the order was changed before the drug was delivered

As far as we know, all rewritten rows should only be used for audit purposes, i.e. these rows correspond to drug doses never being given to patients. We are considering removing this audit trail in the future as its use in research seems limited and it only acts to create a confusing table.

(N.B. strictly speaking, there are a two other order status, 'Reported' and 'Bolus', but they seem to be used extremely rarely if at all).

I haven't investigated CANCELREASON further so I can't comment on that!

[ilana-intensix]

Thanks very much!

In the cases where a medicine was Paused or Stopped, can we understand that the medicine was administered through the ENDTIME? That is, the time of pausing/stopping was the value in the ENDTIME column?

[alistairewj]

No problem. That is exactly what you should assume. Stopping/pausing a drug generates the ENDTIME. In the case of changing a drug, a new row will be created with the new drug dose which has a STARTTIME the same as the previous ENDTIME. It ends up working very nicely as you can select all administrations of the solution using the LINKORDERID and they will overlap perfectly. There's a toy example here: https://mimic.physionet.org/mimicdata/io/#metavision

[ilana-intensix]

Excellent, thanks again.