Documentation: diagrams for derive_var_trtemfl

[kaz462]

Please select a category the issue is focused on?

Function Documentation

Let us know where something needs a refresh or put your idea here!

Clean up the following diagram and place it in the derive_var_trtemfl documentation

flowchart TB

AEITOXGR["AEITOXGR"] -->|"missing AEITOXGR: \n AE not exist before TRTSDT"| arise1["New AE records \n within TEAE window"] --> |New|TEAE[Step 3: Flag records identified above as TEAE]

AEITOXGR -->|"non-missing AEITOXGR:  \n AE exist before TRTSDT"|Worsen1["Step 1: Identify worsened records \n (compare with the previous record) \nafter TRTSDT"]--> Worsen2["Step 2: Get the first worsen date \n after TRTSDT"]-->worsen3["All records after \n the 1st worsen date \n and within TEAE window"]--> |Worsen|TEAE

style TEAE fill:#b3ffb3,stroke:#4CAF50,stroke-width:2px
style Worsen1 fill:#b3ffb3,stroke:#4CAF50,stroke-width:2px
style Worsen2 fill:#b3ffb3,stroke:#4CAF50,stroke-width:2px

Loading

[github-actions]

This issue is stale because it has been open for 90 days with no activity.

[bms63]

@kaz462 do you think you could add these in. I think they are super handy. The Stale Bot got us!!

[bms63]

Hi @kaz462 - do you have time to look over this guidance from Phuse - do we align to it?

https://phuse.s3.eu-central-1.amazonaws.com/Deliverables/Safety+Analytics/WP-087+Recommended+Definition+of++Treatment-Emergent+Adverse+Events+in+Clinical+Trials+.pdf

[bms63]

Any takers for this @pharmaverse/admiral @pharmaverse/admiral_comm? The content is there - just making the diagram fit in our documentation would be very helpful.

I really like Lucid

[github-actions]

This issue is stale because it has been open for 90 days with no activity.

[manciniedoardo]

Possibly a candidate for a potential examples rework working group?

[bms63]

not sure if we need a working group - @rossfarrugia also wanted us to review this paper https://phuse.s3.eu-central-1.amazonaws.com/Deliverables/Safety+Analytics/[WP-087+Recommended+Definition+of++Treatment-Emergent+Adverse+Events+in+Clinical+Trials+.pdf](https://phuse.s3.eu-central-1.amazonaws.com/Deliverables/Safety+Analytics/WP-087+Recommended+Definition+of++Treatment-Emergent+Adverse+Events+in+Clinical+Trials+.pdf)

it is our oldest issue that we should resolve! maybe we can just hammer out in our core calls

[manciniedoardo]

I had a look at the whitepaper and coded up all the examples:

adae <- tribble(
    ~USUBJID, ~TRTSDTM,           ~TRTEDTM,           ~ASTDTM,            ~AENDTM,            ~AEITOXGR, ~AETOXGR,
    # Patient 1: Pre treatment AE
    "1",      "2021-01-01T00:01", "2021-12-31T23:59", "2020-12-20T00:00", "2020-12-21T00:00", "2",       "2",
    # Patient 2: On-treatment AE
    "2",      "2021-01-01T00:01", "2021-12-31T23:59", "2021-12-20T00:00", "2021-12-21T00:00", "2",       "2",
    # Patient 3: Pre-treatment AE, then on-treatment AE at same intensity
    "3",      "2021-01-01T00:01", "2021-12-31T23:59", "2020-12-20T00:00", "2020-12-21T00:00", "2",       "2",
    "3",      "2021-01-01T00:01", "2021-12-31T23:59", "2021-12-20T00:00", "2021-12-21T00:00", "2",       "2",
    # Patient 4: Pre-treatment AE, then on-treatment AE at worsened intensity
    "4",      "2021-01-01T00:01", "2021-12-31T23:59", "2020-12-20T00:00", "2020-12-21T00:00", "2",       "2",
    "4",      "2021-01-01T00:01", "2021-12-31T23:59", "2021-12-20T00:00", "2021-12-21T00:00", "2",       "3",
    # Patient 5: Pre-treatment AE, then on-treatment AE at improved intensity
    "5",      "2021-01-01T00:01", "2021-12-31T23:59", "2020-12-20T00:00", "2020-12-21T00:00", "2",       "2",
    "5",      "2021-01-01T00:01", "2021-12-31T23:59", "2021-12-20T00:00", "2021-12-21T00:00", "2",       "1",
    # Patient 6: AE starting pre-treatment and continuing on-treatment, then second AE at same intensity
    "6",      "2021-01-01T00:01", "2021-12-31T23:59", "2020-12-23T00:00", "2021-01-21T00:00", "2",       "2",
    "6",      "2021-01-01T00:01", "2021-12-31T23:59", "2021-12-20T00:00", "2021-12-21T00:00", "2",       "2",
    # Patient 7: AE starting pre-treatment and continuing on-treatment, then second AE at worsened intensity
    "7",      "2021-01-01T00:01", "2021-12-31T23:59", "2020-12-23T00:00", "2021-01-21T00:00", "2",       "2",
    "7",      "2021-01-01T00:01", "2021-12-31T23:59", "2021-12-20T00:00", "2021-12-21T00:00", "2",       "3",
    # Patient 8: AE starting pre-treatment and continuing on-treatment, then second AE at improved intensity
    "8",      "2021-01-01T00:01", "2021-12-31T23:59", "2020-12-23T00:00", "2021-01-21T00:00", "2",       "2",
    "8",      "2021-01-01T00:01", "2021-12-31T23:59", "2021-12-20T00:00", "2021-12-21T00:00", "2",       "1",
    # Patient 9: AE starting pre-treatment and continuing on-treatment, and no change in intensity
    "9",      "2021-01-01T00:01", "2021-12-31T23:59", "2020-12-23T00:00", "2021-01-21T00:00", "2",       "2",
    # Patient 10: AE starting pre-treatment and continuing on-treatment, and worsening intensity
    "10",     "2021-01-01T00:01", "2021-12-31T23:59", "2020-12-23T00:00", "2021-01-21T00:00", "2",       "4",
    # Patient 11: AE starting pre-treatment and continuing on-treatment, and improving intensity
    "11",     "2021-01-01T00:01", "2021-12-31T23:59", "2020-12-23T00:00", "2021-01-21T00:00", "2",       "1",
    # Patient 12: AE starting pre-treatment, worsening, then improving
    "12",     "2021-01-01T00:01", "2021-12-31T23:59", "2020-12-23T00:00", "2021-01-21T00:00", "3",       "2",
    # Patient 13: AE starting pre-treatment, improving, then worsening
    "13",     "2021-01-01T00:01", "2021-12-31T23:59", "2020-12-23T00:00", "2021-01-21T00:00", "1",       "2",
  ) %>%
  mutate(
    ASTDTM = ymd_hm(ASTDTM),
    AENDTM = ymd_hm(AENDTM),
    TRTSDTM = ymd_hm(TRTSDTM),
    TRTEDTM = ymd_hm(TRTEDTM),
  )

derive_var_trtemfl(
    adae,
    new_var = TRTEMFL,
    trt_end_date = TRTEDTM,
    end_window = 0,
    initial_intensity = AEITOXGR,
    intensity = AETOXGR,
    subject_keys = exprs(USUBJID)
  ) %>%
  select(USUBJID, TRTSDTM, TRTEDTM, ASTDTM, AENDTM, AEITOXGR, AETOXGR, TRTEMFL)

Output:

Excluding the cases where an AE is recorded on multiple lines, I think derive_var_trtemfl() matches what the whitepaper does. If an AE is encoded on multiple lines, then I think it's ambiguous. I can go ahead an make the diagrams as described at the start of the issue, if we are all happy.

Thanks @bundfussr for discussing this with me.