Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Utilization pre/post SQL queries #1831

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Utilization pre/post SQL queries #1831

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
155 changes: 155 additions & 0 deletions mimic-iv/concepts/utilization/post_utilization.sql
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,155 @@
-- -------------------------------------------------------------------------
-- Query: Post-Discharge Healthcare Utilization with ED Visit Classification
-- Description:
-- This query calculates the number of Emergency Department (ED) visits and
-- inpatient readmissions within a specified time window after each hospital
-- discharge. Additionally, it classifies ED visits based on whether they
-- resulted in an admission ("with admission") or not ("without admission").
-- This helps assess the patient's healthcare utilization following
-- discharge from the hospital, with a distinction between different types of ED visits.
--
-- Instructions:
-- - Set the desired time window in days by modifying the `time_window_days`
-- value in the `parameters` CTE (Common Table Expression).
-- - The default time window is 30 days.
--
-- Notes:
-- - The query uses the `admissions` table for inpatient readmissions and the
-- `transfers` table for ED visits.
-- - Only admissions and ED visits occurring after the current discharge are
-- counted.
-- - ED visits are classified into those that result in an admission and those
-- that do not.
-- -------------------------------------------------------------------------

WITH parameters AS (
SELECT 30 AS time_window_days -- Modify this value to set the time window in days
),

current_admissions AS (
SELECT
subject_id,
hadm_id,
admittime,
dischtime
FROM
`physionet-data.mimiciv_hosp.admissions`
),

-- Calculate post-discharge inpatient readmissions
post_inpatient_readmissions AS (
SELECT
ca.subject_id,
ca.hadm_id AS current_hadm_id,
COUNT(DISTINCT na.hadm_id) AS post_inpatient_count
FROM
current_admissions ca
CROSS JOIN
parameters p
LEFT JOIN
`physionet-data.mimiciv_hosp.admissions` na
ON ca.subject_id = na.subject_id
AND na.admittime > ca.dischtime
AND na.admittime <= TIMESTAMP_ADD(ca.dischtime, INTERVAL p.time_window_days DAY)
AND na.hadm_id != ca.hadm_id -- Exclude the current admission
GROUP BY
ca.subject_id,
ca.hadm_id
),

-- Calculate post-discharge ED visits that do NOT result in admission
post_ed_visits_wo_adm AS (
SELECT
ca.subject_id,
ca.hadm_id AS current_hadm_id,
COUNT(DISTINCT t.transfer_id) AS post_ed_wo_adm_count
FROM
current_admissions ca
CROSS JOIN
parameters p
LEFT JOIN
`physionet-data.mimiciv_hosp.transfers` t
ON ca.subject_id = t.subject_id
AND t.eventtype = 'ED'
AND t.intime > ca.dischtime
AND t.intime <= TIMESTAMP_ADD(ca.dischtime, INTERVAL p.time_window_days DAY)
AND (t.hadm_id IS NULL OR t.hadm_id != ca.hadm_id) -- Exclude EDs within the same admission
AND t.hadm_id IS NULL -- Ensure ED visit did not result in an admission
GROUP BY
ca.subject_id,
ca.hadm_id
),

-- Calculate post-discharge ED visits that result in admission
post_ed_visits_w_adm AS (
SELECT
ca.subject_id,
ca.hadm_id AS current_hadm_id,
COUNT(DISTINCT t.transfer_id) AS post_ed_w_adm_count
FROM
current_admissions ca
CROSS JOIN
parameters p
LEFT JOIN
`physionet-data.mimiciv_hosp.transfers` t
ON ca.subject_id = t.subject_id
AND t.eventtype = 'ED'
AND t.intime > ca.dischtime
AND t.intime <= TIMESTAMP_ADD(ca.dischtime, INTERVAL p.time_window_days DAY)
AND t.hadm_id != ca.hadm_id -- Exclude EDs within the same admission
AND t.hadm_id IS NOT NULL -- Ensure ED visit resulted in an admission
GROUP BY
ca.subject_id,
ca.hadm_id
),

-- Combine post-discharge ED visits
post_ed_visits AS (
SELECT
woa.subject_id,
woa.current_hadm_id,
COALESCE(woa.post_ed_wo_adm_count, 0) AS post_ed_wo_adm_count,
COALESCE(wa.post_ed_w_adm_count, 0) AS post_ed_w_adm_count
FROM
post_ed_visits_wo_adm woa
FULL OUTER JOIN
post_ed_visits_w_adm wa
ON woa.subject_id = wa.subject_id
AND woa.current_hadm_id = wa.current_hadm_id
),

-- Combine post-discharge inpatient readmissions and ED visits
post_discharge_utilization AS (
SELECT
ca.subject_id,
ca.hadm_id,
ca.admittime,
ca.dischtime,
COALESCE(pir.post_inpatient_count, 0) AS post_inpatient_count,
COALESCE(pev.post_ed_wo_adm_count, 0) AS post_ed_wo_adm_count,
COALESCE(pev.post_ed_w_adm_count, 0) AS post_ed_w_adm_count
FROM
current_admissions ca
LEFT JOIN
post_inpatient_readmissions pir
ON ca.subject_id = pir.subject_id
AND ca.hadm_id = pir.current_hadm_id
LEFT JOIN
post_ed_visits pev
ON ca.subject_id = pev.subject_id
AND ca.hadm_id = pev.current_hadm_id
)

SELECT
subject_id,
hadm_id,
admittime,
dischtime,
post_inpatient_count,
post_ed_wo_adm_count,
post_ed_w_adm_count
FROM
post_discharge_utilization
ORDER BY
subject_id,
hadm_id;
110 changes: 110 additions & 0 deletions mimic-iv/concepts/utilization/prior_utilization.sql
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,110 @@
-- -------------------------------------------------------------------------
-- Query: Prior Healthcare Utilization
-- Description:
-- This query calculates the number of prior Emergency Department (ED) visits
-- and inpatient admissions within a specified time window before each
-- current hospital admission. It helps assess the patient's healthcare
-- utilization history prior to the current admission.
--
-- Instructions:
-- - Set the desired time window in days by modifying the `time_window_days`
-- value in the `parameters` CTE (Common Table Expression).
-- - The default time window is 365 days (one year).
--
-- Notes:
-- - The query uses the `admissions` table for inpatient admissions and the
-- `transfers` table for ED visits.
-- - Only prior admissions and ED visits occurring before the current
-- admission are counted.
-- -------------------------------------------------------------------------

WITH parameters AS (
SELECT 365 AS time_window_days -- Modify this value to set the time window in days
),

current_admissions AS (
SELECT
subject_id,
hadm_id,
admittime,
dischtime
FROM
`physionet-data.mimiciv_hosp.admissions`
),

-- Calculate prior inpatient admissions
prior_inpatient_admissions AS (
SELECT
ca.subject_id,
ca.hadm_id AS current_hadm_id,
COUNT(DISTINCT pa.hadm_id) AS prior_inpatient_count
FROM
current_admissions ca
CROSS JOIN
parameters p
LEFT JOIN
`physionet-data.mimiciv_hosp.admissions` pa
ON ca.subject_id = pa.subject_id
AND pa.admittime < ca.admittime
AND pa.admittime >= TIMESTAMP_SUB(ca.admittime, INTERVAL p.time_window_days DAY)
AND pa.hadm_id != ca.hadm_id -- Exclude the current admission
GROUP BY
ca.subject_id,
ca.hadm_id
),

-- Calculate prior ED visits
prior_ed_visits AS (
SELECT
ca.subject_id,
ca.hadm_id AS current_hadm_id,
COUNT(DISTINCT t.transfer_id) AS prior_ed_count
FROM
current_admissions ca
CROSS JOIN
parameters p
LEFT JOIN
`physionet-data.mimiciv_hosp.transfers` t
ON ca.subject_id = t.subject_id
AND t.eventtype = 'ED'
AND t.intime < ca.admittime
AND t.intime >= TIMESTAMP_SUB(ca.admittime, INTERVAL p.time_window_days DAY)
AND (t.hadm_id IS NULL OR t.hadm_id != ca.hadm_id) -- Exclude EDs within the same admission
GROUP BY
ca.subject_id,
ca.hadm_id
),

-- Combine prior inpatient admissions and prior ED visits
prior_utilization AS (
SELECT
ca.subject_id,
ca.hadm_id,
ca.admittime,
ca.dischtime,
COALESCE(pia.prior_inpatient_count, 0) AS prior_inpatient_count,
COALESCE(pev.prior_ed_count, 0) AS prior_ed_count
FROM
current_admissions ca
LEFT JOIN
prior_inpatient_admissions pia
ON ca.subject_id = pia.subject_id
AND ca.hadm_id = pia.current_hadm_id
LEFT JOIN
prior_ed_visits pev
ON ca.subject_id = pev.subject_id
AND ca.hadm_id = pev.current_hadm_id
)

SELECT
subject_id,
hadm_id,
admittime,
dischtime,
prior_inpatient_count,
prior_ed_count
FROM
prior_utilization
ORDER BY
subject_id,
hadm_id;