Compute hospital Length of Stay (LoS)
length_of_stay.Rd
length_of_stay
returns Length of Stay for hospital admissions in unit of
hours and days.
Usage
length_of_stay(
ipadmdad,
adm_dtvar = "admission_date_time",
dis_dtvar = "discharge_date_time"
)
Arguments
- ipadmdad
(
data.frame
ordata.table
)
Table with all relevant encounters of interest from DRM table "ipadmdad" (see GEMINI Data Repository Dictionary). Must contain three fields: encounter ID (genc_id
), and two date-time variables corresponding to admission and discharge (typicallyadmission_date_time
anddischarge_date_time
). Date-time variables must be in "yyyy-mm-dd hh:mm" format.- adm_dtvar
(
character
)
Character string defining the column name for admission date-time (typically "admission_date_time").- dis_dtvar
(
character
)
Character string defining the column name for discharge date-time (typically "discharge_date_time").
Value
data.table with the same number of rows as input "ipadmdad", with additional derived numeric fields labelled as "los_hrs_derived" and "los_days_derived".
Details
Length of Stay is defined as the duration of hospital in-patient stay,
thus, calculated as (discharge_date_time - admission_date_time
).
It is currently a major patient outcome in various research studies
and one of the physician performance indicators in MyPracticeReport.
Researchers might consider a few modifications to this metric.
One example will be adjusting for patient who receives palliative care.
If transfer to palliative unit within the hospital is trackable
(e.g. through room transfer data), then it might be good idea to re-define
discharge point to the palliative unit entrance. This will reduce bias in
metric since patient care under palliative order is very different.