HEALTHCARE COST & UTILIZATION PROJECT

User Support

Do Your own analysis
Explore Expert Research & Limited Datasets

Central Distributor SID: Description of Data Elements

 
LOS - Length of stay, cleaned
 
Documentation Sections:
General Notes
Uniform Values
State Specific Notes
General Notes
 

Length of stay (LOS) is calculated by subtracting the admission date (ADATE) from the discharge date (DDATE). Same-day stays are therefore coded as 0. Leave days are not subtracted. Before edit checks are performed, LOS and LOS_X have the same value. If LOS is set to inconsistent (.C), the value of LOS_X is retained.

LOS is not equal to the calculated value in the following cases:

  • LOS is set to the supplied length of stay if the length of stay cannot be calculated (ADATE and/or DDATE is missing or invalid). Note: If the supplied length of stay codes same-day stays as 1 or subtracts leave days, then the supplied length of stay is NOT used.
  • LOS is missing (.) if the length of stay cannot be calculated and the supplied length of stay is missing.
  • LOS is invalid (.A) if
    • it is greater than the maximum value allowed during HCUP processing (the maximum allowed in the 1988-1997 data is 32,767; the maximum allowed beginning in the 1998 data is 20 years)
    • - or -
    • the length of stay cannot be calculated and the supplied length of stay is nonnumeric.
  • An invalid calculated LOS is not replaced by the supplied length of stay.
  • If the data source does not supply either admission date (ADATE) and discharge date (DDATE), or length of stay, then beginning in the 1998 data LOS is not present on the HCUP files. In the 1988-1997 data, LOS is retained on the HCUP files and is set to unavailable from source (.B).
  • LOS is inconsistent (.C) if
    • LOS is negative (ELOS03 beginning in the 1998 data and ED011 in the 1988-1997 data),
    • Excessively long (ELOS04 beginning in the 1998 data and ED601 in the 1988-1997 data), or
    • Charges per day are unjustifiably low (ED911) or high (ED921).

Edit checks ED911 and ED921 are only performed on the 1988-1997 data. No charge per day edit checks are performed on the HCUP data beginning in the 1998 data.

 
Top
 
 
 
Uniform Values
 
VariableDescriptionValueValue Description
LOSLength of stay, cleaned0 - 365 (for HCUP inpatient data), 0-30 (for HCUP outpatient data)Days (Prior to data year 2017, LOS was limited to 0-3 days for outpatient data. In the 1988-1997 inpatient data, LOS can be greater than 365 days)
.Missing
.AInvalid
.BUnavailable from source (coded in 1988-1997 data only)
.CInconsistent: beginning with 1998 data, ELOS03, ELOS04; in 1988-1997 data, ED011, ED601, ED911n, ED921
 
Top
 
 
 
State Specific Notes

Alaska

The reported length of stay was not used when LOS could not be calculated because Alaska coded same-day stays with a value of 1. The appropriate edit check for consistency of reported and calculated length of stay could not be performed.



Arizona

Beginning in 1995, the source reports same-day stays as zero days so the supplied length of stay was used to assign LOS when length of stay could not be calculated from dates. Prior to 1995, the reported length of stay was not used when LOS could not be calculated because Arizona coded same-day stays with a value of 1 and subtracted days of absence from LOS.

Beginning with 2005, reported length of stay is no longer supplied.



Arkansas

Only the calculated length of stay could be used to assign LOS because Arkansas did not report length of stay.



Colorado

Beginning in 2007, the reported length of stay was used when LOS could not be calculated during the HCUP processing.

The reported length of stay was not used when LOS could not be calculated because Colorado:

  • coded same-day stays with the value 1 and
  • subtracted days of absence


Florida

Beginning in 2004, Florida provided admission date (ADATE) and discharge date (DDATE) the supplied length of stay was used when LOS and LOS_X could not be calculated from the supplied dates. In 2000-2003, the supplied length of stay was used to assign LOS and LOS_X because Florida did not provide the admission and discharge date necessary for calculating length of stay. The supplied length of stay was coded according to the HCUP standard that assigns a length of stay of zero (0) to same day stays.

In 1997-1999, the coding of LOS and LOS_X is inconsistent with the coding of length of stay in other states. Florida provided the reported length of stay but not the admission and discharge date necessary for calculating LOS. Florida codes same-day stays as LOS=1; the HCUP standard coding of same-day stays is LOS=0. Usually 2% of a states' discharges are same-day stays.

Prior to 1997, the reported length of stay was not used when LOS could not be calculated because Florida:

  • coded same-day stays with the value 1 and
  • subtracted days of absence.


Georgia

Beginning with the 2001 data, Georgia supplied admission and discharge dates for calculating LOS for almost all discharges. The reported length of stay was only used to assign LOS when dates were unavailable. Prior to 2001, the reported length of stay was not used when LOS could not be calculated because Georgia coded same-day stays with a value of 1.



Hawaii

Starting in 2017, Hawaii supplies reported length of stay.

Prior to 2017, only the calculated length of stay could be used to assign LOS because Hawaii did not supply reported length of stay.



Indiana

Only the calculated length of stay could be used to assign LOS because Indiana codes same day stays with a value of 1.



Iowa

Prior to 2007, the reported length of stay was not used when LOS could not be calculated because Iowa coded same-day stays with a value of 1.

Beginning in 2007, the reported length of stay was used when LOS could not be calculated during the HCUP processing.



Kansas

The reported length of stay was not used when LOS could not be calculated because Kansas coded same-day stays with a value of 1.



Kentucky

Beginning with the 2007 data, Kentucky supplied admission and discharge dates for calculating LOS for all discharges. The reported length of stay was only used to assign LOS when dates were unavailable. Prior to 2007, the reported length of stay was not used when LOS could not be calculated because Kentucky coded same-day stays with a value of 1.



Maine

The supplied length of stay was not used when length of stay could not be calculated because Maine coded same-day stays with a value of 1.



Massachusetts

The supplied length of stay was not used when LOS could not be calculated because Massachusetts:

  • coded same-day stays with the value 1 and
  • subtracted days of absence.


Michigan

Beginning in 2007, when LOS could not be calculated, the data source supplied LOS was used. Note that Michigan assigns the value 365 for any stays greater than 364 days.

Prior to 2001, LOS could not be calculated because Michigan did not report admission or discharge dates. Beginning with the 2001 data, Michigan provided complete dates and LOS could be calculated. In 2003, only the calculated length of stay could be used to assign LOS because Michigan codes same day stays with a value of 1.

Caution: Prior to 2001, if LOS = 365, then the stay may be longer than a year. Michigan uses the value 365 for stays that are greater than equal 364 days.



Missouri

Beginning with the 2007 data, Missouri supplied admission and discharge dates for calculating LOS for all discharges. The reported length of stay was only used to assign LOS when dates were unavailable. Prior to 2007, the reported length of stay was not used when LOS could not be calculated because Kentucky coded same-day stays with a value of 1. The appropriate edit check for consistency of reported and calculated length of stay could not be performed.



Nebraska

The reported length of stay was not used when LOS could not be calculated because Nebraska coded same-day stays with the value 1.



Nevada

Only the calculated length of stay could be used to assign LOS because Nevada codes same day stays with a value of 1.



New Jersey

The proportion of same-day stays (LOS = 0) in 2005 is 4.2% lower than in 2004.



New Mexico

The original New Mexico inpatient file contained outpatient records from HOSPID=35020. During the creation on the 2011 Mew Mexico SID these outpatient records, identified by a length of stay of 0 days, were excluded. This approach also eliminated any inpatient discharges with same day stays for this hospital. Any analysis using the 2011 New Mexico SID should be aware that this hospital is missing information on same day stays.



New York

The assignment of LOS and LOS_X varies by year in New York:

  • Beginning in 2000 data, the length of stay (LOS and LOS_X) in New York was calculated from the admission and discharge dates. Because New York masked the dates on AIDS/HIV* records, the calculated length of stay was missing. During HCUP processing, other information provided by New York was used to determine LOS and LOS_X when the calculated length of stay was missing. The length of stay provided by New York (which did not include leave days), total leave days, and a flag that indicates a same day stay were used to determine a length of stay that was consistent with the coding of length of stay on other HCUP records.


  • In the 1998-1999 data purchased from NTIS, the length of stay (LOS and LOS_X) in New York was calculated from the admission and discharge dates. Because New York masked the admission and discharge dates on AIDS/HIV* records, LOS and LOS_X was missing (.) on these discharges. An updated version of the 1998-1999 data is available through the HCUP Central Distributor with LOS and LOS_X coded on the New York AIDS/HIV* records. The updated version has LOS and LOS_X calculated using the method described for the 2000 data.

    In the 1998-1999 data purchased from HCUP Central Distributor, the length of stay (LOS and LOS_X) in New York was calculated using the method described for the 2000 data.


  • In the 1988-1997 HCUP data, LOS and LOS_X could not be calculated from dates because New York did not report full admission and discharge dates. During HCUP processing, the length of stay provided by New York was used to assign LOS and LOS_X. The length of stay provided by New York was adjusted during HCUP processing to be consistent with the coding of length of stay in other states.

*The HCUP data elements LOS and LOS_X are missing (.) for AIDS/HIV patients. Beginning with the 2005 New York identifies AIDS/HIV records by ICD-9-CM diagnosis code or DRG:

  • An admitting, principal or any secondary diagnosis of "042", "043", "044", "7958", "27910", "27919", "2793", "1363", "79571", "07951", "07952", "07953", "V017" or "V08".
  • A DRG of 488 through 490 or 700 through 716.

*For 1998- 2004 data, New York identifies AIDS/HIV records by ICD-9-CM diagnosis code or DRG:

  • An admitting, principal, or secondary diagnosis of "042" "043" "044" "7958" "27910", "27919", "2793", "1363", "79571", "07951", "07952", "07953" or "V08".
  • A DRG of 488 "HIV with Extensive Operating Room Procedure", 489 "HIV with Major related condition", or 490 "HIV with or without Other Related Condition".

Please note that the admitting diagnosis is not retained in the HCUP databases.



North Carolina

Beginning with the 2007 data, North Carolina supplied admission and discharge dates for calculating LOS for almost all discharges. The reported length of stay was only used to assign LOS when dates were unavailable. Prior to 2007, the reported length of stay was not used when LOS could not be calculated because North Carolina coded same-day stays with a value of 1.



Oregon

Prior to 1994, the reported length of stay was assigned to LOS if dates were not available. However, the coding of same day stay varies: some Oregon hospitals report discharges on the day of admission as one day stay (LOS=1), in addition to reporting same day stay as zero days (LOS=0).

Beginning in 1994, the reported length of stay was not used when LOS could not be calculated from dates because Oregon coded all same-day stays as one day (LOS=1).

Beginning in 2007, the reported length of stay was used if the calculated LOS was missing.



South Carolina

The reported length of stay was not used when LOS could not be calculated because South Carolina coded same-day stays with a value of 1.



South Dakota

Prior to 2007, only the calculated length of stay could be used to assign LOS_X because South Dakota codes same day stays with a value of 1.

Beginning in 2007, regardless if same day stays are coded with a value of 1, the source provided length of stay is used when LOS could not be calculated from dates.



Utah

Prior to 2004, the reported length of stay was not used when LOS could not be calculated because Utah coded same-day stays with a value of 1. Beginning in 2004, Utah coded same day stays with a value of 0.



Vermont

The reported length of stay was not used when LOS could not be calculated because Vermont coded same-day stays with the value 1.



Vermont

After the 2002 Vermont SID file was created, the data source reported that a small number of discharges have incorrect information. The following hospitals are affected.


DSHOSPID Discharges with Correct Information
6950 3 neonatal discharges with incorrect principal diagnosis
6952 1 discharge with incorrect date of birth
6953 1 discharge with incorrect principal diagnosis
2 neonatal discharges missing secondary diagnosis of V3000
6955 3 discharge with incorrect principal diagnosis
6956 1 discharge with incorrect gender
6957 2 discharges with incorrect principal diagnosis
3 discharges with incorrect principal procedure
4 discharges that incorrect patient disposition
8 discharges with a incorrect primary payer
6958 9 discharges with incorrect discharge date and length of stay
6959 1 discharge with incorrect principal diagnosis
6962 1 discharge with incorrect admission date and length of stay
1 discharge with incorrect discharge date and length of stay
2 discharges with incorrect principal procedure
6963 24 discharges missing principal procedure "4524", procedure date, and operating physician
Total 66 discharges


Washington

Beginning in data year 2010, Washington provided only length of stay to HCUP and not the admission and discharge dates. The provided length of stay was used for the HCUP data elements LOS and LOS_X. Washington codes same days stays with the value of 1, instead of 0. This is different than other HCUP databases.

Prior to data year 2012, the length of stay was calculated from admission and discharge dates.



West Virginia

Beginning in 2001, West Virginia provides LOS.

Prior to 2001, only the calculated length of stay was used to assign LOS because West Virginia did not provide the reported length of stay.



Wisconsin

For 1988-1994, the reported length of stay was not used when LOS could not be calculated because Wisconsin subtracted leave days and coded length of stay greater than 999 days as 999 days. Beginning with 1995, length of stay was not supplied. From 1995 - 2006, Wisconsin did not consistently supply length of stay - only the calculated length of stay was used to assign LOS and LOS_X. Beginning in 2007, we used the supplied length of stay to assign LOS and LOS_X when these fields could not be calculated from dates.


Top
 
 
 

Internet Citation: HCUP Central Distributor SID Description of Data Elements - All States. Healthcare Cost and Utilization Project (HCUP). August 2008. Agency for Healthcare Research and Quality, Rockville, MD. www.hcup-us.ahrq.gov/db/vars/siddistnote.jsp?var=los.
Are you having problems viewing or printing pages on this website?
If you have comments, suggestions, and/or questions, please contact hcup@ahrq.gov.
Privacy Notice, Viewers & Players
Last modified 8/12/08