Discharge quarter (DQTR) is derived from either the month of the discharge date or the supplied discharge quarter. If both of those fields are invalid or missing, DQTR is set to zero. For these cases, a temporary discharge quarter = 3 was used for the DRG grouper and ICD-9-CM verification routines because these algorithms require a valid discharge quarter.

For HCUP inpatient data only, a temporary discharge quarter = 3 was used for the DRG grouper and ICD-9-CM verification routines when the discharge quarter was missing or invalid because these algorithms require a valid discharge quarter.

Special consideration was needed to handle the Massachusetts data in the 2006 NIS. Fourth quarter data from sampled hospitals in Massachusetts were unavailable for inclusion in the 2006 NIS. To account for the missing quarter of data, we sampled one fourth of the Massachusetts NIS discharges from the first three quarters and modified the records to represent the fourth quarter. To ensure a representative sample, we sorted the Massachusetts NIS discharges by hospital, discharge quarter, Clinical Classifications Software (CCS) diagnosis group for the principal diagnosis, gender, age, and a random number before selecting every fourth record. The following describes the adjustments made to the selected Massachusetts NIS records:

1. We relabeled the discharge quarter (DQTR) to four and saved the original discharge quarter in a new data element (DQTR_X).
2. We adjusted the admission month (AMONTH) by the number of months corresponding to the change in the discharge quarter.
3. We adjusted the total charges (TOTCHG and TOTCHG_X) using quarter-specific adjustment factors calculated as the mean total charges in the fourth quarter for all Northeastern NIS states (excluding Massachusetts) divided by the mean total charges in the first, second, or third quarter for all Northeastern NIS states (excluding Massachusetts).

We then adjusted the discharge weights for the Massachusetts records to appropriately account for the shifting of quarter one through three discharges to quarter four.

Beginning in 2004, Florida supplied discharge dates for all records. In 1997- 2003, Florida did not supply discharge date. DQTR was assigned from the discharge quarter provided by Florida.

Special consideration was needed to handle the Massachusetts data beginning in the 2006 NIS. Fourth quarter data from sampled hospitals in Massachusetts were unavailable for inclusion in the 2006 and 2007 NIS. To account for the missing quarter of data, we sampled one fourth of the Massachusetts NIS discharges from the first three quarters and modified the records to represent the fourth quarter. To ensure a representative sample, we sorted the Massachusetts NIS discharges by hospital, discharge quarter, Clinical Classifications Software (CCS) diagnosis group for the principal diagnosis, gender, age, and a random number before selecting every fourth record. The following describes the adjustments made to the selected Massachusetts NIS records:

1. We relabeled the discharge quarter (DQTR) to four and saved the original discharge quarter in a new data element (DQTR_X).
2. We adjusted the admission month (AMONTH) by the number of months corresponding to the change in the discharge quarter.
3. We adjusted the total charges (TOTCHG and TOTCHG_X) using quarter-specific adjustment factors calculated as the mean total charges in the fourth quarter for all Northeastern NIS states (excluding Massachusetts) divided by the mean total charges in the first, second, or third quarter for all Northeastern NIS states (excluding Massachusetts).