Common use of Data File Information Clause in Contracts

Data File Information. The 2015 Hospital Inpatient Stays public use data set consists of one event-level data file. The file contains characteristics associated with the STAZ event and imputed expenditure data. The 2015 STAZ public use data set contains variable and frequency distributions for a total of 2,921 hospital inpatient stay records reported during the 2015 portion of Round 3 and Rounds 4 and 5 for Panel 19, as well as Rounds 1, 2, and the 2015 portion of Round 3 for Panel 20 of the MEPS Household Component. This file includes hospital inpatient stay records for all household survey members who resided in eligible responding households and for whom at least one hospital inpatient stay was reported. Hospital inpatient stay records known to have ended before January 1, 2015 or after December 31, 2015 are not included on this file. Some household members may have had multiple hospital inpatient stays reported and, thus, will be represented in multiple records on this file. Other household members may have had reported no hospital inpatient stays and, thus, will have no records on this file. Of the 2,921 hospital inpatient stay records, 2,817 are associated with persons having a positive person-level weight (PERWT15F). The persons represented on this file had to meet the following three criteria: 1. The hospital stay had to have been reported by a household survey respondent as an inpatient hospital stay (regardless of a stay’s length). Thus, the file contains some hospitalizations that were reported as not including an overnight stay. 2. The hospital stay had to have ended during 2015. Stays that began prior to 2015 but ended during 2015 are included on this data file. Stays that began in 2015 but ended during 2016 are excluded from this data file and will be included in a subsequent 2016 IP data file. Persons with no hospital inpatient stay events for 2015 are not included on this event-level IP file but are represented on the person-level 2015 Full Year Population Characteristics file. 3. The persons represented on this file also had to meet either 3a) or 3b): a) Be classified as a key in-scope person who responded for his or her entire period of 2015 eligibility (i.e., persons with a positive 2015 full-year person- level sampling weight (PERWT15F > 0)), or b) Be an eligible member of a family all of whose key in-scope members have a positive person-level weight (PERWT15F > 0). (Such a family consists of all persons with the same value for FAMIDYR.) That is, the person must have a positive full-year family-level weight (FAMWT15F > 0). Note that FAMIDYR and FAMWT15F are variables on the 2015 Full Year Consolidated Data File. One caveat that should be noted is that in the case of a newborn and the hospital inpatient stay associated with the newborn’s birth, a separate hospital inpatient stay record exists on the file only if the newborn was discharged after the mother. Thus, hospital stays associated with a normal birth are generally represented on the file as a single record (i.e., the mother’s hospital inpatient stay record, covering expenditure data for both the mother and baby). In situations where the newborn was discharged after the mother, the birth event will be represented as two records: one record for the mother and one record for the baby. For newborns re-admitted to the hospital during the reference year, each subsequent re-admission will have a separate record. Each inpatient record includes the following: start and end dates of the hospital inpatient stay; number of nights in the hospital; reason entered the hospital; condition(s) associated with the hospital inpatient stay; medicines prescribed at discharge; flat fee information; imputed sources of payment; total payment and total charge for both the facility and physician portions of the hospital inpatient stay expenditure; a full-year person-level weight; variance strata; and variance PSU. To append person-level information such as demographic or health insurance coverage to each event record, data from this file can be merged with 2015 MEPS HC person-level data (e.g. Full Year Consolidated or Full Year Population Characteristics file) using the person identifier, DUPERSID. Hospital inpatient stay events can also be linked to the MEPS 2015 Medical Conditions File and the MEPS 2015 Prescribed Medicines File. Please see Section 5.0 or the MEPS 2015 Appendix File, HC-178I, for details on how to merge MEPS data files.

Data File Information. The 2015 2009 Hospital Inpatient Stays public use data set consists of one event-level data file. The file contains characteristics associated with the STAZ event and imputed expenditure data. The 2015 2009 STAZ public use data set contains variable and frequency distributions for a total of 2,921 3,302 hospital inpatient stay records reported during the 2015 2009 portion of Round 3 and Rounds 4 and 5 for Panel 1913, as well as Rounds 1, 2, and the 2015 2009 portion of Round 3 for Panel 20 14 of the MEPS Household Component. This file includes hospital inpatient stay records for all household survey members who resided in eligible responding households and for whom at least one hospital inpatient stay was reported. Hospital inpatient stay records known to have ended before January 1, 2015 2009 or after December 31, 2015 2009 are not included on this file. Some household members may have had multiple hospital inpatient stays reported and, thus, will be represented in multiple records on this file. Other household members may have had reported no hospital inpatient stays and, thus, will have no records on this file. Of the 2,921 3,302 hospital inpatient stay records, 2,817 3,192 are associated with persons having a positive person-level weight (PERWT15FPERWT09F). The persons represented on this file had to meet the following three criteria: 1. The hospital stay had to have been reported by a household survey respondent as an inpatient hospital stay (regardless of a stay’s length). Thus, the file contains some hospitalizations that were reported as not including an overnight stay. 2. The hospital stay had to have ended during 20152009. Stays that began prior to 2015 2009 but ended during 2015 2009 are included on this data file. Stays that began in 2015 2009 but ended during 2016 2010 are excluded from this data file and will be included in a subsequent 2016 2010 IP data file. Persons with no hospital inpatient stay events for 2015 2009 are not included on this event-level IP file but are represented on the person-level 2015 2009 Full Year Population Characteristics file. 3. The persons represented on this file also had to meet either 3a) 3a or 3b):: a) Be classified as a key in-scope person who responded for his or her entire period of 2015 2009 eligibility (i.e., persons with a positive 2015 2009 full-year person- level sampling weight (PERWT15F PERWT09F > 0)), or b) Be an eligible member of a family all of whose key in-scope members have a positive person-level weight (PERWT15F PERWT09F > 0). (Such a family consists of all persons with the same value for FAMIDYR.) That is, the person must have a positive full-year family-level weight (FAMWT15F FAMWT09F > 0). Note that FAMIDYR and FAMWT15F FAMWT09F are variables on the 2015 Full Year Consolidated Data File2009 Population Characteristics file. One caveat that should be noted is that in the case of a newborn and the hospital inpatient stay associated with the newborn’s birth, a separate hospital inpatient stay record exists on the file only if the newborn was discharged after the mother. Thus, hospital stays associated with a normal birth are generally represented on the file as a single record (i.e., the mother’s hospital inpatient stay record, covering expenditure data for both the mother and baby). In situations where the newborn was discharged after the mother, the birth event will be represented as two records: one record for the mother and one record for the baby. For newborns re-admitted to the hospital during the reference year, each subsequent re-admission will have a separate record. Each inpatient record includes the following: start and end dates of the hospital inpatient stay; number of nights in the hospital; reason entered the hospital; condition(s) associated with the hospital inpatient stay; medicines prescribed at discharge; flat fee information; imputed sources of payment; total payment and total charge for both the facility and physician portions of the hospital inpatient stay expenditure; a full-year person-level weight; variance strata; and variance PSU. To append person-level information such as demographic or health insurance coverage to each event record, data from this file can be merged with 2015 2009 MEPS HC person-level data (e.g. Full Year Consolidated or Full Year Population Characteristics filefiles) using the person identifier, DUPERSID. Hospital inpatient stay events can also be linked to the MEPS 2015 2009 Medical Conditions File and the MEPS 2015 2009 Prescribed Medicines File. Please see Section 5.0 or the MEPS 2015 2009 Appendix File, HC-178IHC-126I, for details on how to merge MEPS data files.

Data File Information. The 2015 2014 Hospital Inpatient Stays public use data set consists of one event-level data file. The file contains characteristics associated with the STAZ event and imputed expenditure data. The 2015 2014 STAZ public use data set contains variable and frequency distributions for a total of 2,921 2,957 hospital inpatient stay records reported during the 2015 2014 portion of Round 3 and Rounds 4 and 5 for Panel 1918, as well as Rounds 1, 2, and the 2015 2014 portion of Round 3 for Panel 20 19 of the MEPS Household Component. This file includes hospital inpatient stay records for all household survey members who resided in eligible responding households and for whom at least one hospital inpatient stay was reported. Hospital inpatient stay records known to have ended before January 1, 2015 2014 or after December 31, 2015 2014 are not included on this file. Some household members may have had multiple hospital inpatient stays reported and, thus, will be represented in multiple records on this file. Other household members may have had reported no hospital inpatient stays and, thus, will have no records on this file. Of the 2,921 2,957 hospital inpatient stay records, 2,817 are associated with persons having a positive person-level weight (PERWT15FPERWT14F). The persons represented on this file had to meet the following three criteria: 1. The hospital stay had to have been reported by a household survey respondent as an inpatient hospital stay (regardless of a stay’s length). Thus, the file contains some hospitalizations that were reported as not including an overnight stay. 2. The hospital stay had to have ended during 20152014. Stays that began prior to 2015 2014 but ended during 2015 2014 are included on this data file. Stays that began in 2015 2014 but ended during 2016 2015 are excluded from this data file and will be included in a subsequent 2016 2015 IP data file. Persons with no hospital inpatient stay events for 2015 2014 are not included on this event-level IP file but are represented on the person-level 2015 2014 Full Year Population Characteristics file. 3. The persons represented on this file also had to meet either 3a) or 3b): a) Be classified as a key in-scope person who responded for his or her entire period of 2015 2014 eligibility (i.e., persons with a positive 2015 2014 full-year person- level sampling weight (PERWT15F PERWT14F > 0)), or b) Be an eligible member of a family all of whose key in-scope members have a positive person-level weight (PERWT15F PERWT14F > 0). (Such a family consists of all persons with the same value for FAMIDYR.) That is, the person must have a positive full-year family-level weight (FAMWT15F FAMWT14F > 0). Note that FAMIDYR and FAMWT15F FAMWT14F are variables on the 2015 2014 Full Year Consolidated Data File. One caveat that should be noted is that in the case of a newborn and the hospital inpatient stay associated with the newborn’s birth, a separate hospital inpatient stay record exists on the file only if the newborn was discharged after the mother. Thus, hospital stays associated with a normal birth are generally represented on the file as a single record (i.e., the mother’s hospital inpatient stay record, covering expenditure data for both the mother and baby). In situations where the newborn was discharged after the mother, the birth event will be represented as two records: one record for the mother and one record for the baby. For newborns re-admitted to the hospital during the reference year, each subsequent re-admission will have a separate record. Each inpatient record includes the following: start and end dates of the hospital inpatient stay; number of nights in the hospital; reason entered the hospital; condition(s) associated with the hospital inpatient stay; medicines prescribed at discharge; flat fee information; imputed sources of payment; total payment and total charge for both the facility and physician portions of the hospital inpatient stay expenditure; a full-year person-level weight; variance strata; and variance PSU. To append person-level information such as demographic or health insurance coverage to each event record, data from this file can be merged with 2015 2014 MEPS HC person-level data (e.g. Full Year Consolidated or Full Year Population Characteristics file) using the person identifier, DUPERSID. Hospital inpatient stay events can also be linked to the MEPS 2015 2014 Medical Conditions File and the MEPS 2015 2014 Prescribed Medicines File. Please see Section 5.0 or the MEPS 2015 2014 Appendix File, HC-178IHC-168I, for details on how to merge MEPS data files.

Data File Information. The 2015 2007 Hospital Inpatient Stays public use data set consists of one event-level data file. The file contains characteristics associated with the STAZ event and imputed expenditure data. The 2015 2007 STAZ public use data set contains variable and frequency distributions for a total of 2,921 3,040 hospital inpatient stay records reported during the 2015 2007 portion of Round 3 and Rounds 4 and 5 for Panel 1911, as well as Rounds 1, 2, and the 2015 2007 portion of Round 3 for Panel 20 12 of the MEPS Household Component. This file includes hospital inpatient stay records for all household survey members respondents who resided in eligible responding households and for whom reported at least one hospital inpatient stay was reportedstay. Hospital inpatient stay records known to have ended before January 1, 2015 2007 or after December 31, 2015 2007 are not included on this file. Some household members respondents may have had multiple hospital inpatient stays reported and, thus, will be represented in multiple records on this file. Other household members respondents may have had reported no hospital inpatient stays and, thus, will have no records on this file. Of the 2,921 3,040 hospital inpatient stay records, 2,817 2,929 are associated with persons having a positive person-level weight (PERWT15FPERWT07F). The persons represented on this file had to meet the following three criteria: 1. The hospital stay had to have been reported by a household survey respondent as an inpatient hospital stay (regardless of a stay’s length). Thus, the file contains some hospitalizations that were reported as not including an overnight stay. 2. The hospital stay had to have ended during 20152007. Stays that began prior to 2015 2007 but ended during 2015 2007 are included on this data file. Stays that began in 2015 2007 but ended during 2016 2008 are excluded from this data file and will be included in a subsequent 2016 2008 IP data file. Persons with no hospital inpatient stay events for 2015 2007 are not included on this event-level IP file but are represented on the person-level 2015 2007 Full Year Population Characteristics file. 3. The persons represented on this file also had to meet either 3a) 3a or 3b):: a) Be classified as a key in-scope person who responded for his or her entire period of 2015 2007 eligibility (i.e., persons with a positive 2015 2007 full-year person- level sampling weight (PERWT15F PERWT07F > 0)), or b) Be an eligible member of a family all of whose key in-scope members have a positive person-level weight (PERWT15F PERWT07F > 0). (Such a family consists of all persons with the same value for FAMIDYR.) That is, the person must have a positive full-year family-level weight (FAMWT15F FAMWT07F > 0). Note that FAMIDYR and FAMWT15F FAMWT07F are variables on the 2015 Full Year Consolidated Data File2007 Population Characteristics file. One caveat that should be noted is that in the case of a newborn and the hospital inpatient stay associated with the newborn’s birth, a separate hospital inpatient stay record exists on the file only if the newborn was discharged after the mother. Thus, hospital stays associated with a normal birth are generally represented on the file as a single record (i.e., the mother’s hospital inpatient stay record, covering expenditure data for both the mother and baby). In situations where the newborn was discharged after the mother, the birth event will be represented as two records: one record for the mother and one record for the baby. For newborns re-admitted to the hospital during the reference year, each subsequent re-admission will have a separate record. Each inpatient record includes the following: start and end dates of the hospital inpatient stay; number of nights in the hospital; reason entered the hospital; condition(s) associated with the hospital inpatient stay; medicines prescribed at discharge; flat fee information; imputed sources of payment; total payment and total charge for both the facility and physician portions of the hospital inpatient stay expenditure; a full-year person-level weight; variance strata; and variance PSU. To append person-level information such as demographic or health insurance coverage to each event record, data from this file can be merged with 2015 2007 MEPS HC person-level data (e.g. Full Year Consolidated or Full Year Population Characteristics filefiles) using the person identifier, DUPERSID. Hospital inpatient stay events can also be linked to the MEPS 2015 2007 Medical Conditions File and the MEPS 2015 2007 Prescribed Medicines File. Please see Section 5.0 or the MEPS 2015 2007 Appendix File, HC-178IHC-110I, for details on how to merge MEPS data files.

Data File Information. The 2015 2006 Hospital Inpatient Stays public use data set consists of one event-level data file. The file contains characteristics associated with the STAZ event and imputed expenditure data. The 2015 2006 STAZ public use data set contains variable and frequency distributions for a total of 2,921 3,315 hospital inpatient stay records reported during the 2015 2006 portion of Round 3 and Rounds 4 and 5 for Panel 1910, as well as Rounds 1, 2, and the 2015 2006 portion of Round 3 for Panel 20 11 of the MEPS Household Component. This file includes hospital inpatient stay records for all household survey members respondents who resided in eligible responding households and for whom reported at least one hospital inpatient stay was reportedstay. Hospital inpatient stay records known to have ended before January 1, 2015 2006 or after December 31, 2015 2006 are not included on this file. Some household members respondents may have had multiple hospital inpatient stays reported and, thus, will be represented in multiple records on this file. Other household members respondents may have had reported no hospital inpatient stays and, thus, will have no records on this file. Of the 2,921 3,315 hospital inpatient stay records, 2,817 3,202 are associated with persons having a positive person-level weight (PERWT15FPERWT06F). The persons represented on this file had to meet the following three criteria: 1. ) The hospital stay had to have been reported by a household survey respondent as an inpatient hospital stay (regardless of a stay’s length). Thus, the file contains some hospitalizations that were reported as not including an overnight stay. 2. ) The hospital stay had to have ended during 20152006. Stays that began prior to 2015 2006 but ended during 2015 2006 are included on this data file. Stays that began in 2015 2006 but ended during 2016 2007 are excluded from this data file and will be included in a subsequent 2016 2007 IP data file. Persons with no hospital inpatient stay events for 2015 2006 are not included on this event-level IP file but are represented on the person-person- level 2015 2006 Full Year Population Characteristics file. 3. ) The persons represented on this file also had to meet either 3a) 3a or 3b):: a) Be classified as a key in-scope person who responded for his or her entire period of 2015 2006 eligibility (i.e., persons with a positive 2015 2006 full-year person- person-level sampling weight (PERWT15F PERWT06F > 0)), or b) Be an eligible member of a family all of whose key in-scope members have a positive person-level weight (PERWT15F PERWT06F > 0). (Such a family consists of all persons with the same value for FAMIDYR.) That is, the person must have a positive full-year family-level weight (FAMWT15F FAMWT06F > 0). Note that FAMIDYR and FAMWT15F FAMWT06F are variables on the 2015 Full Year Consolidated Data File2006 Population Characteristics file. One caveat that should be noted is that in the case of a newborn and the hospital inpatient stay associated with the newborn’s birth, a separate hospital inpatient stay record exists on the file only if the newborn was discharged after the mother. Thus, hospital stays associated with a normal birth are generally represented on the file as a single record (i.e., the mother’s hospital inpatient stay record, covering expenditure data for both the mother and baby). In situations where the newborn was discharged after the mother, the birth event will be represented as two records: one record for the mother and one record for the baby. For newborns re-admitted to the hospital during the reference year, each subsequent re-admission will have a separate record. Each inpatient record includes the following: start and end dates of the hospital inpatient stay; number of nights in the hospital; reason entered the hospital; condition(s) associated with the hospital inpatient stay; medicines prescribed at discharge; flat fee information; imputed sources of payment; total payment and total charge for both the facility and physician portions of the hospital inpatient stay expenditure; a full-year person-level weight; variance strata; and variance PSU. To append person-level information such as demographic or health insurance coverage to each event record, data Data from this file can be merged with 2015 the MEPS HC person-level data (e.g. Full Year Consolidated or 2006 Full Year Population Characteristics file) File using the person identifier, DUPERSID, to append person-level information, such as demographic or health insurance characteristics, to each record. Hospital inpatient stay events can also be linked to the MEPS 2015 2006 Medical Conditions File and the MEPS 2015 2006 Prescribed Medicines File. Please see Section 5.0 or the MEPS 2015 2006 Appendix File, HC-178IHC-102I, for details on how to merge MEPS data files.

Data File Information. The 2015 2012 Hospital Inpatient Stays public use data set consists of one event-level data file. The file contains characteristics associated with the STAZ event and imputed expenditure data. The 2015 2012 STAZ public use data set contains variable and frequency distributions for a total of 2,921 3,045 hospital inpatient stay records reported during the 2015 2012 portion of Round 3 and Rounds 4 and 5 for Panel 1916, as well as Rounds 1, 2, and the 2015 2012 portion of Round 3 for Panel 20 17 of the MEPS Household Component. This file includes hospital inpatient stay records for all household survey members who resided in eligible responding households and for whom at least one hospital inpatient stay was reported. Hospital inpatient stay records known to have ended before January 1, 2015 2012 or after December 31, 2015 2012 are not included on this file. Some household members may have had multiple hospital inpatient stays reported and, thus, will be represented in multiple records on this file. Other household members may have had reported no hospital inpatient stays and, thus, will have no records on this file. Of the 2,921 3,045 hospital inpatient stay records, 2,817 2,940 are associated with persons having a positive person-level weight (PERWT15FPERWT12F). The persons represented on this file had to meet the following three criteria: 1. The hospital stay had to have been reported by a household survey respondent as an inpatient hospital stay (regardless of a stay’s length). Thus, the file contains some hospitalizations that were reported as not including an overnight stay. 2. The hospital stay had to have ended during 20152012. Stays that began prior to 2015 2012 but ended during 2015 2012 are included on this data file. Stays that began in 2015 2012 but ended during 2016 2013 are excluded from this data file and will be included in a subsequent 2016 2013 IP data file. Persons with no hospital inpatient stay events for 2015 2012 are not included on this event-level IP file but are represented on the person-level 2015 2012 Full Year Population Characteristics file. 3. The persons represented on this file also had to meet either 3a) or 3b): a) Be classified as a key in-scope person who responded for his or her entire period of 2015 2012 eligibility (i.e., persons with a positive 2015 2012 full-year person- level sampling weight (PERWT15F PERWT12F > 0)), or b) Be an eligible member of a family all of whose key in-scope members have a positive person-level weight (PERWT15F PERWT12F > 0). (Such a family consists of all persons with the same value for FAMIDYR.) That is, the person must have a positive full-year family-level weight (FAMWT15F WTFMPS12 > 0). Note that FAMIDYR and FAMWT15F WTFMPS12 are variables on the 2015 2012 Full Year Consolidated Data File. One caveat that should be noted is that in the case of a newborn and the hospital inpatient stay associated with the newborn’s birth, a separate hospital inpatient stay record exists on the file only if the newborn was discharged after the mother. Thus, hospital stays associated with a normal birth are generally represented on the file as a single record (i.e., the mother’s hospital inpatient stay record, covering expenditure data for both the mother and baby). In situations where the newborn was discharged after the mother, the birth event will be represented as two records: one record for the mother and one record for the baby. For newborns re-admitted to the hospital during the reference year, each subsequent re-admission will have a separate record. Each inpatient record includes the following: start and end dates of the hospital inpatient stay; number of nights in the hospital; reason entered the hospital; condition(s) associated with the hospital inpatient stay; medicines prescribed at discharge; flat fee information; imputed sources of payment; total payment and total charge for both the facility and physician portions of the hospital inpatient stay expenditure; a full-year person-level weight; variance strata; and variance PSU. To append person-level information such as demographic or health insurance coverage to each event record, data from this file can be merged with 2015 2012 MEPS HC person-level data (e.g. Full Year Consolidated or Full Year Population Characteristics file) using the person identifier, DUPERSID. Hospital inpatient stay events can also be linked to the MEPS 2015 2012 Medical Conditions File and the MEPS 2015 2012 Prescribed Medicines File. Please see Section 5.0 or the MEPS 2015 2012 Appendix File, HC-178IHC-152I, for details on how to merge MEPS data files.

Data File Information. The 2015 Hospital Inpatient Stays This public use data set consists of one 2 event-level data filefiles. The file File 1 contains characteristics associated with the STAZ hospital inpatient stay event and imputed expenditure data. The 2015 STAZ File 2 contains pre-imputed and unimputed expenditure data from both the Household and Medical Provider Components for all hospital inpatient stay events on File 1. Please see the Attachment 1 for definitions of imputed, un- imputed and pre-imputed expenditure variables. Both files 1 and 2 of this public use data set contains variable and frequency distributions for a total of 2,921 2,207 hospital inpatient stay records reported during records. Of the 2015 portion of Round 3 and Rounds 4 and 5 for Panel 192,207 hospital inpatient stay records, as well as Rounds 1, 2, and the 2015 portion of Round 3 for Panel 20 of the MEPS Household Component2,138 are associated with persons having a positive person-level weight (WTDPER96). This file includes These files include hospital inpatient stay records for all household survey members respondents who resided in eligible responding households and for whom reported at least one hospital inpatient stay. Each record represents one household-reported hospital inpatient stay was reportedthat occurred during calendar year 1996. Hospital inpatient stay records stays known to have ended before January 1, 2015 or occurred after December 31, 2015 1996 are not included on this file. Some household members respondents may have had multiple hospital inpatient stays reported and, thus, and thus will be represented in multiple records on this file. Other household members respondents may have had reported no hospital inpatient stays and, thus, and thus will have no records on this file. Of These data were collected during rounds 1, 2, and 3 of the 2,921 hospital inpatient stay records, 2,817 are associated with persons having a positive person-level weight (PERWT15F)MEPS HC. The persons represented on this file had to meet the following three criteria: 1. ) The hospital stay had to have been reported by a household survey respondent as an inpatient hospital stay (regardless of a stay’s length). Thus, the file contains some hospitalizations that were reported as not including an overnight stay. 2. ) The hospital stay had to have ended during 20151996. Stays that began prior to 2015 1996, but ended during 2015 1996, are included on this data file. Stays that began in 2015 1996, but ended during 2016 1997, are excluded from this data file and will be included in represented on a subsequent 2016 IP 1997 data file. Persons Please note that persons with no hospital inpatient stay events stays use for 2015 1996 are not included on this event-level IP file (but are represented on the MEPS person-level 2015 Full Year Population Characteristics filefiles). 3. ) The persons represented on this file also had to also meet either 3a) 3a or 3b):: a) Be classified as a key in-scope person who responded for his or her entire period of 2015 1996 eligibility (i.e., persons with a positive 2015 1996 full-year person- person-level sampling weight (PERWT15F WTDPER96 > 0)), or b) Be classified as either an eligible non-key person or an eligible out-of-scope person who responded for his or her entire period of 1996 eligibility, and belonged to a family (i.e., all persons within a household (DUID) with the same value of FAMID) in which all eligible family members responded for their entire period of 1996 eligibility, and at least one family member has a positive 1996 full-year person weight (i.e., eligible non- key or eligible out-of-scope persons who are members of a family all of whose key in-scope members have a positive person-level weight (PERWT15F > 0). (Such a family consists of all persons with the same value for FAMIDYR.) That is, the person must have a positive 1996 full-year family-level weight (FAMWT15F > WTFAM96 >0)). Note that FAMIDYR Please refer to Attachment 1 for definitions of key, non-key, inscope and FAMWT15F are variables on the 2015 Full Year Consolidated Data Fileeligible. One caveat that should be noted is that in the case of a newborn newborn, and the inpatient hospital inpatient stay associated with the newborn’s birth, a separate hospital inpatient stay record exists on the file only if the newborn was discharged after the mother. Thus, hospital stays associated with a normal birth are generally represented on the file as a single record (i.e., the mother’s hospital inpatient stay record, covering expenditure data for both the mother and baby). In situations where the newborn was discharged after the mother, the birth event will be represented as two records: records (one record for the mother and one record for the baby). For newborns re-admitted to the hospital during the reference year, each subsequent re-admission will have a separate record. Each hospital inpatient stay record on File 1 includes the following: start and end dates of the hospital inpatient stay; number of nights in the hospital; reason entered the hospital; main surgical procedure; condition(s) associated with the hospital inpatient stay; medicines prescribed at dischargedischarge ; flat fee information; , imputed sources of payment; , total payment and total charge for both the facility and physician portions components of the hospital inpatient stay expenditure; and a full-year person-level weight; variance strata; . File 2 of this public use data set is intended for analysts who want to perform their own imputations to handle missing data. . This file contains one set of un-imputed expenditure information from the Medical Provider Component as well as one set of pre-imputed expenditure information from the Household Component. Both sets of expenditure data have been subject to minimal logical editing that accounted for outliers, copayments or charges reported as total payments, and variance PSUreimbursed amounts that were reported as out of pocket payments. To In addition, edits were implemented to correct for misclassifications between Medicare and Medicaid and between Medicare HMO’s and private HMO’s as payment sources. However, missing data was not imputed. Data from these files can be merged with previously released 1996 MEPS HC person level data using the unique person identifier, DUPERSID, to append person-level information person characteristics such as demographic or health insurance coverage characteristics to each event record, data from this file can be merged with 2015 MEPS HC person-level data (e.g. Full Year Consolidated or Full Year Population Characteristics file) using the person identifier, DUPERSID. Hospital inpatient stay events can also be linked to the MEPS 2015 1996 Medical Conditions File (HC-006) and the MEPS 2015 1996 Prescribed Medicines FileFile (HC- 10A). Please see Section 5.0 or the MEPS 2015 The Appendix File, HC-178I, for File contains details on how to merge link MEPS data files.

Data File Information. The 2015 Hospital Inpatient Stays 1998 STAZ public use data set consists of one 2 event-level data filefiles. The file File 1 contains characteristics data associated with the STAZ event characteristics of hospital inpatient stays and imputed expenditure datadata from both the Household and the Medical Provider Component. The 2015 File 2 contains select survey administration and ID variables as well as pre-imputed and unimputed expenditure data from both the Household and Medical Provider Components. Please see Attachment 1 for definitions of imputed, pre-imputed, and unimputed expenditure variables. Both Files 1 and 2 of the STAZ public use data set contains contain variable and frequency distributions for a total of 2,921 2,589 hospital inpatient stay records reported during the 2015 1998 portion of Round 3 and Rounds 4 and 5 for Panel 192, as well as Rounds 1, 2, 2 and the 2015 1998 portion of Round 3 for Panel 20 of 3 (i.e., the Rounds for the MEPS Household Componentpanels covering calendar year 1998). This file includes These files include hospital inpatient stay records for all household survey members respondents who resided in eligible responding households and for whom reported at least one hospital inpatient stay was reportedstay. Hospital inpatient stay records known to have ended before January 1, 2015 or occurred after December 31, 2015 1998, are not included on this file. Some household members respondents may have had multiple hospital inpatient stays reported and, thus, will be represented in multiple records on this file. Other household members respondents may have had reported no hospital inpatient stays and, thus, will have no records on this file. Of the 2,921 2,589 hospital inpatient stay records, 2,817 2,487 are associated with persons having a positive person-level weight (PERWT15FWTDPER98). The persons represented on this file had to meet the following three criteria: 1. ) The hospital stay had to have been reported by a household survey respondent as an inpatient hospital stay (regardless of a stay’s length). Thus, the file contains some hospitalizations that were reported as not including an overnight stay. 2. ) The hospital stay had to have ended during 20151998. Stays that began prior to 2015 1998, but ended during 2015 1998, are included on this data file. Stays that began in 2015 1998, but ended during 2016 1999, are excluded from this data file and will be included in represented on a subsequent 2016 IP 1999 data file. Persons Please note that persons with no hospital inpatient stay events stays use for 2015 1998 are not included on this event-level IP file (but are represented on the MEPS person-level 2015 Full Year Population Characteristics filefiles). 3. ) The persons represented on this file also had to meet either 3a) 3a or 3b):: a) Be classified as a key in-scope person who responded for his or her entire period of 2015 1998 eligibility ([i.e., persons with a positive 2015 1998 full-year person- person-level sampling weight (PERWT15F WTDPER98 > 0))], or b) Be classified as either an eligible non-key person or an eligible out-of-scope person who responded for his or her entire period of 1998 eligibility, and belonged to a family (i.e., all persons within a household (DUID) with the same value of FAMID) in which all eligible family members responded for their entire period of 1998 eligibility, and at least one family member has a positive 1998 full-year person weight [i.e., eligible non-key or eligible out-of-scope persons who are members of a family all of whose key in-scope members have a positive person-level weight (PERWT15F > 0). (Such a family consists of all persons with the same value for FAMIDYR.) That is, the person must have a positive 1998 full-year family-level weight (FAMWT15F > WTFAM98 >0)]. Note that FAMIDYR Please refer to Attachment 1 for definitions of keyness, in-scope and FAMWT15F are variables on the 2015 Full Year Consolidated Data Fileeligibility. One caveat that should be noted is that that, in the case of a newborn and the hospital inpatient stay associated with the newborn’s birth, a separate hospital inpatient stay record exists on the file only if the newborn was discharged after the mother. Thus, hospital stays associated with a normal birth are generally represented on the file as a single record (i.e., the mother’s hospital inpatient stay record, covering expenditure data for both the mother and baby). In situations where the newborn was discharged after the mother, the birth event will be represented as two records: one record for the mother and one record for the baby. For newborns re-admitted to the hospital during the reference year, each subsequent re-admission will have a separate record. Each inpatient record includes the following: start and end dates of the hospital inpatient stay; number of nights in the hospital; reason entered the hospital; condition(s) associated with the hospital inpatient stay; medicines prescribed at discharge; flat fee information; imputed sources of payment; total payment and total charge for both the facility and physician portions of the hospital inpatient stay expenditure; a full-year person-level weight; variance strata; and variance PSU. To append person-level information such as demographic or health insurance coverage to each event record, data from this file can be merged with 2015 MEPS HC person-level data (e.g. Full Year Consolidated or Full Year Population Characteristics file) using the person identifier, DUPERSID. Hospital inpatient stay events can also be linked to the MEPS 2015 Medical Conditions File and the MEPS 2015 Prescribed Medicines File. Please see Section 5.0 or the MEPS 2015 Appendix File, HC-178I, for details on how to merge MEPS data files.

Data File Information. The 2015 2019 Hospital Inpatient Stays public use data set consists of one event-level data file. The file contains characteristics associated with the STAZ event and imputed expenditure data. The 2015 2019 STAZ public use data set contains variable and frequency distributions for a total of 2,921 2,884 hospital inpatient stay records reported during the 2015 2019 portion of Round 3 and Rounds 4 and 5 for Panel 1923, as well as Rounds 1, 2, and the 2015 2019 portion of Round 3 for Panel 20 24 of the MEPS Household Component. This file includes hospital inpatient stay records for all household survey members who resided in eligible responding households and for whom at least one hospital inpatient stay was reported. Hospital inpatient stay records known to have ended before January 1, 2015 2019 or after December 31, 2015 2019 are not included on this file. Some household members may have had multiple hospital inpatient stays reported and, thus, will be represented in multiple records on this file. Other household members may have had reported no hospital inpatient stays and, thus, will have no records on this file. Of the 2,921 2,884 hospital inpatient stay records, 2,817 2,841 are associated with persons having a positive person-level weight (PERWT15FPERWT19F). The persons represented on this file had to meet the following three criteria: 1. The hospital stay had to have been reported by a household survey respondent as an inpatient hospital stay (regardless of a stay’s length). Thus, the file contains some hospitalizations that were reported as not including an overnight stay. 2. The hospital stay had to have ended during 20152019. Stays that began prior to 2015 2019 but ended during 2015 2019 are included on this data file. Stays that began in 2015 2019 but ended during 2016 2020 are excluded from this data file and will be included in a subsequent 2016 2020 IP data file. Persons with no hospital inpatient stay events for 2015 2019 are not included on this event-level IP file but are represented on the person-level 2015 2019 Full Year Population Characteristics file. 3. The persons represented on this file also had to meet either 3a) or 3b): a) Be classified as a key in-scope person who responded for his or her entire period of 2015 2019 eligibility (i.e., persons with a positive 2015 2019 full-year person- level sampling weight (PERWT15F PERWT19F > 0)), or b) Be an eligible member of a family all of whose key in-scope members have a positive person-level weight (PERWT15F PERWT19F > 0). (Such a family consists of all persons with the same value for FAMIDYR.) That is, the person must have a positive full-year family-level weight (FAMWT15F FAMWT19F > 0). Note that FAMIDYR and FAMWT15F FAMWT19F are variables on the 2015 2019 Full Year Consolidated Data File. One caveat that should be noted is that in the case of a newborn and the hospital inpatient stay associated with the newborn’s birth, a separate hospital inpatient stay record exists on the file only if the newborn was discharged after the mother. Thus, hospital stays associated with a normal birth are generally represented on the file as a single record (i.e., the mother’s hospital inpatient stay record, covering expenditure data for both the mother and baby). In situations where the newborn was discharged after the mother, the birth event will be represented as two records: one record for the mother and one record for the baby. For newborns re-admitted to the hospital during the reference year, each subsequent re-admission will have a separate record. Each inpatient record includes the following: start and end dates of the hospital inpatient stay; number of nights in the hospital; reason entered the hospital; condition(s) associated with the hospital inpatient stay; medicines prescribed at discharge; flat fee information; imputed sources of payment; total payment and total charge for both the facility and physician portions of the hospital inpatient stay expenditure; a full-year person-level weight; variance strata; and variance PSU. To append person-level information such as demographic or health insurance coverage to each event record, data from this file can be merged with 2015 2019 MEPS HC person-level data (e.g. Full Year Consolidated or Full Year Population Characteristics file) using the person identifier, DUPERSID. Hospital inpatient stay events can also be linked to the MEPS 2015 2019 Medical Conditions File and the MEPS 2015 2019 Prescribed Medicines File. Please see Section 5.0 or the MEPS 2015 2019 Appendix File, HC-178IHC-213I, for details on how to merge MEPS data files.

Data File Information. ‌ The 2015 2021 Hospital Inpatient Stays public use data set consists of one event-level data file. The file contains characteristics associated with the STAZ event and imputed expenditure data. The 2015 2021 STAZ public use data set contains variable and frequency distributions for a total of 2,921 2,557 hospital inpatient stay records reported during the 2015 2021 portion of Round 3 7, and all of Rounds 8 and 9 for Panel 23; Round 6 and the 2021 portions of Rounds 5 and 7 for Panel 24; the 2021 portion of Round 3, and all of Rounds 4 and 5 for Panel 19, 25; as well as Rounds 1, 2, and the 2015 2021 portion of Round 3 for Panel 20 26 of the MEPS Household Component. This file includes hospital inpatient stay records for all household survey members who resided in eligible responding households and for whom at least one hospital inpatient stay was reported. Hospital inpatient stay records known to have ended before January 1, 2015 2021 or after December 31, 2015 2021 are not included on this file. Some household members may have had multiple hospital inpatient stays reported and, thus, will be represented in multiple records on this file. Other household members may have had reported no hospital inpatient stays and, thus, will have no records on this file. Of the 2,921 2,557 hospital inpatient stay records, 2,817 2,502 are associated with persons having a positive person-level weight (PERWT15FPERWT21F). The persons represented on this file had to meet the following three criteria: 1. The hospital stay had to have been reported by a household survey respondent as an inpatient hospital stay (regardless of a stay’s length). Thus, the file contains some hospitalizations that were reported as not including an overnight stay. 2. The hospital stay had to have ended during 20152021. Stays that began prior to 2015 2021 but ended during 2015 2021 are included on this data file. Stays that began in 2015 2021 but ended during 2016 2022 are excluded from this data file and will be included in a subsequent 2016 2022 IP data file. Persons with no hospital inpatient stay events for 2015 2021 are not included on this event-level IP file but are represented on the person-level 2015 2021 Full Year Population Characteristics file. 3. The persons represented on this file also had to meet either 3a) or 3b): a) Be classified as a key in-scope person who responded for his or her entire period of 2015 2021 eligibility (i.e., persons with a positive 2015 2021 full-year person- level sampling weight (PERWT15F PERWT21F > 0)), or b) Be an eligible member of a family all of whose key in-scope members have a positive person-level weight (PERWT15F PERWT21F > 0). (Such a family consists of all persons with the same value for FAMIDYR.) That is, the person must have a positive full-year family-level weight (FAMWT15F FAMWT21F > 0). Note that FAMIDYR and FAMWT15F FAMWT21F are variables on the 2015 2021 Full Year Consolidated Data File. One caveat that should be noted is that in the case of a newborn and the hospital inpatient stay associated with the newborn’s birth, a separate hospital inpatient stay record exists on the file only if the newborn was discharged after the mother. Thus, hospital stays associated with a normal birth are generally represented on the file as a single record (i.e., the mother’s hospital inpatient stay record, covering expenditure data for both the mother and baby). In situations where the newborn was discharged after the mother, the birth event will be represented as two records: one record for the mother and one record for the baby. For newborns re-admitted to the hospital during the reference year, each subsequent re-admission will have a separate record. Each inpatient record includes the following: start and end dates of the hospital inpatient stay; number of nights in the hospital; reason entered the hospital; condition(s) associated with the hospital inpatient stay; medicines prescribed at discharge; flat fee information; imputed sources of payment; total payment and total charge for both the facility and physician portions of the hospital inpatient stay expenditure; a full-year person-level weight; variance strata; and variance PSU. To append person-level information such as demographic or health insurance coverage to each event record, data from this file can be merged with 2015 2021 MEPS HC person-level data (e.g. Full Year Consolidated or Full Year Population Characteristics file) using the person identifier, DUPERSID. Hospital inpatient stay events can also be linked to the MEPS 2015 2021 Medical Conditions File and the MEPS 2015 Prescribed Medicines File. Please see Section 5.0 or the MEPS 2015 2021 Appendix File, HC-178IHC-229I, for details on how to merge MEPS data files.

Data File Information. The 2015 2013 Hospital Inpatient Stays public use data set consists of one event-level data file. The file contains characteristics associated with the STAZ event and imputed expenditure data. The 2015 2013 STAZ public use data set contains variable and frequency distributions for a total of 2,921 2,862 hospital inpatient stay records reported during the 2015 2013 portion of Round 3 and Rounds 4 and 5 for Panel 1917, as well as Rounds 1, 2, and the 2015 2013 portion of Round 3 for Panel 20 18 of the MEPS Household Component. This file includes hospital inpatient stay records for all household survey members who resided in eligible responding households and for whom at least one hospital inpatient stay was reported. Hospital inpatient stay records known to have ended before January 1, 2015 2013 or after December 31, 2015 2013 are not included on this file. Some household members may have had multiple hospital inpatient stays reported and, thus, will be represented in multiple records on this file. Other household members may have had reported no hospital inpatient stays and, thus, will have no records on this file. Of the 2,921 2,862 hospital inpatient stay records, 2,817 2,739 are associated with persons having a positive person-level weight (PERWT15FPERWT13F). The persons represented on this file had to meet the following three criteria: 1. The hospital stay had to have been reported by a household survey respondent as an inpatient hospital stay (regardless of a stay’s length). Thus, the file contains some hospitalizations that were reported as not including an overnight stay. 2. The hospital stay had to have ended during 20152013. Stays that began prior to 2015 2013 but ended during 2015 2013 are included on this data file. Stays that began in 2015 2013 but ended during 2016 2014 are excluded from this data file and will be included in a subsequent 2016 2014 IP data file. Persons with no hospital inpatient stay events for 2015 2013 are not included on this event-level IP file but are represented on the person-level 2015 2013 Full Year Population Characteristics file. 3. The persons represented on this file also had to meet either 3a) or 3b): a) Be classified as a key in-scope person who responded for his or her entire period of 2015 2013 eligibility (i.e., persons with a positive 2015 2013 full-year person- level sampling weight (PERWT15F PERWT13F > 0)), or b) Be an eligible member of a family all of whose key in-scope members have a positive person-level weight (PERWT15F PERWT13F > 0). (Such a family consists of all persons with the same value for FAMIDYR.) That is, the person must have a positive full-year family-level weight (FAMWT15F FAMWT13F > 0). Note that FAMIDYR and FAMWT15F FAMWT13F are variables on the 2015 2013 Full Year Consolidated Data File. One caveat that should be noted is that in the case of a newborn and the hospital inpatient stay associated with the newborn’s birth, a separate hospital inpatient stay record exists on the file only if the newborn was discharged after the mother. Thus, hospital stays associated with a normal birth are generally represented on the file as a single record (i.e., the mother’s hospital inpatient stay record, covering expenditure data for both the mother and baby). In situations where the newborn was discharged after the mother, the birth event will be represented as two records: one record for the mother and one record for the baby. For newborns re-admitted to the hospital during the reference year, each subsequent re-admission will have a separate record. Each inpatient record includes the following: start and end dates of the hospital inpatient stay; number of nights in the hospital; reason entered the hospital; condition(s) associated with the hospital inpatient stay; medicines prescribed at discharge; flat fee information; imputed sources of payment; total payment and total charge for both the facility and physician portions of the hospital inpatient stay expenditure; a full-year person-level weight; variance strata; and variance PSU. To append person-level information such as demographic or health insurance coverage to each event record, data from this file can be merged with 2015 2013 MEPS HC person-level data (e.g. Full Year Consolidated or Full Year Population Characteristics file) using the person identifier, DUPERSID. Hospital inpatient stay events can also be linked to the MEPS 2015 2013 Medical Conditions File and the MEPS 2015 2013 Prescribed Medicines File. Please see Section 5.0 or the MEPS 2015 2013 Appendix File, HC-178IHC-160I, for details on how to merge MEPS data files.

Data File Information. The 2015 Hospital Inpatient Stays This public use data set consists of one 2 event-level data filefiles. The file File 1 contains characteristics associated with the STAZ hospital inpatient stay event and imputed expenditure data. The 2015 STAZ File 2 contains select survey administration and ID variables as well as pre-imputed and unimputed expenditure data from both the Household and Medical Provider Components. Please see the Attachment 1 for definitions of imputed, pre-imputed, and unimputed expenditure variables. Both Files 1 and 2 of this public use data set contains contain variable and frequency distributions for a total of 2,921 3710 hospital inpatient stay records reported during the 2015 1997 portion of Round 3 and Rounds 4 and 5 for Panel 191, as well as Rounds 1, 2, 2 and the 2015 1997 portion of Round 3 for Panel 20 of 2 (i.e., the Rounds for the MEPS Household Componentpanels covering calendar year 1997). This file includes These files include hospital inpatient stay records for all household survey members respondents who resided in eligible responding households and for whom reported at least one hospital inpatient stay was reportedstay. Hospital inpatient stay records known to have ended before January 1, 2015 or occurred after December 31, 2015 1997, are not included on this file. Some household members respondents may have had multiple hospital inpatient stays reported and, thus, will be represented in multiple records on this file. Other household members respondents may have had reported no hospital inpatient stays and, thus, will have no records on this file. Of the 2,921 3710 hospital inpatient stay records, 2,817 3574 are associated with persons having a positive person-level weight (PERWT15FWTDPER97). The persons represented on this file had to meet the following three criteria: 1. ) The hospital stay had to have been reported by a household survey respondent as an inpatient hospital stay (regardless of a stay’s length). Thus, the file contains some hospitalizations that were reported as not including an overnight stay. 2. ) The hospital stay had to have ended during 20151997. Stays that began prior to 2015 1997, but ended during 2015 1997, are included on this data file. Stays that began in 2015 1997, but ended during 2016 1998, are excluded from this data file and will be included in represented on a subsequent 2016 IP 1998 data file. Persons Please note that persons with no hospital inpatient stay events stays use for 2015 1997 are not included on this event-level IP file (but are represented on the MEPS person-level 2015 Full Year Population Characteristics filefiles). 3. ) The persons represented on this file also had to also meet either 3a) 3a or 3b):: a) Be classified as a key in-scope person who responded for his or her entire period of 2015 1997 eligibility (i.e., persons with a positive 2015 1997 full-year person- person-level sampling weight (PERWT15F WTDPER97 > 0)), or b) Be classified as either an eligible non-key person or an eligible out- of-scope person who responded for his or her entire period of 1997 eligibility, and belonged to a family (i.e., all persons within a household (DUID) with the same value of FAMID) in which all eligible family members responded for their entire period of 1997 eligibility, and at least one family member has a positive 1997 full- year person weight (i.e., eligible non-key or eligible out-of-scope persons who are members of a family all of whose key in-scope members have a positive person-level weight (PERWT15F > 0). (Such a family consists of all persons with the same value for FAMIDYR.) That is, the person must have a positive 1997 full-year family-level weight (FAMWT15F > WTFAM97 >0)). Note that FAMIDYR Please refer to Attachment 1 for definitions of keyness, in-scope and FAMWT15F are variables on the 2015 Full Year Consolidated Data Fileeligibility. One caveat that should be noted is that that, in the case of a newborn and the hospital inpatient stay associated with the newborn’s birth, a separate hospital inpatient stay record exists on the file only if the newborn was discharged after the mother. Thus, hospital stays associated with a normal birth are generally represented on the file as a single record (i.e., the mother’s hospital inpatient stay record, covering expenditure data for both the mother and baby). In situations where the newborn was discharged after the mother, the birth event will be represented as two records: one record for the mother and one record for the baby. For newborns re-admitted to the hospital during the reference year, each subsequent re-admission will have a separate record. Each inpatient record includes the following: start and end dates of the hospital inpatient stay; number of nights in the hospital; reason entered the hospital; condition(s) associated with the hospital inpatient stay; medicines prescribed at discharge; flat fee information; imputed sources of payment; total payment and total charge for both the facility and physician portions of the hospital inpatient stay expenditure; a full-year person-level weight; variance strata; and variance PSU. To append person-level information such as demographic or health insurance coverage to each event record, data from this file can be merged with 2015 MEPS HC person-level data (e.g. Full Year Consolidated or Full Year Population Characteristics file) using the person identifier, DUPERSID. Hospital inpatient stay events can also be linked to the MEPS 2015 Medical Conditions File and the MEPS 2015 Prescribed Medicines File. Please see Section 5.0 or the MEPS 2015 Appendix File, HC-178I, for details on how to merge MEPS data files.

Data File Information. The 2015 2004 Hospital Inpatient Stays public use data set consists of one event-level data file. The file contains characteristics associated with the STAZ event and imputed expenditure data. The 2015 2004 STAZ public use data set contains variable and frequency distributions for a total of 2,921 3,456 hospital inpatient stay records reported during the 2015 2004 portion of Round 3 and Rounds 4 and 5 for Panel 198, as well as Rounds 1, 2, and the 2015 2004 portion of Round 3 for Panel 20 9 of the MEPS Household Component. This file includes hospital inpatient stay records for all household survey members respondents who resided in eligible responding households and for whom reported at least one hospital inpatient stay was reportedstay. Hospital inpatient stay records known to have ended before January 1, 2015 2004 or after December 31, 2015 2004 are not included on this file. Some household members respondents may have had multiple hospital inpatient stays reported and, thus, will be represented in multiple records on this file. Other household members respondents may have had reported no hospital inpatient stays and, thus, will have no records on this file. Of the 2,921 3,456 hospital inpatient stay records, 2,817 3,304 are associated with persons having a positive person-level weight (PERWT15FPERWT04F). The persons represented on this file had to meet the following three criteria: 1. ) The hospital stay had to have been reported by a household survey respondent as an inpatient hospital stay (regardless of a stay’s length). Thus, the file contains some hospitalizations that were reported as not including an overnight stay. 2. ) The hospital stay had to have ended during 20152004. Stays that began prior to 2015 2004 but ended during 2015 2004 are included on this data file; H85DCB.PDF provides the file codebook. Stays that began in 2015 2004 but ended during 2016 2005 are excluded from this data file and will be included in a subsequent 2016 2005 IP data file. Persons with no hospital inpatient stay events for 2015 2004 are not included on this event-level IP file but are represented on the person-level 2015 2004 Full Year Population Characteristics file. 3. ) The persons represented on this file also had to meet either 3a) 3a or 3b):: a) Be classified as a key in-scope person who responded for his or her entire period of 2015 2004 eligibility (i.e., persons with a positive 2015 2004 full-year person- person-level sampling weight (PERWT15F PERWT04F > 0)), or b) Be an eligible member of a family all of whose key in-scope members have a positive person-level weight (PERWT15F PERWT04F > 0). (Such a family consists of all persons with the same value for FAMIDYR.) That is, the person must have a positive full-year family-level weight (FAMWT15F FAMWT04F > 0). Note that FAMIDYR and FAMWT15F FAMWT04F are variables on the 2015 Full Year Consolidated Data File2004 Population Characteristics file. One caveat that should be noted is that in the case of a newborn and the hospital inpatient stay associated with the newborn’s birth, a separate hospital inpatient stay record exists on the file only if the newborn was discharged after the mother. Thus, hospital stays associated with a normal birth are generally represented on the file as a single record (i.e., the mother’s hospital inpatient stay record, covering expenditure data for both the mother and baby). In situations where the newborn was discharged after the mother, the birth event will be represented as two records: one record for the mother and one record for the baby. For newborns re-admitted to the hospital during the reference year, each subsequent re-admission will have a separate record. Each inpatient record includes the following: start and end dates of the hospital inpatient stay; number of nights in the hospital; reason entered the hospital; condition(s) associated with the hospital inpatient stay; medicines prescribed at discharge; flat fee information; imputed sources of payment; total payment and total charge for both the facility and physician portions of the hospital inpatient stay expenditure; a full-year person-level weight; variance strata; and variance PSU. To append person-level information such as demographic or health insurance coverage to each event record, data Data from this file can be merged with 2015 the MEPS HC person-level data (e.g. Full Year Consolidated or 2004 Full Year Population Characteristics file) File using the unique person identifier, DUPERSID, to append person- level information, such as demographic or health insurance characteristics, to each record. Hospital inpatient stay events can also be linked to the MEPS 2015 2004 Medical Conditions File and the MEPS 2015 2004 Prescribed Medicines File. Please see Section 5.0 or the MEPS 2015 2004 Appendix File, HC-178IHC-085I, for details on how to merge MEPS data files.

Data File Information. The 2015 2010 Hospital Inpatient Stays public use data set consists of one event-level data file. The file contains characteristics associated with the STAZ event and imputed expenditure data. The 2015 2010 STAZ public use data set contains variable and frequency distributions for a total of 2,921 2,843 hospital inpatient stay records reported during the 2015 2010 portion of Round 3 and Rounds 4 and 5 for Panel 1914, as well as Rounds 1, 2, and the 2015 2010 portion of Round 3 for Panel 20 15 of the MEPS Household Component. This file includes hospital inpatient stay records for all household survey members who resided in eligible responding households and for whom at least one hospital inpatient stay was reported. Hospital inpatient stay records known to have ended before January 1, 2015 2010 or after December 31, 2015 2010 are not included on this file. Some household members may have had multiple hospital inpatient stays reported and, thus, will be represented in multiple records on this file. Other household members may have had reported no hospital inpatient stays and, thus, will have no records on this file. Of the 2,921 2,843 hospital inpatient stay records, 2,817 2,737 are associated with persons having a positive person-level weight (PERWT15FPERWT10F). The persons represented on this file had to meet the following three criteria: 1. The hospital stay had to have been reported by a household survey respondent as an inpatient hospital stay (regardless of a stay’s length). Thus, the file contains some hospitalizations that were reported as not including an overnight stay. 2. The hospital stay had to have ended during 20152010. Stays that began prior to 2015 2010 but ended during 2015 2010 are included on this data file. Stays that began in 2015 2010 but ended during 2016 2011 are excluded from this data file and will be included in a subsequent 2016 2011 IP data file. Persons with no hospital inpatient stay events for 2015 2010 are not included on this event-level IP file but are represented on the person-level 2015 2010 Full Year Population Characteristics file. 3. The persons represented on this file also had to meet either 3a) 3a or 3b):: a) Be classified as a key in-scope person who responded for his or her entire period of 2015 2010 eligibility (i.e., persons with a positive 2015 2010 full-year person- level sampling weight (PERWT15F PERWT10F > 0)), or b) Be an eligible member of a family all of whose key in-scope members have a positive person-level weight (PERWT15F PERWT10F > 0). (Such a family consists of all persons with the same value for FAMIDYR.) That is, the person must have a positive full-year family-level weight (FAMWT15F FAMWT10F > 0). Note that FAMIDYR and FAMWT15F FAMWT10F are variables on the 2015 Full Year Consolidated Data File2010 Population Characteristics file. One caveat that should be noted is that in the case of a newborn and the hospital inpatient stay associated with the newborn’s birth, a separate hospital inpatient stay record exists on the file only if the newborn was discharged after the mother. Thus, hospital stays associated with a normal birth are generally represented on the file as a single record (i.e., the mother’s hospital inpatient stay record, covering expenditure data for both the mother and baby). In situations where the newborn was discharged after the mother, the birth event will be represented as two records: one record for the mother and one record for the baby. For newborns re-admitted to the hospital during the reference year, each subsequent re-admission will have a separate record. Each inpatient record includes the following: start and end dates of the hospital inpatient stay; number of nights in the hospital; reason entered the hospital; condition(s) associated with the hospital inpatient stay; medicines prescribed at discharge; flat fee information; imputed sources of payment; total payment and total charge for both the facility and physician portions of the hospital inpatient stay expenditure; a full-year person-level weight; variance strata; and variance PSU. To append person-level information such as demographic or health insurance coverage to each event record, data from this file can be merged with 2015 2010 MEPS HC person-level data (e.g. Full Year Consolidated or Full Year Population Characteristics filefiles) using the person identifier, DUPERSID. Hospital inpatient stay events can also be linked to the MEPS 2015 2010 Medical Conditions File and the MEPS 2015 2010 Prescribed Medicines File. Please see Section 5.0 or the MEPS 2015 2010 Appendix File, HC-178IHC-135I, for details on how to merge MEPS data files.

Data File Information. The 2015 2008 Hospital Inpatient Stays public use data set consists of one event-level data file. The file contains characteristics associated with the STAZ event and imputed expenditure data. The 2015 2008 STAZ public use data set contains variable and frequency distributions for a total of 2,921 2,821 hospital inpatient stay records reported during the 2015 2008 portion of Round 3 and Rounds 4 and 5 for Panel 1912, as well as Rounds 1, 2, and the 2015 2008 portion of Round 3 for Panel 20 13 of the MEPS Household Component. This file includes hospital inpatient stay records for all household survey members respondents who resided in eligible responding households and for whom reported at least one hospital inpatient stay was reportedstay. Hospital inpatient stay records known to have ended before January 1, 2015 2008 or after December 31, 2015 2008 are not included on this file. Some household members respondents may have had multiple hospital inpatient stays reported and, thus, will be represented in multiple records on this file. Other household members respondents may have had reported no hospital inpatient stays and, thus, will have no records on this file. Of the 2,921 2,821 hospital inpatient stay records, 2,817 2,690 are associated with persons having a positive person-level weight (PERWT15FPERWT08F). The persons represented on this file had to meet the following three criteria: 1. The hospital stay had to have been reported by a household survey respondent as an inpatient hospital stay (regardless of a stay’s length). Thus, the file contains some hospitalizations that were reported as not including an overnight stay. 2. The hospital stay had to have ended during 20152008. Stays that began prior to 2015 2008 but ended during 2015 2008 are included on this data file. Stays that began in 2015 2008 but ended during 2016 2009 are excluded from this data file and will be included in a subsequent 2016 2009 IP data file. Persons with no hospital inpatient stay events for 2015 2008 are not included on this event-level IP file but are represented on the person-level 2015 2008 Full Year Population Characteristics file. 3. The persons represented on this file also had to meet either 3a) 3a or 3b):: a) Be classified as a key in-scope person who responded for his or her entire period of 2015 2008 eligibility (i.e., persons with a positive 2015 2008 full-year person- level sampling weight (PERWT15F PERWT08F > 0)), or b) Be an eligible member of a family all of whose key in-scope members have a positive person-level weight (PERWT15F PERWT08F > 0). (Such a family consists of all persons with the same value for FAMIDYR.) That is, the person must have a positive full-year family-level weight (FAMWT15F FAMWT08F > 0). Note that FAMIDYR and FAMWT15F FAMWT08F are variables on the 2015 Full Year Consolidated Data File2008 Population Characteristics file. One caveat that should be noted is that in the case of a newborn and the hospital inpatient stay associated with the newborn’s birth, a separate hospital inpatient stay record exists on the file only if the newborn was discharged after the mother. Thus, hospital stays associated with a normal birth are generally represented on the file as a single record (i.e., the mother’s hospital inpatient stay record, covering expenditure data for both the mother and baby). In situations where the newborn was discharged after the mother, the birth event will be represented as two records: one record for the mother and one record for the baby. For newborns re-admitted to the hospital during the reference year, each subsequent re-admission will have a separate record. Each inpatient record includes the following: start and end dates of the hospital inpatient stay; number of nights in the hospital; reason entered the hospital; condition(s) associated with the hospital inpatient stay; medicines prescribed at discharge; flat fee information; imputed sources of payment; total payment and total charge for both the facility and physician portions of the hospital inpatient stay expenditure; a full-year person-level weight; variance strata; and variance PSU. To append person-level information such as demographic or health insurance coverage to each event record, data from this file can be merged with 2015 2008 MEPS HC person-level data (e.g. Full Year Consolidated or Full Year Population Characteristics filefiles) using the person identifier, DUPERSID. Hospital inpatient stay events can also be linked to the MEPS 2015 2008 Medical Conditions File and the MEPS 2015 2008 Prescribed Medicines File. Please see Section 5.0 or the MEPS 2015 2008 Appendix File, HC-178IHC-118I, for details on how to merge MEPS data files.

Data File Information. The 2015 2011 Hospital Inpatient Stays public use data set consists of one event-level data file. The file contains characteristics associated with the STAZ event and imputed expenditure data. The 2015 2011 STAZ public use data set contains variable and frequency distributions for a total of 2,921 2,810 hospital inpatient stay records reported during the 2015 2011 portion of Round 3 and Rounds 4 and 5 for Panel 1915, as well as Rounds 1, 2, and the 2015 2011 portion of Round 3 for Panel 20 16 of the MEPS Household Component. This file includes hospital inpatient stay records for all household survey members who resided in eligible responding households and for whom at least one hospital inpatient stay was reported. Hospital inpatient stay records known to have ended before January 1, 2015 2011 or after December 31, 2015 2011 are not included on this file. Some household members may have had multiple hospital inpatient stays reported and, thus, will be represented in multiple records on this file. Other household members may have had reported no hospital inpatient stays and, thus, will have no records on this file. Of the 2,921 2,810 hospital inpatient stay records, 2,817 2,709 are associated with persons having a positive person-level weight (PERWT15FPERWT11F). The persons represented on this file had to meet the following three criteria: 1. The hospital stay had to have been reported by a household survey respondent as an inpatient hospital stay (regardless of a stay’s length). Thus, the file contains some hospitalizations that were reported as not including an overnight stay. 2. The hospital stay had to have ended during 20152011. Stays that began prior to 2015 2011 but ended during 2015 2011 are included on this data file. Stays that began in 2015 2011 but ended during 2016 2012 are excluded from this data file and will be included in a subsequent 2016 2012 IP data file. Persons with no hospital inpatient stay events for 2015 2011 are not included on this event-level IP file but are represented on the person-level 2015 2011 Full Year Population Characteristics file. 3. The persons represented on this file also had to meet either 3a) or 3b): a) Be classified as a key in-scope person who responded for his or her entire period of 2015 2011 eligibility (i.e., persons with a positive 2015 2011 full-year person- level sampling weight (PERWT15F PERWT11F > 0)), or b) Be an eligible member of a family all of whose key in-scope members have a positive person-level weight (PERWT15F PERWT11F > 0). (Such a family consists of all persons with the same value for FAMIDYR.) That is, the person must have a positive full-year family-level weight (FAMWT15F FAMWT11F > 0). Note that FAMIDYR and FAMWT15F FAMWT11F are variables on the 2015 Full Year Consolidated Data File2011 Population Characteristics file. One caveat that should be noted is that in the case of a newborn and the hospital inpatient stay associated with the newborn’s birth, a separate hospital inpatient stay record exists on the file only if the newborn was discharged after the mother. Thus, hospital stays associated with a normal birth are generally represented on the file as a single record (i.e., the mother’s hospital inpatient stay record, covering expenditure data for both the mother and baby). In situations where the newborn was discharged after the mother, the birth event will be represented as two records: one record for the mother and one record for the baby. For newborns re-admitted to the hospital during the reference year, each subsequent re-admission will have a separate record. Each inpatient record includes the following: start and end dates of the hospital inpatient stay; number of nights in the hospital; reason entered the hospital; condition(s) associated with the hospital inpatient stay; medicines prescribed at discharge; flat fee information; imputed sources of payment; total payment and total charge for both the facility and physician portions of the hospital inpatient stay expenditure; a full-year person-level weight; variance strata; and variance PSU. To append person-level information such as demographic or health insurance coverage to each event record, data from this file can be merged with 2015 2011 MEPS HC person-level data (e.g. Full Year Consolidated or Full Year Population Characteristics file) using the person identifier, DUPERSID. Hospital inpatient stay events can also be linked to the MEPS 2015 2011 Medical Conditions File and the MEPS 2015 2011 Prescribed Medicines File. Please see Section 5.0 or the MEPS 2015 2011 Appendix File, HC-178IHC-144I, for details on how to merge MEPS data files.

Data File Information. The 2015 2016 Hospital Inpatient Stays public use data set consists of one event-level data file. The file contains characteristics associated with the STAZ event and imputed expenditure data. The 2015 2016 STAZ public use data set contains variable and frequency distributions for a total of 2,921 2,765 hospital inpatient stay records reported during the 2015 2016 portion of Round 3 and Rounds 4 and 5 for Panel 1920, as well as Rounds 1, 2, and the 2015 2016 portion of Round 3 for Panel 20 21 of the MEPS Household Component. This file includes hospital inpatient stay records for all household survey members who resided in eligible responding households and for whom at least one hospital inpatient stay was reported. Hospital inpatient stay records known to have ended before January 1, 2015 2016 or after December 31, 2015 2016 are not included on this file. Some household members may have had multiple hospital inpatient stays reported and, thus, will be represented in multiple records on this file. Other household members may have had reported no hospital inpatient stays and, thus, will have no records on this file. Of the 2,921 2,765 hospital inpatient stay records, 2,817 2,692 are associated with persons having a positive person-level weight (PERWT15FPERWT16F). The persons represented on this file had to meet the following three criteria: 1. The hospital stay had to have been reported by a household survey respondent as an inpatient hospital stay (regardless of a stay’s length). Thus, the file contains some hospitalizations that were reported as not including an overnight stay. 2. The hospital stay had to have ended during 20152016. Stays that began prior to 2015 2016 but ended during 2015 2016 are included on this data file. Stays that began in 2015 2016 but ended during 2016 2017 are excluded from this data file and will be included in a subsequent 2016 2017 IP data file. Persons with no hospital inpatient stay events for 2015 2016 are not included on this event-level IP file but are represented on the person-level 2015 2016 Full Year Population Characteristics file. 3. The persons represented on this file also had to meet either 3a) or 3b): a) Be classified as a key in-scope person who responded for his or her entire period of 2015 2016 eligibility (i.e., persons with a positive 2015 2016 full-year person- level sampling weight (PERWT15F PERWT16F > 0)), or b) Be an eligible member of a family all of whose key in-scope members have a positive person-level weight (PERWT15F PERWT16F > 0). (Such a family consists of all persons with the same value for FAMIDYR.) That is, the person must have a positive full-year family-level weight (FAMWT15F FAMWT16F > 0). Note that FAMIDYR and FAMWT15F FAMWT16F are variables on the 2015 2016 Full Year Consolidated Data File. One caveat that should be noted is that in the case of a newborn and the hospital inpatient stay associated with the newborn’s birth, a separate hospital inpatient stay record exists on the file only if the newborn was discharged after the mother. Thus, hospital stays associated with a normal birth are generally represented on the file as a single record (i.e., the mother’s hospital inpatient stay record, covering expenditure data for both the mother and baby). In situations where the newborn was discharged after the mother, the birth event will be represented as two records: one record for the mother and one record for the baby. For newborns re-admitted to the hospital during the reference year, each subsequent re-admission will have a separate record. Each inpatient record includes the following: start and end dates of the hospital inpatient stay; number of nights in the hospital; reason entered the hospital; condition(s) associated with the hospital inpatient stay; medicines prescribed at discharge; flat fee information; imputed sources of payment; total payment and total charge for both the facility and physician portions of the hospital inpatient stay expenditure; a full-year person-level weight; variance strata; and variance PSU. To append person-level information such as demographic or health insurance coverage to each event record, data from this file can be merged with 2015 2016 MEPS HC person-level data (e.g. Full Year Consolidated or Full Year Population Characteristics file) using the person identifier, DUPERSID. Hospital inpatient stay events can also be linked to the MEPS 2015 2016 Medical Conditions File and the MEPS 2015 2016 Prescribed Medicines File. Please see Section 5.0 or the MEPS 2015 2016 Appendix File, HC-178IHC-188I, for details on how to merge MEPS data files.