Common use of Details on Person Weight Construction Clause in Contracts

Details on Person Weight Construction. The person-level weight PERWT12F was developed in several stages. First, person-level weights for Panel 16 and Panel 17 were created separately. The weighting process for each panel included adjustments for nonresponse over time and calibration to independent population totals. The calibration was initially accomplished separately for each panel by raking the corresponding sample weights for those in-scope at the end of the calendar year to Current Population Survey (CPS) population estimates based on five variables. The five variables used in the establishment of the initial person-level control figures were: census region (Northeast, Midwest, South, West); MSA status (MSA, non-MSA); race/ethnicity (Hispanic; Black, non-Hispanic; Asian, non- Hispanic; and other); sex; and age. A 2012 composite weight was then formed by multiplying each weight from Panel 16 by the factor .49 and each weight from Panel 17 by the factor .51. The choice of factors reflected the relative sample sizes of the two panels, helping to limit the variance of estimates obtained from pooling the two samples. The composite weight was raked to the same set of CPS-based control totals. When the poverty status information derived from income variables became available, a final raking was undertaken on the previously established weight variable. Control totals were established using poverty status (five categories: below poverty, from 100 to 125 percent of poverty, from 125 to 200 percent of poverty, from 200 to 400 percent of poverty, at least 400 percent of poverty), the other five variables previously used in the weight calibration, as well as age categories cross-classified with categories associated with numbers of office-based visits and age categories cross-classified with categories reflecting the number of prescribed medicines purchased.

Details on Person Weight Construction. The person-level weight PERWT12F PERWT11F was developed in several stages. First, personPerson-level weights for Panel 16 15 and Panel 17 16 were created separately. The weighting process for each panel included adjustments an adjustment for nonresponse over time and calibration to independent population totalsfigures. The calibration was initially accomplished separately for each panel by raking the corresponding sample weights for those in-scope at the end of the calendar year to Current Population Survey (CPS) population estimates based on five six variables. The five six variables used in the establishment of the initial person-level control figures were: census region (Northeast, Midwest, South, West); MSA status (MSA, non-MSA); race/ethnicity (Hispanic; Black, non-Hispanic; Asian, non- Hispanic; and other); sex; education level; and age. A 2012 2011 composite weight was then formed by multiplying each weight from Panel 16 15 by the factor .49 .43 and each weight from Panel 17 16 by the factor .51.57. The choice of factors reflected the relative sample sizes of the two panels, helping to limit the variance of estimates obtained from pooling the two samples. The composite weight was again raked to the same set of CPS-based control totals. When the poverty status information derived from income variables became available, a final raking was undertaken on the previously established weight variable. Control totals were established using poverty status (five categories: below poverty, from 100 to 125 percent of poverty, from 125 to 200 percent of poverty, from 200 to 400 percent of poverty, at least 400 percent of poverty), ) to replace education level as one of the other five six variables previously used in the weight calibration, as well as age categories cross-classified with categories associated with numbers of office-based visits and age categories cross-classified with categories reflecting the number of prescribed medicines purchased.

Details on Person Weight Construction. The person-level weight PERWT12F PERWT13F was developed in several stages. First, person-level weights for Panel 16 17 and Panel 17 18 were created separately. The weighting process for each panel included adjustments for nonresponse over time and calibration to independent population totals. The calibration was initially accomplished separately for each panel by raking the corresponding sample weights for those in-scope at the end of the calendar year to Current Population Survey (CPS) population estimates based on five variables. The five variables used in the establishment of the initial person-level control figures were: census region (Northeast, Midwest, South, West); MSA status (MSA, non-MSA); race/ethnicity (Hispanic; Black, non-Hispanic; Asian, non- Hispanic; and other); sex; and age. A 2012 2013 composite weight was then formed by multiplying each weight from Panel 16 17 by the factor .49 .506 and each weight from Panel 17 18 by the factor .51.494. The choice of factors reflected the relative sample sizes of the two panels, helping to limit the variance of estimates obtained from pooling the two samples. The composite weight was raked to the same set of CPS-based control totals. When the poverty status information derived from income variables became available, a final raking was undertaken on the previously established weight variable. Control totals were established using poverty status (five categories: below poverty, from 100 to 125 percent of poverty, from 125 to 200 percent of poverty, from 200 to 400 percent of poverty, at least 400 percent of poverty), the other five variables previously used in the weight calibration, as well as age categories cross-classified with categories and a variable associated with numbers of office-based visits and age categories cross-classified with categories reflecting the number of prescribed medicines purchasedhospital stays for those under the age of 65.

Details on Person Weight Construction. The person-level weight PERWT12F PERWT14F was developed in several stages. First, person-level weights for Panel 16 18 and Panel 17 19 were created separately. The weighting process for each panel included adjustments an adjustment for nonresponse over time and calibration to independent population totals. The calibration was initially accomplished separately for each panel by raking the corresponding sample weights for those in-scope at the end of the calendar year to Current Population Survey (CPS) population estimates based on five variables. The five variables used in the establishment of the initial person-level control figures were: census region (Northeast, Midwest, South, West); MSA status (MSA, non-MSA); race/ethnicity (Hispanic; Black, non-non- Hispanic; Asian, non- non-Hispanic; and other); sex; and age. A 2012 2014 composite weight was then formed by multiplying each weight from Panel 16 18 by the factor .49 .500 and each weight from Panel 17 19 by the factor .51.500. The choice of factors reflected the relative sample sizes of the two panels, helping to limit the variance of estimates obtained from pooling the two samples. The composite weight was raked to the same set of CPS-based control totals. When the poverty status information derived from income variables became available, a final raking was undertaken on the previously established weight variable. Control totals were established using poverty status (five categories: below poverty, from 100 to 125 percent of poverty, from 125 to 200 percent of poverty, from 200 to 400 percent of poverty, at least 400 percent of poverty), ) as well as the other five variables previously used in the weight calibration, as well as age categories cross-classified with categories associated with numbers of office-based visits and age categories cross-classified with categories reflecting the number of prescribed medicines purchased.

Details on Person Weight Construction. The person-level weight PERWT12F PERWT15F was developed in several stages. First, personPerson-level weights for Panel 16 19 and Panel 17 20 were created separately. The weighting process for each panel included adjustments an adjustment for nonresponse over time and calibration to independent population totalsfigures. The calibration was initially accomplished separately for each panel by raking the corresponding sample weights for those in-scope at the end of the calendar year to Current Population Survey (CPS) population estimates based on five variables. The five variables used in the establishment of the initial person-level control figures were: census region (Northeast, Midwest, South, West); MSA status (MSA, non-MSA); race/ethnicity (Hispanic; Black, non-Hispanic; Asian, non- Hispanic; and other); sex; and age. A 2012 2015 composite weight was then formed by multiplying each weight from Panel 16 19 by the factor .49 .460 and each weight from Panel 17 20 by the factor .51.540. The choice of factors reflected the relative sample sizes of the two panels, helping to limit the variance of estimates obtained from pooling the two samples. The composite weight was raked to the same set of CPS-based control totals. When the poverty status information derived from income variables became available, a final raking was undertaken on the previously established weight variable. Control totals were established using poverty status (five categories: below poverty, from 100 to 125 percent of poverty, from 125 to 200 percent of poverty, from 200 to 400 percent of poverty, at least 400 percent of poverty), ) as well as the other five variables previously used in the weight calibration, as well as age categories cross-classified with categories associated with numbers of office-based visits and age categories cross-classified with categories reflecting the number of prescribed medicines purchased.

Details on Person Weight Construction. The person-level weight PERWT12F PERWT16F was developed in several stages. First, person-level weights for Panel 16 20 and Panel 17 21 were created separately. The weighting process for each panel included adjustments an adjustment for nonresponse over time and calibration to independent population totals. The calibration was initially accomplished separately for each panel by raking the corresponding sample weights for those in-scope at the end of the calendar year to Current Population Survey (CPS) population estimates based on five variables. The five variables used in the establishment of the initial person-level control figures were: census region (Northeast, Midwest, South, West); MSA status (MSA, non-MSA); race/ethnicity (Hispanic; Black, non-non- Hispanic; Asian, non- non-Hispanic; and other); sex; and age. A 2012 2016 composite weight was then formed by multiplying each weight from Panel 16 20 by the factor .49 .510 and each weight from Panel 17 21 by the factor .51.490. The choice of factors reflected the relative sample sizes of the two panels, helping to limit the variance of estimates obtained from pooling the two samples. The composite weight was raked to the same set of CPS-based control totals. When the poverty status information derived from income variables became available, a final raking was undertaken on the previously established weight variable. Control totals were established using poverty status (five categories: below poverty, from 100 to 125 percent of poverty, from 125 to 200 percent of poverty, from 200 to 400 percent of poverty, at least 400 percent of poverty), ) as well as the other five variables previously used in the weight calibration. In developing the final person-level weight for 2016 (PERWT16F), as well as age categories cross-classified with categories associated with an additional raking dimension was included beyond those based on the usual six variables. This dimension was added to adjust the distribution of inpatient hospital utilization among the elderly to reflect trends in other data sources. The table below shows ratios of weighted numbers for those 65 and older that were used to establish this additional raking dimension, modifying the corresponding estimates obtained without the additional dimension. Ratio of office-based visits Adjusted to Unadjusted Weights (Cases where AGE16X>=65 and age categories cross-classified with categories reflecting the number INSC1231=1) # of prescribed medicines purchased.Inpatient Discharges (IPDIS16) # of Nights in Hospital for Discharges (IPNGTD16) Ratio 1+ 5 - 9 1.1548

Details on Person Weight Construction. The person-level weight PERWT12F PERWT16F was developed in several stages. First, personPerson-level weights for Panel 16 20 and Panel 17 21 were created separately. The weighting process for each panel included adjustments an adjustment for nonresponse over time and calibration to independent population totalsfigures. The calibration was initially accomplished separately for each panel by raking the corresponding sample weights for those in-scope at the end of the calendar year to Current Population Survey (CPS) population estimates based on five variables. The five variables used in the establishment of the initial person-level control figures were: census region (Northeast, Midwest, South, West); MSA status (MSA, non-MSA); race/ethnicity (Hispanic; Black, non-Hispanic; Asian, non- Hispanic; and other); sex; and age. A 2012 2016 composite weight was then formed by multiplying each weight from Panel 16 20 by the factor .49 .510 and each weight from Panel 17 21 by the factor .51.490. The choice of factors reflected the relative sample sizes of the two panels, helping to limit the variance of estimates obtained from pooling the two samples. The composite weight was raked to the same set of CPS-based control totals. When the poverty status information derived from income variables became available, a final raking was undertaken on the previously established weight variable. Control totals were established using poverty status (five categories: below poverty, from 100 to 125 percent of poverty, from 125 to 200 percent of poverty, from 200 to 400 percent of poverty, at least 400 percent of poverty), ) as well as the other five variables previously used in the weight calibration. In developing the final person-level weight for 2016 (PERWT16F), as well as age categories cross-classified with categories associated with an additional raking dimension was included beyond those based on the usual six variables. This dimension was added to adjust the distribution of inpatient hospital utilization among the elderly to reflect trends in other data sources. The table below shows ratios of weighted numbers for those 65 and older that were used to establish this additional raking dimension, modifying the corresponding estimates obtained without the additional dimension. Ratio of office-based visits Adjusted to Unadjusted Weights (Cases where AGE16X>=65 and age categories cross-classified with categories reflecting the number INSC1231=1) # of prescribed medicines purchased.Inpatient Discharges (IPDIS16) # of Nights in Hospital for Discharges (IPNGTD16) Ratio 1+ 0 - 4 1.1227 1+ 5 - 9 1.1548 1+ 10+ 1.3597

Details on Person Weight Construction. The person-level weight PERWT12F PERWT16F was developed in several stages. First, personPerson-level weights for Panel 16 20 and Panel 17 21 were created separately. The weighting process for each panel included adjustments an adjustment for nonresponse over time and calibration to independent population totalsfigures. The calibration was initially accomplished separately for each panel by raking the corresponding sample weights for those in-scope at the end of the calendar year to Current Population Survey (CPS) population estimates based on five variables. The five variables used in the establishment of the initial person-level control figures were: census region (Northeast, Midwest, South, West); MSA status (MSA, non-MSA); race/ethnicity (Hispanic; Black, non-Hispanic; Asian, non- Hispanic; and other); sex; and age. A 2012 2016 composite weight was then formed by multiplying each weight from Panel 16 20 by the factor .49 .510 and each weight from Panel 17 21 by the factor .51.490. The choice of factors reflected the relative sample sizes of the two panels, helping to limit the variance of estimates obtained from pooling the two samples. The composite weight was raked to the same set of CPS-based control totals. When the poverty status information derived from income variables became available, a final raking was undertaken on the previously established weight variable. Control totals were established using poverty status (five categories: below poverty, from 100 to 125 percent of poverty, from 125 to 200 percent of poverty, from 200 to 400 percent of poverty, at least 400 percent of poverty), ) as well as the other five variables previously used in the weight calibration. In developing the final person-level weight for 2016 (PERWT16F), as well as age categories cross-classified with categories associated with an additional raking dimension was included beyond those based on the usual six variables. This dimension was added to adjust the distribution of inpatient hospital utilization among the elderly to reflect trends in other data sources. The table below shows ratios of weighted numbers for those 65 and older that were used to establish this additional raking dimension, modifying the corresponding estimates obtained without the additional dimension. Ratio of office-based visits Adjusted to Unadjusted Weights (Cases where AGE16X>=65 and age categories cross-classified with categories reflecting the number INSC1231=1) # of prescribed medicines purchased.Inpatient Discharges (IPDIS16) # of Nights in Hospital for Discharges (IPNGTD16) Ratio 1+ 5 - 9 1.1548

Details on Person Weight Construction. The person-level weight PERWT12F PERWT15F was developed in several stages. First, person-level weights for Panel 16 19 and Panel 17 20 were created separately. The weighting process for each panel included adjustments an adjustment for nonresponse over time and calibration to independent population totals. The calibration was initially accomplished separately for each panel by raking the corresponding sample weights for those in-scope at the end of the calendar year to Current Population Survey (CPS) population estimates based on five variables. The five variables used in the establishment of the initial person-level control figures were: census region (Northeast, Midwest, South, West); MSA status (MSA, non-MSA); race/ethnicity (Hispanic; Black, non-non- Hispanic; Asian, non- non-Hispanic; and other); sex; and age. A 2012 2015 composite weight was then formed by multiplying each weight from Panel 16 19 by the factor .49 .460 and each weight from Panel 17 20 by the factor .51.540. The choice of factors reflected the relative sample sizes of the two panels, helping to limit the variance of estimates obtained from pooling the two samples. The composite weight was raked to the same set of CPS-based control totals. When the poverty status information derived from income variables became available, a final raking was undertaken on the previously established weight variable. Control totals were established using poverty status (five categories: below poverty, from 100 to 125 percent of poverty, from 125 to 200 percent of poverty, from 200 to 400 percent of poverty, at least 400 percent of poverty), ) as well as the other five variables previously used in the weight calibration, as well as age categories cross-classified with categories associated with numbers of office-based visits and age categories cross-classified with categories reflecting the number of prescribed medicines purchased.

Details on Person Weight Construction. The person-level weight PERWT12F PERWT14F was developed in several stages. First, person-level weights for Panel 16 18 and Panel 17 19 were created separately. The weighting process for each panel included adjustments an adjustment for nonresponse over time and calibration to independent population totals. The calibration was initially accomplished separately for each panel by raking the corresponding sample weights for those in-scope at the end of the calendar year to Current Population Survey (CPS) population estimates based on five variables. The five variables used in the establishment of the initial person-level control figures were: census region (Northeast, Midwest, South, West); MSA status (MSA, non-MSA); race/ethnicity (Hispanic; Black, non-non- Hispanic; Asian, non- non-Hispanic; and other); sex; and age. A 2012 2014 composite weight was then formed by multiplying each weight from Panel 16 18 by the factor .49 .500 and each weight from Panel 17 19 by the factor .51.500. The choice of factors reflected the relative sample sizes of the two panels, helping to limit the variance of estimates obtained from pooling the two samples. The composite weight was raked to the same set of CPS-based control totals. When the poverty status information derived from income variables became available, a final raking was undertaken on the previously established weight variable. Control totals were established using poverty status (five categories: below poverty, from 100 to 125 percent of poverty, from 125 to 200 percent of poverty, from 200 to 400 percent of poverty, at least 400 percent of poverty), as well as the other five variables previously used in the weight calibration, as well as age categories cross-classified with categories associated with numbers of office-based visits and age categories cross-classified with categories reflecting the number of prescribed medicines purchased.

Details on Person Weight Construction. The person-level weight PERWT12F PERWT13F was developed in several stages. First, person-level weights for Panel 16 17 and Panel 17 18 were created separately. The weighting process for each panel included adjustments for nonresponse over time and calibration to independent population totals. The calibration was initially accomplished separately for each panel by raking the corresponding sample weights for those in-scope at the end of the calendar year to Current Population Survey (CPS) population estimates based on five variables. The five variables used in the establishment of the initial person-level control figures were: census region (Northeast, Midwest, South, West); MSA status (MSA, non-MSA); race/ethnicity (Hispanic; Black, non-Hispanic; Asian, non- Hispanic; and other); sex; and age. A 2012 2013 composite weight was then formed by multiplying each weight from Panel 16 17 by the factor .49 .51 and each weight from Panel 17 18 by the factor .51.49. The choice of factors reflected the relative sample sizes of the two panels, helping to limit the variance of estimates obtained from pooling the two samples. The composite weight was raked to the same set of CPS-based control totals. When the poverty status information derived from income variables became available, a final raking was undertaken on the previously established weight variable. Control totals were established using poverty status (five categories: below poverty, from 100 to 125 percent of poverty, from 125 to 200 percent of poverty, from 200 to 400 percent of poverty, at least 400 percent of poverty), the other five variables previously used in the weight calibration, as well as age categories cross-classified with categories and a variable associated with numbers of office-based visits and age categories cross-classified with categories reflecting the number of prescribed medicines purchasedhospital stays for those under the age of 65.

Details on Person Weight Construction. The person-level weight PERWT12F PERWT16F was developed in several stages. First, personPerson-level weights for Panel 16 20 and Panel 17 21 were created separately. The weighting process for each panel included adjustments an adjustment for nonresponse over time and calibration to independent population totalsfigures. The calibration was initially accomplished separately for each panel by raking the corresponding sample weights for those in-scope at the end of the calendar year to Current Population Survey (CPS) population estimates based on five variables. The five variables used in the establishment of the initial person-level control figures were: census region (Northeast, Midwest, South, West); MSA status (MSA, non-MSA); race/ethnicity (Hispanic; Black, non-Hispanic; Asian, non- Hispanic; and other); sex; and age. A 2012 2016 composite weight was then formed by multiplying each weight from Panel 16 20 by the factor .49 .510 and each weight from Panel 17 21 by the factor .51.490. The choice of factors reflected the relative sample sizes of the two panels, helping to limit the variance of estimates obtained from pooling the two samples. The composite weight was raked to the same set of CPS-based control totals. When the poverty status information derived from income variables became available, a final raking was undertaken on the previously established weight variable. Control totals were established using poverty status (five categories: below poverty, from 100 to 125 percent of poverty, from 125 to 200 percent of poverty, from 200 to 400 percent of poverty, at least 400 percent of poverty), ) as well as the other five variables previously used in the weight calibration. In developing the final person-level weight for 2016 (PERWT16F), as well as age categories crossan additional raking dimension was included beyond those based on the usual six variables. This dimension was added to adjust the distribution of inpatient hospital utilization among the elderly to reflect trends in other data sources. The table below shows ratios of weighted numbers for those 65 and older that were used to establish this additional raking dimension, modifying the corresponding estimates obtained without the additional dimension. Ratio of Adjusted to Unadjusted Weights (Cases where AGE16X>=65 and INSC1231=1) # of Inpatient Discharges (IPDIS16) # of Nights in Hospital for Discharges (IPNGTD16) Ratio 1+ 5-classified with categories associated with numbers of office-based visits and age categories cross-classified with categories reflecting the number of prescribed medicines purchased.9 1.1548

Details on Person Weight Construction. The person-level weight PERWT12F PERWT11F was developed in several stages. First, personPerson-level weights for Panel 16 15 and Panel 17 16 were created separately. The weighting process for each panel included adjustments an adjustment for nonresponse over time and calibration to independent population totalsfigures. The calibration was initially accomplished separately for each panel by raking the corresponding sample weights for those in-scope at the end of the calendar year to Current Population Survey (CPS) population estimates based on five six variables. The five six variables used in the establishment of the initial person-level control figures were: census region (Northeast, Midwest, South, West); MSA status (MSA, non-MSA); race/ethnicity (Hispanic; Black, non-Hispanic; Asian, Asian non- Hispanic; and other); sex; education level; and age. A 2012 2011 composite weight was then formed by multiplying each weight from Panel 16 15 by the factor .49 .43 and each weight from Panel 17 16 by the factor .51.57. The choice of factors reflected the relative sample sizes of the two panels, helping to limit the variance of estimates obtained from pooling the two samples. The composite weight was again raked to the same set of CPS-based control totals. When the poverty status information derived from income variables became available, a final raking was undertaken on the previously established weight variable. Control totals were established using poverty status (five categories: below poverty, from 100 to 125 percent of poverty, from 125 to 200 percent of poverty, from 200 to 400 percent of poverty, at least 400 percent of poverty), ) ) to replace education level as one of the other five six variables previously used in the weight calibration, as well as age categories cross-classified with categories associated with numbers of office-based visits and age categories cross-classified with categories reflecting the number of prescribed medicines purchased.

Details on Person Weight Construction. The person-level weight PERWT12F PERWT14F was developed in several stages. First, person-level weights for Panel 16 18 and Panel 17 19 were created separately. The weighting process for each panel included adjustments and adjustment for nonresponse over time and calibration to independent population totals. The calibration was initially accomplished separately for each panel by raking the corresponding sample weights for those in-scope at the end of the calendar year to Current Population Survey (CPS) population estimates based on five variables. The five variables used in the establishment of the initial person-level control figures were: census region (Northeast, Midwest, South, West); MSA status (MSA, non-MSA); race/ethnicity (Hispanic; Black, non-non- Hispanic; Asian, non- non-Hispanic; and other); sex; and age. A 2012 2014 composite weight was then formed by multiplying each weight from Panel 16 18 by the factor .49 .500 and each weight from Panel 17 19 by the factor .51.500. The choice of factors reflected the relative sample sizes of the two panels, helping to limit the variance of estimates obtained from pooling the two samples. The composite weight was raked to the same set of CPS-based control totals. When the poverty status information derived from income variables became available, a final raking was undertaken on the previously established weight variable. Control totals were established using poverty status (five categories: below poverty, from 100 to 125 percent of poverty, from 125 to 200 percent of poverty, from 200 to 400 percent of poverty, at least 400 percent of poverty), ) as well as the other five variables previously used in the weight calibration, as well as age categories cross-classified with categories associated with numbers of office-based visits and age categories cross-classified with categories reflecting the number of prescribed medicines purchased.

Details on Person Weight Construction. The person-level weight PERWT12F PERWT16F was developed in several stages. First, person-level weights for Panel 16 20 and Panel 17 21 were created separately. The weighting process for each panel included adjustments for nonresponse over time and calibration to independent population totals. The calibration was initially accomplished separately for each panel by raking the corresponding sample weights for those in-scope at the end of the calendar year to Current Population Survey (CPS) population estimates based on five variables. The five variables used in the establishment of the initial person-level control figures were: census region (Northeast, Midwest, South, West); MSA status (MSA, non-MSA); race/ethnicity (Hispanic; Black, non-Hispanic; Asian, non- non-Hispanic; and other); sex; and age. A 2012 2016 composite weight was then formed by multiplying each weight from Panel 16 20 by the factor .49 .510 and each weight from Panel 17 21 by the factor .51.490. The choice of Using such factors reflected the relative sample sizes of the two panels, helping to form composite weights serves to limit the variance of estimates obtained from pooling the two samples. The resulting composite weight was raked to the same set of CPS-based control totals. When Then, when the poverty status information (derived from the MEPS income variables variables) became available, a final another raking was undertaken on undertaken, using dimensions reflecting poverty status in addition to the previously established weight variablementioned five variables. Control totals were established using poverty status (five categories: below poverty, from 100 to 125 percent of poverty, from 125 to 200 percent of poverty, from 200 to 400 percent of poverty, at least 400 percent of poverty), ) as well as the other five variables previously used in the weight calibration. Thus, the raking for the final weight reflected poverty status as well as age categories crossthe other five variables previously used in the weight calibration. In developing the final person-classified with categories associated with level weight for 2016 (PERWT16F), an additional raking dimension was included beyond those based on the usual six variables. This dimension was added to adjust the distribution of inpatient hospital utilization among the elderly to reflect trends in other data sources. The table below shows ratios of weighted numbers for those 65 and older that were used to establish this additional raking dimension, modifying the corresponding estimates obtained without the additional dimension. Ratio of officeAdjusted to Unadjusted Weights (Cases where AGE16X>=65 and INSC1231=1) # of Inpatient Discharges (IPDIS16) # of Nights in Hospital for Discharges (IPNGTD16) Ratio 1+ 0-based visits and age categories cross4 1.1227 1+ 5-classified with categories reflecting the number of prescribed medicines purchased.9 1.1548 1+ 10+ 1.3597

Details on Person Weight Construction. The person-level weight PERWT12F PERWT14F was developed in several stages. First, personPerson-level weights for Panel 16 18 and Panel 17 19 were created separately. The weighting process for each panel included adjustments an adjustment for nonresponse over time and calibration to independent population totalsfigures. The calibration was initially accomplished separately for each panel by raking the corresponding sample weights for those in-scope at the end of the calendar year to Current Population Survey (CPS) population estimates based on five variables. The five variables used in the establishment of the initial person-level control figures were: census region (Northeast, Midwest, South, West); MSA status (MSA, non-MSA); race/ethnicity (Hispanic; Black, non-Hispanic; Asian, non- Hispanic; and other); sex; and age. A 2012 2014 composite weight was then formed by multiplying each weight from Panel 16 18 by the factor .49 .500 and each weight from Panel 17 19 by the factor .51.500. The choice of factors reflected the relative sample sizes of the two panels, helping to limit the variance of estimates obtained from pooling the two samples. The composite weight was raked to the same set of CPS-based control totals. When the poverty status information derived from income variables became available, a final raking was undertaken on the previously established weight variable. Control totals were established using poverty status (five categories: below poverty, from 100 to 125 percent of poverty, from 125 to 200 percent of poverty, from 200 to 400 percent of poverty, at least 400 percent of poverty), ) as well as the other five variables previously used in the weight calibration, as well as age categories cross-classified with categories associated with numbers of office-based visits and age categories cross-classified with categories reflecting the number of prescribed medicines purchased.

Details on Person Weight Construction. The person-level weight PERWT12F PERWT15F was developed in several stages. First, person-level weights for Panel 16 19 and Panel 17 20 were created separately. The weighting process for each panel included adjustments an adjustment for nonresponse over time and calibration to independent population totals. The calibration was initially accomplished separately for each panel by raking the corresponding sample weights for those in-scope at the end of the calendar year to Current Population Survey (CPS) population estimates based on five variables. The five variables used in the establishment of the initial person-level control figures were: census region (Northeast, Midwest, South, West); MSA status (MSA, non-MSA); race/ethnicity (Hispanic; Black, non-non- Hispanic; Asian, non- non-Hispanic; and other); sex; and age. A 2012 2015 composite weight was then formed by multiplying each weight from Panel 16 19 by the factor .49 .46 and each weight from Panel 17 20 by the factor .51.54. The choice of factors reflected the relative sample sizes of the two panels, helping to limit the variance of estimates obtained from pooling the two samples. The composite weight was raked to the same set of CPS-based control totals. When the poverty status information derived from income variables became available, a final raking was undertaken on the previously established weight variable. Control totals were established using poverty status (five categories: below poverty, from 100 to 125 percent of poverty, from 125 to 200 percent of poverty, from 200 to 400 percent of poverty, at least 400 percent of poverty), as well as the other five variables previously used in the weight calibration, as well as age categories cross-classified with categories associated with numbers of office-based visits and age categories cross-classified with categories reflecting the number of prescribed medicines purchased.