A Interpreting QA MIL Package output

The QA MIL Package output datasets l3_mil_aggregate and l3_mil_aggregate_deliv are large files containing multiple levels corresponding to different combinations of stratifying variables. Sections A.1 and A.2 are intended to help users navigate these datasets.

Please note that the unit of analysis for each dataset is slightly different. The unit of analysis for the l3_mil_aggregate is each row in the MIL table, while the unit of analysis for the l3_mil_aggregate_deliv table is deliveries only.

A.1 l3_mil_aggregate Level References

The QA MIL Package’s l3_mil_aggregate output dataset contains 1023 levels corresponding to different combinations of stratifying variables. Table A.1 functions as a reference for users seeking a specific variable combination.

Table A.1: l3_mil_aggregate levels with one stratifying variable
Level Number	Level Description	Overall	MatchMethod	DaysDiff	sex	EncType	AgeGroup	ICD_Ver	YearMonth	Year	Birth_Type	LinkageStatus
000	Overall	X
001	MatchMethod		X
002	DaysDiff			X
004	sex				X
008	EncType					X
016	AgeGroup						X
032	ICD_Ver							X
064	YearMonth								X
128	Year									X
256	Birth_Type										X
512	LinkageStatus											X

To find the level number for a multiple variable stratification, users should add the level numbers for each variable of interest in Table A.1 together—e.g., the level number for Year is 64 and the Level for YearMonth is 128. The resulting level number for the datasets stratified by Year YearMonth is 192. Similarly, to stratify by Sex, AgeGroup, and MatchMethod, users should add 004 + 016 + 001 together to get 021. Table A.2 demonstrates this principle:

Table A.2: Example l3_mil_aggregate levels with two stratifying variables
Level Number	Level Description	Overall	MatchMethod	DaysDiff	sex	EncType	AgeGroup	ICD_Ver	YearMonth	Year	Birth_Type	LinkageStatus
021	AgeGroup sex MatchMethod		X		X		X
192	Year YearMonth								X	X

A.2 l3_mil_aggregate_deliv Level References

The QA MIL Package’s l3_mil_aggregate_deliv output dataset contains 255 levels corresponding to different combinations of stratifying variables. The Table A.3 functions as a reference for users seeking to stratify this table by a single variable.

Table A.3: l3_mil_aggregate_deliv levels with one stratifying variable
Level Number	Level Description	Overall	MatchMethod	DaysDiff	sex	EncType	AgeGroup	ICD_Ver	YearMonth	Year
000	Overall	X
001	InfantsLinked		X
002	EncType			X
004	AgeGroup				X
008	ICD_Ver					X
016	YearMonth						X
032	Year							X
064	Birth_Type								X
128	LinkageStatus									X

To find the level number for a multiple variable stratification, users should add the level numbers for each variable of interest in Table A.3 together—e.g., the level number for Year is 016 and the Level for BirthType is 064. The resulting level number for the datasets stratified by Year BirthType is 080. Similarly, to stratify by InfantsLinked, EncType, and AgeGroup users should add 001 + 002 + 004 together to get 007. Table A.4 demonstrates this principle:

Table A.4: l3_mil_aggregate_deliv levels with one stratifying variable
Level Number	Level Description	Overall	MatchMethod	DaysDiff	sex	EncType	AgeGroup	ICD_Ver	YearMonth	Birth_Type	LinkageStatus	Year
007	EncType AgeGroup InfantsLinked		X	X	X
016	YearMonth						X