1.5 Medical Product Use During Pregnancy Study Summary Tables
In studies of medical product use during pregnancy, the QRP Reporting Tool outputs tables summarizing either pregnancy episodes or medical product use episodes by trimester or gestational week of pregnancy.
If the user creates a matched non-pregnant cohort (by specifying T4NOPREG in qrp.Stratafile.TABLEID
) and has chosen to output that non-pregnant cohort (by specifying T4NOPREG in the Table File parameter TABLESUBSTRAT), then the QRP Reporting Tool will output said cohort(s) in the summary tables. All pregnant cohorts will be clustered together under a “Pregnant Cohort” header row, while non-pregnant cohorts will be clustered together under a “Matched Non-Pregnant Cohort” header row. In the case when a matched non-pregnant cohort does not exist, this level of clustering does not apply. Thus, throughout this section, reference to pregnancy episodes includes matched non-pregnant episodes where applicable.
1.5.1 Output
In studies of medical product use during pregnancy, users specify a list of medical product of interest (MOI) codes using standardized terminologies like NDC or RxNorm. These MOI codes are used to identify and construct MOI episodes during and around pregnant episodes. In QRP Report, users can summarize the number of pregnancy episodes with evidence of an MOI episode, the number of MOI episodes, and the number of times a medical product code was identified (both before and after adjustment for same-day dispensings). Each of the aforementioned quantities is summarized either by trimester or by gestational week of pregnancy. Summary tables output by trimester and MOI are derived from the QRP t4preg and t4nopreg output datasets, while summary tables output by gestational week and MOI are derived from t4preggestwk and t4nopreggestwk datasets. No further stratification (outside of trimester or gestational week) is available in tables, regardless of user-specified stratification in qrp.UserStrataFile.
Regardless of how users choose to stratify the summary tables or what quantity they choose to summarize, summary tables in studies of medical product use during pregnancy will output the pregnancy cohort(s) followed by the matched non-pregnant cohort(s) if applicable. Formatting of header rows before the medical products of interest depends on how the user has chosen to band together cohorts (see User Options). In summary tables stratified by trimester, all header rows for cohorts will include the number of pregnant (or matched non-pregnant) episodes in that particular cohort in parentheses next to the cohort label.
1.5.1.1 Summary Tables by Trimester
For each pregnant cohort in the study, users can summarize four different quantities by MOI and trimester. Table 1.6 below shows the available quantities to summarize and the titles of those tables.
Note that table T1 is the only summary table by trimester that outputs percent in addition to count, since percentages are calculated as the proportion of pregnant episodes with evidence of the MOI in the given period out of the total number of pregnant episodes. Tables T2 through T4 summarize MOI episodes or codes instead of pregnant episodes, so no percentages are output.
Table Reference | Summarized Quantity | Table Title |
---|---|---|
T1 | Number of pregnancy episodes (or matched non-pregnant episodes) with evidence of any medical product of interest episode | Pregnancy Episodes [and Matched Non-Pregnant Episodes] with [ReportTitle] in the [Database] from [StartDate] to [EndDate] |
T2 | Number of medical product of interest episodes | [ReportTitle] Episodes Among Pregnant [and Matched Non-Pregnant] Cohort(s) in the [Database] from [StartDate] to [EndDate] |
T3 | Number of medical product of interest codes | [ReportTitle] Codes Among Pregnant [and Matched Non-Pregnant] Cohort(s) in the [Database] from [StartDate] to [EndDate], without Adjusting for Same-Day Dispensings |
T4 | Number of medical product of interest codes, adjusted for same-day dispensings | [ReportTitle] Codes Among Pregnant [and Matched Non-Pregnant] Cohort(s) in the [Database] from [StartDate] to [EndDate], Adjusting for Same-Day Dispensings |
For each quantity, the table includes nine columns, each defining a different period during or around the pregnancy episode. Table 1.7 in this documentation defines these nine periods.6
Column | Meaning |
---|---|
[EVIDENCE] in the Pre-Pregnancy Period | Number of [QUANTITY] overlapping the user-specified pre-pregnancy period. Use during the pre-pregnancy period does not constitute use during a trimester of pregnancy. |
[EVIDENCE] in Any Trimester | Number of [QUANTITY] overlapping any trimester of pregnancy. |
[EVIDENCE] in the First Trimester | Number of [QUANTITY] overlapping the first trimester of pregnancy, which is defined as the first 97 days after the estimated pregnancy start. |
[EVIDENCE] in the Second Trimester | Number of [QUANTITY] overlapping the second trimester of pregnancy, which is defined as days 98-185 after the estimated pregnancy start. |
[EVIDENCE] in the Third Trimester | Number of [QUANTITY] overlapping the second trimester of pregnancy, which is defined as day 186 after the estimated pregnancy start through the delivery date*. |
[EVIDENCE] in All Trimesters | Number of [QUANTITY] overlapping at least one day in each trimester of pregnancy. |
[EVIDENCE] Only During First Trimester | Number of [QUANTITY] overlapping at least one day in the first trimester of pregnancy, but no days in either the second or third trimesters. |
[EVIDENCE] Only During Second Trimester | Number of [QUANTITY] overlapping at least one day in the second trimester of pregnancy, but no days in either the first or third trimesters. |
[EVIDENCE] Only During Third Trimester | Number of [QUANTITY] overlapping at least one day in the third trimester of pregnancy, but no days in either the first or second trimesters. |
1.5.1.2 Summary Tables by Gestational Week
Whereas users can summarize four different quantities in MOI summary tables by trimester, only two quantities can be summarized by gestational week for each pregnant cohort in the study: number of pregnancy episodes with evidence of an MOI episode and number of MOI episodes.
Tables summarized by gestational week contain one column for each gestational week during which MOI use is assessed, beginning with the first week and ending with the last week an MOI episode is identified. For instance, consider the example shown in 1.2 below where a user specifies that MOIs should be assessed starting 90 days before pregnancy (Week -12), and the longest pregnancy in the cohort is 282 days (Week 41) in duration. If the earliest MOI episode overlapping an assessed period starts in Week -18 before a pregnancy start and the latest MOI episode overlapping an assessed period ends in Week 59 after a pregnancy start, the summary table by gestational week would start in Week -18 and end in Week 41, since assessment ends at delivery.
Each quantity is summarized in its own table, numbered sequentially. Table 1.8 below shows the available quantities to summarize and the titles of those tables. In this table, there will be a footnote indicated after [ReportTitle] in the title that reads “Displayed percentages represent the number of pregnancy (or matched non-pregnant) episodes with evidence of the exposure of interest as a proportion of all pregnancy (or matched non-pregnant) episodes.” When non-pregnant episodes are not displayed in the table, that parenthetical is removed from the footnote. Finally, in the case when [ReportTitle] is not specified in QRP Report Label File parameter LABEL where QRP Report Label File parameter LABELTYPE = ReportTitle, it will default to “Exposure(s) of Interest”.
Table Reference | Summarized Quantity | Table Title |
---|---|---|
T5 | Number of pregnancy episodes (or matched non-pregnant episodes) with evidence of any medical product of interest episode | Pregnancy Episodes [and Matched Non-Pregnant Episodes] with [ReportTitle] in the [Database] from [StartDate] to [EndDate], by Gestational Week |
T6 | Number of medical product of interest episodes | [ReportTitle] Episodes Among Pregnant [and Matched Non-Pregnant] Cohort(s) in the [Database] from [StartDate] to [EndDate] ,by Gestational Week |
Note that table “T5” is the only summary table by gestational week that outputs percent in addition to count, since percentages are calculated as the proportion of pregnant episodes with evidence of the MOI in the given period out of the total number of pregnant episodes.
1.5.2 User Options
Users can leverage qrp_report.TableFile to specify which table(s) to output based on the quantity they are interested in summarizing and whether they’d like the summary output by trimester or gestational week. Users can leverage qrp_report.TableColumnsFile to specify which columns to include in the table(s).
1.5.2.1 Labeling
As in other study designs, users may specify labels for cohorts (both pregnant and non-pregnant) using the Label File parameter LABEL where the Label File parameter LABELTYPE is set to GroupLabel
. In the absence of a user-specified cohort label, the QRP Reporting Tool will populate the group name as specified in QRP.
In addition to labeling cohorts, users may also label MOIs using the Label File parameter LABEL where the Label File parameter LABELTYPE is set to MOILabel
. When no labels are specified, the QRP Reporting Tool will use the cohort and/or MOI names specified in the QRP.
1.5.2.2 Clustering Together Cohorts and MOIs
Users have the option to cluster together pregnant (or matched non-pregnant) cohorts and/or MOIs. There are 3 levels of clustering that users may choose to implement in a summary table for a study of medical product use during pregnancy:
- User-specified cohorts clustered under a header
- When a study of medical product use in pregnancy has defined multiple pregnant cohorts, the user may choose to cluster those cohorts under a combined header by specifying the appropriate parameter in the Label File.
- MOIs clustered together under an MOI header.
- In the case when a user wants to cluster several MOIs under a single header, they may specify which MOIs to group using the Label File parameter LABELVAR and set the MOI Header label using Label File parameter LABEL where LABELTYPE =
MOIHEADER
- In the case when a user wants to cluster several MOIs under a single header, they may specify which MOIs to group using the Label File parameter LABELVAR and set the MOI Header label using Label File parameter LABEL where LABELTYPE =
Figure 1.3 below shows an example of a summary table by trimester where the user has implemented all levels of clustering, whereas Figure 1.4 shows an example of a summary table by gestational week in a study where there is no matched non-pregnant cohort and the user hasn’t clustered any cohorts or MOIs.
1.5.3 Calculations
Counts of pregnancy (or matched non-pregnant) and MOI episodes are taken from the distributed QRP output in msoc.[RUNID]_t4_cida_preg (or correspondingly from msoc.[RUNID]_t4_cida_nopreg) to populate summary tables by trimester and from msoc.[RUNID]_t4_cida_preg_gestwk (or correspondingly from msoc.[RUNID]_t4_cida_nopreg_gestwk) for summary tables by gestational week. A gestational week is calculated relative to the pregnancy start where Week 1 represents Days 0-6 and Week -1 represents Days -7 to -1. There is no “Gestational Week 0.”
In Tables T1 and T5 where percentages are shown, the denominator is the number of pregnancy episodes lasting to that period. In the case of table T1, the percentage corresponds to the number of pregnancy episodes with MOI use at least one day in the given period divided by the total number of pregnancy episodes. The only exception is when the given period is the third trimester, when the denominator is the total number of pregnancy episodes lasting at least one day into the third trimester. In the case of table T5, the percentage corresponds to the number of pregnancy episodes with MOI use at least on day in the given gestational week divided by the total number of pregnancy episodes with at least one day in that gestational week.
[QUANTITY] is a user-specified quantity among: 1. Number of pregnancy episodes (or matched non-pregnant episodes) with evidence of any medical product of interest episode; 2. Number of medical product of interest episodes; 3. Number of medical product of interest codes; or 4. Number of medical product of interest codes, after bridging gaps and adjusting dates for Same-Day Dispensings.
[EVIDENCE] changes based on [QUANTITY]. When [QUANTITY] = 1, [EVIDENCE] = “Use”; when [QUANTITY] = 2, [EVIDENCE] = “Exposure Episodes”; and when [QUANTITY] = 3 or 4, [EVIDENCE] = “Codes”.
Users may specify whether evidence of medical products on the delivery date should constitute an exposure in
qrp.Type4File.REMOVEMOIADATE
.↩︎