← SASnRSummarize Age by Treatment Group with Descriptive Statistics

Introduction

Clinical summary tables almost always need descriptive statistics for each treatment group plus a Total column covering all treatments combined.
Standard request: n, nmiss, mean, sd, min, Q1, median, Q3, max — per treatment level and overall.
In SAS we lean on proc summary with a class statement, preloadfmt, and completetypes so every planned level appears in the output even when it has zero rows.
In tidyverse we group_by() the treatment and call summarize(), duplicating records in advance to add the Total band.
In base R we use aggregate() with a custom function, then unpack the matrix column and rename the results.
This lesson summarizes Age for the flagged analysis population (fasfl = "Y") and shows the pattern in all three languages.

Summarize Age by Treatment Group with Descriptive Statistics

SAS (Base SAS)+

data adsl;
  input studyid $ usubjid $ trt01pn age fasfl $ trt01p :$20.;
  datalines;
MYCSG MYCSG-1001 1 23 Y "Dose level 1"
MYCSG MYCSG-1002 3 68 Y "Dose level 3"
MYCSG MYCSG-1003 3 .  Y "Dose level 3"
MYCSG MYCSG-1004 3 35 Y "Dose level 3"
MYCSG MYCSG-1006 3 54 Y "Dose level 3"
MYCSG MYCSG-1007 1 63 N "Dose level 1"
;
run;

proc format;
    value treatment
    1=1
    2=2
    3=3
    4=4
    ;
run;

data adsl02;
    set adsl;
    where fasfl="Y";
    treatment=trt01pn;
    output;
    treatment=4;
    output;
run;

proc summary data=adsl02 nway completetypes;
    class treatment/preloadfmt;
    var  age;
    output out=stats01 n= nmiss= mean= std= min= q1= median= q3= max= /autoname;
    format treatment treatment.;
run;

data adsl02; filters to the flagged rows (where fasfl = "Y") and duplicates each record — once with its real treatment, once with treatment = 4 (the Total band).
proc format; value treatment 1=1 2=2 3=3 4=4; declares that 1-4 are the valid treatment codes; this matters because we ask PROC SUMMARY to use the format later.
proc summary data=adsl02 nway completetypes; class treatment / preloadfmt; tells SAS to include every formatted level in the output, even ones with no rows.
var age; names the analysis variable; the output statement lists the statistics we want.
output out=stats01 n= nmiss= mean= std= min= q1= median= q3= max= / autoname; writes one statistic per column with auto-generated names like age_n, age_mean.
After running, STATS01 has one row per treatment band — 1, 3, and 4 (Total) in this sample; adjust adsl02 or the format if other levels appear in the production data.

R (tidyverse)+

library(tidyverse)
adsl <-
tribble(
~studyid, ~usubjid, ~trt01pn, ~age, ~fasfl, ~trt01p,
"MYCSG", "MYCSG-1001", 1, 23, "Y", "Dose level 1",
"MYCSG", "MYCSG-1002", 3, 68, "Y", "Dose level 3",
"MYCSG", "MYCSG-1003", 3, NA, "Y", "Dose level 3",
"MYCSG", "MYCSG-1004", 3, 35, "Y", "Dose level 3",
"MYCSG", "MYCSG-1006", 3, 54, "Y", "Dose level 3",
"MYCSG", "MYCSG-1007", 1, 63, "N", "Dose level 1"
)

stats01 <- adsl %>%
group_by(trt01pn, trt01p) %>%
summarize(
nrecs = n(),
nmiss = sum(is.na(age)),
n = nrecs - nmiss,
mean = mean(age, na.rm = TRUE),
stddev = sd(age, na.rm = TRUE),
min = min(age, na.rm = TRUE),
q1 = quantile(age, 0.25, type = 2, na.rm = TRUE),
median = median(age, na.rm = TRUE),
q3 = quantile(age, 0.75, type = 2, na.rm = TRUE),
max = max(age, na.rm = TRUE) 
)

adsl is built with tribble() so the sample data reads like a table right in the code.
group_by(trt01pn, trt01p) sets each treatment level as its own group; grouping on both the numeric code and the label keeps both in the final tibble.
summarize() computes the full set in one call: nrecs, nmiss, n, mean, stddev, and the five-number summary.
quantile(age, 0.25, type = 2, na.rm = TRUE) uses type = 2 because that matches the SAS default definition of Q1 and Q3 — the R default (type = 7) would give slightly different values.
na.rm = TRUE lets mean, sd, min, and max skip missing values rather than returning NA for the whole group.
We run and compare row-for-row against SAS STATS01 — numbers should match when the same type 2 quantile definition is in play.

Happy learning!

Search lessons

Introduction

Summarize Age by Treatment Group with Descriptive Statistics