Technical note: Comparison between three methods for evaluating Clinical Mastitis
frequency in dairy cows: cow-level risk, cow level-rate and quarter-level rate

SAS v. 9.4 (SAS Institute Inc., Cary, NC, USA). Presentation of Code

**The %macro statement tells SAS that this is the beginning of the macro**;
%macro risk(year=, beginmonth=, endmonth=);
%do month = &beginmonth %to &endmonth;

**Creating new variable**;
title "data &month &year";
data nrate_&month._&year;
set work.nrate1;
if month NE &month then delete;
if year NE &year then delete;
if dim_mast >=0 then status_mastitis=1;
else if dim_mast='.' then status_mastitis=0;
run;

**Change date format**;
data nrate_&month._&year;
set nrate_&month._&year;
date1=input(date_mast,mmddyy10.);
format date1 date9.;
drop date_mast;

**Selecting beginning and end date for each month**;
if month = 5 and year=2016 then date_mast_begin = mdy(&month,16,&year);
else date_mast_begin = mdy(&month,1,&year);
if month =2 then date_mast_end = mdy(&month,28,&year);
else if month in (4,6,9,11)then date_mast_end = mdy(&month,30,&year);
else if month in (1,3,5,7,8,10,12) then date_mast_end = mdy(&month,31,&year);
rename date1 = date_mast;
run;

**Creating table 1 (from Figure 1)**;
proc sql;
create table rate_mast1 as
select id, date_mast , max_date_dry_or_cull, month, year
from nrate_&month._&year
where (date_mast BETWEEN date_mast_begin AND date_mast_end);
quit;

**Creating table 2 (from Figure 2)**;
data plusdays (keep = id date_mast max_date_dry_or_cull);
set rate_mast1;
ft_days= intnx('day', date_mast,14);
run;

proc print data=plusdays;
format date_mast max_date_dry_or_cull date10.;
run;

proc sql;
create table days_at_risk1 as
select
id,
min(date_mast) as CM_date,
max(max_date_dry_or_cull) as Exit_date,
max(max_date_dry_or_cull) - min(date_mast) as Length_of_stay
from plusdays
group by id;
quit;

proc print data=days_at_risk1;
format CM_date Exit_date date10.;
run;


title "Quarters affected in &month &year";
proc sql;
create table have as
select id, date_mast, MQ_1,MQ_2,MQ_3,MQ_4, pathog_1MQ, pathog_2MQ,
pathog_3MQ, pathog_4MQ
from nrate_&month._&year
;
quit;


data need1;
set have;
quarters_affected = sum((TRIM(LEFT(mq_1)) ne '.'),(TRIM(LEFT(mq_2))  ne
'.'),(TRIM(LEFT(mq_3))  ne  '.'),(TRIM(LEFT(mq_4))  ne '.'));
run;

proc print data=need1;
run;

**Creating table 3 (from Figure 2); Frequency quarters**;
proc freq data=need1;
tables quarters_affected;
run;

data nrate_&month._&year._dar;
set rate_mast1;
date1 = date_mast;
if month in (4) then date2 = mdy(&month,30,&year);
else if  month in (1,3,5)then date2 = mdy(&month,31,&year);
else if month = 2 then date2 = mdy(&month,28,&year);
no_of_days  = intck ('days', date1, date2);
format date1 date2 date10.;
run;

proc print data = nrate_&month._&year._dar;
run;

proc print data=rate_mast1;
run;

proc freq data=nrate_&month._&year._dar;
tables no_of_days;
run;

proc freq data=days_at_risk1;
tables length_of_stay;
run;

*Same code BUT healthy cows*;
data nrate_&month._&year._healthycows;
set nrate_&month._&year;
if status_mastitis=1 then delete;
run;

**Creating table 4 (From Figure 1)**;
proc sql;
create table days_at_risk_healthy as
select
id,
min(calv_date) as Calving_date,
max(max_date_dry_or_cull) as Exit_date,
max(max_date_dry_or_cull) - min(calv_date) as Length_of_stay
from nrate_&month._&year._healthycows
group by id;
quit;

proc print data=days_at_risk_healthy;
format Calving_date Exit_date date10.;
run;

**The %mend statement signals the end of the macro**;
%end;
%mend risk;
%risk(year=2016, beginmonth=5, endmonth=12);
%risk(year=2017, beginmonth=1, endmonth=5);