topic Re: Find 3 consecutive "None" values after last not-None value. Flag that value and flag v in SAS Programming

Find 3 consecutive "None" values after last not-None value. Flag that value and flag value above it

Aryyyan — Tue, 25 Jan 2022 21:42:49 GMT

I need help to calculate these 2 flags value.

Flag1 = first day of 3 consecutive days with observed severity “None”. A “Missing” grade in the middle of “None” grades disrupts the sequence and the count starts over at the next “None”. If never any symptoms on-study, or never 3 consecutive days with “None” after a grade > 0, set to 998.

Flag2 = If a symptom was reported with grade > “None”, set to last day a symptom was reported before 3 consecutive “None” days. If no symptom before end of diary, set to "No value".

SUBJECT	covday	SYCD	SYSEV	Flag1	Flag2
715307978	0	ABD	None
715307978	1	ABD	None
715307978	2	ABD	None
715307978	3	ABD	Mild		Y
715307978	4	ABD	None	Y
715307978	5	ABD	None
715307978	6	ABD	None
715307978	7	ABD	None
715307978	8	ABD	None
715307978	9	ABD	None
715307978	0	ANOSMIA	None
715307978	1	ANOSMIA	None
715307978	2	ANOSMIA	None
715307978	3	ANOSMIA	None
715307978	4	ANOSMIA	Mild		Y
715307978	5	ANOSMIA	None	Y
715307978	6	ANOSMIA	None
715307978	7	ANOSMIA	None
715307978	8	ANOSMIA	None
715307978	9	ANOSMIA	None
715307978	0	SOB	None
715307978	1	SOB	None
715307978	2	SOB	None
715307978	3	SOB	None
715307978	4	SOB	None
715307978	5	SOB	None
715307978	6	SOB	None
715307978	7	SOB	None
715307978	8	SOB	None
715307978	9	SOB	None

Re: Find 3 consecutive "None" values after last not-None value. Flag that value and flag v

Patrick — Wed, 26 Jan 2022 00:48:46 GMT

Please provide sample data that covers all the cases you describe. This will help to avoid misunderstandings.

Flag1 = first day of 3 consecutive days with observed severity “None”. A “Missing” grade in the middle of “None” grades disrupts the sequence and the count starts over at the next “None”. If never any symptoms on-study, or never 3 consecutive days with “None” after a grade > 0, set to 998.

Please also provide sample data as a working SAS data step as done below for the data you already posted.

data have;
  infile datalines4
    dlm='|'
    truncover
    dsd;
  input
    subject          : $9.
    covday           : best32.
    sycd             : $7.
    sysev            : $4.
    flag1            : $1.
    flag2            : $1.;
  datalines4;
715307978|0|ABD|None| | 
715307978|1|ABD|None| | 
715307978|2|ABD|None| | 
715307978|3|ABD|Mild| |Y
715307978|4|ABD|None|Y| 
715307978|5|ABD|None| | 
715307978|6|ABD|None| | 
715307978|7|ABD|None| | 
715307978|8|ABD|None| | 
715307978|9|ABD|None| | 
715307978|0|ANOSMIA|None| | 
715307978|1|ANOSMIA|None| | 
715307978|2|ANOSMIA|None| | 
715307978|3|ANOSMIA|None| | 
715307978|4|ANOSMIA|Mild| |Y
715307978|5|ANOSMIA|None|Y| 
715307978|6|ANOSMIA|None| | 
715307978|7|ANOSMIA|None| | 
715307978|8|ANOSMIA|None| | 
715307978|9|ANOSMIA|None| | 
715307978|0|SOB|None| | 
715307978|1|SOB|None| | 
715307978|2|SOB|None| | 
715307978|3|SOB|None| | 
715307978|4|SOB|None| | 
715307978|5|SOB|None| | 
715307978|6|SOB|None| | 
715307978|7|SOB|None| | 
715307978|8|SOB|None| | 
715307978|9|SOB|None| | 
;;;;

Re: Find 3 consecutive "None" values after last not-None value. Flag that value and flag v

Reeza — Wed, 26 Jan 2022 01:17:52 GMT

Do you always have only 9 days? If so, this may be one of the few cases where a wide format would be easier to work with.

Re: Find 3 consecutive "None" values after last not-None value. Flag that value and flag v

Reeza — Wed, 26 Jan 2022 01:21:01 GMT

Also your criteria in the post doesn't match what's in the subject line....which one is it? Assuming subject line as it matches the data more?

Re: Find 3 consecutive "None" values after last not-None value. Flag that value and flag v

Aryyyan — Wed, 26 Jan 2022 04:10:28 GMT

Thank you @Patrick and @Reeza.

In datset below, I need one observation per subject*symptom where

R_T1 = first day of 3 consecutive days with observed severity “None”. A “Missing” grade in the middle of “None” grades disrupts the sequence and the count starts over at the next “None”. If never any symptoms on-study, or never 3 consecutive days with “None” after a grade > 0, set to 998.

For example below, for sycd ABD, I need covday 15 as R_T1, for sycd ANOSMIA covday 17, for COUGH covday 17, and for SOB since there is never 3 consecutive none after Severe so I need value 998.

R_T0 = If a symptom was reported with grade > “None”, set to last day a symptom was reported in [0, R_T1] before 3 consecutive “None” days. If no symptom before end of diary, set to 998.

For example below, for sycd ABD, I need covday 14 as R_T1, for sycd ANOSMIA covday 15, for COUGH covday 18, and for SOB since there is never 3 consecutive none after Severe so I need value 998.

data have;

infile datalines4

dlm='|'

truncover

dsd;

input

subject : $9.

covday : best32.

sycd : $7.

sysev : $8.

sysevn : 1.;

datalines4;

775575874|0|ABD|None|0

775575874|1|ABD|None|0

775575874|2|ABD|None|0

775575874|3|ABD|None|0

775575874|4|ABD|None|0

775575874|5|ABD|None|0

775575874|6|ABD|None|0

775575874|7|ABD|None|0

775575874|8|ABD|None|0

775575874|9|ABD|None|0

775575874|10|ABD|None|0

775575874|11|ABD|None|0

775575874|12|ABD|None|0

775575874|13|ABD|None|0

775575874|14|ABD|Mild|1

775575874|15|ABD|None|0

775575874|16|ABD|None|0

775575874|17|ABD|None|0

775575874|18|ABD|None|0

775575874|19|ABD|None|0

775575874|20|ABD|None|0

775575874|0|ANOSMIA|None|0

775575874|1|ANOSMIA|None|0

775575874|2|ANOSMIA|None|0

775575874|3|ANOSMIA|None|0

775575874|4|ANOSMIA|None|0

775575874|5|ANOSMIA|None|0

775575874|6|ANOSMIA|None|0

775575874|7|ANOSMIA| |.

775575874|8|ANOSMIA|None|0

775575874|9|ANOSMIA|None|0

775575874|10|ANOSMIA|None|0

775575874|11|ANOSMIA|None|0

775575874|12|ANOSMIA|None|0

775575874|13|ANOSMIA|None|0

775575874|14|ANOSMIA| |.

775575874|15|ANOSMIA|Mild|1

775575874|16|ANOSMIA| |.

775575874|17|ANOSMIA|None|0

775575874|18|ANOSMIA|None|0

775575874|19|ANOSMIA|None|0

775575874|20|ANOSMIA|None|0

775575874|0|COUGH|None|0

775575874|1|COUGH|None|0

775575874|2|COUGH|None|0

775575874|3|COUGH|None|0

775575874|4|COUGH|None|0

775575874|5|COUGH|None|0

775575874|6|COUGH|None|0

775575874|7|COUGH|None|0

775575874|8|COUGH|None|0

775575874|9|COUGH|Mild|1

775575874|10|COUGH|Mild|1

775575874|11|COUGH|Mild|1

775575874|12|COUGH|Mild|1

775575874|13|COUGH|Mild|1

775575874|14|COUGH| |.

775575874|15|COUGH|Moderate|2

775575874|16|COUGH| |.

775575874|17|COUGH|Severe|3

775575874|18|COUGH|None|0

775575874|19|COUGH|None|0

775575874|20|COUGH|None|0

775575874|0|SOB|None|0

775575874|1|SOB|None|0

775575874|2|SOB|None|0

775575874|3|SOB|None|0

775575874|4|SOB|None|0

775575874|5|SOB|None|0

775575874|6|SOB|None|0

775575874|7|SOB|None|0

775575874|8|SOB|None|0

775575874|9|SOB|None|0

775575874|10|SOB|Mild|1

775575874|11|SOB|None|0

775575874|12|SOB|Mild|1

775575874|13|SOB|Mild|1

775575874|14|SOB| |.

775575874|15|SOB|Moderate|2

775575874|16|SOB| |.

775575874|17|SOB| |.

775575874|18|SOB|Moderate|2

775575874|19|SOB| |.

775575874|20|SOB|Severe|3

;;;;

 data have;
  infile datalines4
    dlm='|'
    truncover
    dsd;
  input
  subject          : $9.
    covday           : best32.
    sycd             : $7.
    sysev            : $8.
    sysevn			 : 1.;
  datalines4;
775575874|0|ABD|None|0
775575874|1|ABD|None|0
775575874|2|ABD|None|0
775575874|3|ABD|None|0
775575874|4|ABD|None|0
775575874|5|ABD|None|0
775575874|6|ABD|None|0
775575874|7|ABD|None|0
775575874|8|ABD|None|0
775575874|9|ABD|None|0
775575874|10|ABD|None|0
775575874|11|ABD|None|0
775575874|12|ABD|None|0
775575874|13|ABD|None|0
775575874|14|ABD|Mild|1
775575874|15|ABD|None|0
775575874|16|ABD|None|0
775575874|17|ABD|None|0
775575874|18|ABD|None|0
775575874|19|ABD|None|0	
775575874|20|ABD|None|0
775575874|0|ANOSMIA|None|0
775575874|1|ANOSMIA|None|0
775575874|2|ANOSMIA|None|0
775575874|3|ANOSMIA|None|0
775575874|4|ANOSMIA|None|0
775575874|5|ANOSMIA|None|0
775575874|6|ANOSMIA|None|0
775575874|7|ANOSMIA| |.
775575874|8|ANOSMIA|None|0
775575874|9|ANOSMIA|None|0
775575874|10|ANOSMIA|None|0
775575874|11|ANOSMIA|None|0
775575874|12|ANOSMIA|None|0
775575874|13|ANOSMIA|None|0
775575874|14|ANOSMIA| |.
775575874|15|ANOSMIA|Mild|1
775575874|16|ANOSMIA| |.
775575874|17|ANOSMIA|None|0
775575874|18|ANOSMIA|None|0
775575874|19|ANOSMIA|None|0
775575874|20|ANOSMIA|None|0
775575874|0|COUGH|None|0
775575874|1|COUGH|None|0
775575874|2|COUGH|None|0
775575874|3|COUGH|None|0
775575874|4|COUGH|None|0
775575874|5|COUGH|None|0
775575874|6|COUGH|None|0
775575874|7|COUGH|None|0
775575874|8|COUGH|None|0
775575874|9|COUGH|Mild|1
775575874|10|COUGH|Mild|1
775575874|11|COUGH|Mild|1
775575874|12|COUGH|Mild|1
775575874|13|COUGH|Mild|1
775575874|14|COUGH| |.
775575874|15|COUGH|Moderate|2
775575874|16|COUGH| |.
775575874|17|COUGH|Severe|3
775575874|18|COUGH|None|0
775575874|19|COUGH|None|0
775575874|20|COUGH|None|0
775575874|0|SOB|None|0
775575874|1|SOB|None|0
775575874|2|SOB|None|0
775575874|3|SOB|None|0
775575874|4|SOB|None|0
775575874|5|SOB|None|0
775575874|6|SOB|None|0
775575874|7|SOB|None|0
775575874|8|SOB|None|0
775575874|9|SOB|None|0
775575874|10|SOB|Mild|1
775575874|11|SOB|None|0
775575874|12|SOB|Mild|1
775575874|13|SOB|Mild|1
775575874|14|SOB| |.
775575874|15|SOB|Moderate|2
775575874|16|SOB| |.
775575874|17|SOB| |.
775575874|18|SOB|Moderate|2
775575874|19|SOB| |.
775575874|20|SOB|Severe|3
;;;;

 data have;
  infile datalines4
    dlm='|'
    truncover
    dsd;
  input
  subject          : $9.
    covday           : best32.
    sycd             : $7.
    sysev            : $8.
    sysevn			 : 1.;
  datalines4;
775575874|0|ABD|None|0
775575874|1|ABD|None|0
775575874|2|ABD|None|0
775575874|3|ABD|None|0
775575874|4|ABD|None|0
775575874|5|ABD|None|0
775575874|6|ABD|None|0
775575874|7|ABD|None|0
775575874|8|ABD|None|0
775575874|9|ABD|None|0
775575874|10|ABD|None|0
775575874|11|ABD|None|0
775575874|12|ABD|None|0
775575874|13|ABD|None|0
775575874|14|ABD|Mild|1
775575874|15|ABD|None|0
775575874|16|ABD|None|0
775575874|17|ABD|None|0
775575874|18|ABD|None|0
775575874|19|ABD|None|0	
775575874|20|ABD|None|0
775575874|0|ANOSMIA|None|0
775575874|1|ANOSMIA|None|0
775575874|2|ANOSMIA|None|0
775575874|3|ANOSMIA|None|0
775575874|4|ANOSMIA|None|0
775575874|5|ANOSMIA|None|0
775575874|6|ANOSMIA|None|0
775575874|7|ANOSMIA| |.
775575874|8|ANOSMIA|None|0
775575874|9|ANOSMIA|None|0
775575874|10|ANOSMIA|None|0
775575874|11|ANOSMIA|None|0
775575874|12|ANOSMIA|None|0
775575874|13|ANOSMIA|None|0
775575874|14|ANOSMIA| |.
775575874|15|ANOSMIA|Mild|1
775575874|16|ANOSMIA| |.
775575874|17|ANOSMIA|None|0
775575874|18|ANOSMIA|None|0
775575874|19|ANOSMIA|None|0
775575874|20|ANOSMIA|None|0
775575874|0|COUGH|None|0
775575874|1|COUGH|None|0
775575874|2|COUGH|None|0
775575874|3|COUGH|None|0
775575874|4|COUGH|None|0
775575874|5|COUGH|None|0
775575874|6|COUGH|None|0
775575874|7|COUGH|None|0
775575874|8|COUGH|None|0
775575874|9|COUGH|Mild|1
775575874|10|COUGH|Mild|1
775575874|11|COUGH|Mild|1
775575874|12|COUGH|Mild|1
775575874|13|COUGH|Mild|1
775575874|14|COUGH| |.
775575874|15|COUGH|Moderate|2
775575874|16|COUGH| |.
775575874|17|COUGH|Severe|3
775575874|18|COUGH|None|0
775575874|19|COUGH|None|0
775575874|20|COUGH|None|0
775575874|0|SOB|None|0
775575874|1|SOB|None|0
775575874|2|SOB|None|0
775575874|3|SOB|None|0
775575874|4|SOB|None|0
775575874|5|SOB|None|0
775575874|6|SOB|None|0
775575874|7|SOB|None|0
775575874|8|SOB|None|0
775575874|9|SOB|None|0
775575874|10|SOB|Mild|1
775575874|11|SOB|None|0
775575874|12|SOB|Mild|1
775575874|13|SOB|Mild|1
775575874|14|SOB| |.
775575874|15|SOB|Moderate|2
775575874|16|SOB| |.
775575874|17|SOB| |.
775575874|18|SOB|Moderate|2
775575874|19|SOB| |.
775575874|20|SOB|Severe|3
;;;;

 data have;
  infile datalines4
    dlm='|'
    truncover
    dsd;
  input
  subject          : $9.
    covday           : best32.
    sycd             : $7.
    sysev            : $8.
    sysevn			 : 1.;
  datalines4;
775575874|0|ABD|None|0
775575874|1|ABD|None|0
775575874|2|ABD|None|0
775575874|3|ABD|None|0
775575874|4|ABD|None|0
775575874|5|ABD|None|0
775575874|6|ABD|None|0
775575874|7|ABD|None|0
775575874|8|ABD|None|0
775575874|9|ABD|None|0
775575874|10|ABD|None|0
775575874|11|ABD|None|0
775575874|12|ABD|None|0
775575874|13|ABD|None|0
775575874|14|ABD|Mild|1
775575874|15|ABD|None|0
775575874|16|ABD|None|0
775575874|17|ABD|None|0
775575874|18|ABD|None|0
775575874|19|ABD|None|0	
775575874|20|ABD|None|0
775575874|0|ANOSMIA|None|0
775575874|1|ANOSMIA|None|0
775575874|2|ANOSMIA|None|0
775575874|3|ANOSMIA|None|0
775575874|4|ANOSMIA|None|0
775575874|5|ANOSMIA|None|0
775575874|6|ANOSMIA|None|0
775575874|7|ANOSMIA| |.
775575874|8|ANOSMIA|None|0
775575874|9|ANOSMIA|None|0
775575874|10|ANOSMIA|None|0
775575874|11|ANOSMIA|None|0
775575874|12|ANOSMIA|None|0
775575874|13|ANOSMIA|None|0
775575874|14|ANOSMIA| |.
775575874|15|ANOSMIA|Mild|1
775575874|16|ANOSMIA| |.
775575874|17|ANOSMIA|None|0
775575874|18|ANOSMIA|None|0
775575874|19|ANOSMIA|None|0
775575874|20|ANOSMIA|None|0
775575874|0|COUGH|None|0
775575874|1|COUGH|None|0
775575874|2|COUGH|None|0
775575874|3|COUGH|None|0
775575874|4|COUGH|None|0
775575874|5|COUGH|None|0
775575874|6|COUGH|None|0
775575874|7|COUGH|None|0
775575874|8|COUGH|None|0
775575874|9|COUGH|Mild|1
775575874|10|COUGH|Mild|1
775575874|11|COUGH|Mild|1
775575874|12|COUGH|Mild|1
775575874|13|COUGH|Mild|1
775575874|14|COUGH| |.
775575874|15|COUGH|Moderate|2
775575874|16|COUGH| |.
775575874|17|COUGH|Severe|3
775575874|18|COUGH|None|0
775575874|19|COUGH|None|0
775575874|20|COUGH|None|0
775575874|0|SOB|None|0
775575874|1|SOB|None|0
775575874|2|SOB|None|0
775575874|3|SOB|None|0
775575874|4|SOB|None|0
775575874|5|SOB|None|0
775575874|6|SOB|None|0
775575874|7|SOB|None|0
775575874|8|SOB|None|0
775575874|9|SOB|None|0
775575874|10|SOB|Mild|1
775575874|11|SOB|None|0
775575874|12|SOB|Mild|1
775575874|13|SOB|Mild|1
775575874|14|SOB| |.
775575874|15|SOB|Moderate|2
775575874|16|SOB| |.
775575874|17|SOB| |.
775575874|18|SOB|Moderate|2
775575874|19|SOB| |.
775575874|20|SOB|Severe|3
;;;;

Re: Find 3 consecutive "None" values after last not-None value. Flag that value and flag v

Ksharp — Wed, 26 Jan 2022 12:24:36 GMT

Assuming I understand your question.

data have;
  infile datalines4
    dlm='|'
    truncover
    dsd;
  input
    subject          : $9.
    covday           : best32.
    sycd             : $7.
    sysev            : $4.
    flag1            : $1.
    flag2            : $1.;
  datalines4;
715307978|0|ABD|None| | 
715307978|1|ABD|None| | 
715307978|2|ABD|None| | 
715307978|3|ABD|Mild| |Y
715307978|4|ABD|None|Y| 
715307978|5|ABD|None| | 
715307978|6|ABD|None| | 
715307978|7|ABD|None| | 
715307978|8|ABD|None| | 
715307978|9|ABD|None| | 
715307978|0|ANOSMIA|None| | 
715307978|1|ANOSMIA|None| | 
715307978|2|ANOSMIA|None| | 
715307978|3|ANOSMIA|None| | 
715307978|4|ANOSMIA|Mild| |Y
715307978|5|ANOSMIA|None|Y| 
715307978|6|ANOSMIA|None| | 
715307978|7|ANOSMIA|None| | 
715307978|8|ANOSMIA|None| | 
715307978|9|ANOSMIA|None| | 
715307978|0|SOB|None| | 
715307978|1|SOB|None| | 
715307978|2|SOB|None| | 
715307978|3|SOB|None| | 
715307978|4|SOB|None| | 
715307978|5|SOB|None| | 
715307978|6|SOB|None| | 
715307978|7|SOB|None| | 
715307978|8|SOB|None| | 
715307978|9|SOB|None| | 
;;;;



data want;
 merge have have(keep=subject sycd sysev rename=(subject=_s sycd=_c sysev=_sev) firstobs=2);
 if subject=lag(subject) and sysev='None' and lag(sysev) ne 'None' then new_flag1='Y';
 if subject=_s and sycd=_c and _sev='None' and sysev ne 'None' then new_flag2='Y';
drop _:;
run;

Re: Find 3 consecutive "None" values after last not-None value. Flag that value and flag v

Aryyyan — Wed, 26 Jan 2022 15:21:43 GMT

Hi @Ksharp ,

I need one observation per subject*symptom where

R_T1 (Flag1) = first day of 3 consecutive days with observed severity “None”. A “Missing” grade in the middle of “None” grades disrupts the sequence and the count starts over at the next “None”. If never any symptoms on-study, or never 3 consecutive days with “None” after a grade > 0, set to 998.

For example below, for sycd ABD, I need covday 15 as R_T1, for sycd ANOSMIA covday 17, for COUGH covday 17, and for SOB since there is never 3 consecutive none after Severe so I need value 998.

R_T0 (Flag2)= If a symptom was reported with grade > “None”, set to last day a symptom was reported in [0, R_T1] before 3 consecutive “None” days. If no symptom before end of diary, set to 998.

For example below, for sycd ABD, I need covday 14 as R_T1, for sycd ANOSMIA covday 15, for COUGH covday 18, and for SOB since there is never 3 consecutive none after Severe so I need value 998.

Please check new dataset above that I posted.

Thank you for your help.

Re: Find 3 consecutive "None" values after last not-None value. Flag that value and flag v

Ksharp — Thu, 27 Jan 2022 12:59:42 GMT

OK. My bad. You want "3 consecutive days with observed severity “None”. " . Try this :

data have;
  infile datalines4
    dlm='|'
    truncover
    dsd;
  input
    subject          : $9.
    covday           : best32.
    sycd             : $7.
    sysev            : $4.
    flag1            : $1.
    flag2            : $1.;
  datalines4;
715307978|0|ABD|None| | 
715307978|1|ABD|None| | 
715307978|2|ABD|None| | 
715307978|3|ABD|Mild| |Y
715307978|4|ABD|None|Y| 
715307978|5|ABD|None| | 
715307978|6|ABD|None| | 
715307978|7|ABD|None| | 
715307978|8|ABD|None| | 
715307978|9|ABD|None| | 
715307978|0|ANOSMIA|None| | 
715307978|1|ANOSMIA|None| | 
715307978|2|ANOSMIA|None| | 
715307978|3|ANOSMIA|None| | 
715307978|4|ANOSMIA|Mild| |Y
715307978|5|ANOSMIA|None|Y| 
715307978|6|ANOSMIA|None| | 
715307978|7|ANOSMIA|None| | 
715307978|8|ANOSMIA|None| | 
715307978|9|ANOSMIA|None| | 
715307978|0|SOB|None| | 
715307978|1|SOB|None| | 
715307978|2|SOB|None| | 
715307978|3|SOB|None| | 
715307978|4|SOB|None| | 
715307978|5|SOB|None| | 
715307978|6|SOB|None| | 
715307978|7|SOB|None| | 
715307978|8|SOB|None| | 
715307978|9|SOB|None| | 
;;;;


data _have;
n=0;
do  until(last.sysev);
 set have;
 by subject sycd sysev  notsorted;
 n+1;
end;
do until(last.sysev);
 set have;
 by  subject sycd sysev notsorted;
 output;
end;
run;


data want;
 merge _have _have(keep=subject sycd sysev n rename=(subject=_s sycd=_c sysev=_sev n=_n) firstobs=2);
 if subject=lag(subject) and sysev='None' and lag(sysev) ne 'None' and n>3 then new_flag1='Y';
 if subject=_s and sycd=_c and _sev='None' and sysev ne 'None'  and _n>3 then new_flag2='Y';
drop _: n;
run;

Re: Find 3 consecutive "None" values after last not-None value. Flag that value and flag v

Aryyyan — Thu, 27 Jan 2022 16:46:53 GMT

Thank you @Ksharp. It is very helpful but I want "3 consecutive days with observed severity “None” after "Last Not-None (Missing, Mild, Moderate, or Severe)" severity.

Re: Find 3 consecutive "None" values after last not-None value. Flag that value and flag v

Reeza — Thu, 27 Jan 2022 17:33:59 GMT

@Aryyyan wrote:

Thank you @Ksharp. It is very helpful but I want "3 consecutive days with observed severity “None” after "Last Not-None (Missing, Mild, Moderate, or Severe)" severity.

As I noted a few days ago, that criteria is in your subject line but you weren't including it in your actual requirements.

Re: Find 3 consecutive "None" values after last not-None value. Flag that value and flag v

Ksharp — Fri, 28 Jan 2022 12:26:38 GMT

OK . You want this ?

data have;
  infile datalines4
    dlm='|'
    truncover
    dsd;
  input
    subject          : $9.
    covday           : best32.
    sycd             : $7.
    sysev            : $4.
    flag1            : $1.
    flag2            : $1.;
  datalines4;
715307978|0|ABD|None| | 
715307978|1|ABD|None| | 
715307978|2|ABD|None| | 
715307978|3|ABD|Mild| |Y
715307978|4|ABD|None|Y| 
715307978|5|ABD|None| | 
715307978|6|ABD|None| | 
715307978|7|ABD|None| | 
715307978|8|ABD|None| | 
715307978|9|ABD|None| | 
715307978|0|ANOSMIA|None| | 
715307978|1|ANOSMIA|None| | 
715307978|2|ANOSMIA|None| | 
715307978|3|ANOSMIA|None| | 
715307978|4|ANOSMIA|Mild| |Y
715307978|5|ANOSMIA|None|Y| 
715307978|6|ANOSMIA|None| | 
715307978|7|ANOSMIA|None| | 
715307978|8|ANOSMIA|None| | 
715307978|9|ANOSMIA|None| | 
715307978|10|ANOSMIA|Mild| |Y
715307978|11|ANOSMIA|None|Y| 
715307978|12|ANOSMIA|None| | 
715307978|13|ANOSMIA|None| | 
715307978|0|SOB|None| | 
715307978|1|SOB|None| | 
715307978|2|SOB|None| | 
715307978|3|SOB|None| | 
715307978|4|SOB|None| | 
715307978|5|SOB|None| | 
715307978|6|SOB|None| | 
715307978|7|SOB|None| | 
715307978|8|SOB|None| | 
715307978|9|SOB|None| | 
;;;;


data _have;
n=0;
do  until(last.sysev);
 set have;
 by subject sycd sysev  notsorted;
 n+1;
 if last.sycd then last_group=1;
end;
do until(last.sysev);
 set have;
 by  subject sycd sysev notsorted;
 output;
end;
run;


data want;
 merge _have _have(keep=subject sycd sysev n last_group
 rename=(subject=_s sycd=_c sysev=_sev n=_n last_group=_last_group) firstobs=2);
 if subject=lag(subject) and sysev='None' and lag(sysev) ne 'None' and n>2 and last_group then new_flag1='Y';
 if subject=_s and sycd=_c and _sev='None' and sysev ne 'None'  and _n>2 and _last_group  then new_flag2='Y';
drop _: n last_group;
run;

Re: Find 3 consecutive "None" values after last not-None value. Flag that value and flag v

Aryyyan — Fri, 28 Jan 2022 17:40:36 GMT

Yup, thank you so much for your help!!!