About SAS_USER_928

SAS_USER_928

Hi Tom, Yes, my data is currently sorted by SITE and START_TIME. Also, if I have two records from the same PT and with the same DATEPART(START_TIME), I am keeping both but do not need to calculate their time difference (i.e., time difference is to be calculated only if PTs are different and have the same DATEPART(START_TIME) and SITE). Thanks

SAS_USER_928

Thanks, Tom. I still see values populated for each observation which is not required. I want to see the time difference ONLY if the datepart(start time) for the two PTs at a site is exactly the same. Thanks

SAS_USER_928

Hi there, I have the same dataset as above but with a minor change in the requirements. I want to calculate: 1. the difference (in minutes) between start times (ONLY of the observations that have the same Start Dates) per site. This will be saved as Time1 (as above). Goal is to know how closely/far apart PTs at each site are opening their forms on each day (avg. difference). 2. the difference (in minutes) between previous observation's save time and current observation's start time (ONLY of the observations that have the same Start Dates) per site. This will be saved as Time2 as above. Goal is to know how closely/far apart PTs at each site are saving and opening their forms on each day (avg. difference). Note: if there are no two observations with same Start Time at a site, time1 and time2 will be blank for both the observations. Thanking you in advance for the help.

SAS_USER_928 · ‎05-03-2024

Thank you so much, Tom. Much appreciate it.

SAS_USER_928 · ‎05-03-2024

Yes, I like the HOURS to include fraction of an hour, such as 23.5 for 23 hours 30 seconds. So which line in the above code should be modified to address this? Thanks

SAS_USER_928 · ‎05-03-2024

Thanks, Tom. The code seems to work fine but a few things to add or needs modification for me to use it. You have used datalines but my data set is in sas7bdat format. Also the two data variables are in DATETIME20. format. The difference in your code below is in hours:minutes:seconds format. I need it numeric such as 25 or 30 so that I can use it further to calculate average of hours per site. Thank so much in advance for your time and help.

SAS_USER_928 · ‎05-03-2024

Hi, My data set is in a sas7bdat format and the two date variables are in DATETIME20. The entire data set is already pre-processed (cleaned and sorted by Site, PT, and Start Time). The only two things I want to do are: 1. to calculate the difference in hours between the current start and previous save time for the same patient. and 2. to calculate the difference in hours between the current save and previous save time for the same patient. I already have the following code but it is not populating some values correctly. DATA TIME_GAP; SET FORM; BY SITE PT START_TIME; /* Initialize below variables */ IF FIRST.SITE OR FIRST.PT THEN DO; PREV_START_TIME = .; PREV_SAVE_TIME = .; END; ELSE DO; PREV_PT = LAG(PT); PREV_SAVE_TIME = LAG(SAVE_TIME); FORMAT PREV_SAVE_TIME DATETIME20.; PREV_START_TIME = LAG(START_TIME); FORMAT PREV_START_TIME DATETIME20.; IF NOT MISSING(PREV_SAVE_TIME) AND PT = PREV_PT THEN START_SAVE_DIFF = ROUND(((START_TIME - PREV_SAVE_TIME) / 3600), 0.01); ELSE START_SAVE_DIFF = ""; IF NOT MISSING(PREV_START_TIME) AND PT = PREV_PT THEN START_START_DIFF = ROUND(((START_TIME - PREV_START_TIME) / 3600), 0.01); ELSE START_START_DIFF = ""; END; RUN; Thanks

SAS_USER_928 · ‎05-02-2024

Hi, I need some help with writing a SAS code to handle the following scenario. I have a data set that has multiple sites; each site has multiple patients; and each patient can have data of up to 7 days (one row represents one day). The data set has form start and save time. Data set is also setup to have no duplicates. First day data of each patient should always have the previous save and open time (both are user defined variables) as NULL. e.g., sr.1, 8, and 9 will have previous open and save time variables as NULL. And using these variables, I want to calculate the the following for each patient at a site: number of hours a patient took between today's and yesterday's start times (i.e., from today's start to previous start) and number of hours a patient took between today's open and yesterday's start times (i.e., from today's start to previous day save) Sr. PT SITE START_TIME SAVE_TIME 1 001 01 4/9/2024 16:44 4/9/2024 16:46 2 001 01 4/10/2024 16:39 4/10/2024 16:40 3 001 01 4/11/2024 16:07 4/11/2024 16:08 4 001 01 4/12/2024 16:05 4/12/2024 16:06 5 001 01 4/13/2024 20:05 4/13/2024 20:06 6 001 01 4/14/2024 19:31 4/14/2024 19:52 7 001 01 4/15/2024 16:25 4/15/2024 16:25 8 002 01 4/15/2024 16:19 4/15/2024 16:20 9 003 01 4/15/2024 16:21 4/15/2024 16:22 10 002 02 4/3/2024 18:34 4/3/2024 18:41 11 002 02 4/4/2024 18:03 4/4/2024 18:09 12 002 02 4/6/2024 21:11 4/6/2024 21:14 13 002 02 4/6/2024 21:14 4/6/2024 21:16 14 002 02 4/7/2024 18:17 4/7/2024 18:20 15 002 02 4/8/2024 18:12 4/8/2024 18:14 16 002 02 4/9/2024 18:06 4/9/2024 18:08 17 004 02 4/5/2024 18:03 4/5/2024 18:10 18 004 02 4/6/2024 16:27 4/6/2024 16:32 19 004 02 4/7/2024 19:18 4/7/2024 19:22 20 004 02 4/8/2024 18:01 4/8/2024 18:07 21 004 02 4/9/2024 18:06 4/9/2024 18:16 22 004 02 4/10/2024 18:08 4/10/2024 18:10 23 004 02 4/11/2024 19:22 4/11/2024 19:26 24 005 02 4/8/2024 18:05 4/8/2024 18:07 25 005 02 4/9/2024 19:32 4/9/2024 19:34 26 005 02 4/10/2024 18:45 4/10/2024 18:47 27 005 02 4/11/2024 18:02 4/11/2024 18:04 28 005 02 4/12/2024 18:04 4/12/2024 18:07 29 005 02 4/13/2024 18:01 4/13/2024 18:03 30 005 02 4/14/2024 18:06 4/14/2024 18:08 Thanks so much in advance for all the help.

SAS_USER_928 · ‎03-01-2022

Thank you @PaigeMiller. There is only 1 scenario left out. See PT3 which has only one observation and I do not want to exclude that and want the Duration to be calculated as Extraction Date - (Lab) Date to see if the lab is past-due. Note: Extraction Date is hard-code as 2022-03-01. Thank you

SAS_USER_928 · ‎03-01-2022

Hi @PaigeMiller, This time I have a different situation and thought if you could please help... using SAS or Proc SQL. I have a list of pts. that have to complete a lab every N (say between 4 to 5) weeks and outside of this window will be considered NON COMPLIANT. Here is the list of PTs and their lab dates: PT Date Duration in Weeks 1 1/5/2021 1 2/24/2021 1 4/7/2021 1 5/19/2021 2 1/14/2021 2 3/2/2021 3 2/25/2022 My data is sorted by PT and Date in ascending order. For PT1, I need to take the difference between every two labs and calculate the difference (duration) in weeks. If there is only 1 lab performed so far (as is the case with PT3) then use today's date to calculate the duration in weeks. Thanking you in advance for your help.

SAS_USER_928 · ‎02-22-2022

Thank you so much both of you for your prompt responses. Actually both solutions worked for me and wanted to mark both as accepted solutions but the platform did not let me do so. Again, thank you so so much.

SAS_USER_928 · ‎02-21-2022

Hi, I have the following scenario and am stuck with creating the output. A patient is required to perform one of two labs per visit. Therefore, I have 2 observations per patient per visit, one as DONE and other as NOT DONE, in the input data set. I'm trying to flag patients' lab status per visit based on whether the lab was performed or not. If only one lab was DONE (at any visit) then flag it as DONE, else NOT DONE. Therefore, The output data set will have only 1 observation per patient per visit flagged either as DONE or NOT DONE depending on the above criteria. E.g., below is the input data set with 2 patients: PatientID VisitID Date LabStatus A 1 1/1/2022 Done A 1 1/1/2022 Not Done A 2 1/5/2022 Done A 2 1/5/2022 Not Done B 1 1/15/2022 Done B 1 1/15/2022 Not Done B 2 1/20/2022 Not Done B 2 1/20/2022 Not Done OUTPUT Data Set: PatientID VisitID Flag A 1 Done A 2 Done B 1 Done B 2 Not Done Thank you all for the help.

SAS_USER_928 · ‎11-26-2021

Hi Tom, I tried that as well already but for some reason it did not give me correct results, e.g., formatted difference 11:58 was displayed as 11.96 instead of 11.58. Thanks

SAS_USER_928 · ‎11-26-2021

Hi @Astounding , I am having the same issue as well. Tried below solution but could not make it to work for my case. I have only two variables, DOSE_DT_TIME and PT. Every PT is supposed to take 2 doses daily with 12 hours difference. Paige Miller helped to calculate the difference between every two doses per PT and the difference is perfect. So with the above two original variables I have two additional variables now: Difference between the two dates formatted as TIME and non-formatted as difference in SECONDS (i.e., 12:09:00 and 43740). I need a new variable to calculate the difference in hours as number (i.e., 12.09 as 12 hours and 9 minutes). Last column below is calculating wrong number of hours. Any help will be greatly appreciated. Thank you so much. PT DOSE_DT_TIME DIFF_FRMTD DIFF_IN_SEC DIFF_IN_HRS 1 10/1/2021 21:52 12:52:00 46320 12 1 10/2/2021 10:30 12:38:00 45480 12 1 10/2/2021 22:36 12:06:00 43560 12 1 10/3/2021 10:34 11:58:00 43080 11 1 10/3/2021 21:42 11:08:00 40080 11 2 10/2/2021 11:25 2 10/2/2021 20:00 8:35:00 30900 8 2 10/3/2021 8:00 12:00:00 43200 12 2 10/3/2021 20:00 12:00:00 43200 12 2 10/4/2021 8:00 12:00:00 43200 12 2 10/4/2021 20:00 12:00:00 43200 12 Her is part of the code. DIFF_IN_SEC = DOSE_DT_TIME- LAG(DOSE_DT_TIME); DIFF_IN_HRS = INT(DIFF_IN_SEC /3600) + 0.01 * MOD(DIFF_IN_SEC , 60);

SAS_USER_928 · ‎11-26-2021

Hi @PaigeMiller , No not in SAS application. I create a data set which I then import to a third party application (part of my job). They provide their own set of tests based on the types of variables. Mean was disabled for this character TIME formatted variable. That's the reason I need a numeric variable to hold the number of hours (as number and not as time). Any help in the matter will be greatly appreciated. I am trying to get it to work but failing miserably.

Online Status	Offline
Date Last Visited	Friday

Re: How to flag first observation of a Patient as NULL?

Re: How to flag first observation of a Patient as NULL?

Re: How to flag first observation of a Patient as NULL?

Re: How to flag first observation of a Patient as NULL?

Re: How to flag first observation of a Patient as NULL?

Re: How to flag first observation of a Patient as NULL?

Re: How to flag first observation of a Patient as NULL?

How to flag first observation of a Patient as NULL?

Re: How to calculate the time difference between two observations of t...

Re: How to calculate the time difference between two observations of t...

Re: How to flag a variable based on a condition

Re: How to flag a variable based on a condition

Re: How to calculate the time difference between two observations of t...

Re: How to flag a variable based on a condition

Re: How to flag first observation of a Patient as NULL?

Re: How to flag first observation of a Patient as NULL?

Re: How to flag first observation of a Patient as NULL?

Re: How to flag first observation of a Patient as NULL?

Re: How to flag first observation of a Patient as NULL?

Re: How to flag first observation of a Patient as NULL?

Re: How to flag first observation of a Patient as NULL?

How to flag first observation of a Patient as NULL?

Re: How to calculate the time difference between two observations of t...

Re: How to calculate the time difference between two observations of t...

Re: How to flag a variable based on a condition

How to flag a variable based on a condition

Re: Convert time format to integer value

Re: Convert time format to integer value

Re: How to calculate the time difference between two observations of t...