Datetime Variable Creation

[Gregorytus07]

Hi everyone,

I have a dataset containing four datetime variables (RGL_time, RPH_time, RAGAP_time, and RBCARB_time), each indicating when corresponding values (glucose, pH, AGAP, and Bcarb) were resolved. The overall resolution time is determined by the earliest point when glucose and any two other values have been resolved. For instance, if glucose was resolved at 10 am, pH at 2 pm, AGAP at 4 pm, and BCARB at 6 pm on the same day, the overall resolution time would be 4 pm. This is because 4 pm is the earliest time when glucose and any two additional values have been resolved. However, if pH resolved at 2 pm, AGAP at 4 pm, BCARB at 6 pm, and glucose at 8 pm on the same day, the overall resolution time would be 8 pm. This is because glucose must always be included in the values that have resolved. 

I have attached a sample dataset for additional insight into the dataset.

Thanks for your help.

[ballardw]

Many users here don't want to download Excel files because of virus potential, others have such things blocked by security software. Also if you give us Excel we have to create a SAS data set and due to the non-existent constraints on Excel data cells the result we end up with may not have variables of the same type (numeric or character) and even values.

Instructions here: https://communities.sas.com/t5/SAS-Communities-Library/How-to-create-a-data-step-version-of-your-dat... will show how to turn an existing SAS data set into data step code that can be pasted into a forum code box using the </> icon or attached as text to show exactly what you have and that we can test code against.

I read your description a few times and I can't derive actual programmable rules, partially because I can't open XLSX from unknown sources. Examples without rules are very hard to program. Especially with something like medical tests that may be done at all hours. What happens when some of these are performed at 10 or 11 PM and the remainders at 12 or 01 AM? Is the calendar date the "day" for same day purposes?

Are any of these tests repeated on a given "day" (again consider midnight in the discussion).

How do we know if something is resolved (again, I can't look at your spreadsheet to make a guess)?

[Gregorytus07]

@ballardw Thanks for your feedback.

The dataset only contains the first recorded value and the corresponding date and time when each of the lab tests met the criteria for resolution. This is a screenshot of the sample dataset for additional context.

Thanks for your help.

[ballardw]

Since your very first observation has the issue about different calendar dates already asked about for clarification you do have to provide some sort of rule(s) that are used to define "day" with the way you described your need.

Also, I don't see anything related to patient identity so have to assume that each observation of the data set is treated on its own.

[Gregorytus07]

@ballardw Yes, each study_id has only one row of data. For example: study ID 1 had glucose value resolved on 2/28/17 at 7 pm, while its AGAP and Bcarb resolved on 2/27/17 at 7 pm. Therefore, the overall resolution time would be 2/28/17 at 7 pm. Because the overall resolution time is determined by the earliest point when glucose and any two other values have resolved. 

[Tom]

I don't understand what is difficult about this.  Perhaps I have the wrong impression about what your variables mean?

You appear to have 4 datetime variables.

You appear to want the MAX of the GLUCOSE datetime and 2nd smallest other datetime .

want = max(glucose,smallest(2,of testA testb testc));