Hi mkeintz, Thank you so much for the help. I got what you said about the CRE filtering. For the criterion, sorry for not making it clear. The criteria are as follows: if we have 2 obs: we keep the first obs and remove the second if it's < 12 months from the first. else (>12 months difference) we keep it. if we have 3 obs: we keep the first and remove the second and last if both are < 12 months from the first (ex 1/1/2018, 1/3/2018, 3/9/2018). Else, we keep the first and last if the last is >12 months from the first and the second is < 12 months from either the first or last such as #344. Else we keep the three of them if the second is > 12 months from the first and >12 months from the last such as # 912.  Hope this is clear now. Thank you. ... View more

Hi mkeintz, Thank you for your reply. I ran the code but it did not work as intended. The result deleted all obs for diseases (Acino and Auris) other than CRE. Also for CRE, there are some obs that should've been deleted but they're still in the result (bold italic). Thank you This is what I got as a result: Obs Id Disease DOB Date12345678910111213 123 CRE 01/08/1961 09/02/2020 344 CRE 02/12/1956 08/06/2019 344 CRE 02/12/1956 03/03/2022 323 CRE 07/01/1993 01/06/2019 323 CRE 07/01/1993 09/06/2020 323 CRE 07/01/1993 09/30/2020 912 CRE 03/01/2012 03/03/2018 912 CRE 03/01/2012 05/06/2019 912 CRE 03/01/2012 09/06/2020 409 CRE 08/07/1987 03/03/2018 409 CRE 08/07/1987 05/06/2019 409 CRE 08/07/1987 09/06/2019 409 CRE 08/07/1987 10/06/2021   ... View more

Sure. Thank you ... View more

Thank you for your reply. Yes, I noticed the dates. It should be 04/30/2020 and 09/30/2020, sorry about that. What I meant is obs with "date" var. which is the testing date. So if the first test was done on 01/02/2019 it will be the first event. Any other test/s for CRE before 01/02/2020 should be removed and the first test will be kept only. If this same person had more tests after 01/02/2020, such as 02/07/2021 obs will be kept. If he had another one on 04/08/2021, we should keep this last test and keep 02/07/2021. Hope this clarify 🙂 ... View more

Hi mkeintz,   Yes, it is a typographical error on my part. But, thinking of what you've mentioned about having matching rules, it definitely should be considered in my second step of merging these data with other sources of data. For the data sample that I've attached, I noticed that the dates are all formatted in a weird way. I tried entering data using datalines but I had similar issues as well! Is there a specific way to enter dates in datalines?   Thank you, Razina ... View more

Hi mkeintz, You have a very eye-opening discussion. I really appreciate it. Here are my answers to the questions you asked: How do you want to treat instances of missing value in collection_date?  You've already implied that you want to keep them.  Should they enter into your subsequent evaluation of date gaps? If a collection date is missing I'm gonna use Episode_date or result_date instead. If all data are missing, the case should be deleted. Also, your sample data seems to be already ordered by collection_date (except when collection_date is missing).  Is this the actual condition of your data?   No, that is not the actual condition of the data. Also, when reading your original post I thought you wanted to remove instances with gaps of LESS THAN 60 days.  But your example suggests the opposite.  Do you want to remove instances of MORE THAN 60 days?     Are you just saying to keep the earliest date plus all dates within 60 days of it (plus missing dates)? Sorry, for the confusion. My fault.  The rules for removing data are: The same patient with two different diseases -no matter what collection_date is, both events will be retained. Same patient with the same disease and different collection_date, the first event will be retained according to the date difference between the two "collection_date" and the "disease": For example, C-Auris is one event for the whole life, so for any difference, only the first case would be retained. For Acinetobacter, on the other hand, if the difference is more than 60 days, both events will be retained (if less the second event will be deleted). For CRE, the difference is more than 90 days instead of 60. If a patient has 3 collection dates, the first event will be retained. The second event will be retained according to the disease rule above. The third event will be retained if its collection_date difference from the second is more than 45 or 60 days, if less, it will be removed.  So what is the rule when you have two separate date ranges of less than 60 days.  I.e., what if your have a given individuals with collection dates   on DAYS 1, 5  and DAYS 101,105. These dates have two gaps of 4 days and 1 gap of 96 da.ys. In this case, only events on days 1 and 105 will be retained. I've attached an excel sheet with a sample of the data. Not sure if this is what you asked for! Thank you     ... View more

Hi mkeintz,   Thank you for the quick reply. I tried the code but it removed the observations which had missing values for the collection_date such as Minnie Mouse obs. What is needed is as follows:  Remove the 4th obs with collection_date 5/11/2021 > 60 days than the first  collection_date of 12/21/2019.   SSN DOB Disease Name Collection_Date 123456789 1/1/1900 Flu John Doe 4/4/2019 123456789 1/1/1900 Flu John Doe 4/5/2019 223344556 2/2/1901 Flu Mickey Mouse 12/21/2019 223344556 2/2/1901 Flu Mickie Mouse 5/11/2021 765432199 2/3/1903 Flu Minnie Mouse     Thank you ... View more