Hi there,   I have a dataset with medication claims for different patients. If a certain medication is claimed, this shows up in a specific varbiable (yes=1, no=.). Now if within a single patient_id, the wanted variable is "1" once or more, I would like to be it "1" in every single row of that patient_id.   My data looks like this:   data mydata; input patient_id $ medication $ claim; datalines; ID1 medication . ID1 medication . ID1 medicationX 1 ID1 medication . ID2 medication . ID2 medication . ID2 medication . ID3 medicationX 1 ID3 medication . ID3 medicationX 1 ID3 medication . ; run; I would like to obtain a dataset with filled "claim" variable for each patient_id, if "1" is present once or more within a patient_id, that looks like this: data mydata; input patient_id $ medication $ claim; datalines; ID1 medication 1 ID1 medication 1 ID1 medicationX 1 ID1 medication 1 ID2 medication . ID2 medication . ID2 medication . ID3 medicationX 1 ID3 medication 1 ID3 medicationX 1 ID3 medication 1 ; run; I hope it is understandable. Thank you for your help. ... View more

Hello,   I have a dataset with claims of medication, each line / observation is one claim. I have the patient_ID, the date of the claim, and the medication. Now I created a variable, which defines the event (claim of a specific medication x, newuse=1). Here is a sample dataset:   patient_ID date_claim medication newuse 1 17.06.2016 Medication 0 1 30.08.2019 Medication 0 1 30.08.2019 Medication 0 1 23.09.2019 Medication 0 1 23.09.2019 Medication 0 1 23.09.2019 Medication 0 1 23.09.2019 Medication 0 1 24.10.2019 Medication X 1 1 22.11.2019 Medication 0 1 03.03.2020 Medication 0 1 15.07.2020 Medication 0 1 02.03.2021 Medication 0 1 15.11.2021 Medication 0 2 27.01.2017 Medication 0 2 24.08.2018 Medication 0 2 14.09.2018 Medication 0 2 24.09.2018 Medication 0 2 19.10.2018 Medication 0 2 16.11.2018 Medication 0 2 01.02.2019 Medication 0 2 01.02.2019 Medication X 1 2 08.04.2019 Medication 0 2 06.05.2019 Medication 0 2 04.11.2019 Medication 0 2 04.11.2019 Medication 0 2 27.03.2020 Medication 0 2 21.06.2021 Medication 0 2 14.12.2021 Medication 0 3 25.06.2018 Medication 0 3 27.11.2018 Medication 0 3 16.04.2019 Medication 0 3 15.04.2020 Medication 0 3 15.04.2020 Medication 0 3 15.04.2020 Medication 0 3 24.04.2020 Medication 0 3 12.10.2020 Medication 0 3 12.10.2020 Medication 0 3 20.11.2020 Medication 0 3 20.11.2020 Medication X 1   Now I would like to create a new variable, which counts the number of days before and after this claim, i.e. counting number of days backwards and forward from this event (newuse=1) for each patient_ID. How can I do this / which command do I have to apply?   I would be very happy for your help, thank you in advance. ... View more

Hi   I have a simple dataset, which contains several demographic variables such as sex and age of persons.   I would like to display the mean std median p25 p65 p75 p85 and p95 of the variable "age".   I use the proc univariate command, which does not give me p65, p85 and p95 automatically.   Is there an easy way to provide them with proc univariate?     Thank you! ... View more

Hello I am working with claims data from a health insurance. My dataset provides information about every claims from their insured patients, which means patients may appear more than once in the dataset if they have more than one claim. Patients are identified by their patient-ID   I would like to find out how many different patients I have in my dataset, meaning I would like to know how many different patient-IDs I have in my dataset.   If I try  proc freq data=dataset; tables patient_id; run; it just shows me how often a certain patient-ID appears in the dataset, but not how many different patient-IDs there are.   Any suggestions, how I could easily figure this out with a simple command?   Thank you! ... View more