Thanks so much, everyone, But so far, no go. The mkeintz novinosrin solutions deleted appropriate records for any patient who had had a positive test, and also all records for patients who had not had any positive test. Those latter records need to be maintained in the dataset. The hhinohar solution yielded a dataset with 33 records with missing values for every variable on every record. There was a problem reading in PATIENT_ID. I'll look more into it. The solution, using the data provided, should delete the 18 records marked with "delete," yielding a dataset with 82 records. It should maintain all the records for patient ID that never had a positive result, as well as the records with the positive results, and the records for patients with a positive result for which the specimen was collected before the date of the positive result. Thanks again for you efforts!
... View more
