In the adnimerge table, when COLPROT indicates ADNI3, the the only records that have a non-missing value for DX are those for which VISCODE equals "bl". Is this as intended? Many of these records seem to have a non-missing value for DIAGNOSIS in the dxsum_pdxconv_adniall table (when matched based on RID and VISCODE2).
Similarly, in the adnimerge table, when COLPROT indicates ADNI3 and COLPROT is not "bl", ADASQ4 is missing. When looking at what I believe to be the matching record in the adas_adnigo23 table, Q4TASK and Q4SCORE are not missing. Why do these values not appear in the adnimerge table?
The R code that creates the ADNIMERGE table is being corrected to take care of the missing baseline diagnoses. That fix should come through in the next day or two. DX.bl will now have the following categories (CN, SMC, EMCI, LMCI, MCI, AD; where MCI is E- or L- as recruited in ADNI3.)
