pandas.io.gbq.to_gbq creates duplicate rows, pandas==0.19.2

[existeundelta]

df.to_gbq(data, dataset, chunksize=10000, verbose=False, reauth=False, if_exists='append', private_key=private_key)

Problem description

We only insert DataFrame one time
It seems is in insert because we also see duplicates in Bigquery WebUIConsole
can be related with #5840?

pandas==0.19.2

[jreback]

w/o a reproducible example this is not useful

if you would like to add an example great