>>> import pandas as pd
>>> book1 = pd.read_csv('C:\\scratch\\book1.csv')
>>> book2 = pd.read_csv('C:\\scratch\\book2.csv')
>>> book1
0:        ID  tran_num  user_id     ...      sequence provider_id  collections
0   18322    222768        5     ...           3.0          21            1
1   18323    222769        5     ...           3.0          21            1
2   18324    222770        5     ...           3.0          21            1
3   18325    222771        5     ...           3.0          21            1
4   18326    222772        5     ...           3.0           1            1
5   18327    222773        5     ...           NaN           1            1
6   18328    222774        5     ...           3.0           1            1
7   18329    222775        5     ...           3.0           1            1
8   18330    222776        5     ...           3.0           1            1
9   18331    222777        5     ...           3.0           1            1
10  18332    222778        5     ...           3.0           1            1

[11 rows x 13 columns]
>>> book2
1:       ID  service_code ada_code  ...   submit_on  status  notes
0  16179          1201     1201  ...         NaN       I    112
1  16178          1120    D1120  ...         NaN       A    111
2  16177          1112     1112  ...         NaN       I    463
3  16176          1111     1111  ...         NaN       I    432
4  16175          1110    D1110  ...         NaN       A    110
5  16174           999    D0999  ...         NaN       A    492

[6 rows x 10 columns]
>>> comb = book1.merge(book2, on='service_code', how='inner')

---- select all columns ----

>>> comb
3:     ID_x  tran_num  user_id type  ...   taxable_sale  submit_on  status  notes
0  18322    222768        5    S  ...            NaN        NaN       A    110

[1 rows x 22 columns]


---- select only desired columns ----


>>> comb[['service_code','sequence','provider_id','collections','ada_code','description']]
4:    service_code  sequence     ...       ada_code  description
0          1110       3.0     ...          D1110    Prophylax

[1 rows x 6 columns]


---- output to CSV ----

>>> pruned = comb[['service_code','sequence','provider_id','collections','ada_code','description']]
>>> pruned.to_csv('C:\\scratch\\output.csv')