我有一个表，该表的列应该只有数字。但是，源数据可能具有不正确的数据，例如该列中的单词。
因此该列表示为VARCHAR，这样从CSV导入的数据不会崩溃。
我只需要在该列中找到单词，就可以创建一个更新，用正确的数字替换它们。
我开始创建排除某些数字的查询，但是查询的时间太长并且行太多。
SELECT ID,LinkedAccountId,Engagement FROM billing_info WHERE Engagement != '803000006530' AND Engagement != '807000000401' AND Engagement != '403000001956' AND Engagement != '28211756' AND Engagement != '807000000345' AND Engagement != '800000026680' AND Engagement != '807000000351' AND Engagement != '807000000356' AND Engagement != '3000139146' AND Engagement != '807000000394' AND Engagement != '807000000041' AND Engagement != '808000000220' AND Engagement != '807000000396' AND Engagement != '807000000397' AND Engagement != '807000000340' AND Engagement != '807000000398' AND Engagement != '870000000403' AND Engagement != '28200598' AND Engagement != '403000002065' AND Engagement != '807000000398' AND Engagement != '2000549001' AND LinkedAccountId = '12345678919';
数据会像这样返回，我需要继续添加数字以排除：
| 3696724 | 123456789101 | 807000000381 |
| 3696817 | 123456789101 | 800000032764 |
| 3697085 | 123456789101 | complexity |
| 3697088 | 123456789101 | complexity |
| 3697089 | 123456789101 | complexity |
| 3697100 | 123456789101 | complexity |
| 3697107 | 123456789101 | complexity |
| 3697115 | 123456789101 | complexity |
| 3697137 | 123456789101 | complexity |
| 3697155 | 123456789101 | complexity |
| 3697194 | 123456789101 | complexity |
| 3697205 | 123456789101 | 808000000239 |
| 3697236 | 123456789101 | complexity |
+---------+-----------------+--------------+
我只能选择“复杂性”一词，但其中可能还有数百个我不知道的其他词。
如何创建一个查询，在此文本字段中选择仅单词？