Lead with Nothing
Medical claims extracts will frequently drop the leading zeros from codes as shown in this dataset. It is important to add leading zeros back to your codes before using crosswalks and code sets.
Revenue center codes should always be zero-padded to 4 digits (‘0000’). HCPCS codes can be zero-filled to 5 digits (‘00000’). DRG and bill type codes should be zero-filled to 3 digits (‘000’). Working with ICD codes is trickier (see “International Classification of Disease, Period!”).
When working with an extract that has dropped leading zeros it is critical to retain the ‘.’ (and vice versa). Consider our dataset above; a copy of the original dataset was converted to Excel without maintaining the leading zero (as frequently happens when importing from .csv files). The first row shows an ICD-CM (version 9) diagnosis code value of 320. Since leading zeros and‘.’ have been dropped it is impossible to know whether this value code indicates ‘003.20’ (Localized salmonella infection, unspecified) or ’03.20’ (Faucial diphtheria).
All extracts you work with should include either the ‘.’ or all leading zeros for ICD codes (with the exception of ICD-PCS revision 10). If not, go back to your extract provider and request a new copy.
Dr. Olmsted has a Ph.D. in Economics and has been working with healthcare data for over twenty years for companies including IHCIS (now part of Optum), RTI International, and Health Dialog.