There is no coding convention in ICD-9 that states that you automatically substitute "chronic" for not specified. I also checked AHA Coding Clinic and Faye Brown's ICD-9, just in case I missed anything.
This is one of those urban legends, I think. That's why you always look up your term in the alpha index first. If the documentation doesn't state acute or chronic, you'd go with your NOS code, and then look at the tabular index. That certainly may state 'chronic', and would be appropriate for a NOS condition only if it's part of the code description, but I would not code either acute or chronic from the alpha index only, and certainly not if the documentation is non-specific.
For example, Hepatitis NOS is 573.3. If you look that up in your alpha, it falls directly next to the Hepatitis heading (as do all NOS codes, typically). In looking at the tabular index, you would read, Hepatitis, unspecified. If, because there was no notation of 'chronic' or 'acute' by the provider, you went directly to chronic, you'd append a different code--571.41, which would be incorrect based on the coding guidelines.
A different example is Sinusitis. The modifiers in parentheses following the alpha heading include Chronic, which if you read the code, ends in .9 (usually indicative of a NOS code) When you read in your tabular index, you see that although sinusitis 473.9 is defined as 'chronic', it's also appropriate to append that code for unspecified sinusitis.
This is why you can't automatically assume a coding convention will work for every scenario without reviewing both the alpha and the tabular indices. It's common for a NOS condition to also be the same code as 'chronic', but I would never code a NOS condition as acute.
Hope this helps.