There's no standard way to do this. It's just important to be clear what you mean, and how the error was calculated. In general your error should have the same precision as your value.

For instance, if the value you obtained is 15.0 mg with an error of 10.0%, you could write

15.0 ± 1.5 mg

Or you could just write 15.0 mg with an error of 10.0%. But be advised that this can be ambiguous, because do you mean there is a total variation of 10% (5% above and 5% below) or do you mean a total variation of 20% with 10% above and 10% below the mean. One way I do NOT like is 15.0 mg ± 10%, simply because this can be ambiguous also at times; for example if your value has a unit of percent (e.g., % crystallinity).

Although I wrote it above, it bears repeating that you need to define what "% error" even means, and how many replicates you calculated it over.