Professor Friedman at confrontationright blog has an illustration of why it may be important to have the “real” analyst at trial:
The defendant was charged with possession with intent to distribute a quantity of methamphetamine of 4 grams or more but less than 200 grams. In Texas, this is a first-degree felony with a punishment range of 5 to 99 years or life and a fine up to $10,000.00. The official lab report, signed by the analyst who performed the actual test, reflected the presence in the tested substance of methamphetamine with an aggregate weight of 4.51 grams. The analyst testified at trial. Using his personal notes to refresh his memory, he testified that the aggregate weight of the methamphetamine was 1.51 grams. The prosecutor then asked the analyst to review the lab report. After a pause the analyst testified that report was in error and that his notes reflected the correct amount. (Apparently, a clerk prepared the report from the notes of the analyst and made the error.)
This was a happy result for the accused, but not so for a co-accused who’d plead guilty on the basis of the same report.
Of course we are unlikely to find similar situations at USACIL or military drug labs (I had my fingers crossed when I wrote that sentence) – I expect the rapture soon.