When inspecting unit-dose packaged drugs, there would be some possibility o
f inspection errors and ultimately dispensing errors due to pharmacist's in
correct memory of drug identification. Therefore, this study was aimed to e
stablish a computer-aided system for inspecting drugs of unit dose packages
more accurately and efficiently. First, we analyzed the identifiability of
5846 tablets and capsules using drug codes, in order to define an identifi
cation problem for unit dose-packaged drugs. It was shown that as much as 3
6% of the drugs do not have any codes, 4.0% of the code-marked drugs have i
dentical codes with others, and that 9.7 % of the drugs with codes on both
sides of the surface share the same code with other drugs. Thus, it was cle
arly shown that in many cases it is impossible to identify pharmaceuticals
exactly using drug codes only. On the other hand, it was indicated that app
roximately 80% of the drugs which were not identified with drug codes only,
could be identified when additional information on drugs' color, size, for
m, and splitting line was provided. Therefore, in this study, we designed a
n inspection-supporting system to present promptly all the information nece
ssary for the drug identification by displaying real drug images as linked
with prescribing information on the monitor. Retrieval of prescription data
from the order entry host was automatically performed by entering patient'
s ID number or by selecting a patient from the name list of patients at the
inspection terminal computer. Moreover, the system was equipped with abili
ties to automatically check drug interactions and duplicate prescriptions,
to provide various information of prescribed drugs and to monitor patients'
drug history. We used the developed system and evaluated that it is a usef
ul tool for accurate and efficient inspection of unit dose-packaged drugs.
In addition, the quality of drug inspection increased by utilizing system f
unctions such as checking drug interactions and providing drug information.