Abstract
The Development of the database system for detecting and preventing drug-drug interactions in Songklanagarind HospitalThis study is aimed to develop the database system and algorithm for detecting drug-drug interactions by starting from Songklanagarind Hospital. Four databases for drug interaction detection were developed using Access 2000 Thai edition (Microsoft() operated on Windows 98, i.e. standard drug name database, trade name database, drug group database, and drug interaction database. Standard drug name database was a two-field table including the names of drugs available in Songklanagarind Hospital and their equivalent standard drug names. Trade name database was also a two-field table consisting of a list of standard drug names, appeared as trade names in the standard drug name database, and their equivalent generic names. The generic names were defined based on TIMS and Songklanagarind Hospital Formulary. All generic names existing in the standard drug name database and trade name database were set in to their corresponding pharmacological groups and collected in the drug group database. Certain generic drugs were arranged in to more than one group. Any drug group was also put in the supergroup if possible. The groups were defined based on a pharmacological textbook. The name of the groups must be similar to those called by the drug interaction references using in database development. The drug interaction database was an eleven-field table involved pairs of interacting drugs, significance (significance level, onset, severity, and documentation), effect, mechanism, management, title and page of reference referred. Knowledge of drug interactions were based on three widely known reference books, i.e. Drug Interaction Facts, Evaluation of Drug Interactions, and Management Clinically Important Drug Interactions. Only the details of effect, mechanism, and management of drug interactions were translated into Thai language. The rest were quoted directly from the references. Algorithm for detection of drug interactions, which was written using structured query language (SQL) for Access, were comprised of two major steps, i.e. i) an expansion of prescription, and ii) an extraction of information. An expansion of prescription involved with relating any drug name in a prescription to the corresponding standard name, generic name, and group name based on the databases. After the expansion, drug interactions were extracted directly by inner-joining the drug interaction table with two copies of expanded list. Then a list of pairs of interacting drugs and related details was shown when an interaction was present. The developed databases and algorithm were tested at Naratiwatrachanagarind Hospital in Naratiwat in order to evaluate that this database system is applicable at the other hospitals. The concept of detection algorithm was similar to the one used at Songklanagarind Hospital. The steps of detection was written using SQL for FoxPro operated on DOS. The testing result showed that the database system was applicable in this hospital.In conclusion, the database system developed originally at Songklanagarind Hospital can be used as a tool to detect drug-drug interactions in hospital. It is possible that the system is applicable for being use among the other hospitals in Thailand.
