The Application of an Institutional Clinical Data Warehouse to the Assessment of Adverse Drug Reactions - Evalutation of Aminoglycoside and Cephalosporin Associated Nephrotoxicity

Q. Zhang, Y. Matsumura, T. Teratani, S. Yoshimoto, T. Mineno, K. Nakagawa, M. Nagahama, H. Takeda.

Department of Integrated Medicine, Medical Informatics, Graduated School of Medicine, Osaka University, Osaka, Japan

S. Kuwata

Medical Informatics Division, Tottori University Hospital, Tottori, Japan

The Application of an Institutional Clinical Data Warehouse to the Assessment of Adverse Drug Reactions - Evaluation of Aminoglycoside and Cephalosporin Associated Nephrotoxicity

This article serves to demonstrate the usefulness of storing clinical information in a data warehouse for the purposes of assessing adverse drug reactions. Aminoglycoside and Cephalosporin associated nephrotoxicity in pediatric inpatients was chosen as a means to exemplify the specific utility of this method of assessment because of Aminoglycoside's high propensity for leading to drug induced renal toxicity and Cephalosporins being listed in the top 25 drug classes frequently associated with adverse drug reactions (ADRs). The pediatric inpatient environment was chosen because most of the previous studies on this subject have utilized adult populations as their samples. The limited existence of studies related to the utility of clinical research is noted by the authors and this article examines one specific way clinical research may be utilized.

The authors of this article indicate that in order not to place clinical response time at risk, separate clinical data warehouses (CDW) need to be established, optimizing data for aggregate analysis so that it may be utilized for the purposes of research. Eligibility for the study conducted by the authors was based on an average length of stay greater than or equal to two days between 01.01.2000 and 12.31.2004 and exposure to the targeted drugs. Lab test results were used to further identify and select patients for the study and sCr results greater than or equal to 1.2 mg\dl at least once during the inpatient admit period was the defined criteria for renal toxicity.

Ultimately 2,060 patients were selected for the study; 1,030 in the group exposed to the targeted drugs and 1,030 in the non-exposed group. The exposure group was sub-categorized into patients exposed to aminoglycoside, patients exposed to cephalosporin and patients exposed to a combination of aminoglycoside and cephalosporin. Within the exposure group, 2.91% of the patients exhibited sCr levels greater than or equal to the determined threshold of 1.2 mg/dl, while in the non-exposure group only 1.17% of the patients had sCr levels above 1.2 mg/dl. The relative risk between the two groups was calculated to be 2.5%. The rate of observed nephrotoxicity within the exposure group was 4.55% in the amimoglycoside group (A), 2.37% in the cephalosporin group (C) and 8.64% in the combination aminoglycoside and cephalosporin group (A+C). Statistically significant differences were observed between the C group and the non-exposure group and within the A+C group and the non-exposure group. While no statistically significant difference was found between the A group and the non-exposure group, this was attributed by the authors to the small size of the A group (22 patients). No statistically significant difference was noted between the A group and the A+C group however, a significant difference was noted between the C group and the A+C group.

The results of this study are consistent with past findings cited by the authors, in that nephrotoxicity is contributed to by the treatment with either cephalosporins alone or in conjunction with aminoglycosides. Due to the volume of prescribed drugs and associated potential toxicity in hospital environments there is a persistent need to assess ADRs. The authors used aminoglycoside and cephalosporin associated toxicity as a specific means of demonstrating the utility of clinical data warehouses in assessing ADRs and found that the relative risk in this instance was 2.04 at a 95% confidence interval it is thought by the authors that this would provide a relevant reference to clinicians.

While several limitations of this study are well noted by the authors, the overall conclusion is that CDWs can be useful in the assessment of ADRs, particularly in Japan where "a trusted third party policy for gathering clinical data from diverse data sources is still in its infancy..."

Reference: 1) The Application of an Institutional Clinical Data Warehouse to the Assessment of Adverse Drug Reactions (ADRs) - Evaluation of Aminoglycoside and Cephalosporin Associated NephrotoxicityQ. Zhang, Y. Matsumura, T. Teratani, S. Yoshimoto, T. Mineno, K. Nakagawa, M. Nagahama, S. Kuwata, H. Takeda Methods of Information in Medicine 2007; 46: 516-522.

Dan Davis