Adverse drug reactions related hospitalization identified by discharge ICD-9 codes in a univeristy hospital in Riyadh

OBJECTIVE: To assess the feasibility of using International Classification of Disease code (ICD-9) to ascertain the prevalence, seriousness, and preventability of adverse drug reaction (ADRs).

METHODS: A retrospective study between the years 1982 and 2005 was conducted at King Khaled University Hospital (KKUH), Riyadh, Saudi Arabia to examine the ICD-9 codes assigned on discharge to identify ADRs. A list of the ICD-9 codes related to ADRs were identified. These codes were entered into the hospital computer program at the study site to identify corresponding patients' medical records. The total number of patients admitted to the hospital each year was identified to calculate the prevalence of ADRs, and descriptive analysis was also conducted.

RESULTS: A total of 89 patients were identified and included. Drug classes commonly associated with ADR include hormones and synthetic substitutes (14.6%), followed by primary systemic agents (13.5%). Almost 50% of cases had chronic conditions and use other drugs when the ADR had occurred. The majority of the ADRs were type A (54%) preventable, while 39% were type B non preventable, and only 6% were type C occur with long term use. The prevalence per year ranged from 0.07% in 1993 to 0.003% in 1999.

CONCLUSION: Identifying ADRs causing hospital admission by using ICD-9 coding system is easy and practical. However, under or inaccurate recording of ICD-9 codes may be a limitation to the use of such an important tool.