Program reduces medication errors during MR and CT procedures

May 27, 2009

Medication errors may arise less often in a busy hospital radiology department than in other inpatient services, but they can cause more serious damage when they do happen. Radiologists at Boston-based Beth Israel Deaconess Medical Center came to these conclusions after evaluating 27 months of high-tech medical imaging experience. They used their data to identify the causes of errors and devise strategies to address them.

Medication errors may arise less often in a busy hospital radiology department than in other inpatient services, but they can cause more serious damage when they do happen. Radiologists at Boston-based Beth Israel Deaconess Medical Center came to these conclusions after evaluating 27 months of high-tech medical imaging experience. They used their data to identify the causes of errors and devise strategies to address them.

The effort is part of a hospital-wide effort to eliminate medication-related mistakes by 2012 and to comply with a Joint Commission on Accreditation of Healthcare Organizations program aimed at medical error reduction.

Chief investigator Dr. Gerwin Williams discovered from his computer-assisted evaluation of 38,268 contrast-enhanced CT and 26,888 enhanced MRI procedures that from one perspective, the radiology department had performed well. He uncovered 286 medication-related errors that arose during 0.4% of the studies. By comparison, the 2006 edition of the United States Pharmacopeia reported results from a five-year study showing a 5% medication error rate in radiology departments.

But from another angle, the Beth Israel Deaconess experience raised concerns. The harm rate of CT- and MR-related medication errors was about 6%. Fifteen (5%) of the medication-related errors involved temporary harm requiring hospitalization, and four (1%) cases may have caused permanent harm, Williams said in a presentation at the 2008 RSNA meeting.

The rate was highest among incidents that did not involve a contrast agent. Only two cases in this category were identified, but both involved possible permanent damage.

Williams noted that radiology's 6% overall harm rate was half the 12% rate predicted by the Pharmacopeia, and was generally in the range of harm rated for other types of medical departments reported in the medical literature. A Canadian study involving all departments at four hospitals, for example, came up with 7.5% medication-adverse event rate (CMAJ 2004;170[11]:1678), though Williams warned that department-to-department comparisons are problematic because of differing levels of care.

Four types of problems were classified for the 258 intravenous contrast-related medication errors at Beth Israel Deaconess:
• 119 CT contrast extravasations, with 46 requiring plastic surgery consults
• 95 CT contrast reactions, with 10 leading to possible temporary harm and two to permanent injury
• eight MR extravasations, with one needing a plastic surgery consult
• 36 MR contrast reactions, with nine associated with possible temporary harm

The automated database search uncovered 28 medication errors that did not involve intravenous contrast administration. Eleven involved administration of the wrong drug without causing harm. Seven were medication reactions, with one possibly involving permanent harm.

One of six cases involving wrong dosages led to possibly permanent harm. For three patients, administration was halted when symptoms of a known allergy were observed, and in one case, gadolinium contrast was erroneously administered to a patient with a low estimated glomerular filtration rate, but it caused only temporary harm.

These findings served as the basis for a root cause analysis that helped radiology staff identify preventive measures. It recommended eliminating all but one vial option for epinephrine administration to address dosing problems. Electronic flagging techniques were implemented to help identify patients with a history of contrast reactions and to identify discrepant protocol-study requests.

Additional protection was secured by labeling all medications on procedural trays, with a designated timeout period during each case to verify that the correct drug was selected. Point-of-service creatinine testing in the MRI and CT suites was established to avoid nephrotoxic reactions.