Detection of Adverse Mediation-Related Events

From Clinfowiki
Revision as of 22:56, 18 November 2006 by Hbc001 (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

Detection of Adverse Medication-Related Events

In the landmark Harvard Medical Practice Study published in 1991, it was estimated that 3.7% of all hospitalized patients experienced an adverse event. Some of the first reports to quantify the significance of medication-related adverse drug events in the U.S. Healthcare system in the 1990s, found the adverse event incidence rate associated with medications ranged from 2 to 7 per 100 admissions. However, it was not until the Institute of Medicine (IOM) Report To Err is Human in 1999 that there was heightened interest in medication event analysis. A subsequent IOM report, Crossing the Quality Chasm, identified patient safety as a key dimension of quality and identified information technology as a critical means of achieving this goal, including the implementation of computerized provider order entry (CPOE) and clinical decision support systems (CDSS). CPOE and CDS systems have been shown to reduce the incidence of medication errors and improve patient safety. However, the outcomes research to date has been completed at highly-resourced, large academic medical centers with internally developed CPOE and CDS systems and may not necessarily be generalizable to commercially available systems being implemented by most organizations today. Also, because the definition of CDSS in the industry seems to be open to broad interpretation, not all CPOE and CDS systems are developed or deployed equally. The combination of these factors, make it essential for broad scale development and deployment of computerized surveillance technologies to assist organizations in detecting medication-related events and to measure the impacts of these commercially available CPOE and CDS systems on patient safety, quality and outcomes over time.

A variety of methods have been shown to be successful in detecting medication-related adverse events. Manual chart review was the first method identified and it is still considered to be the ‘gold standard’ for detecting adverse events. Though manual chart review is the most comprehensive of all of the detection methods, it is also the most resource-intensive and expensive to deploy and therefore not feasible for wide-scale, long-term use. Due to these limitations, computerized surveillance methods for medication-related event detection were later developed using the results of manual chart review studies as their basis. The most published computerized surveillance methods described by Classen et al, Kilbridge et al and Jha et al used logic-based rules, including panic value lab rules (digoxin > 2ng/mL), drug-lab rules (warfarin and INR > 4) and use of drug antidotes (naloxone for opiate overdose, dextrose 50% for drug-induced hypoglycemia) to detect potential medication adverse events. Though successful, logic-based rules are only able to detect a limited set of events due to their requirement for codified medication and laboratory data. Additional studies by Jha et al and Murff et al further explored the use of natural language processing (NLP) technologies by text scanning clinician progress notes and discharge summaries using a lexicon developed from prior methods. NLP has also shown to be a proven method for detecting medication-related adverse events that may not be captured by the logic-based rules method, such as allergic reactions and oversedation from opiates where the event is not severe enough to warrant an antidote (naloxone). However there are various barriers to utilizing NLP surveillance technology for broad-scale deployment, including no standard medical lexicon or standard terminology for identifying adverse medication events, variation in individual clinician and organizational documentation methods and NLP technologies are fairly expensive to purchase, implement and maintain. Medication adverse effect ICD-9 codes (M Codes) is additional method that has been studied for detecting medication-related events. However, because ICD-9 codes are primarily utilized for healthcare reimbursement, they are not commonly and reliably used by most providers for coding adverse effects to medications, making this method of limited value in most organizations. Though logic-based rules, NLP technologies and ICD-9 codes are not as comprehensive in detection of events when compared to manual chart review, these computerized surveillance technologies have clearly been proven to be collaboratively beneficial in detecting a broad range of medication-related adverse events.

As organizations continue to purchase and implement commercially available CPOE and CDS systems with a primary focus on improving the quality and safety of patient care, an equal focus may also need to be placed on measuring the performance and impact of these systems on patient outcomes over time. Computerized surveillance methods can play a significant role in reliably and efficiently detecting medication-related adverse events and for providing the opportunity to capture the metrics necessary to measure organizational impacts of quality improvement initiatives and CPOE and CDSS systems over the life time. Because these computerized surveillance technologies require fewer resources to maintain post development when compared to manual chart review, there is significant opportunity for their continued maturity in the coming years which may in turn, provide additional opportunity for computerized surveillance technology to be shared across organizationsand even possibly ‘served-up’ by academic centers to support more wide-scale deployment and significant long-term benefits from these systems in the future. by hbc001

1. A Roadmap for National Action on Clinical Decision Support: June 12, 2006. AMIA Workgroup
2. Bates DW, Cullen DJ, et al. Incidence of Adverse Drug Events and Potential Adverse Drug Events. JAMA. 1995;274:29-34.
3. Bates DW, Evans RS, et al. Detecting Adverse Events Using Information Technology. JAMIA. 2003;10:115–128.
4. Bates DW, Leape LL, et al. Effect of Computerized Physician Order Entry and a Team Intervention on Prevention of Serious Medication Errors. JAMA 1998;280:1311-1316.
5. Brennan TA, Leape LL, et al. Incidence of Adverse Events and Negligence in Hospitalized Patients: Results of the Harvard Medical Practice Study I. NEJM. 1991;324:370-376.
6. Campbell EM, Sittig DF, et al. Types of Unintended Consequences Related to Computerized Provider Order Entry. JAMIA. 2006;13:547-556.
7. Classen DC, Pestotnik SL, et al. Computerized Surveillance of Adverse Drug Events in Hospital Patients. JAMA. 1991;266:2847-2851.
8. Institute of Medicine. To err is human. Building a safer health system. Washington, DC, National Academy Press, 1999.
9. Institute of Medicine. Crossing the Quality Chasm: A New Health System for the 21st Century. Washington, DC, National Academy Press, 2001.
10. Jha AK, Kuperman GJ, et al. Identifying Adverse Drug Events: Development of a Computer-based Monitor and Comparison with Chart Review and Stimulated Voluntary Report. JAMIA. 1998;5:305-314.
11. Kaushal R, et al. Effects of Computerized Physician Order Entry and Clinical Decision Support Systems on Medication Safety. Arch Intern Med. 2006;163:1409-1416.
12. Kilbridge PM, Alexander L, et al. Implementation of a System for Computerized Adverse Drug Event Surveillance and Intervention at an Academic Medical Center. JCOM. 2006;13:94-100.
13. Kilbridge PM, Campbell UC, Cozart HC, et al. Automated Surveillance for Adverse Drug Events at a Community Hospital and an Academic Medical Center. JAMIA. 2006;13:372-377.
14. Murff HJ, Forster AJ, et al. Electronically Screening Discharge Summaries for Adverse Medical Events. JAMIA. 2003;10:339-350.