Patient Safety Tip of the Week


May 5, 2009     Adverse Drug Events in the ICU


Medication errors and adverse drug events, of course, occur in every part of the healthcare system. But there are some unique adverse drug events that occur in special areas. We have previously discussed medication errors and adverse drug events occurring in the radiology suite and in the operating room and ambulatory surgery suite. Recently there have been several good papers about adverse drug events occurring in the ICU.


In many respects, the ICU environment has many of the crucial factors that can allow for “the perfect storm”. Patients are critically ill, often with multiple organ failure. As such they are more vulnerable to adverse drug events than healthier patients. The number of drugs patients in the ICU receive is higher than in most other areas of the hospital and the drugs used more often fall into the category of “high-risk” drugs. And often side effects of drugs are not immediately recognized because they are masked by or confused with changes related to the underlying medical conditions.


The ICU environment is arguably the most complex of any healthcare environment. It tends to have the most expansive human-technology interface, further enhancing the risk of errors. Because many of the patients are coming from other areas of the hospital or even from other hospitals, the potential for fumbled handoffs is very high. And we have previously discussed the many problems that may arise when ICU patients must be transported to other areas of the hospital for diagnostic testing, etc.


When the Harvard Work Hours and Health Study was performed, one offshoot was the Critical Care Safety Study (Rothschild et al 2005). This was a study that looked prospectively at adverse events and serious errors occurring in the coronary and medical ICU’s in a large academic teaching hospital. Data were collected over a year, using multiple different methods to identify adverse events and serious errors. They identified adverse events at a rate of 80.5 per 1000 patient-days and serious errors 149.7 per 1000 patient-days. Around 12% of both types were potentially life-threatening. Significantly, almost half of the events were considered to be potentially preventable. About half of the adverse events and 78% of the serious errors involved medications. This included not only drugs used as treatments but also drugs used in prophylaxis, diagnosis, and monitoring of medications. Cardiovascular drugs, anticoagulants, and antiinfective drugs were the categories most often involved.


A new paper in the BMJ (Valentin et al 2009) looked at errors in administration of parenteral drugs in intensive care units in 113 units in 23 countries. One-third of the 1328 patients had an error. They found 74.5 events per 100 patient days, including 0.9% that caused permanent harm or death. Cardiovascular drugs, anticoagulants, insulin, sedative/analgesic drugs and antimicrobial drugs were the categories most often involved. Wrong time and missed dose were the two most common types of error seen. Note that more than half the errors resulting in serious harm were errors of omission.


In analyzing the risk factors for such errors, they found increased risk with severity of illness, the number of organ failures present, the number of parenteral administrations, higher level of care, and the number of patients per nurse. Trainees were involved in a substantial number of cases with errors. Increased workload, stress and fatigue were contributory factors in a third of all events. Interestingly, there was a strong “protective” relationship with the existence of a clinical incident reporting system.


However, one very interesting finding, which tends to run counter to prevailing thinking, was that error rates were higher for medications previously prepared by a pharmacist and reduced when nurses labeled syringes of medications they themselves had prepared.


Both the Critical Care Safety Study and the BMJ papers noted that the majority of errors occurred during routine care rather than during emergencies.


A recent excellent review of medication errors in critical care appeared in the Canadian Medical Association Journal (Camire et al 2009). They point out that the number of preventable and potential adverse drug events in ICU’s is double that seen on general care units but this is likely due to the fact that ICU patients receive twice as many medications. Medication errors are also considerably more frequent in medical ICU’s than surgical ICU’s. They noted risk factors similar to those described above and also included problems with medication reconciliation (failure to document a patient’s usual medication list) and housestaff hours worked.


We’d like to emphasize the need for medication reconciliation. Not only are some important drugs inadvertently omitted when patients are transferred to ICU’s but there are often many medications begun in ICU’s that are never discontinued. The classic examples are proton pump inhibitors and other drugs used as prophylaxis against GI bleeding. While there is evidence such prophylaxis is effective in at-risk patients in the ICU, there is little evidence to support their use in non-ICU patients. These agents are frequently continued not only on transfer to the general care units but may be continued past discharge as well. In fact, we recommend regular surveillance for such occurrence either manually or electronically.


We’d also like to speculate that the type of ICU coverage may be an important risk factor for medication errors. Though most hospitals today have dedicated ICU coverage (if not by intensivists, at least by a dedicated ICU team each month) there are still vestiges of older systems where ICU coverage may be by housestaff who have responsibilities for other areas. We all know that verbal orders are high risk for errors. If staff are busy elsewhere and have to give verbal orders to the ICU staff, risks are high. While CPOE may help minimize those risks, we have also given examples of how remote order entry may be dangerous.


The biggest value of the CMAJ paper is it bibliography and list of strategies demonstrated to prevent medication errors in the ICU. These include things like CPOE with clinical decision support, use of barcoding/bedside medication verification, and smart pumps. Medication reconciliation, as noted above, is extremely important. Having a clinical pharmacist round with the medical team is especially valuable in ICU’s. Standardized orders or standardized medication protocols are also useful.


Lastly, the CMAJ article also has good recommendations about disclosure to patients and families about medication errors that have occurred.






Rothschild JM, Landrigan CP, Cronin JW, et al. The Critical Care Safety Study: The incidence and nature of adverse events and serious medical errors in intensive care. Critical Care Medicine 2005; 33(8): 1694-1700;jsessionid=J1CQc0fVd0482vk2Gy0PhJbZJG5pjHwlgTh2ZRJt14TsnYVtlnqD!928310026!181195629!8091!-1?index=1&database=ppvovft&results=1&count=10&searchid=1&nav=search



Valentin A, Capuzzo M, Guidet B, Moreno R, Metnitz B, Bauer P, Metnitz P, on behalf of the Research Group on Quality Improvement of the European Society of Intensive Care Medicine (ESICM) and the Sentinel Events Evaluation (SEE) Study Investigators. Errors in administration of parenteral drugs in intensive care units: multinational prospective study. BMJ  2009; 338:b814, doi: 10.1136/bmj.b814 (Published 12 March 2009)



Camiré E, Moyen E, Stelfox HT. Medication errors in critical care: risk factors, prevention and disclosure. CMAJ 2009; 180: 936-943







Patient Safety Tip of the Week Archive


What’s New in the Patient Safety World Archive