The below information represents a small portion of data and information contained in USP's most recent MEDMARX Data Report: A Chartbook of 2000–2004 Findings from Intensive Care Units and Radiological Services.  For complete data findings related to medication errors in these clinical areas see: http://www.usp.org/products/medMarx/

Intensive care units (ICUs) are one of the most expensive components of U.S. healthcare representing 10% of acute care beds, yet comprising approximately 30% of acute care costs.^1-3  More than 4 million people are admitted to ICUs annually and the number of ICU patients and their proportion of hospital admissions are expected to grow given the increasing number of elderly people and the increasing acuity of illness of hospitalized patients.  

Patient care in ICU areas is complex, in part because of the broad scope of acute illnesses and pre-existing conditions present among ICU patients. On average, mortality rates for these patients range between 12 and 17%.⁴  Previous studies have shown that medical errors are common in ICU areas⁵ and one recent study estimates that there are approximately 1.5 serious medical errors per 10 critical care beds per day.⁹ 

USP's Center for the Advancement of Patient Safety analyzed 38,371 records submitted to MEDMARX by 503 facilities during the 5-year period between January 1, 2000 and December 31, 2004 for error events that occurred in ICU areas.  For purposes of this report, records specifically documenting the neonatal or pediatric ICU as the location of the medication error were excluded because of the special character of these units and their patient population. 

The time of day when the largest number of medication errors were reported to have occurred was between the hours of 8 a.m. and 12 noon (n=6,744) with a peak around 9 a.m. (n=1,647) (Figure 1).  Additional spikes in error occurrence were seen around 6 p.m., 8 p.m., and 12 a.m. 

Physicians commonly conduct patient care rounds in the early morning hours, followed by preparing orders for lab tests, medications, and other treatments and/or procedures.  Therefore, it is not surprising that the greatest number of ICU medication errors occurred between the hours of 8 a.m. and 12 noon.  One of the implications for healthcare facilities is to recognize that there are variations in medication use volume throughout the entire day (and the corresponding opportunities for error) and perhaps better plan for and allocate the necessary resources to accommodate such fluctuations.

Figure 1.  ICU Medication Errors by Time of Day When Error Occurred

A little over 7% (n=2,819) of the ICU medication errors were categorized as potential errors whereas 89.3% (n=34,282) were non-harmful and 3.3% (n=1,270) resulted in some level of patient harm (Figure 2).  Among the harmful errors, 83.7% (n=1,063) resulted in temporary harm whereas 68 were sentinel events including 14 fatalities.  The percentage of harmful ICU errors (3.3%) is nearly twice the overall harm threshold of 1.7% for all MEDMARX records for the same 5-year period.   

Figure 2.  Severity of Medication Errors in ICUs

Nearly half of the medication errors originated in the Prescribing and Transcribing/Documenting nodes (i.e., phases) of the medication use process (Table 1).  Errors originating in the Administering node represented the largest percentage of actual errors (42.5%), the largest percentage of harmful errors (61.1%), and the largest percentage of sentinel events (57.4%).  Errors originating in the Prescribing node resulted in the second highest percentage of sentinel events (29.4%), a result that is disproportionately high considering the small number (n=168) of harmful errors in the Prescribing node. 

 Table 1.  Errors in ICUs by Node:  Actual Compared to Harmful and Sentinel Events

a. Node is not applicable for Category A error records.
b. Percents may not add to 100% as a result of rounding.

Five types of error exceeded the overall ICU harm threshold of 3.3%.  As a percentage, Wrong administration technique errors ranked first with nearly 11% resulting in patient harm (Table 2).  This percentage is substantially higher when compared to the 6.3% incidence of Wrong administration technique errors seen in the general MEDMARX data set.  Examples of Wrong administration technique include administering an IV drug too rapidly, mixing a drug in an incompatible solution so that a precipitate forms, not activating the drug product chamber into the main IV solution chamber, not flushing an IV line, and crushing sustained-released tablets.

Table 2.  Leading Types of Error for ICUs by Non harmful and Harmful Outcomes ^a 

a.  Represents a cross-tabulation of Types of Error by Error Category for the 5-year period 2000-2004.
b.  Data based on 36,790 records with 39,894 selections.
c.  Exceeds the 3.3% of harm calculated for all ICU error records combined.

Combining related, but somewhat infrequent causes of error identified several common problems.  IV pumps and other equipment were cited a total of 1,237 times of which 10.8% were harmful (Table 3).  Causes related to drug product packaging/labeling were cited a combined total of 949 times, of which 7.4% were harmful.  Six communication-related causes totaled 5,021 selections, of which 5.4% were associated with harm.  The use of IV pumps and other medical equipment is more extensive within ICUs than in most other patient care areas.  The percentage of harm associated with errors involving pumps/equipment, drug packaging/labeling, and communication point to problem areas that warrant further attention. 

Table 3.  Selected Causes of Error Related to Equipment, Product Packaging/Labeling, and Communication in ICUs 

Selected Error Cases

Case #1–IV Pump Programming: A diabetic patient was receiving an IV of regular insulin one unit/mL at a rate of 10 units/hour titrated per sliding scale. Upon changing to a new bag of insulin, the IV pump was reset manually to clear prior totals and to enter the new volume to be infused. Shortly after the new bag was hung, a nurse noticed that the infusion pump was incorrectly set at 150 units (i.e., 150 mL/hour). The infusion was stopped and the patient was given Dextrose 50% in water and closely monitored for the next 8 hours. If the total volume of the bag (100 mL) had been infused at the rate of 150 units/hour, it would have taken only 40 minutes for the patient to receive 100 units of insulin, potentially causing irreversible brain damage and/or death from cerebral edema and insulin shock. The type of error was identified as Improper dose/quantity.

Case #2–Communication Failures:  A patient who presented in the emergency room was given a heparin bolus and started on a heparin infusion.  The patient was transferred to the coronary care unit where a physician, unaware that the patient was on a heparin infusion, ordered enoxaparin.  Later, an on-call physician, unaware that the patient was receiving the enoxaparin, ordered another dose of the heparin infusion.  The nurse who received the physician's call did not inform him of the patient's other medications.  The patient received both heparin and enoxaparin for 15 hours, leading to a drop in the patient's hemoglobin and hematocrit, shortness of breath, and rales.  The patient was given a blood transfusion and placed on a ventilator. The causes of error were reported as failures in communication and not following procedures and protocols.

Case #3–Dosage Form Confusion, Product Labeling Issues:  A loading dose of 500 mcg/kg of esmolol followed by a titrated IV infusion of 5,000 mg/500 mL was ordered for a surgical intensive care patient experiencing acute arrhythmias. Based on the patient's weight of 264 pounds (120 kg), the nurse correctly calculated the loading dose to be 60 mg. However, the noise and activity of the surgical intensive care unit distracted the nurse, and the nurse inadvertently retrieved an ampul of concentrated esmolol (250 mg/mL) instead of the ready-to-use product (10 mg/mL). The nurse drew up 3 mL (750 mg), mixed this amount with a small volume of normal saline, and administered the dose to the patient. The patient's condition soon deteriorated and he went into cardiac arrest resulting in brain hypoxia and death.

The complexity of ICUs, combined with the high acuity of the patients treated in these areas, creates an environment that is more susceptible to harmful patient outcomes when medication errors occur.  The Agency for Healthcare Research and Quality (AHRQ) Evidence Report/Technology Assessment —Making Healthcare Safer: A Critical Analysis of Patient Safety Practices, identified several interventions that can reduce mortality in ICU patients.¹⁰ These include the following: reporting adverse events; staffing ICUs with physicians trained in critical care (i.e., intensivists); improving communications and the culture of safety among caregivers; and controlling blood glucose in critically ill patients.  Other approaches to minimizing medication errors in ICUs include:

• Assigning a dedicated pharmacist to ICU areas
• Implementing a CPOE and bar-code system
• Reducing medical residents' work hours
• Ensuring adequate patient staffing patterns
• Simplifying and standardizing IV pumps, monitors, and IV catheters and 
• Allowing adequate input on and training for new technology (e.g., IV pumps)

1. Joint position statement: Essential provisions for critical care in health system reform. Society of Critical Care Medicine. American Association of Critical Care Nurses. Critical Care Medicine 1994;22(12):2017-9.
2. Groeger JS, Guntupalli KK, Strosberg M, et al. Descriptive analysis of critical care units in the United States: Patient characteristics and intensive care unit utilization. Critical Care Medicine 1993;21(2):279-91.
3. Halpern NA, Bettes L, Greenstein R. Federal and nationwide intensive care units and healthcare costs: 1986-1992. Critical Care Medicine 1994;22(12):2001-7.
4. Al-Asadi L, Dellinger R, Deutch J, Nathan S. Clinical impact of closed versus open provider care in a medical intensive care unit. American Journal of Respiratory & Critical Care Medicine 1996;153:A360.
5. Bracco D, Favre JB, Bissonnette B, et al. Human errors in a multidisciplinary intensive care unit: a 1-year prospective study. Intensive Care Medicine 2001;27(1):137-45.
6. Donchin Y, Gopher D, Olin M, et al. Look into the nature and causes of human errors in the intensive care unit. Critical Care Medicine 1995;23(2):294-300.
7. Herout PM, Erstad BL. Medication errors involving continuously infused medications in a surgical intensive care unit. Critical Care Medicine 2003;31(12 Suppl):S678-S86.
8. Osman S, Harris CB, Dunagan C, Prentice D, Fraser VJ, Kollef MH. Reporting of medical errors: An intensive care experience. Critical Care Clinics 2004;32(3):727-33.
9. 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):1694-700.