Intravenous Infusion Safety Initiative Prevents Medication Errors, Leading to Cost Savings and High Nurse Satisfaction

Problem Addressed

• Common problem with major health consequences: Medication errors account for almost 20 percent of in-hospital medical injuries; errors during IV infusion, a commonly used technique that often involves high-risk medications, are frequently among the most egregious of these errors.¹ Since IV medication is delivered directly into the patient's bloodstream there is more likelihood that it will cause patient injury if there is an error. For example, IV medications are associated with 54 percent of potential adverse drug events,² and one study found that 40 percent of patient deaths from such events are due to administration of the wrong dose, while 16 percent are due to administration of the wrong drug.³ 
• Most error-prone part of the process: Medication administration by bedside nurses is the most error-prone part of the IV medication process, primarily because there are few built-in safeguards and redundancies (unlike during other parts of the process). Approximately 38 percent of medication errors that cause preventable injuries occur at this stage, and only 2 percent of these errors are caught before they cause harm.⁴ Additionally, general purpose IV devices are programmed at the bedside, and the range of programming options is so large that it is relatively easy to inadvertently deliver incorrect doses of medication (for example, a missing decimal point or an accidental double key press can lead to a 10- to 100-fold overdose). Finally, the potential for error is also increased due to the large variability in drug names, dosing concentrations, dose limits, maximum infusion rates, weight limits, and volume limits. For example, the drug magnesium sulfate has 10 different dosing units available.⁴
• Additional risk for patients using patient-controlled analgesia (PCA) devices: Although many patients benefit from PCA use for opioid administration, individual patient responses to opioids vary, placing some patients at risk for respiratory depression and a decrease in heart rate/blood pressure. If detected early, patients with respiratory depression can be effectively treated by adjusting the PCA or continuous infusion dose, or using naloxone, but in severe cases narcotic oversedation can be fatal. The risk of patient harm due to PCA-related medication errors is 3.5 times greater than for other medication errors.⁴

Description of the Innovative Activity

• Standardized medication delivery mechanism: Saint Joseph's/Candler installed a computerized IV infusion system designed to avert programming errors and provide actionable data on various aspects of the averted errors. The main components of this "all-in-one" system include a central computer attached to the IV pole with PCA pump modules, and portable "point-of-care" devices that nurses use to input information to get correct information on dosing and connect infusion devices. The new system reduced the number of different infusion devices used at Saint Joseph's/Candler from five to one.
• Decision-support drug library: The system uses software called "DERS" (Dose Error Reduction Software) that acts as a decision-support drug library to standardize concentrations, dosing units, and dosing limits for IV infusion medications. The software provides customized information for different patient care types (e.g., critical care, pediatrics, surgery, oncology), which serves to decrease complexity and minimize opportunities for error. The software generates alerts (e.g., indicating that a dosing limit has been exceeded) and automatically averts errors by prompting the nurse to reprogram or cancel an infusion as appropriate. Data analysis functions within the software help to identify opportunities for improving IV medication safety and best practices.
• PCA respiratory monitors: Monitoring devices for capnography (to measure exhaled carbon dioxide and breathrate) and oximetry (to measure blood oxygen and pulse rate) are used for non-intubated patients receiving PCA therapy in critical care and general nursing units. These devices identify the early warning signs of respiratory depression by providing trend data at the bedside, thus helping clinicians assess respiratory response to PCA therapy.
• Expanded role for respiratory therapists: The hospital revised its policy to require respiratory therapists to round on every patient receiving PCA therapy at least once during every 12-hour shift. When continuous capnography is being used and an issue arises that a nurse cannot resolve, respiratory therapists function as "first responders," assessing the patient, reviewing trended capnography data and the amount and type of medication received, helping the nurse identify the cause of the patient's change in status, determining the appropriate intervention, and working with the physician to improve outcomes.
• Wireless networking to monitor patients, update systems: The hospital's wireless networking system allows pharmacists to remotely monitor patients receiving infusions outside pre-established limits and quickly install software upgrades, revise best-practice datasets, and gather data for analysis.
• Ongoing analysis to stimulate quality improvement: Analysis of data from software logs allows Saint Joseph's/Candler to make continual targeted improvements based on the needs identified. The most numerous software alerts were related to the drugs heparin and propofol, which allowed Saint Joseph's/Candler to create more effective protocols for these drugs.

References/Related Articles

Contact the Innovator

Results

• Many errors prevented during standard infusion: The system, which automatically documents averted errors, found that 245 errors were prevented between October 2002 and July 2003 (the first 9 months after the program was implemented), including 166 averted overdoses. Between January and June 2006 (after additional components were added), 967 errors were prevented, 328 of which were potential overdoses that were greater than 1.5 times the maximum dose and thus highly likely to cause harm. 
• Many errors averted during PCA infusion: During the PCA system's first 4 months of use in 2004, 52 programming errors were identified and corrected. In addition, continuous respiratory monitoring helped clinicians identify numerous cases requiring intervention, even when programming was correct and a patient received appropriate dosing. For example, between July 2004 and March 2007, respiratory therapists identified and intervened with 16 patients with declining physiologic status, thus avoiding possible transfer to the intensive care unit.
• Cost savings: Using a conservative estimate of $6,000 for the cost of an adverse drug event,⁴  the 328 averted overdoses from January to June 2006 saved the hospital system nearly $2 million. As noted earlier, heparin and propofol accounted for 73 percent of the highest risk potential overdoses; the changes to the protocol for propofol alone yielded a more than 50-percent reduction in propofol dosing alarms, saving more than $1 million (with estimated costs falling from $1,774,395 to $650,330).
• High nurse satisfaction: Anecdotal reports from nurses indicate that the program makes them feel more comfortable in aggressively managing patients' pain and less reluctant to give additional medication. This is particularly important with sickle cell patients, who often require high doses of opioids. Nurses also report feeling more satisfied with the IV medication system because it demonstrates the hospital's commitment to support and protect them when administering high-risk medications.

Evidence Rating (What is this?)

Context of the Innovation

Planning and Development Process

• Identifying program goals: The multidisciplinary medication safety team (consisting of physicians, pharmacists, nurses, respiratory therapists, risk managers, and others) decided that its first priority should be averting errors that pose the greatest risk of harm. To that end, the team established the following four goals in 2001: 
  • Increase detection and prevention of IV medication administration errors, resulting in improved patient care and decreased mortality and morbidity.
  • Increase identification and documentation of errors and prevented errors, including the types of errors, where and when they occur, and the most error-prone drugs
  • Implement an error-detection system with built-in feedback loops, so that opportunities for improvement could be identified and implemented over time. 
  • Decrease the complexity of existing infusion technology.
• Selecting an IV infusion system: Recognizing that computerized IV infusion safety systems (also known as "smart pump" systems) are specifically designed to avert IV infusion programming errors, the team decided in 2002 that implementation of such a system was the best approach to safeguarding patients against high-risk medication errors. After reviewing published reports and comparing and evaluating available systems (including testing system demos and visiting other hospitals), the team selected a system. 
• Pilot testing: Saint Joseph's/Candler chose a pilot site to evaluate the PCA and respiratory monitoring modules for 6 months.
• Training: Clinical experts from various patient care areas spent several weeks training nurses and other staff to use the new system. Training included expert sessions, skills labs, hands-on exposure, and an internet-based computer training module provided by the vendor.
• Program roll-out: Based on the success of the pilot test described above, the new infusion system was installed on all units in the two hospitals on one day in October 2002. Installation of 584 point-of-care units and 760 infusion pump modules began at 5 a.m. and finished within 8 hours.
• Adding components: Additional features were assessed and, if found to be useful, added as they became available. For example, a wireless network was added in 2004.

Resources Used and Skills Needed

• Staffing: The program has not required the hiring of additional staff, even with the change in responsibilities for respiratory therapists described earlier. 
• Costs: The initial purchase and implementation of the IV infusion system cost about $2.5 million. In addition, the contract with the vendor to maintain and update the equipment and software runs about $65,000 annually.

Funding Sources

Tools and Other Resources

Getting Started with This Innovation

• Take a multidisciplinary approach: A collaborative approach is key to improving patient and medication safety. At St. Joseph's/Candler, having a multidisciplinary team research and present technological alternatives helped in making a strong case to senior hospital officials. Similarly, involving nurses in research and implementation of the technology resulted in a high level of acceptance and compliance.
• Choose a system with wireless capability: Not all systems allow for software updates via a wireless system. This feature is important, even if it involves an additional expense, since it eliminates the need for a technician to manually load revised drug safety limits. It also allows rapid access to infusion data and implementation of process improvements.

Sustaining This Innovation

• Track data: Analysis of actual and averted errors on an ongoing basis is critical to identifying opportunities for improvement. 
• Stay abreast of new technology: Because infusion technology continues to evolve rapidly, hospitals need to stay on top of these changes and critically evaluate the safety and financial benefits and drawbacks of adding new features to the system.