Marla M Husch
Pharmacist Research Coordinator
Patient Safety Department
Northwestern Memorial Hospital
Intravenous (IV) medications contribute to untold millions of healthcare success stories every year – healing, ameliorating pain and symptoms, or sustaining life. However, they also have the potential to cause patient harm. In recent years, “smart” IV pumps with dose error reduction systems have been introduced to the market with high hopes of preventing most of these errors. Smart pump technology is currently available in conjunction with general-infusion pumps, patient-controlled analgesia (PCA) pumps and some syringe pumps. IV pumps that are not electronically integrated with other medication-use information systems such as computerised physician order entry (CPOE) have limited potential to reduce medication errors and harm to the patient.
What are “smart” pumps?
Smart pumps are specifically designed to reduce IV medication errors that are the result of a “slip-of-the-finger” programming mistake. For example, if a clinician programmes the pump for 100ml/h of heparin instead of 10ml/h, the results could be fatal for the patient. Smart pump software checks programmed doses against preset limits determined by each institution for specific care areas. For example, intensive care units (ICU) can allow higher dose limits of dopamine because most beds have a cardiac monitor that is quickly accessible. Most vendors offer software that has already been configured according to evidence-based practice but also allows an institution to create its own parameters. The software is designed to perform a final check, which compares the programmed parameters to the preset limits.
Smart pump technology essentially serves as a forcing function to draw the clinician’s attention to a potential dose error and to prevent at least some normal human errors from reaching the patient. If the dose entered by the clinician falls outside of the preset limits, the clinician will be alerted. Depending on how the limits have been set up, the clinician may then be permitted to override the alert (“soft” limit) or not be allowed to continue at all (“hard” limit).
The entire safety process, however, depends on the clinician’s accurate selection of the correct medication from the pump’s menu, so that the software is checking the correct dose limits in its stored table. Therefore, for some high-risk medications such as heparin or insulin, a mandatory independent double check (eg, second nurse) of drug selection and programming may still be advisable before the drug is administered.
Another feature of most smart pumps is the ability to record and download all alerts and clinician responses, commonly referred to as the continuous quality improvement report, or CQI. Through CQI, all potential programming errors caught by the software can be further analysed. However, these reports do not explicitly address the question of whether the smart pump technology was responsible for the identification of these errors or whether they would have been caught just as well with traditional pumps along with good clinical practice. It is difficult to draw conclusions regarding the impact of smart pumps on patient outcomes without such an analysis, because wrong programming is frequently intercepted by the combination of nurse vigilance and experience, monitoring and double checks, as determined by each institution’s policies. Without taking this into consideration, CQI reports may overestimate the risk of errors and harm associated with pumps that do not have integrated safety software.
What is the overall impact of smart pumps?
Errors related to IV medication have been studied but not well quantified due to varying definitions and methodologies. Most studies limit the definition of an IV medication error to an IV pump programming error. In a recent study conducted at Northwestern University by Husch et al,(1) the definition of an IV pump error was expanded to include all errors that may cause or lead to inappropriate medication use via an IV pump. Such events could be related to professional practice, products, procedures and systems, including order communication, product labelling, compounding, dispensing, administration, education, monitoring and medication or equipment use. The study concluded that, although smart pumps are definitely a part of a comprehensive safety system, they will not generate meaningful improvements in patient safety without an electronic real-time interface with other systems, such as CPOE, pharmacy information systems (PIS), electronic medical records (EMR) and barcoded medication administration systems (BCMA).
Current smart pumps without an interface to other systems do not have the capacity to alert clinicians to errors involving a wrong dose or wrong medication (ie, one that does not match the order), wrong patient, wrong time or wrong indication. They only have the capacity to protect patients from errors related to programming, and this is further limited to preventing only those errors that exceed routine dose limits. From a practical perspective, the dose limits will need to be set at reasonably large ranges so that clinicians are not faced with excessively frequent alerts. If alerts occur too often, “alert fatigue” sets in and clinicians begin to use the “override” function without sufficient vigilance. However, if ranges are set fairly generously, the possibility of unrecognised error necessarily increases.
These pumps cannot detect changes in the patient’s condition, intended changes in orders or unintended concentration changes of refilled continuous infusions. Patients will not be protected from the many other types of IV medication errors that can lead to patient harm. For example, in the study by Husch et al,(1) an error was identified involving a dopamine drip that was infusing at four times the prescribed amount. This was not due to a programming error – the pump was set as though the drip was the correct concentration – but rather to a dispensing error in which the concentration that was dispensed for a refill infusion of dopamine was quadruple that in the previously dispensed bag. The bag was correctly labelled, but the nurse did not verify the label and the pump had no capability to “read” the IV bag’s barcode label to verify that the correct medication had been loaded into the pump.
In a study by Rothschild et al,(2) smart pumps were found to have no measurable impact on serious IV medication error rates. The authors attributed this finding to the way in which the nursing staff were using the technology. They concluded that, although smart pumps have great promise, technological and nursing behavioural factors must be addressed if these pumps are to achieve their potential for improving medication safety.
Nevertheless, if utilised correctly, smart pumps will avert some life-threatening errors due to some programming mistakes. For example, if a clinician intends to set a pump for 12.7ml/h to infuse heparin but inadvertently programmes 127ml/h, the pump will not allow the infusion to continue, protecting the patient from a possibly fatal overdose of heparin.
Vendors are improving smart pump technology to address some of the current limitations. Some pumps can only be updated with new software and dose limits by physically upgrading each piece of equipment; most vendors are now addressing this significant inconvenience, cost and risk issues with wireless updates from a central server.
Other local patient care improvement may be achieved through software to alert the clinician if a patient becomes oversedated while receiving ï¿½narcotics via a PCA. Some vendors now offer integrated PCA monitoring via end-tidal CO(2) monitoring (EtCO(2)) and pulse oximetry (SpO(2)).
The most important future development will be full integration of the smart pump with the patient’s medication orders (CPOE), pharmacy information system (PIS) and patient/drug/order identification and matching (barcode medication administration device), or BCMA. Without this full integration, the potential of smart pumps is limited to the prevention of a serious but small range of errors that arise in pump programming.
With this integration, the power of information technology will be able to address many of the IV medication errors that are generated during ordering, documentation, dispensing and ï¿½monitoring phases of medication administration. All major vendors are currently working in this direction.
With the addition of any new technology, expect the emergence of new types of errors. This potential has been demonstrated with other technologies such as CPOE. There is little research available at present to quantify new types of errors made possible by smart pumps. For example, complex drug libraries will necessarily have many selections, making it perhaps more likely that clinicians will choose the wrong item from a lengthy list. Because of this, an institution may decide that the library poses a bigger problem than no library at all and remove most of the medications from it. Excessive alerts and associated “alert fatigue” may reduce the effectiveness of “soft” (overridable) dose limits.
What makes sense regarding smart pump technology and your institution?
With extensive emerging technology in patient care (CPOE, BCMA, EMR, etc), and pressure to address medication safety from Leapfrog and Joint Commission on Accreditation of Healthcare Organizations (JCAHO), hospitals may feel a need to purchase new IT systems quickly. Unfortunately, there is little information about actual projected error prevention, annual costs and software maintenance, and few independent studies that offer a detailed analysis and comparison of implementation, training and ongoing maintenance and costs for all of the different types of technology available. One good source for unbiased research and comparisons is ECRI (see Resource).
To weigh the costs and benefits of different systems, hospitals will first need to understand their own most common, harmful and costly medication errors. This effort should be led by a dedicated multidisciplinary task force representing major clinical areas of the institution (pharmacy, medicine, surgery, paediatrics, intensive care, emergency department, anaesthesia and operating rooms, etc). Medication error analysis can be accomplished by analysing incident reports, performing small observational studies and carrying out staff interviews. This will provide a good feel for the rate and cause of medication errors, which will serve as the baseline.
Next, the claims of the various vendors will need to be evaluated carefully in order to find the right system that will fit the particular needs of an institution. Researching the independent literature and conducting site visits to other institutions that have already purchased and implemented the technology can be helpful in the decision-making process.
For those who recognise that their IV pumps are near obsolescence and need replacement, but who are not satisfied with the current technology or unsure of whether the new technology will decrease error rates, a short-term lease versus purchase may be a good option until the IV pump market matures. A cost analysis should be performed comparing buying versus leasing. If the decision is made to purchase smart pumps prior to full market maturity, ask the vendor for information regarding feasibility and cost to upgrade when available.
Although they would be more effective if interfaced with other information systems such as EMR, BMCA, CPOE and PIS, smart pumps with dose error reduction software can decrease the likelihood of a small number of harmful IV medication errors. Advancing technology to reduce medication errors is a giant gain for the Patient Safety movement. Keep in mind that, from the perspective of harm, type and cause of errors, not all hospitals are created equal and, therefore, when it comes to technology there is no “one size fits all” solution. The implementation of the correct system can prove a fruitful first step, but make sure that you research thoroughly, choose thoughtfully and have the appropriate staff involved in the decision process (see Box 1). It is also critical to recognise that smart pumps are only part of a larger patient safety journey that will require continuous attention and enhancement in years to come.
- Husch M, Sullivan C, Rooney D,et al. Insights from the sharp end of intravenous medication errors: implications for infusion pump technology.Qual Saf Health Care 2005;14:80-6.
- Rothschild JM, Keohane CA, Cook EF, et al. A controlled trial of smartinfusion pumps to improve medication safety in critically ill patients.Crit Care Med 2005;33:533-40.