Thousands of people die each year as a result of medication errors. Medication errors can be attributed to faults in both humans and medication use systems. Therefore, it is necessary to address resolutions to both of these predicaments. The anticoagulant heparin is amongst the most implicated medications. Thus, it has been documented in the top five high-alert medications. Two notable events that triggered recent interest in this topic are the heparin overdoses that occurred in California, associated with actor Dennis Quaid’s newborn twins, and those affecting neonates in an Indiana hospital. The Failure Mode Effect Analysis (FMEA) is a proactive approach to error prevention. Implementation of an FMEA system would serve as a crucial method that will help to recognize potential failures of a product or process before adverse events occur. FMEA can help identify where the use of technology can be implemented to facilitate the reduction of medication errors, especially pertaining to heparin as in this case. Studies have shown how technology, such as computerized heparin nomagram system (HepCare), smart pump infusion technology, computerized physician order entry (CPOE), and the bar coding system, can reduce medication errors. Expanding nationwide awareness of these methods should result in a significant decline of medication errors.
Introduction
Errors are unavoidable in today highly complex and technologically advanced medical treatment facilities and hospital. Recent studies have shown that over 1.3 million people suffer from unintended injuries in United States hospitals as a result of medical errors. As hospital medicine becomes more complex, the frequency of medical errors is increasing. For instance, one study found that medication prescription errors increased in hospitals by almost 250% between 1983 and 1998.1 The Institute of Medicine estimates that thousands of people die each year as a result of medication errors. The top five offending high-alert medications are heparin, insulin, morphine, potassium chloride and warfarin. Common errors that occur with heparin include wrong dose, drug mix up, and administering it to the wrong patient. Failure to recognize change in a patient’s condition due to error can cause potential complications such as an increase in length of stay at the hospital, blood clots if dosing is too low (could lead to stroke or heart attack) or bleeding if dosing is too high, and death.2 Implementing precautionary measures is critical to enhancing patient safety. Hospital personnel, including doctors, nurses and pharmacists play a pivotal role in medication safety. Pharmacists and technicians have the responsibility of placing the correct medications into automated dispensing cabinets, which are password protected. When nurses need to administer a drug, they simply enter their password into the computer system and retrieve the drug from an assigned drawer.3 Pre-measured heparin vials are often kept in these automated dispensing drug cabinets. Failure Mode Effect Analysis (FMEA) is a proactive error prevention system that identifies potential failures of a product or process before adverse events occurs. By recognizing the potential failures, one can implement actions to reduce their chances of occurring.1 One example of use of the FMEA utility in healthcare were for system improvements in the field of anesthesiology, where mortality has been reduced.4 Successful implementation of FMEA by the Veteran Affairs caught the attention of the United States Joint Commission on Accreditation of Healthcare Organizations (JCAHO). In July 2001, JCAHO introduced a new leadership standard that requires department heads in healthcare organizations to perform a FMEA certification on at least one clinical process yearly.1
A multi-disciplinary team should be formed to complete the FMEA analysis, since an FMEA developed by a team of people is more comprehensive and valuable than one developed by a single person.5 These representatives should consist of various medical disciplines that participate in the selected healthcare process for analysis, while maintaining a healthy distance from the direct process being evaluated in order to provide a unbiased perspective. In an intensive care unit setting, an FMEA team may be composed of an intensivist, a senior executive, a staff nurse, a respiratory therapist or other therapists, a pharmacist, an information specialist and a mixture of clinical consultants with proficiency in selected components of the process.1
The FMEA team begins by identifying all of the usual exceptional steps, those that are included selectively or occasionally in response to unfolding events, in the process to analyze for potential errors. These errors are then flagged as failure modes, since a product or process is unsuccessful at producing its desired outcome.1 All identified failure modes are rated on a scale of 1 to 10 based on the anticipated severity of the consequence, likelihood of occurrence, and ability to detect the failure; where 10 is the most severe consequence, the greatest likelihood of occurrence, and the least likely chance of detecting the failure. The severity, occurrence, and detection ratings are multiplied together to get the risk priority number (RPN). The combinations with the highest RPNs are the potential failures that improvement efforts should be focused on.5 This process helps to prioritize changes that need to be made. Next, the team develops new procedures that would decrease the likelihood of the failures. New checks may be implemented to detect emerging failures as early as possible. Healthcare systems may possibly need additional personnel in order to execute the preventative measures and corrections recommended. In addition to new personnel, new or modified equipment, new documentation template and new hospitals policies or procedures may be required as well.1 As the new process is being executed to reduce errors, RPNs are recalculated to measure the progress and prioritize additional changes. Once the team concurs that the process has been optimized, process modifications are accompanied by additional staff education and training as required. As part of the continuous quality improvement, actual errors and “near misses” are recorded, as are the clinical outcomes of the process. New modifications are implemented based on these periodic assessments. Finally, it is important to continually monitor the process to ensure that the systematic improvements are fully operational.1 Various methods of technology can be implemented to facilitate the reduction of medication errors, especially with the high-alert anticoagulant heparin. Some examples include a computerized heparin nomagram system (HepCare), smart infusion technology, computerized physician order entry (CPOE), and the bar coding system.
The Joint Commission on Accreditation of Healthcare Organizations (JCAHO) recommends using at least two patient identifiers whenever taking blood samples or administering medications or blood products. Appropriate identifiers include patient’s name, an assigned identification number, telephone number, or other patient-specific identifier. Bar coding consisting of two or more patient-specific identifiers, complies with this JCAHO recommendation.6 Upon admission to the hospital, patients would be issued a wristband with a bar-coded unique patient identifier. The hospital would have bedside scanners linked to: the admission, discharge, and transfer database; the pharmacy information system; the personnel system; and software to supply active decision support (i.e. a rules engine).6 The bar coding system checks prescription orders against electronic medical records and then sends them to the hospital pharmacy where the medications are labeled with the patient’s name and unique barcode.4 When a patient is to receive a medication, the nurse would scan his/her bar-coded employee identifier, the patient’s wristband to confirm patient identity, and then each package of medications to be administered at the bedside. The system would verify the dispensing authority of the caregiver, confirm the patient’s identity with his/her medication profile in the pharmacy information system, check the rules engine for any alerts or reminders for the nurse, electronically record the action in an online medication administration record (MAR), and store data for later aggregate analysis.6 The principles behind bar coding suggest that errors leading to patient harm could be significantly reduced through the effective use of this technology. Utilizing the bar coding system helps verify that the right drug is being administered to the right patient at the right dose by the right route and at the right time.6 Studies show that the Veteran Affairs (VA) hospitals implementation of bar coding nationwide have seen an 86% reduction in medication errors. Therefore, utilization of bar coding technology with other forms of technology can significantly reduce medication error system wide.4 Medication errors are not essentially the fault of the clinician, but are due to inadequate procedures and systems used in the care of patients. It is thus necessary to put effective systems in place to avoid mistakes, particularly for high-risk medications. Unfractionated heparin (UFH) is highlighted as a high-risk medication and has the potential to cause errors in many ways, including miscalculations, dose ordering, and infusion pump-related errors.7 Factors that contribute to heparin being a high risk medication are that it is expressed in units rather than milligrams, has a narrow therapeutic index and its adverse effect profile.8 To overcome errors in prescribing, calculating doses, administering, and monitoring intravenous heparin, a computerized heparin nomagram system (HepCare) was developed by the Mayo Clinic pharmacy. Using interactive cues between the prescriber, nurse, pharmacist, and the laboratory, this system largely improves heparin safety. It tracks the medication use and administration schedules. In addition, the system is used to standardize practice protocols and provide a quality assurance program which facilitates UFH dosing and safety improvements. The core system is set up to interact with other Mayo Clinic computerized systems so that patient demographic information, medications, laboratory results, microbiology results, and nutritional data are interfaced into the pharmacy computer system.7 First, the physician or clinician puts in a written intravenous heparin order and is then sent through the hospital computer system where both the pharmacist and nurse can access the order. Then, the nurse who is taking primary care of the patient receives the order and accesses HepCare through the computer, unless the pharmacist intervenes to institute corrections to the order. This system prompts the nurse to determine whether administration of UFH is appropriate for the patient. Patients are prohibited from entering into the protocol if they are allergic to UFH or currently receiving anticoagulants such as direct thrombin inhibitors or low-molecular-weight heparin (LMWH). If a patient is an appropriate candidate for the protocol selected, the HepCare system calculates the UFH bolus and infusion rate clinically based upon the nomagram and the patient’s weight. Afterward, the nurse reviews the recommended UFH dosage via the HepCare system. If the nurse accepts the recommendation, the computer submits an electronic order for the next activated partial thromboplastin time (aPTT) to be drawn in six hours. Subsequently, the nurse administers the UFH loading dose and continuous infusion drip as calculated through the computerized HepCare system. The laboratory value of the follow-up aPTT is transferred directly into the HepCare system. If the aPTT laboratory result is not accessed by the system within a defined time, both the pharmacist and nurse are notified every hour of the necessity of obtaining the follow-up aPTT value until the result is available.7 The HepCare computer system internally determines whether or not the follow-up aPTT is within goal. If the aPTT is within goal, the computer will submit the order for the next aPTT; however, if the aPTT is not within goal, the HepCare system will then recalculate the dosage and send the nurse the recalculated dose. If a reported aPTT value is inconsistent with prior aPTTs, without UFH dosing changes, the system requires a validation aPTT before suggesting dose changes. This cycle is repeated until the UFH is discontinued by the clinician.7 There are several safety measures built into the HepCare system. Measures include the detection of drug-drug interactions, automatic aPTT level ordering and monitoring, validation of the patient weight, and continuous assessment of laboratory values. In addition, the clinician is notified when there is a drop in hemoglobin of greater then 2g/dL or the platelets drop to less than 100,000. Utilizing the HepCare system has decreased the frequency of medication errors from 0.5 per patient to 0.006 per patient.7 The major goal of this system is to allow more efficient and accurate use of intravenous heparin and to prevent medication errors. This computer system dramatically reduces the amount of work and effort by the clinician, allows standardization of practice, and promotes methods for continuous improvement through real-time reporting capabilities and quality assurance.12 Some disadvantages with this system include propagating terrors, technology downtime, physician opposition and complexity. Smart pumps differ from older infusion pumps since they can be programmed to incorporate a drug library, a directory of parental medications with their admixture concentrations and they have software functionality that provides point-of-care decision support for overly high or low intravenous infusion rates. The libraries are arranged and managed by the hospital pharmacy department with input from the pharmacy and therapeutics committee, as well as patient care areas. The device prompts the user to choose a medication from the library, confirm the selection, input a volume to be infused, and input an infusion rate or dose. For all medications selected from the library, the keypad entry of an infusion rate in milliliters will automatically calculate the equivalent dose in units, milligrams, or micrograms.9 Safeguards integrated into the device include not permitting the patient’s weight to be used if the drug is not dosed according to weight which provides less room for error, identifying whether the drug is already infusing on another channel, and alerting the user if the dose exceeds or is less than the institution-established predetermined limit. In one study, unfractionated heparin administration generated a significant number of alerts versus other high alert medications. For example, there were a total of 501 alerts generated, of which 250 were overdose alerts and 178 were under dose alerts. The results also displayed 73 duplicate drug therapies or same drug infusing alerts. When the alerts were created, the smart pump users either reprogrammed the device or cancelled the infusion. These outcomes demonstrate that the smart pump infusion facilitated in the prevention of heparin dosing errors; however, in many cases the alerts were ignored and the infusion was continued.9 Implementation of a computerized physician order entry (CPOE) system is also another proposed solution to reduce medication errors. Often illegible handwritten orders, overlooked drug allergies and incorrect dosage lead to medication errors.4 These systems provide feedback by giving clinical decision support such as dose checking, alerts and reminders. Using CPOE can prevent these transcription errors, adverse drug interactions, mis-dosages, and prescription errors. By utilizing this program, it allows physicians to select the drug they want to prescribe along with the correct dose from a computerized menu. CPOE programs are often used alone in healthcare institutions; however, integrating this system with other forms of technology can significantly reduce medication errors. Although technology has been proven beneficial towards the reduction of medication errors, it can also have a negative impact when the safety alerts become excessive. When the frequency of reminders and alerts become too much, alert fatigue may develop. This can lead clinicians to override both significant and insignificant alerts which compromise the desired safety effect. Studies show that clinicians override drug safety alerts in 49 – 96% of cases and it is recommended that valuable alerts should be distinguished from those that ought to be ignored.10 It is imperative to institute practices to minimize inappropriate warnings, such as high specificity, information content, sensitivity workflow, and efficient handling. In regards to specificity, alerts should be clinically relevant for each individual patient and should not be of minor importance. The content of the warning should be clear and concise. In addition, sensitivity workflow should be implemented to decrease disruptions in the clinical environment. Examples of sensitivity workflow include directing alerts to the appropriate personnel to take action against to the annoying repetitive alerts. Lastly, efficient handling refers to minimizing the number of keystrokes, mouse clicks, and screen changes.10 The aforementioned practices should help reduce the amount of alert fatigue currently seen.
Conclusion
As a result of heparin overdose cases such as these and other less public ones, major hospitals all across the nation have implemented measures for preventing heparin medication errors. Implementation of an FMEA system would serve as a crucial method to help recognize potential failures of a product or process before adverse events occur.1 An FMEA can help identify where the use of technology can be implemented to facilitate the reduction of medication errors, especially pertaining to heparin as in this case. Various solutions to the problems pertaining to the similar vials of heparin include using standardized doses, pre-filled syringes, pre-mixed intravenous bags, differentiating labels, and adding warning labels. Although, these solutions have proven beneficial they may present with limitations of space constraints in the pharmacy and automated dispensing cabinets, limited concentrations available, as well as increased industry and health-system costs. Solutions to problems pertaining to human errors include a multidisciplinary team, multiple checkpoints, as well as additional training and personnel. These solutions may prove to be valuable; however; they can be time consuming, budget constraints and can lead to miscommunication between healthcare professionals. Errors related to high-alert medications are often attributed to variations in the prescribing, dispensing, and administration of the drug. Solutions to prescribing and dosing errors include using bar coding technology, HepCare software, the smart infusion pump, and the computerized physician order entry program. Disadvantages to using these methods of technology include propagating errors, system failure, physician opposition, technology downtime, cost-related issues and complex systems. Implementation of technology can be used as a tool, among many other, that needs to be effectively used in order to achieve safe medication use by health institutions. Ultimately, the understanding of human factor elements is crucial in order for these systems to perform well when dealing with the interfaces between people and technology.6 Medication errors can be contributed to human error as well as a system-wide dilemma. Thus, it is necessary to address resolutions to both of these predicaments. Implementation of an FMEA would serve as a crucial method to recognize potential failures before adverse events occur. Various methods of technology should also be implemented to facilitate the reduction of medication errors, especially with the high-alert anticoagulant heparin. Employment of these preventative steps will help to reduce errors, such that the incidences experienced by the Indiana and Quaid babies should not be repeated. Expanding nationwide awareness of these methods should result in a significant decline of heparin medication errors.
References
1. Duwe B, Fuchs BD, Hansen-Flaschen J. Failure mode and effects analysis application to critical care medicine. Critical Care Clinics. 2005; 21(1).
2. Runy LA. A guide to safer use of dangerous medications: high-alert medications. Hospital and Health Network. 2004
3. TheIndyChannel.com. Hospital changes procedures after babies’ fatal overdoses. Indiana News. www.theindychannel.com/news/9879402/detail.html (accessed 2008 Feb 13).
4. Crane J, Crane F. Preventing medication errors in hospitals through a systems approach and technological innovation: a prescription for 2010. Hospital Topics. 2006; 84(4): 3-7.
5. Resource Engineering, Inc. What is failure mode and effects analysis? FMEA Overview. www.reseng.com/cveroverviews/fmea_overview.htm (accessed 2008 Jan 31).
6. American Hospital Association, Health Research and Educational Trust, and the Institute for Safe Medication Practices. Pathways for medication safety: assessing bedside bar-coding readiness. http://www.ismp.org/Tools/PathwaySection3.pdf (accessed 2008 Apr 8).
7. Oyen L, Nishimura R, Ou NN et al. Effectiveness of a computerized system for intravenous heparin administration: using information technology to improve patient care and patient safety. Am Heart Hosp J. 2005; 3: 75-81.
8. Food and Drug Administration. Heparin Sodium Injection, USP. www.fda.gov/cder/foi/label/2007/017037s161lbl.pdf (accessed on 2008 Feb 10).
9. Fanikos J, Fiumara K, Baroletti S et al. Impact of smart infusion technology on administration of anticoagulants (Unfractioned Heparin, Argatroban, Lepirudin, and Bivalirudin). The American Journal of Cardiology. 2007; 99(7): 1002-05.
10. Van der Sijs H, Aarts J, Vulto A, Berg M. Overriding of Drug Safety Alerts in Computerized Physician Order Entry. Journal of the American Medical Informatics Association. 2006; 13(2).
