Karen M. Hunter RN, BSN
This article was made possible by an educational grant from
Chamberlain College of Nursing
Hunter, K. (June 2011). Implementation of an Electronic Medication Administration Record and Bedside Verification System. Online Journal of Nursing Informatics (OJNI), 15 (2), Available at http://ojni.org/issues/?p=672
Technology designed to facilitate the medication administration process in healthcare facilities promises to reduce medication errors. This article describes one hospital’s journey on the road to medication safety.
Due to staffing shortages and the expectations placed on staff today, human errors are becoming more prevalent in the administration of medications. Several articles have been written to discuss the percentage of incidents and the amount of money spent on medication errors. The first introduction was with the 1999 Institute of Medicine report which stated 44,000 to 98,000 Americans die annually due to medical mistakes. (Cohen, 1999; Cousins, 1998) Other statistics include:
7,000 die each year due to medication errors (Hughes, 2005)
400,000 preventable drug-related injuries every year (Mullan 2005)
There are 3.5 incidents per 1000 medications administered (Medical News today, 2008)
Medication errors result in at least one death per day and 1.3 million injuries per year in the US.(Patel, 2009)
Serious injuries reported to FDA because of hospital drug errors have increased from about 35,000 in 1998 to about 90,000 in 2005.
(Medical Errors, 2010)
Information technology has been shown to decrease medication errors (Figge, 2009). Among the types of technology is barcoding of medications. The percentage of the decrease of errors related to barcoding ranges from 60 to 97% (Medical News Today, 2008, Patel 2009, Figge, 2009). Unfortunately, only 10 to 12% of hospitals have barcoding in place. (Stencel 2006) The use of technology, with its checks and double checks can be a valuable tool for today’s nurse. Had this technology been in place, some of the more publicized medication errors in recent years may have been prevented, with harm to the patient reduced and lives saved. In 2007, twins of actor Dennis Quaid were given the wrong strength of Heparin, catapulting medication errors into the national spotlight. Quaid has appeared before Congress and on CBS news to raise awareness of this issue.(Kaushal, 2001) In 2009, two pregnant women at a Florida hospital received Prostin, a drug that induces labor instead of the ordered progesterone suppositories the doctor ordered to prevent premature labor. The results were the death of two twins and the premature birth of a baby that suffered severe brain damage. (Inglesby, 2006)
This paper will review the implementation of the Electronic Medication Administration Record (eMAR) and the Bedside Medication Verification (BMV) module at Chilton Memorial Hospital, a 256 bed hospital in Pompton Plains, New Jersey. The hospital has nearly all of the Meditech HCIS modules installed in the inpatient and outpatient areas. Meditech, based out of Boston, MA and has over 2,200 customers worldwide. Through implementation of the eMAR and BMV, the 5 rights of medication administration are maintained (right patient, right medication, right dose, right route and right time), notifications are at the nurse’s fingertips, errors and warnings are readily available, allergy checking is automatically performed, dose checking and other relevant clinical data are accessible. This paper will discuss one hospital’s journey on the road to medication safety.
A medication error can be described best as an “unintended act or as an act that does not achieve its intended outcome.” (Wideman, 2010) The delivery of a single dose of a medication is the end result of a complicated process involving 10 to 15 steps, each of which offers an opportunity for error. (Wideman, 2010) With 26-38% of errors occurring at administration, (LaDuke. 2009, Stencel, 2006) Chilton Memorial embarked on the mission to improve patient safety in 2007. One of their main initiatives was to implement Meditech’s Bedside Medication Verification (BMV) system. One of the first tasks to be done was flowcharting the current process flow. An integral part of planning and implementing an online medication system is to know where the facility is and where they want to go. Flowcharting the current process shows the benefits and risks of what is in place. Flowcharting the new process shows what changes need to be made. Flowcharting will help teams to understand all steps in the process by serving to identify ineffiencies, misunderstandings, redundancies and areas of neglect while providing insights into how a process should be performed. (Anderson, 2009)
A core group was created of 8 members including an IS (Information Systems) nursing analyst, IS pharmacy analyst, IS project manager, IS consultant, IS/Nursing Liaison, nurse manager, pharmacist and quality assurance nurse. Eight hour work sessions were utilized on a weekly and as needed basis. These sessions were comprised of dictionary building, testing and process improvement.
Many processes that were manually done prior to an online system can now be automatically completed through computerized documentation screens, lab view groups, access groups and verb strip functions. Documentation screens can have one or more queries in them and can be made required. This ensures consistency of charting and fulfillment of regulatory and facility mandates. (Figure 1)
Lab view groups provide results when a nurse selects a medication for administration. For instance, when the nurse documents on Coumadin, the PT and INR will display on the screen. In the Meditech HCIS system, the display can be for a certain number of results or days. The results can be marked for high, low, critical and delta checks as defined in the Laboratory module. (Figure 2)
Warnings and alerts like the lab view groups, while helpful to the staff, can be overused at times, resulting in what is known as “Alert Fatigue.” This occurs when the end user, after receiving too many alerts, begins to ignore and/or override the alerts. A recent review stated that safety alerts are overridden by clinicians 49-96% of the time (Bar Code Technology, 2010). Some of the alerts this system has include patient identification warnings, dose warnings and allergy warnings. (Figures 3, 4, 5) An occurrence happened where the nurse pulled the correct medications for the correct patient, pulled up the electronic MAR and scanned the medications, but then entered the wrong room. When she scanned the patient and received the error warning that the medications were not on this patient’s MAR, she tried again. After 3 attempts to scan the patient and receiving the error message, the nurse, frustrated by “the system” exited out of the MAR and gave the patient the medications, including ones for high blood pressure and insulin. While no ill effects were found, this patient had to be closely monitored. It is vital that the system is seen, not as a nuisance, but as a benefit and that the warnings are not being ignored (Cummings, 2005).
Access groups can help to define which users have certain abilities. Information can be displayed or restricted per user group. Within each access group, verb strip functions are allowed. Each verb strip function handles a different action in the system. Some examples are Scan Med, Edit Doc, Undo Doc, and Full Doc. Restriction of what types of orders are seen as well as how the medications are listed may also be done.
Searching for the right equipment can be a daunting task. It is important to keep in mind that a variety of tools may be a better fit depending on the individual needs of the facility. Many vendors will allow the facility a testing period for their equipment and this is most beneficial when trying to compare the pros and cons of barcode scanners, laptops and computers on wheels. This facility used Artromick computers on wheels, Metrologic scanners and the Aruba wireless system. Pyxis is the dispensing machine system that is utilized on each nursing unit, Emergency room and Operating suite. It is important to remember, when selecting equipment, “one size does not fit all!”
Most of the dictionary building was done in the Pharmacy module. Originally, the module was built by a pharmacist, for a pharmacist. Now adding the eyes of Nursing and Physicians, several cosmetic changes needed to be made to the drug dictionary regarding descriptions and strength fields, as well as adding administration criteria.
Parallel testing is vital to the successful implementation of any IS project. The purpose is to test as many scenarios as possible to find any problems, shortcomings, or areas for improvement. Documentation is done in the current process and in the future process. For this hospital, this meant double documenting – once on paper, and once in the computer system. While 2 weeks of parallel was suggested, the unit manager decided on 4 weeks of parallel testing for the nurses after 8-hour training was completed. Having such a long parallel run assisted the core team in tweaking the system and working out issues prior to the live date. It also gave the unit staff time to practice, make mistakes, and get additional help before the documentation online was kept as part of the official medical record. Implementing a system without proper testing would be detrimental to the project. Thorough analysis of the system through testing will display the weak areas and opportunities for workarounds that the staff will be sure to discover.
While other units did not get a 4-week parallel run, they did receive additional training on the job. After each unit was trained in the 8-hour class, each nurse was required to go to the unit that was already live with the system to perform one or more medication passes with an experienced nurse.
Starting the new system on a pilot unit is recommended before performing a full implementation. This allows for studying the results on a small scale and being able to make adjustments as needed. (Anderson, 2009) When the pilot unit went live, coverage was around the clock for 2 weeks. Superusers manned the units along with IS analysts. Once the pilot unit live was over, consecutive units came live every 2 weeks. Reports were run on a daily, weekly and shift-by-shift basis. The manager of the pilot unit closely reviewed these reports and went over discrepancies with the staff to make corrections immediately. This quick on-the-unit instruction helped to reinforce their classroom education. As each concurrent unit came live, the pilot unit manager would meet with the manager of the new unit and review the reports and audits. Due to the diligence of this manager, the scan rate for the inpatient units is in the high 90’s. Reasons for not scanning continue to be reviewed for satisfactory responses.
A quarter of all Americans have received, or know a family member who has received the wrong medication from a healthcare professional. (Cousins, 1998) A survey showed the 61% of Americans were “very concerned” about being given the wrong medicine. (Cohen, 1999) This shows that lay people are becoming more aware of medication errors through personal experience and through various media coverage. Studies have found that a well educated patient is a strong defense against errors and this will result in better outcomes (Medical Errors, 2010, Anderson, 2003). Educating the patient on the barcode system for medications as well as the armband barcode can only decrease the safety risk. Errors in patient identification often occur because of distraction and interruptions. (Wideman, 2010) When the system first went live, a handout was given to each patient explaining the importance of their safety to the hospital. It explained the purpose of scanning the medications and the patient armband and empowered the patient to question the nurse if she tried to administer a medication without scanning their armband. The patients became a co-manager of their care and even began holding out their armband to be scanned when medications were administered. This gave them a feeling of not only safety in their care, but also participation in their treatment.
Scanning rates for medications and patients began and have remained mainly in the 90’s. This was due to the diligence of the managers monitoring reports and providing quick feedback to the nurses. One month after go-live on the initial unit, a post-live survey was prepared. Sixty-two percent of nurses stated they felt “safer” using the system because it made them “follow the 5 rights of medication administration” and gave them pertinent information including vitals signs, lab results and notifications for timely medication administration. The most significant finding was that 62% of the nurses taking the survey stated the new system had actually prevented them from making a medication error. To the senior management, that spoke volumes. The system benefits far outweighed any cost to the facility.
As monitoring continued, patterns began to emerge. Certain medications that the staff were unable to scan were reviewed for correct dictionary entry, barcode readability and printer functionality. If the Pharmacy started receiving a medication from a different manufacturer, the barcode was not always checked to see that it matched what was currently in the system. Scanners needed recalibrated at times and are now on a routine maintenance schedule. The vendor for armbands was replaced and printers realigned to prevent smudging of the barcode.
While a computerized system can show multiple benefits, there are vulnerabilities that must be discussed. One of the hazards of using computerized systems is the dependency the staff now have on computers. Using barcode scanning can produce a blind faith in technology and can produce a false sense of security. (LaDuke, 2009, Anderson, 2009) While the system displays numerous warnings and errors, staff are inclined to press the Enter key and bypass the information. Because of this, it is important to not put too many stops in the systems while the nurse is trying to administer medications. One study showed that nurses override technology 10.3% for medications charted and 4.2% of the time for patients charted. (LaDuke, 2009) It must be stressed in education that the system is there to be in adjunct to the nurse, it is not a replacement for the nurse. The staff is still accountable for the five rights of medication administration.
Workarounds to the system are often found and utilized by staff. These may include administering doses from floor stock, borrow doses from another patient, and obtaining doses on override. (Wideman, 2010) Staff may also place the patient ID on another object instead of the patient. (Koppel, 2008) This was found to be happening on a regular basis on some units. The staff did not want to disturb the patient and sometimes family members were adamant about not waking the patient when asleep. Education of the staff as well as the patient was needed to correct this problem. Another issue found was that pharmacy would send up a single dose in a baggie that had a label on the outside for a previous medication. Nurses would mistakenly scan the barcode on the label on the baggie instead of the barcode on the medication.
Some of the causes documented for workarounds are unreadable barcodes, malfunctioning scanners, lost wireless connectivity, just to name a few. (LaDuke, 2009) Managers must keep on top of these situations and work to resolve these issues in a timely manner. The pilot unit manager reviewed real time data and reports daily to give nurses quick feedback. This helped to reduce workarounds and increased the scan percentage of her unit. When this was not done by other managers, the percentage for their scan rates was not as high. The longer this is allowed to continue without correction, the harder it is to fix.
Since the initial implementation of the inpatient units, the facility has put the MAR and/or BMV modules into practice in the Operating Room, Recovery Room, Outpatient infusion/chemotherapy unit, Radiology, Catheterization Lab and the Emergency Department. As the Emergency Department has now implemented their module, including documentation, computerized physician order entry and the use of the online MAR, the journey continues for patient safety and medication administration improvement. Processes in the ED pose new challenges that often require creative thinking indeed. One point of education was for stat medications. At Chilton Memorial, a stat is defined in the following way: If the medication is not given within 5 minutes, the patient will deteriorate. Many times, the nurse will override a non-stat medication before the doctor has placed the order. This bypasses the safety of the system and can put the patient at risk. An issue was also identified when the physician entered an order that did not match what was available in the dispensing machine. Therefore, the nurse was not scanning the correct vial or tablet. Changes had to occur to allow the doctor to know how medications were stocked in Pyxis so that what he ordered could be scanned. This will become a bigger issue as the hospital plans to go house wide with implementation of Physician Order Management. As always, the safety of the patient must dominate the changes in process.
A successful implementation of bedside medication verification can aid the staff in quicker access to results and other pertinent information supporting better decision-making. It provides safety to the patient, as well as to the nurse. To ensure a positive implementation, the core group must test thoroughly and parallel as long as needed to discover errors, workarounds and barriers. Even in the best systems, errors will inevitably occur. (Wideman, 2010) It is important for upper management to know that the system is not going to make errors disappear completely. What the system will do if used properly is drastically reduce medication errors before reaching the patient.
Alert Fatigue (n.d.). Available at http://www.informatics-review.com/wiki/index.php/Alert_Fatigue, Accessed March 15, 2010.
Anderson,J.G., Jay,S.J., Anderson,M., Hunt,T. (2003). Evaluating the Impact of Information Technology on Medication Errors: A Simulation. JAMIA. Retrieved from http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=342060. Accessed March 6, 2010.
Baker,D.E. (2009). Medication Alert Fatigue: The Potential for Compromised Patient Safety. Hospital Pharmacy Volume 44, Number 6, 460-462.
Bar Code Technology Could Help Curb Drug Errors at Hospitals (2007). iHealthBeat. Available at http://www.ihealthbeat.org/articles/2007/11/26/Bar-Code-Technology-Could-Help-Curb-Drug-Errors-at-Hospitals.aspx?topicID=59. Accessed March 5, 2010.
Bar Code Technology Saves Money and Lives in Health Care Applications (n.d). Integrated Software Design, Inc. White Paper.
Cohen, M.R. (1999). Medication Errors Causes, Prevention, and Risk Management. Boston: Jones and Bartlett.
Cousins, D.D (1998). Medication Use: A Systems Approach to Reducing Errors. Illinois: Joint Commission on Accreditation of Healthcare Organizations.
Cummings,J., Ratko,T., Matuszewski,K (2005). Barcoding to Enhance Patient Safety. College of Medicine. Available at http://www.psqh.com/sepoct05/barcodingrfid1.html. Accessed February 2, 2010.
Dennis Quaid Gives Congress An Earful (2008). Available at http://www.cbsnews.com/stories/2008/05/14/health/main4096047.shtml . Accessed February 2, 2010.
Figge,H.L. (2009). Reducing Medication Errors Using Technology Innovations. U.S. Pharmacist. Available at http://www.uspharmacist.com/content/tabid/92/d/pharmacy_and_technology/c/12633 . Accessed March 5, 2010.
Goth,G. (2006). Raising the Bar. Barcoding has the potential to dramatically reduce medication errors. Healthcare Informatics. Available at http://www.healthcare-informatics.com/ME2/dirmod.asp?sid=9B6FFC446FF7486981EA3C0C3CCE4943&nm=Articles%2FNews&type=Publishing&mod=Publications%3A%3AArticle&mid=8F3A7027421841978F18BE895F87F791&tier=4&id=E4E1FB00EF4C4434A7081D01BFDE42C5 . Accessed 3/5/2010.
Hughes, R.G., Ortiz,E. (2005). Medication Errors. Why they happen and how they can be prevented. American Journal of Nursing Supplement, 14-24.
Inglesby,T. (2006). Reading Within the Lines. Patient Safety & Quality Healthcare. Available at http://www.psqh.com/novdec06/barcoding.html. Accessed February 2, 2010.
Kaushal,R., Barker,K.N., Bates,D.W. (2001). How Can Information Technology Improve Patient Safety and Reduce Medication Errors in Children’s Health Care? Archives of Pediatric Adolescent Medicine Vol. 155, Sep 2001, 1002-1007.
Koppel,R, Wetterneck,T., Telles,J.L., Karsh,B.T. (2008). Workarounds to Barcode Administration Systems: Their Occurrences, Causes, and Threats to Patient Safety. JAMIA, 408-423.
LaDuke,S. (2009). Playing it safe with bar code medication administration. Nursing2009, 32-34.
Medical Errors: The Scope of the Problem. Available at http://www.ahrq.gov/qual/errback.htm. Accessed February 2, 2010.
Medical News Today. Barcoded Technology Used to Reduce Medication Administration Has Flaws (7/2/2008). Available at http://www.medicalnewstoday.com/articles/113498.php . Accessed February 2, 2010.
Mullan,J. (2005). Technology as an Aid to the Nurse – Patient Interaction at the Bedside. American Journal of Nursing Supplement, 39-41.
Patel, A. (2009). Tragic Medication Errors Result in Accidental Abortions and Premature Birth. Available at http://www.abcnews.go.com/print?id=8383062. Accessed March 11, 2010.
Stencel,C., Dobbins,C. (2006). Office of News and Public Information. Available at http://www.allhealth.org/briefingmaterials/MedicationErrors-IOM-1249.pdf . Accessed March 11, 2010.
Wideman,M.V., Whittler,M.E., Anderson,T.M., (n.d.) Barcode Medication Administration: Lessons Learned from an Intensive Care Unit Implementation. Advances in Patient Safety: Vol 3, 437-451, 2010.
Karen M Hunter received a bachelor’s degree at Alderson-Broaddus College in 1989 and began a career in Nursing in West Virginia. Through the years, she has worked on Cardiac Stepdown, ICU, Staff Development and Information Systems. In 2000, she received a certification in Nursing Informatics and has spoken internationally for several years. She has spent over 5 years consulting and is currently working on a full system implementation in Parkersburg, WV.