For the past several months we had been collecting some recently published studies on medication safety to include in our What’s New in the Patient Safety World column. However, there have been so many of them that are clinically relevant that we thought we needed to include them in one of our Patient Safety Tips of the Week where we discuss them in more detail. Hence, today’s column.
A study from the UK (Seden 2012) looked at prescribing errors in admission and discharge prescriptions. One or more errors was found in 43.8% of prescriptions, over half of which were considered significant. They found no significant difference in error rates by level of experience of the prescribing physician. The strongest predictor of error was the number of items per prescription. The risk of error increased 14% for each additional item in the prescription. Medication omission errors accounted for 26.9% of errors, followed by writing errors and dosing errors at 20.7% and 20.6%, respectively. Medication omissions were very common in patients acutely admitted. The authors recommend early use of clinical pharmacists in the medication reconciliation process and better use of IT systems.
IT systems, however, are not the full answer. Another recent study (Linsky 2013) looked at medication discrepancies in the electronic medical record system of ambulatory care patients in the VA system in Boston. They found 60% of patients had at least one medication discrepancy. Prevalence rates of commissions were 36%, omissions 27%, alterations 19%, and duplications 11%. Interestingly, they found an increased number of medications correlated with more commissions and duplications but fewer omissions. The authors conclude that relying on EHR’s alone will not ensure an accurate medication list.
Dose omission incidents may potentially lead to patient harm. A study from ISMP Canada’s reporting database found 159 incidents of patient harm from dose omissions across different hospital settings over a 2.5-year period (ISMP Canada 2013). Most involved mild or moderate harm and there were no cases of severe harm or death. As you might expect, drugs we consider high-alert medications were most frequently involved. Insulin was involved in 18.9% of incidents and heparin in 13.2%. Other drugs frequently involved were potassium chloride, metoprolol, aspirin, HYDROmorphone, metformin, and warfarin. The authors did qualitative analysis of 79 incidents and found several themes:
· The most frequently implicated medications are used in acute clinical situations (eg. acute coronary syndrome or MI, diabetic ketoacidosis, etc.)
· Many were high-alert medications with narrow therapeutic ranges
· Medications associated with negative outcomes were often omitted or delayed
They provide multiple examples of system-based factors contributing to harmful dose omission incidents, such as delivery of medications to a prior area after a patient had been transferred, misplacement of the MAR, lack of understanding of nomograms, etc. But they also identified several themes of at-risk patient care processes:
· Patient transfers
· Complex medication orders
· Medication administration record (MAR) issues
When we do an introduction to patient safety during incoming residents’ week for all the new residents in the Buffalo academic consortium of hospitals we especially focus on medication errors. One of the patient safety practices we strongly encourage them to use is including the indication for the medication any time they order a medication. We provide them numerous examples of look-alike sound-alike (LASA) medication errors that might have been prevented had the prescriber included the indication for the medication. Unfortunately, many CPOE systems do not include a field for indication. A recent paper (Wei 2013) showed how a publicly available, computable resource that links medications with their indications as represented by concepts and billing codes might be used to benefit clinical EMR applications. Another recently published study looked at use of indications tied to the patient’s problem list during CPOE (Galanter 2013). Over a 6-year period over 100,000 alerts fired and resulted in 32 intercepted wrong-patient errors (an interception rate of 0.25 per 1000 alerts). They were able to determine that the prescriber had both patient charts open in almost 60% of those instances (see our comments below about the problem of having multiple charts open simultaneously).
Speaking of wrong-patient medication errors, the Pennsylvania Patient Safety Authority (PPSA) has just put out an analysis of 813 such occurrences over a 6-month period from their data base (Yang 2013). Almost 30% were associated with high-alert medications (insulin, opioids, anticoagulants, etc.). Fortunately, few of these errors resulted in patient harm. But such errors occurred in multiple healthcare settings and involved every phase of the medication process, though the majority occurred during the medication administration phase. Inadequate patient identification checks were common contributory factors. Sometimes nurses (incorrectly) relied on the patient or family’s confirmation of the patient’s name to verify identity. In other cases room numbers (which you’ll recall from the Joint Commission NPSG’s should not be used as a patient identifier) were involved or the medications of two patients in a room were confused. Failure to correctly use the MAR was another contributory factor identified. Transcription errors were also noted, such as transcribing an order into the wrong chart or affixing the wrong patient label to an order. Verbal orders were also identified as contributing factors. Prescribing errors were also common, with the prescriber entering the order on the wrong chart or giving verbal orders on the wrong patient. Errors during the dispensing phase included both filling and delivery errors. Errors in the monitoring phase most often related to laboratory values.
The most common contributing factor they identified was the same medication being ordered (but different doses for each of the two patients). This was considered a factor in 6.4% of the events. Verbal orders and similar patient names were considered factors in about 3% each. And 1.4% were related to confusion with a discharged patient (eg. a medication intended for a patient who had previously been in that bed).
The PPSA article offers several risk reduction strategies. Following the Joint Commission NPSG for accurate patient identification (at least 2 identifiers, which are not to include the patient’s room number or location) is an obvious one. So is limited use of verbal orders. Proper storage of medications needs to be addressed (eg. correct labeling of bins, ensuring that medications are removed from beens and ADC’s when patients are discharged, etc.). Empowering the patient and family, encouraging them to question and speak up, are important as well since many such potential errors were intercepted by patients or families.
Technology may be important. Proper use of barcoding systems is important. Similarly, using all the capabilities of automated dispensing cabinets (ADC’s) is important. Transitioning from paper to CPOE is recommended, though they appropriately point out that such errors may occur during CPOE as well (see also our Patient Safety Tips of the Week May 20, 2008 “CPOE Unintended Consequences: Are Wrong Patient Errors More Common?”, June 26, 2012 “Using Patient Photos to Reduce CPOE Errors”, and July 17, 2012 “More on Wrong-Patient CPOE”).
Despite the numerous examples of errors introduced by CPOE, the overall effect of CPOE on medication errors has been beneficial. A recent systematic review and meta-analysis (Radley 2013) estimates that processing a prescription drug order through a CPOE system decreases the likelihood of error on that order by 48% and that, given this effect size and the degree of CPOE adoption and use in hospitals in 2008, an overall 12.5% reduction in medication errors in the US annually. The authors do, however, acknowledge that it remains unknown whether this error reduction results in a reduction in harm to patients or not.
Another recent study demonstrates the value of the Leapfrog CPOE evaluation tool in monitoring and reducing the occurrence of preventable adverse drug events (Leung 2013). A 43% relative reduction in the rate of preventable ADE was predicted for every 5% increase in Leapfrog scores. We previously described use of the Leapfrog tool in our July 27, 2010 Patient Safety Tip of the Week “EMR’s Still Have a Long Way to Go” and our June 2012 What’s New in the Patient Safety World column “Leapfrog CPOE Simulation: Improvement But Still Shortfalls”.
Many of the drugs noted in several of the studies noted above and another recent review of factors associated with preventable adverse events (Beckett 2012) were high-alert medications. These are drugs associated with a high likelihood of patient harm if given incorrectly or omitted incorrectly. So timely is a new ISMP Medication Safety Alert that points out that many hospitals have high-alert medication lists that are ineffective because they don’t have associated risk-reduction strategies in place (ISMP 2013). It has a great list of key strategies for organizations to implement at multiple levels to reduce the risks of harm from drugs on those high-alert medication lists.
One of those recommendations is doing a FMEA (failure mode and effects analysis). A good example of how to do a FMEA related to medications was also recently published (Lago 2012). They did a FMEA on medication prescribing and administration on pediatric wards and found 37 higher-priority potential failure modes and 71 associated causes and effects. So they were able to identify multiple targets for improvement from a safety perspective.
It is important to recognize that medication errors may vary across different settings. For example, in the pediatric study noted above errors in calculating drug doses and concentrations had the highest risk priority numbers (RPN’s). Similarly, in adults there are differences in medication errors between ICU and non-ICU settings (Latif 2013). The authors, reviewing data from the MEDMARX database found ICU medication errors to be more likely associated with patient harm compared to those in non-ICU settings. While errors occurred most often in the administration phase in both settings, administration errors were more frequent in ICU settings. Leading sources of errors identified in the ICU were deficits in both knowledge and performance, procedure or protocol not being followed, communication, and dispensing device errors. Sources in non-ICU settings were similar except that inaccurate or omitted transcription was more frequent than dispensing device errors. The authors ascribe the greater risk of harm in ICU patients to a variety of factors, including patients being older and sicker with more comorbidities, more medications, more potent drugs, more infusions, more necessary dose calculations, and others.
Also, somewhat surprising in an era of disclosure and apology, when medication errors did occur patients and families were apparently seldom informed about them. This was also seen more often in ICU patients.
Another striking finding in the Latif study was that the staff member responsible for the error was informed of the error a third of the time or less. While feedback should be done in a constructive rather than punitive manner in most circumstances, such feedback is critical if we really expect to reduce errors. A recent study providing feedback regarding prescribing errors to prescribers in a NICU demonstrated a significant improvement in narcotic prescribing errors but not in antibiotic prescribing errors (Sullivan 2013). The authors in the NICU study speculate that narcotic prescribing errors were more likely due to “slips” and “lapses” and were more amenable to feedback than antibiotic prescribing errors, which were much more complex. The best part of the Sullivan study, however, is a great description of the barriers encountered and development of the best way to deliver the feedback. They found that delivering the feedback via carefully structured emails that were short, personal, informative and constructive worked best. They were more likely to be read when the message was in the body of the email rather than in an attachment. And the inclusion of a disclaimer reinforcing the non-punitive nature of the feedback was felt to be very important.
Another study (Eijsbroek 2013) from the Netherlands looked at medication related problems occurring in patients that had been in an ICU during a hospitalization. While the overall occurrence of such problems was relatively infrequent, they did identify cases where chronic medications had not been continued and others where medications intended for short-term use had been inadvertently continued. Our August 30, 2011 Patient Safety Tip of the Week “Unintentional Discontinuation of Medications After Hospitalization” discussed a Canadian study (Bell 2011) showing quite similar issues. The Netherlands study attributed the relatively infrequent occurrence of problems to good medication reconciliation in hospital and to use of a followup clinic that specifically addressed medication issues.
Lastly, the issue of dissemination of medication safety alerts arose again. In our February 26, 2013 Patient Safety Tip of the Week “Insulin Pen Re-Use Incidents: How Do You Monitor Alerts?” we focused on the problem of providers not seeing important safety alerts. However, our focus was primarily at the institutional level. Of course, failure to see (or heed) such alerts is just as important at the individual provider level. An abstract just presented at the American Academy of Neurology annual meeting showed just how widespread and significant that problem may be (Bell 2013). The authors surveyed U.S. neurologists on their knowledge of four recently announced safety risks regarding antiepileptic drugs and determined whether they altered patient care as a result. They found approximately 20% of neurologists were not aware of four major drug safety risks. Many others were aware of such alerts but had not altered their practices as a result of those alerts. Though neurologists learned about drug safety risks from many sources only notifications from specialty organizations were associated with an accurate knowledge of safety risks.
Alerts come from multiple authoritative sources (FDA, ISMP, CDC) and are often also noted in resources like Medscape, specialty journals, trade publications, manufacturers and others. But that may actually also be part of the problem. The alerts come from so many sources, are often duplicative, and may be included with many other alerts that are not relevant to all providers. Often when we get an email from the FDA it has alerts on medications, foods, and devices all interspersed. There are often so many that we miss the important ones as we rapidly scroll through the document(s). Having the alerts come through in a filtered manner would obviously be helpful. Neurologists receive drug safety information non-systematically from multiple sources. Most would prefer implementing "a formal warning process via specialty organizations" and emails with updated product insert warnings directed to specialists.
Seden K, Kirkham JJ, Kennedy T, et al. Cross-sectional study of prescribing errors in patients admitted to nine hospitals across North West England. BMJ Open 2013; 3: 1-14 e002036 doi:10.1136/bmjopen-2012-002036
Linsky A, Simon SR. Medication discrepancies in integrated electronic health records. BMJ Qual Saf 2013; 22(2): 103-109 Published Online First: 25 October 2012 doi:10.1136/bmjqs-2012-001301
ISMP Canada. Aggregate Analysis of Dose Omission Incidents Reported as Causing Harm. ISMP Canada Safety Bulletin 2013; 13(2): 1-6 March 27, 2013
Wei W-Q, Cronin RM, Xu H, et al. Development and evaluation of an ensemble resource linking medications to their indications. J Am Med Inform Assoc 2013; Published Online First: 10 April 2013 doi:10.1136/amiajnl-2012-001431
Galanter W, Falck S, Burns M, et al. Indication-based prescribing prevents wrong-patient medication errors in computerized provider order entry (CPOE). J Am Med Inform Assoc 2013; 20(3): 477-481 Published Online First: 9 February 2013 doi:10.1136/amiajnl-2012-001555
Yang A, Grissinger M. Wrong-Patient Medication Errors: An Analysis of Event Reports in Pennsylvania and Strategies for Prevention. Pa Patient Saf Advis 2013 [prepublication]
Radley DC, Wasserman MR, Olsho LEW, et al. Reduction in medication errors in hospitals due to adoption of computerized provider order entry systems. J Am Med Inform Assoc 2013; Published Online First 20 February 2013 doi:10.1136/amiajnl-2012-001241
Leung AA, Keohane C, Lipsitz S, et al. Relationship between medication event rates and the Leapfrog computerized physician order entry evaluation tool. J Am Med Inform Assoc 2013; Published Online First 18 April 2013 doi:10.1136/amiajnl-2012-001549
Beckett RD, Sheehan, AH, Reddan JG. Factors Associated with Reported Preventable Adverse Drug Events: A Retrospective, Case-Control Study. Ann Pharmacother 2012; 46: 634-641; published ahead of print April 17, 2012, doi:10.1345/aph.1Q785
ISMP (Institute for Safe Medication Practices). Your high-alert medication list—Relatively useless without associated risk-reduction strategies. ISMP Medication Safety Alert! Acute Care Edition 2013; April 4, 2013
Lago P, Bizzarri G, Scalzotto F, et al. Use of FMEA analysis to reduce risk of errors in prescribing and administering drugs in paediatric wards: a quality improvement report. BMJ Open 2012; 2:6 Published 18 December 2012 e001249 doi:10.1136/bmjopen-2012-001249
Latif A, Rawaat N, Pustavoitau A, et al. National Study on the Distribution, Causes, and Consequences of Voluntarily Reported Medication Errors Between the ICU and Non-ICU Settings. Crit Care Med. 2013; 41(2): 389-398
Sullivan KM, Suh S, Monk H, Chuo J. Personalised performance feedback reduces narcotic prescription errors in a NICU. BMJ Qual Saf 2013; 22(3): 256-262 Published Online First: 4 October 2012 doi:10.1136/bmjqs-2012-001089
Eijsbroek H, Howell DCJ, Smith F, Shulman R. Medication issues experienced by patients and carers after discharge from the intensive care unit. Journal of Critical Care 2013; 28(1): 46-50
Bell CM, Brener SS, Gunraj N, et al. Association of ICU or Hospital Admission With Unintentional Discontinuation of Medications for Chronic Diseases. JAMA 2011; 306(8): 840-847
Bell S, Matsumoto M, Shaw S, et al. (Abstract) Evaluating U.S. Neurologists' Knowledge of New Anti-Epileptic Drug Safety Risks (S48.005). Neurology February 12, 2013; 80(Meeting Abstracts 1): S48.005
Print PDF version