Patient Safety Tip of the Week

May 7, 2013          Drug Errors in the Home



Probably the four most common safety issues in ambulatory care are (1) diagnostic errors (2) medication errors (3) failure to follow up on test results and (4) missed opportunities to prepare patients for hospitalizations. We’ve often discussed diagnostic errors and failure to follow up on test results and done several columns on the changing nature of the pre-surgical evaluation. But what about medication issues?


Our April 12, 2011 Patient Safety Tip of the Week “Medication Issues in the Ambulatory Setting” discussed a wide variety of medication errors and other medication issues occurring in patients being followed in the ambulatory setting. But most of the studies we noted relied upon chart review, patient surveys, or data on hospitalizations and ED visits resulting from medication issues. None though really addressed how often medication errors actually occur in the home.


A new study in a pediatric population should open eyes about how often medication errors occur in the home in all patients, pediatric and adult. Walsh and colleagues (Walsh 2013) used review of medical records and prescription orders but added direct observational methods to determine how often medication errors occurred at home in patients who were followed in 3 pediatric oncology clinics. They found an overall error rate of 70.2 per 100 patients. The error rate with potential to cause harm was 36.3 per 100 patients and actual harm was seen in 3.6 per 100 patients. These error rates are higher than those typically seen in comparable hospitalized patients. Importantly, nonchemotherapy agents were more often involved in errors than chemotherapy agents.


The group had previously done a multisite study that demonstrated high outpatient medication error rates in both children and adults with cancer but children were particularly vulnerable to home medication errors (Walsh 2009). This prior study was a chart review study and did not employ the direct observation methodology used in the more recent study. Hence it likely underreported medication errors. The overall medication error rate was 8.1 per 100 clinic visits and over half had the potential to cause harm. The rate in adults was 7.1 per 100 visits, with 61% having potential to cause harm. In children the rate was 18.8 errors per 100 visits, with 41% having potential to cause harm. However, for children over half the errors that had the potential to cause harm or actually caused harm occurred at home.


So what are the implications of this study for patients other than pediatric oncology patients? Though the current Walsh study was done in a pediatric oncology population and they caution that the results might not be generalizable to other populations, we think that many of their findings are highly likely to apply to adults as well. Below are some of the issues we think are take-home lessons that might apply equally well to all patients.


The number of medications may be important. The children in the current study by Walsh et al. were on a median of 10 medications at home. That is similar to many adult patients with chronic diseases. A prospective cohort study (Gandhi 2003) in ambulatory practices which found 25% of outpatients had adverse drug events, 13% of which were serious, noted the number of medications was significantly associated with adverse events.


A frequent root cause of errors in both Walsh studies was a change in the dose or frequency between the time the medication was first ordered and the day it was administered. For example, a chemotherapy regimen might be ordered at the start of the treatment plan but typically gets altered along the way due to changes in clinical or laboratory parameters. So a label on a medication container may tell the patient (or parent or other caregiver) to take it differently than the physician intends it to be administered that particular day. In fact, one of the suggested potential interventions in the 2009 Walsh paper was simply not writing the orders until the day of administration after all lab data has been reviewed. In adults there are a variety of medications (eg warfarin, insulin, anticonvulsants) where we are frequently making changes to the regimens that will not be reflected on the label. There is also often a disparity between what the physician thinks the patient is taking and what the patient is actually taking. In one study (Schillinger 2005) 50% of patients reported taking warfarin doses that were discordant with what the physician reported, often resulting in either over- or under-anticoagulation.


The parents in the current Walsh study were particularly well educated and 97% scored adequately on a test of health literacy. But health literacy has often been a factor in home medication errors. Moreover, there is often a disparity between parents’ reading abilities and their numerical literacy (also known as numeracy). Parents with low numeracy may be especially prone to make errors in tasks requiring dose measurement or measurement conversions (see our June 2012 What’s New in the Patient Safety World column Parents' Math Ability Matters”). This highlights the need to address numeracy skills of parents when communicating medication instructions (we suspect the same is likely to apply to adult medication errors as well).


Language may also be a contributing factor to medication errors in the home. One of the earliest studies on the frequency and impact of drug complications in outpatients (Gandhi 2000) found that number of medical problems, failure to explain side effects, and language other than English or Spanish were factors associated with drug complications.


The complexity of the medication regimen is also an important contributory factor. Particularly for pediatric oncology patients many regimens require dose calculations based on weight or body surface area or are altered dependent upon results of white cell counts or other laboratory parameters. Also some drugs are given on very irregular schedules. But adults are also impacted by the complexity of their prescribed regimens. One study (Wolf 2011) gave well-educated volunteers prescriptions for seven drugs and watched them try to figure out how and when to take them all. They could theoretically be consolidated to be taken in 4 dosing sets per day. Yet only 15% were able to consolidate the regimen to 4 times daily or less. Most ended up with regimens taking medications 6 or 7 times daily. Even the instructions “twice daily” and “every 12 hours” resulted in medications being taken at different times. Another study (Choudhry 2011) extended findings in the literature on the negative impact that regimen complexity has on medication adherence.


Drugs that may be prescribed differently for different conditions are also an issue. One of the potentially life-threatening errors in the current Walsh study involved a label on methotrexate to give 8 tablets daily rather than weekly. That error with methotrexate, by the way, is a serious error that we’ve encountered multiple times (see our July 2010 What’s New in the Patient Safety World columnMethotrexate Overdose Due to Prescribing Error”).


The perceived importance of the medications may also be important. In the current Walsh study 2 of the errors resulting in harm had to do with underdosing or failure to fill prescriptions for antacids. Unfortunately, we as providers often do a poor job of emphasizing the importance of certain OTC drugs. Best example is the patient being discharged after an MI. We given them a sheaf of prescriptions for medications and then also say at the end “By the way, take an aspirin daily.” Guess which medication they are most likely to forget!


Handoffs may also be factors in home medication errors. Intervening hospital admissions, emergency department visits, appointments with multiple physicians, and other transitions of care are all opportunities where medication reconciliation may fail. And we are not just talking about provider-to-provider handoffs. Rather handoffs from one parent to another parent or from a parent to another caregiver are also opportunities for miscommunication and consequent errors in the home.


Perhaps a very important contributory factor is lack of feedback about the errors. That applies to both the providers ordering and the parents administering the medications. Most of the errors in the current study were uncovered by direct observation by the study nurse. Neither the prescriber nor the parent would have known about the errors otherwise. We discussed the issue of lack of feedback in recent columns, both in relation to medication errors (April 23, 2013 “Plethora of Medication Safety Studies”) and diagnostic errors (May 2013 “Scope and Consequences of Diagnostic Errors”).


In both Walsh studies, the administration phase was most often associated with medication errors. Most prior studies of ambulatory medication errors (eg. Gandhi 2003, Gurwitz 2003) found problems in the prescribing and monitoring phases were most common. But those studies did not utilize the direct observational methods used in the current Walsh study.



So what are the potential solutions? Most of the articles emphasize patient/parent education and good medication reconciliation. But we all know that the impact of both to date has been suboptimal.


One simple strategy mentioned by Walsh and colleagues (Walsh 2009) was not to write orders for medications until the actual day of administration after all clinical and laboratory parameters have been reviewed. While that may be reasonable for chemotherapy agents, it is not very practical for the vast majority of medications our adult patients are taking at home.


Pharmacist involvement is one potential solution. Quite a few years ago CMS initiated its Medication Therapy Management (MTM) program for Medicare patients. It involved someone (usually a pharmacist or a specially trained nurse) to interview Medicare patients, often in their homes, about their medications and how they were taking them and doing medication reconciliation. Patients were usually chosen based upon their taking a specified number of medications or meeting a specified dollar threshold for spending on medications. We’ve used MTM programs in several other settings (an ACO setting, an Alzhiemer’s Disease assistance program, a managed care organization) and found such to be extremely useful. Whether done by pharmacists or nurses they almost always discover potentially useful changes in the medication regimen that they then take to the provider for potential action. Most often they find therapeutic duplications, medications that were originally begun as temporary prophylaxis but never discontinued (proton pump inhibitors being the biggest offenders), medications meeting Beers’ criteria for potentially inappropriate medication use in older persons, or use of brand medications when an equally effective but less expensive generic formulation is available. Our experience is that almost every such visit identifies an average of two such potential changes. Patients are usually quite happy – they end up with fewer medications and lower copays. Though MTM was likely originally developed with financial goals in mind, its potential impact for quality and patient safety is probably much higher. When we implemented such in an Alzheimer’s disease population many of the recommended medication changes were for medications that were potentially leading to increased confusion or falls. It you look at the costs of potential complications avoided by such programs the cost of program implementation is probably fully recovered. So as we move to new reimbursement systems (ACO’s, global budgets, etc.) more frequent use of such programs makes sense.


Where are the IT solutions? Barcoding has arguably been the most successful patient safety intervention on the inpatient level but has not been adopted in most outpatient settings, let alone the home. But why not? If we can use our iPhones to scan barcodes of products and have them register at the cashier’s station why hasn’t someone figured out how to integrate barcoding with medication administration in the home?


In our October 16, 2012 Patient Safety Tip of the Week “What is the Evidence on Double Checks?” we noted an interesting application of the double check in a homecare setting via televideo monitoring (Bradford 2012). Basically, with a desktop PC and a webcam one can verify the drug name, dose, and gradations on syringes greater than 1 unit with close to 100% accuracy. However, reading expiration dates on vials proved more difficult, with rates of 63%. We suspect that much of the direct observation in the home done in the current Walsh study could be done via such televideo monitoring.


We can’t continue to think we are responsible for our patients only when they come in to our offices and clinics. We need to use a combination of technology and old-fashioned face-to-face visits in the home to help reduce the errors occurring in the home that we are currently not even aware of.





Some of our prior columns on medication errors in other ambulatory settings:

June 12, 2007              Medication-Related Issues in Ambulatory Surgery

August 14, 2007           More Medication-Related Issues in Ambulatory Surgery

March 24, 2009           Medication Errors in the OR

October 16, 2007        Radiology as a Site at High-Risk for Medication Errors

January 15, 2008          Managing Dangerous Medications in the Elderly

April 2010                    Medication Incidents Related to Cancer Chemotherapy

September 2010           Beers List and CPOE

October 19, 2010        Optimizing Medications in the Elderly

April 12, 2011 Medication Issues in the Ambulatory Setting

June 2012                    Parents' Math Ability Matters








Walsh KE, Roblin DW, Weingart SN, et al. Medication Errors in the Home: A Multisite Study of Children With Cancer. Pediatrics 2013; 131: e1405-e1414



Walsh KE, Dodd KS, Seetharaman K, et al. Medication errors among adults and children with cancer in the outpatient setting. J Clin Oncol. 2009; 27(6): 891896



Gandhi TK, Weingart SN, Borus J, et al. Adverse Drug Events in Ambulatory Care.

N Engl J Med 2003; 348: 1556-1564



Schillinger D, Machtinger E, Wang F, Rodriguez M, Bindman A. Preventing medication errors in ambulatory care: the importance of establishing regimen concordance. In: Henriksen K, Battles J, Lewin DI, Marks E, eds. AHRQ Peer-Reviewed Publication: Advances in Patient Safety: From Research to Implementation, Vol. 2. Rockville, MD; 2005.



Gandhi TK, Burstin HR, Cook EF, et al. Drug complications in outpatients.

J Gen Intern Med. 2000; 15(3): 149-154



Wolf MS, Curtis LM, Waite K, et al. Helping Patients Simplify and Safely Use Complex Prescription Regimens. Arch Intern Med. 2011; 171(4): 300-305



Choudhry NK, Fischer MA, Avorn J, et al. The Implications of Therapeutic Complexity on Adherence to Cardiovascular Medications. Arch Intern Med. 2011; 171(9): 814-822



Gurwitz JH, Field TS, Harrold LR, et al. Incidence and preventability of adverse drug

events among older persons in the ambulatory setting. JAMA 2003;289:1107-16



Bradford N, Armfield NR, Young J, Smith AC. Feasibility and accuracy of medication checks via Internet video. Journal of Telemedicine & Telecare 2012; 18(3): 128-132






Print PDF version












Tip of the Week Archive


What’s New in the Patient Safety World Archive