Abstract
OBJECTIVE: To investigate out-of-hospital medication errors among young children in the United States.
METHODS: Using data from the National Poison Database System, a retrospective analysis of out-of-hospital medication errors among children <6 years old from 2002 through 2012 was conducted.
RESULTS: During 2002–2012, 696 937 children <6 years experienced out-of-hospital medication errors, averaging 63 358 episodes per year, or 1 child every 8 minutes. The average annual rate of medication errors was 26.42 per 10 000 population. Cough and cold medication errors decreased significantly, whereas the number (42.9% increase) and rate (37.2% increase) of all other medication errors rose significantly during the 11-year study period. The number and rate of medication error events decreased with increasing child age, with children <1 year accounting for 25.2% of episodes. Analgesics (25.2%) were most commonly involved in medication errors, followed by cough and cold preparations (24.6%). Ingestion accounted for 96.2% of events, and 27.0% of medication errors were attributed to inadvertently taking or being given medication twice. Most (93.5%) cases were managed outside of a health care facility; 4.4% were treated and released from a health care facility; 0.4% were admitted to a non–critical care unit; 0.3% were admitted to a critical care unit; and 25 children died.
CONCLUSIONS: This is the first comprehensive study to evaluate the epidemiologic characteristics of out-of-hospital medication errors among children <6 years of age on a national level. Increased efforts are needed to prevent medication errors, especially those involving non–cough and cold preparations, among young children.
- AAPCC —
- American Association of Poison Control Centers
- HCF —
- health care facility
- NPDS —
- National Poison Data System
- PCC —
- poison control center
What’s Known on This Subject:
Medication errors involving children represent a frequently occurring public health problem. Since 2003, >200 000 out-of-hospital medication errors have been reported to US poison control centers annually, and ∼30% of these involve children <6 years of age.
What This Study Adds:
During 2002–2012, an average of 63 358 children <6 years experienced out-of-hospital medication errors annually, or 1 child every 8 minutes. There was a significant increase in the number and rate of non–cough and cold medication errors during the study period.
Medication errors have been identified as an important preventable cause of morbidity, mortality, and health care cost, but previous research has struggled to fully describe the scope of the problem.1 The importance of this problem in hospitals is widely recognized; however, the magnitude of medication errors outside of hospitals may be even larger.2
A medication error is defined as “any preventable medication event that may cause or lead to inappropriate medication use or patient harm while the medication is in the control of the health care professional, patient, or consumer.”3 Out-of-hospital medication errors involving children represent a frequently occurring public health issue. Since 2003, >200 000 out-of-hospital medication errors have been reported to poison control centers (PCCs) in the United States annually, and ∼30% of these involve children <6 years of age.4
In contrast to hospital-based medication errors, comparatively few studies have examined medication errors in the out-of-hospital setting.5–7 In addition, published research has often reported medication errors among adults or all age groups, rather than focusing on children.6,8,9 Studies have examined errors associated with parental health literacy and numeracy and their ability to correctly measure doses of medications.10–14 Numerous studies have investigated the dosing inaccuracies of various medication administration devices, such as an oral syringe, dosing spoon, dropper, and dosing cup,11,15–19 whereas others have examined the variability of labeling and packaging of medications.12,20,21 Most studies have been based on relatively small convenience samples, and only a few have used statewide or national databases.6,8,9,22
To our knowledge, this is the first comprehensive study to evaluate the epidemiologic characteristics of out-of-hospital medication errors among children <6 years of age on a national level.
Methods
Design of Study and Study Population
This study retrospectively analyzes data from the National Poison Data System (NPDS) to evaluate characteristics and trends of out-of-hospital unintentional therapeutic pharmaceutical errors among US children <6 years of age reported to the NPDS during the 11-year period, 2002 through 2012. The NPDS is a proprietary database maintained by the American Association of Poison Control Centers (AAPCC) consisting of all poison exposure data from each of the regional PCCs in the United States. PCCs receive calls regarding exposures to potentially toxic materials and provide professional medical advice based on the specific circumstances. The NPDS is the only comprehensive poison exposure surveillance database in the United States, and has extensive quality control measures to ensure accuracy and completeness. The AAPCC defines an unintentional therapeutic pharmaceutical error as “an unintentional deviation from a proper therapeutic regimen that results in the wrong dose, incorrect route of administration, administration to the wrong person, or administration of the wrong substance.”4 These types of errors are referred to as “medication errors” in this article. Specifically excluded are exploratory ingestions, intentional misuse or abuse of pharmaceuticals, and malicious or suicidal exposures.
For reported events involving exposure to >1 pharmaceutical substance, the NPDS includes data regarding each substance, which results in the number of medication error exposures exceeding the total number of children. For calls involving multiple pharmaceuticals, each substance is ranked in order of its contribution to the reported clinical effects as determined by the PCC specialist managing the call. We limited data analysis in this study to the first-ranked pharmaceutical, which was judged to be the primary contributor to any clinical signs or symptoms experienced by the child. Therefore, unless noted otherwise, findings reported in this study will refer to the individual children who experienced the medication error and the first-ranked pharmaceutical agent involved.
Case Selection Criteria
The study population was limited to exposures from the 50 US states and the District of Columbia that involved only pharmaceutical substances with the reason for exposure coded as “unintentional therapeutic error.” Pharmaceutical category was determined by the AAPCC generic coding system, which is maintained by Micromedex Poisindex System (Micromedex Healthcare Series [Internet database]; Truven Health Analytics, Inc, Greenwood Village, CO).4 Single- and multiple-substance nonpharmaceuticals, such as batteries and chemicals, were excluded from the study. Cases also were excluded if the site of exposure was a health care facility (HCF), nursing home, or prison, or if on follow-up by a PCC professional, the event was confirmed as a nonexposure or the clinical effects were considered unrelated to the pharmaceutical agent.
Study Variables
Data were analyzed by child age and gender, year and month of exposure, exposure site, management site, type of medication error, category of pharmaceutical, number of pharmaceuticals involved in exposure, type of formulation, route of exposure, level of health care received, type of therapeutic intervention, chronicity, and medical outcome. Management at an HCF included treated and released, admitted, lost to follow-up, and left against medical advice. Medical outcome was classified by PCC specialists as minor, moderate, or major according to standard NPDS definitions.4 Minor effect results in the development of minimally bothersome symptoms; moderate effect results in more pronounced, more prolonged, or systemic symptoms, usually requiring some form of treatment; and major effect results in life-threatening symptoms or having significant residual disability or disfigurement. In this study, cases with serious outcomes were defined as those with moderate or major effects or death. Chronicity of exposure was classified as acute (a single, repeated, or continuous exposure occurring over a period ≤8 hours), acute-on-chronic (a single exposure that was preceded by a continuous, repeated, or intermittent exposure occurring over a period >8 hours), or chronic (a continuous, repeated, or intermittent exposure to the same substance lasting >8 hours).
Statistical Analysis and Ethical Statement
Data were analyzed by using SPSS 21.0 (IBM SPSS Statistics, IBM Corporation, Chicago, IL) and SAS 9.3 (SAS Institute, Inc, Cary, NC) statistical software. National rates were calculated based on July 1 US intercensal and postcensal estimates for residents <6 years from 2002 through 2012.23 Linear regression was used to test for trends in the number and rate of medication errors. Pearson correlation analysis was used to determine the correlation between age and medication error rate, with r representing the Pearson product-moment correlation coefficient. Statistical significance was established at α = 0.05. Any information identifying the exposed individuals or reporting PCCs was redacted by the AAPCC before receipt of the dataset by the investigators. This study was reviewed by the institutional review boards of the authors’ institutions and judged to be exempt.
Results
General Characteristics
From 2002 through 2012, US PCCs reported 696 937 out-of-hospital medication error exposures among children <6 years of age, which averages 63 358 exposures per year or 26.44 exposures per 10 000 population. The number of substances involved in each exposure episode ranged from 1 to 11. Single-substance exposures accounted for 651 039 (93.4%) cases and 42 517 (6.1%) cases were 2-substance exposures. There was seasonal variation in the incidence of reported medication errors, with the peak during winter months (Fig 1). This pattern was influenced most by seasonal variation in cough and cold medication errors. The same seasonal pattern also was seen for analgesics, asthma therapies, hormones, antimicrobial agents, and eye/ear/nose/throat preparations. Boys were involved in 54.1% of exposures (Table 1). Most (96.9%) children were exposed at their own residence, and most cases were managed on-site (93.5%) without referral to an HCF.
Number of cases of out-of-hospital medication errors among children <6 years of age by month, NPDS 2002–2012.
Characteristics of Cases of Out-of-Hospital Medication Errors by Child Age, NPDS 2002–2012
Medication error events (r = –0.994, P < .001) and rates (r = –0.994, P < .001) were negatively correlated with child age. Children <1 year of age accounted for more than one-fourth (25.2%) of exposures, whereas 5-year-olds accounted for 9.7%. The rate of medication errors was highest among children <1 year of age (39.64 per 10 000 population) and decreased with increasing child age, with 5-year-olds experiencing a rate of 15.45 per 10 000 population (Fig 2).
Average annual number and rate per 10 000 population of cases of out-of-hospital medication errors by child age, NPDS 2002–2012.
Characteristics of Medication Errors
Most medication errors involved liquid formulations (81.9%), followed by tablets/capsules/caplets (14.9%) (Table 1). Among errors involving 1-year-olds, 91.7% were liquid formulations compared with 63.1% in 5-year-olds. Exposures were acute in 81.9% of cases, and 96.2% of exposures were by ingestion (Table 2). Medication errors were commonly attributed to “inadvertently taking or being given medication twice” (27.0%), “other incorrect dose” (17.8%), “confused units of measure” (8.2%), and “wrong medication taken or given” (7.8%) (Table 2). The frequencies of “confused units of measure” and “wrong medication taken or given” increased by 67.0% and 83.5%, respectively, whereas the frequency of “inadvertently taking or being given medication twice” decreased by 17.2% during the study period (data not shown).
Characteristics of Route of Exposure and Type of Medication Error by Child Age, NPDS 2002–2012
Medication Categories
The major categories of medications involved with medication errors included analgesics (25.2%), cough and cold preparations (24.6%), antihistamines (15.0%), and antimicrobial agents (11.8%) (Table 3). The frequency of episodes increased for many of the pharmaceutical categories during the study period, most notably dietary supplements/herbals/homeopathics (765.3% increase), followed by cardiovascular drugs (87.5% increase), analgesics (69.6% increase), anticonvulsants (63.6% increase), antihistamines (61.8% increase), and muscle relaxants (55.6% increase). In contrast, medication errors involving cough and cold preparations and asthma therapies decreased by 59.1% and 32.3%, respectively (Fig 3).
Number of Cases of Out-of-Hospital Medication Errors Among Children <6 Years of Age by Major Pharmaceutical Category, Medical Outcome, HCF Level of Care, and Age Group, NPDS 2002–2012
Number of cases of out-of-hospital medication errors among children <6 years of age by year for selected major pharmaceutical categories, NPDS 2002–2012.
A subset of medication categories accounted for a large proportion of observed adverse medical outcomes. Analgesics accounted for 40.0% (n = 10) of deaths, 7.2% of cases with major or moderate medical effects, and 13.5% of ICU admissions. Cough and cold preparations accounted for 20.0% of deaths (n = 5), 18.7% of major or moderate medical effects, and 7.9% of ICU admissions, whereas cardiovascular drugs accounted for 4.0% of fatalities (n = 1), 11.8% of major or moderate medical effects, and 21.6% of ICU admissions (Table 3).
Level of Care Received, Therapy Received, and Medical Outcome
Most (94.1%) medication error episodes did not receive treatment at an HCF; 4.4% were treated and released; 0.4% were admitted to a non–critical care unit; and 0.3% were admitted to a critical care unit (Table 1). During the study period, the number of cases managed at an HCF increased by 17.8%, from 3154 cases in 2002 to 3717 cases in 2012.
More than two-thirds (67.3%) of cases did not receive a therapeutic intervention. Among those receiving an intervention, single therapies included dilution/irrigation/wash (62.8% of cases), activated charcoal (0.8%), and emetics (0.4%). Most dilution/irrigation/wash was performed on-site, usually at the child’s residence (data not shown).
Most exposure cases were not followed after initial contact with the PCC (73.2%) or resulted in no clinical effects (19.9%) (Table 1). Serious medical outcomes resulting from medication errors occurred in 0.7% of the cases (4549 cases with moderate effects, 348 cases with major effects, and 25 deaths) (Table 3). The number of exposures with moderate or major medical outcomes or death decreased by 19.5% during the study period, from 497 cases in 2002 to 400 cases in 2012.
Among pharmaceutical categories with >1000 exposures, muscle relaxants, cardiovascular drugs, and sedatives/hypnotics/antipsychotics had the highest proportions of hospitalization, with 11.5%, 7.7%, and 6.1%, respectively (data not shown). These 3 also had the highest proportion of serious medical outcomes (9.4% of muscle relaxants, 6.5% of sedatives/hypnotics/antipsychotics, and 5.6% of cardiovascular drug exposures). Medical outcome varied by child age, with children <1 year of age experiencing major or moderate medical outcomes 1.39 times more frequently than children 1 to 5 years of age. In addition, children <1 year had a 2.18 times higher proportion of ICU admissions and a 2.33 times higher proportion of fatal outcomes than older children (Table 3).
Secular Trends in Medication Errors
From 2002 through 2012, there was no statistically significant increase in the overall annual number or rate of medication errors. However, there were significant opposing trends in errors associated with cough and cold preparations versus non–cough and cold medications (Fig 4). From 2002 through 2005, there was no significant trend in the annual number (P = .053) or rate (P = .117) of cough and cold medication errors; however, the number and rate decreased significantly by 66.1% (P < .001) and 66.6% (P < .001), respectively, from 2005 through 2012. This decrease was primarily accounted for by the significant decrease in medication errors and rates among children <4 years of age. The decrease in rates ranged from 38.8% (P < .001) for 5-year-olds to 80.4% (P < .001) for children <1 year.
Number and rate of cases of out-of-hospital medication errors among children <6 years of age for cough and cold preparations and non–cough and cold preparations, NPDS 2002–2012.
During the study period, the number and rate of non–cough and cold medication errors increased significantly by 42.9% (P < .001) and 37.2% (P < .001), respectively. Most of this increase was contributed by medication errors associated with analgesics and antihistamines (Fig 3). The rate of non–cough and cold medication errors increased significantly for all ages, with the increases ranging from 46.4% (P < .001) among 1-year-olds to 31.1% (P < .001) among 2-year-olds.
Discussion
Out-of-hospital medication errors among young children result in more than 63 000 calls to US PCCs annually, or 1 child every 8 minutes. Most exposures are managed on-site at a non–health care facility and do not result in serious medical outcomes. These errors follow a seasonal pattern, peaking during the winter months. This is most likely attributable to the increased use of cough and cold preparations, analgesics, and other medications to treat viral illnesses among young children during that time of year. There was a significant increase in the number and rate of non–cough and cold medication errors during the study period. The reason for this observed increase could not be determined from study data; however, it may be associated with increasing use of analgesics and antihistamines among young children.
The number and rate of medication errors increased with decreasing child age, with children <1 year accounting for more than one-fourth of episodes. This is of concern because the proportions of children who died or who were admitted to the ICU were >2 times higher among children <1 year of age compared with older children. It is unclear why younger children had a higher frequency and rate of reported medication error exposures than older children, but it is likely multifactorial. Older children are able to communicate better with caregivers, indicating whether they have already taken the medication, and thereby avoiding taking medication twice. Children <2 years old experienced a larger proportion of medication errors associated with liquid formulations. These liquid medications may lead to medication errors not experienced with other formulations, such as confusing units of measure or an incorrect amount dispensed. Parents also may have a lower threshold for calling a PCC when a medication error involves a young child compared with older children.
Contrary to the observed overall pattern, some medication categories demonstrated increased frequency of errors with increasing child age, including antidepressants, sedatives/hypnotics/antipsychotics, and stimulants. These observed trends are likely due to age-related prescribing practices.12
From 2005 through 2012, there was a significant decline of >60% in both the frequency and rate of errors involving cough and cold preparations. This decline has been well-documented in other research,9,22 and has largely been attributed to the combined actions of the US Food and Drug Administration, manufacturers, and other groups, such as the American Academy of Pediatrics. The recommendation against the routine use of cough and cold preparations among young children was based on the lack of evidence of effectiveness and the common occurrence of side effects.23
A similar case could be made against the routine use of a number of the other medications associated with pediatric medication errors. For example, analgesics, which was the leading medication category implicated in medication errors in this study (accounting for more than one-fourth of these errors, as well as 10 of the 25 deaths), are most often given to children for their antipyretic effect. However, it is widely known that parents overtreat fever during childhood.24–26 Therefore, an opportunity exists to decrease unnecessary exposure to analgesics and the medication errors that occur with analgesic use. Furthermore, it is also well established that antimicrobial agents are overprescribed by medical professionals, and antihistamines are frequently used inappropriately to treat nonallergic upper respiratory congestion.27–30 The largest increase in medication errors was observed for dietary supplements, herbals, and homeopathic agents, the use of which is generally not based on strong scientific evidence. The use of all pharmaceuticals, especially among young children, should be based on the best available medical evidence, weighing effectiveness against the potential for adverse effects, including those from unintentional medication error. We should apply to other categories of medications what we learned from the success of efforts to decrease unnecessary use of cough and cold preparations among young children.
Prevention
The Centers for Disease Control and Prevention, in collaboration with leaders from industry, academia, professional organizations, and health agencies, developed the PROTECT Initiative, which has outlined priorities aimed at preventing both medication errors and unsupervised ingestions.31,32 Priorities addressing medication errors include an educational campaign and the refining of dosing measures and instructions on medication packaging and labeling to reduce errors made by parents and child caregivers. Recommendations include using only milliliters as a measure for liquid medications. Leading zeros before the decimal point are recommended; trailing zeros after a decimal point are not. The Institute for Safe Medication Practices has similar recommendations and also advocates using only kilograms for patient weights.33 To reduce potential dosing errors, manufactures of over-the-counter single-ingredient liquid pediatric acetaminophen medications voluntary converted these medications to 1 concentration.34 A study of adherence to recommendations by the Food and Drug Administration and Consumer Healthcare Products Association regarding dosing instructions and dosing devices for oral liquid medications found relatively high adherence but room for improvement, particularly in the expression of decimals and fractions in instructions.35 This study was limited to over-the-counter analgesic/antipyretic, cough/cold, and allergy medications. Yin and colleagues36 evaluated the effect of both advanced counseling and provision of the dosing instruments in an emergency department setting. Both interventions were necessary to reduce dosing errors, and among low-literacy caregivers and Spanish-speaking caregivers, neither intervention nor the combination affected a change in dosing errors.
In this study, more than one-fourth of all medication errors were due to “inadvertently took/given medication twice.” Possible prevention strategies might include a programmable timer application for a smart phone that receives a signal from the medication container when it is accessed. Lower-technology strategies might use a simple child-resistant scheduling box for nonliquid medications.
Other prevention efforts have included the introduction of less-toxic medications, which decreases the consequences of medication errors, such as the substitution of selective serotonin reuptake inhibitors for tricyclic antidepressants as appropriate.37 Future research should garner parent and child-caregiver feedback about the ways packaging, labeling, and dosing devices contribute to errors35 and address health care provider communication to low-literacy and non–English speaking caregivers.36
Limitations
This study has a number of limitations. The numbers reported in this study underestimate the true magnitude of medication errors involving young children in the United States, because the NPDS only captures cases reported to member PCCs. It is likely that many medication errors are made without recognition, and even errors that are recognized may still not be reported to PCCs because the individuals involved may decline to call a PCC or seek information from other sources, such as online resources. Other cases may be treated in HCFs or private physician offices without a report being made to a PCC, and some children may not receive any medical attention. Information is based on self-reports from parents, child caregivers, and health care professionals, and cannot be completely verified by the AAPCC or reporting PCC. Exposures do not necessarily represent a poisoning or overdose, and some cases involve exposure to multiple medications, leading to the potential for error when PCC specialists rank drugs as being most responsible for clinical effects. However, because 93.4% of calls are single-substance reports, the overall error in identifying the most responsible medication is likely to be very small.
NPDS data are available only as categorized variables, without a descriptive narrative available to more completely describe the circumstances of the event. Despite these limitations, the NPDS is a nationally representative, cumulative database with extensive information about exposures to a wide array of substances, including those associated with medication errors. It is the most comprehensive and accurate database available for investigation of pediatric out-of-hospital medication errors in the United States.
Conclusions
To our knowledge, this is the first comprehensive study to evaluate the epidemiologic characteristics of out-of-hospital medication errors among children <6 years of age on a national level. Although medication errors involving young children occur frequently, serious medical outcomes are rare. Cough and cold medication errors declined significantly starting in 2005; however, non–cough and cold medication errors increased significantly during the study period. Learning from the success in decreasing cough and cold medication errors, increased efforts are needed to prevent non–cough and cold medication errors among young children.
Acknowledgments
The authors acknowledge the support of Krista Kurz Wheeler in preparing the manuscript revision.
Footnotes
- Accepted August 28, 2014.
- Address correspondence to Huiyun Xiang, MD, MPH, PhD, Center for Injury Research and Policy, The Research Institute at Nationwide Children’s Hospital, 700 Children’s Dr, Columbus, OH 43205. E-mail: huiyun.xiang{at}nationwidechildrens.org
Mr Smith conducted the data analysis, and drafted and revised the manuscript; Mr Chounthirath assisted in data analysis, and revised the manuscript; Drs Spiller and Xiang contributed to conceptualization of the study, assisted in data analysis, and critically reviewed the manuscript; Dr Casavant contributed to conceptualization of the study, assisted in data access and analysis, and critically reviewed the manuscript; Mr Brophy assisted in data analysis, and critically reviewed the manuscript; and all authors approved the final manuscript.
The inferences and conclusions expressed by the authors of this study do not necessarily represent those of the American Association of Poison Control Centers or its member centers.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: No external funding.
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
References
- Copyright © 2014 by the American Academy of Pediatrics