Abstract
OBJECTIVE. The goal was to describe the change in the rate of epinephrine dosing errors in the treatment of pediatric patients in prehospital cardiopulmonary arrest after the Los Angeles County Emergency Medical Services Agency instituted a program in which paramedics were required to use the Broselow tape and to report color zone categories to the base station and base stations were given and instructed formally in the use of the color-coded drug dosing chart.
METHODS. An observational analysis of a natural experiment was performed. Children ≤12 years of age who were determined to be in prehospital cardiopulmonary arrest and who received prehospital epinephrine treatment by paramedics, in the periods of 1994 to 1997 and 2003 to 2004, were included in the study.
RESULTS. In the 1994 to 1997 cohort, we identified 104 subjects in prehospital cardiopulmonary arrest who received epinephrine with a documented weight and route of administration. Only 29 of 104 subjects in the 1994 to 1997 cohort received the correct dose, whereas 46 of 104 subjects received a first dose within 20% of the correct dose. In the 2003 to 2004 cohort, we identified 41 children ≤12 years of age who were in cardiopulmonary arrest and received prehospital epinephrine treatment but 4 children were excluded, leaving 37 subjects. Twenty-one of 37 subjects received the correct dose, whereas 24 of 37 subjects received a dose within 20%. The odds ratio for obtaining the correct epinephrine dose after the system changes versus before was 3.0, and that for obtaining a dose within 20% of the correct dose was 2.5.
CONCLUSIONS. The program seems to have resulted in reduction of the rate of epinephrine dosing errors in the prehospital treatment of children in cardiopulmonary arrest in Los Angeles County.
Although there is some controversy regarding the accuracy of the figures included in the 1998 Institute of Medicine report that stated that each year 44000 to 98000 patients die in the United States as a result of medical errors,1,2 there is no question that medical errors are a substantial cause of morbidity and death in pediatric clinical practice. Kaushal et al3 demonstrated a medication error rate of 5.7 errors per 100 orders in 2 pediatric academic inpatient institutions, similar to error rates observed in the treatment of adult patients.4 Error rates in the emergency department (ED) may be higher. For example, investigators in Toronto, Canada, showed that, for 1532 children who were treated in the ED of a pediatric tertiary care hospital, medication errors were identified in 10% of cases.5 Physicians in training were more likely to commit prescribing errors, and the most seriously ill patients were more likely to be subjected to prescribing errors.5 Although overall error rates in the treatment of adult and pediatric patients are similar, it is estimated that errors with potential to cause harm are 3 times more likely to occur among pediatric patients.6
Incorrect recording of patient weight, leading to incorrect dosing, is one of the most frequently reported errors in pediatric EDs.7 When clinicians are stressed during resuscitation efforts, tools that can reduce reliance on memory and error-prone calculations have the potential to decrease medication error rates and improve outcomes. Since the 1980s, the Broselow tape, a body length-based, color-coded tool, has been used to aid clinicians in rapidly estimating body weight, weight-based drug doses, and the correct size of resuscitation equipment for children.8 In 1988, Lubitz et al9 demonstrated that, among 937 patients who presented to a busy, urban, pediatric ED, use of the Broselow tape resulted in a weight assessment that was within 15% of the exact weight 79% of the time. Several other studies demonstrated improved precision of weight estimates and medication dose calculations by resident physicians and nurses with use of the Broselow tape.10,11 Vilke et al12 showed that, when paramedics estimated pediatric weights with the Broselow tape in a classroom setting, they were able to determine epinephrine doses accurately 95% of the time. Although Vilke et al12 postulated that the tape would be useful for determining drug dosages during actual pediatric resuscitations, there is little published research regarding use of the Broselow tape during prehospital pediatric resuscitations.
Paramedics have infrequent experience with critically ill children. In one study, only 7% of total paramedic calls involved patients <14 years of age.12 In fact, 87% of paramedics in the state of Maine evaluate ≤3 pediatric patients per month.13 Few studies of medication errors and error prevention strategies have been performed in prehospital settings, let alone in the stressful setting of prehospital resuscitation of pediatric patients. With preliminary data from the 1994 to 1997 experience that is a component of the current study, Gausche et al14 demonstrated a 62% error rate in the dosing of epinephrine in a prehospital setting during a prospective comparative trial of airway management strategies for children, despite the fact that paramedics in the study were instructed in the use of the Broselow tape and each ambulance was supplied with a tape at the beginning of the study. More than 87% of the errors were calculation errors, whereas 13% were drug concentration errors.
In 2001, the Los Angeles (LA) County Emergency Medical Services (EMS) Agency instituted a program called “LA Kids.” Before 2001, paramedics were encouraged and instructed to use the Broselow tape, but the LA Kids program mandated that paramedics use the Broselow tape and report the color zones to the base station. Base stations were given a color-coded, drug-dosing chart, and base station personnel were instructed formally in the use of the color-coded system. The objective of our study was to determine whether the LA Kids program reduced epinephrine dosing error rates in the prehospital treatment of pediatric patients in cardiopulmonary arrest, by comparing epinephrine dosing error rates before and after implementation of the LA Kids program.
METHODS
Study Design
This study was an observational before/after evaluation of a natural experiment. The data from 1994 to 1997 were collected during a prospective controlled trial of pediatric airway management that enrolled 830 patients from March 15, 1994, to January 1, 1997. Critically ill or injured children were assigned randomly to receive bag-mask ventilation or endotracheal intubation. The methods and primary results of this interventional study were described previously.15,16 Data collected from 2003 to 2004 were obtained from the LA County EMS Agency for a consecutive cohort of children <13 years of age who experienced prehospital cardiopulmonary arrest between January 1, 2003, and September 30, 2004. Institutional review board approval for this study was obtained from the institutional review board at Harbor-University of California LA Medical Center and from the Western Regional Institutional Review Board, which reviews research protocols for the LA County EMS Agency.
Setting
The setting of the pre- and postintervention data collection was the large, urban EMS system of LA County, California, which uses a 2-tiered response of basic life support and advanced life support.16
System Changes for Determination of Epinephrine Doses for Children
In the pre-LA Kids period, paramedics were trained in the use of the Broselow tape and were encouraged to use it for determination of drug doses and equipment sizes. Paramedics could deliver the first dose of epinephrine before base hospital contact, and base hospitals had information about dosing epinephrine in milligrams per kilogram but were not supplied uniformly with precalculated drug charts or Broselow tapes (Table 1).
System Changes for Determination of Epinephrine Doses for Children
The LA Kids training contained cognitive and psychomotor skill objectives for all paramedic and base personnel regarding use of the Broselow tape and color zone charts for field drug dose determinations. After institution of the LA Kids program, paramedics were required to use the Broselow tape to determine the weight and color zone of the patient and were required to report the color zone to the base station. They were also provided with precalculated drug dosing charts organized according to the same color zones (Table 1). Base hospital personnel were trained in the color-coded system and were provided with the same precalculated drug dosing charts (Fig 1). A quality-improvement process was instituted to track the performance of prehospital and base personnel in documenting the color zone for each child evaluated in the field. During both study periods, paramedics were allowed to administer the first dose of epinephrine before base contact.
Color-coded drug doses. A page from the precalculated drug dosing charts organized in color zones (pink and red zones) is shown. B/P indicates blood pressure; IM, intramuscularly; IV, intravenous(ly); PR, rectally; NS, normal saline; SQ, subcutaneously; HHN, handheld nebulizer; mcgtt, microdrip; ACLS, advanced cardiac life support.
The standard intravenous epinephrine dose was 0.01 mg/kg, at a concentration of 1:10000, whereas the standard endotracheal epinephrine dose was 0.1 mg/kg, at a concentration of 1:1000. For the pre-LA Kids cohort, paramedics were instructed to administer the first dose of epinephrine by using the standard concentration and dose for the route used, but “high-dose” epinephrine, with a 10-fold increase in epinephrine dose, was considered acceptable for second and subsequent doses. In contrast, for the post-LA Kids cohort, high-dose epinephrine was not considered acceptable. Because it was unclear whether administration of a 10-fold increase in epinephrine dose for the pre-LA Kids cohort was an error or was intended for a second or subsequent dose, we elected to compare only the first dose of epinephrine, for which only a standard dose was considered correct. After 1997, endotracheal intubation for children <13 years of age was removed from the paramedic scope of practice, on the basis of results of the Pediatric Airway Management Study.15
Selection of Participants
The pre-LA Kids subjects were selected from the Pediatric Airway Management Study cohort, which included 830 critically ill children <13 years of age from LA and Orange counties. The study group included children in prehospital cardiopulmonary arrest who received prehospital epinephrine treatment with a specified route of epinephrine administration and whose weight was documented with the Broselow tape in the field, was reported by parents, or was measured in the ED. Only children from LA County were included. The post-LA Kids cohort was selected from all LA paramedic calls involving children <13 years of age who were in prehospital cardiopulmonary arrest, received prehospital epinephrine treatment, and had a weight documented with the Broselow tape in the field, reported by parents, or measured in the ED, between January 1, 2003, and September 30, 2004.
Observations
Data collection forms from the original interventional study, EMS data forms, and coroner reports were reviewed for each patient in the 1994 to 1997 group.14–16 For the 2003 to 2004 cohort, EMS data forms and the original paramedic run sheets were collected and reviewed. Variables recorded for both study cohorts included age, patient weight determined with the Broselow tape, gender, arrest rhythm, and actual dose, concentration, and route of the first dose of epinephrine administered. The correct dose of epinephrine that should have been administered to each patient was calculated according to LA County prehospital care treatment guidelines, on the basis of the route of administration (intravenous versus endotracheal tube) and the weight determined with the Broselow tape (used preferentially, if available), the parental weight estimate recorded in the field, or the weight recorded in the ED. In LA County, intraosseous needle insertion was not within the scope of practice during either time period.
Data Analyses
The 1994 to 1997 study database was created by using Paradox 3.5 software (Borland, Scotts Valley, CA), and the 2003 to 2004 cohort data were entered into a Microsoft Excel spreadsheet (Microsoft, Redmond, WA). A final combined database was then translated into native SAS format by using DBMS/Copy 8 (Data Flux, Cary, NC).
With SAS 8.1 software (SAS Institute, Cary, NC), we computed the Mantel-Haenszel χ2, stratified according to route of administration, to compare epinephrine dosing error rates between the 2 cohorts. The primary outcome was receiving the exact correct dose; the secondary outcome was receiving a dose within 20% of the correct dose. Secondary exploratory analysis with logistic modeling was also performed with SAS 8.1 software. A priori, we set the α level at .05 and the power at .8 to find a reduction in the absolute error rate of 20% in the second cohort, assuming an error rate of 50% in the first cohort. To achieve this power, we determined that we would need to enroll retrospectively 65 post-LA Kids subjects in cardiopulmonary arrest who received prehospital epinephrine treatment, given that 104 such children existed in the 1994 to 1997 cohort. The sample size calculations were performed with PASS 2002 (NCSS, Kayesville, UT). However, in an unplanned interim analysis in which we evaluated the error rates for 37 subjects from 2003 to 2004 and compared the findings with rates for the original intervention cohort, we found a statistical and clinical difference in error rates that far exceeded our expectations, and we elected not to continue with additional data collection.
RESULTS
To define our pre-LA Kids cohort, we first identified 601 of 830 total subjects from the 1994 to 1997 Pediatric Airway Management Study who were in cardiopulmonary arrest. Of those 601 children, 446 were from LA County; the 155 children from Orange County were removed from the sample. Of the 110 patients who received prehospital epinephrine treatment, 6 had to be excluded (Fig 2). One subject received epinephrine through the sublingual route, 4 had no documented weight, and the dose administered was unclear for 1, which left 104 subjects for analysis (Fig 2). Epinephrine dosing routes were intravenous for 47 subjects (45%) and endotracheal for 57 (55%) (Table 2). Sixty-one (59%) of the 104 subjects had been measured with the Broselow tape. In the 1994 to 1997 cohort, only 29 (28%) of 104 subjects received the correct first dose, and 46 (44%) of 104 subjects received a first dose within 20% of the correct dose (Table 2 and Fig 3A). The doses were exactly correct for 13 (28%) of 47 subjects who received intravenously administered epinephrine and 16 (28%) of 57 subjects who received endotracheally administered epinephrine. The doses were close (within 20% of the exact dose) for 16 (34%) of 47 subjects who received intravenously administered epinephrine and 23 (40%) of 57 subjects who received endotracheally administered epinephrine.
Enrollment of the pre- and post-LA Kids cohorts. The 1994 to 1997 cohort data were collected previously during a prospective study of pediatric airway management methods in the prehospital setting, whereas the 2003 to 2004 cohort data were collected solely for this study, from a consecutive cohort of children in prehospital cardiopulmonary arrest (CPA). SL indicates sublingual.
Actual versus correct epinephrine doses administered. In the pre-LA Kids cohort, 28% of the children received the correct dose, whereas 44% received a dose within 20% of the correct dose (solid lines) (A). The number of subjects who received a 10-fold error in dose (data points falling on the dotted lines) demonstrates the difficulty of memorizing a different concentration for the endotracheal route. In the post-LA Kids cohort, 54% received the correct dose and 65% received a dose within 20% of the correct dose (B). There might have been fewer 10-fold errors in the post-LA Kids cohort because most of the subjects received epinephrine through the intravenous route, which decreased the errors resulting from needing to memorize a different concentration for the endotracheal route. E indicates administration through an endotracheal tube; I, intravenous administration.
Epinephrine Dosing Errors According to Route
In the 2003 to 2004 cohort, there were a total of 144 children <13 years of age in prehospital cardiopulmonary arrest. Of those subjects, 41 children (28%) received prehospital epinephrine treatment. Two children who did not have a color code recorded and 2 children who did not have a documented weight were excluded, leaving 37 subjects for analysis (Fig 2). Epinephrine dosing routes were intravenous for 34 subjects (92%) and endotracheal for 3 (8%). In total, 21 (57%) of 37 subjects received the correct dose, and 24 (65%) of 37 subjects received a dose within 20% of the exact dose (Table 2 and Fig 3B). The doses were exactly correct for 20 (59%) of 34 subjects who received intravenously administered epinephrine and 1 (33%) of 3 who received endotracheally administered epinephrine.
The Mantel-Haenszel odds ratio, stratified according to route of administration, for obtaining the correct epinephrine dose after the system changes versus before was 3.0 (95% confidence interval [CI]: 1.4–6.6), and that for obtaining a dose within 20% of the correct dose was 2.5 (95% CI: 1.1–5.6). In addition, secondary exploratory analyses using multivariate logistic modeling, adjusting for age, weight, and route of administration, demonstrated effect estimates that were qualitatively similar to the Mantel-Haenszel odds ratios.
Not infrequently, dosing errors were incorrect by a factor of 10, as shown in Fig 3A. Such overdosing and underdosing errors might be attributable to the added difficulty of memorizing the different doses and concentrations that are used when epinephrine is administered through different routes. Tenfold errors were seen less frequently for the post-LA Kids cohort (Fig 3B), possibly because endotracheal intubation was removed from the paramedic scope of practice and therefore there were far fewer children who received epinephrine through the endotracheal route.
Baseline characteristics for the preintervention and postintervention groups were compared, with assessment of the following possible confounding factors: weight, age, whether the Broselow tape was used, gender, presenting rhythm, and route of epinephrine administration (Table 3). The following characteristics were statistically significantly different between the 2 cohorts: median weight, 12 kg (interquartile range [IQR]: 6–18 kg) vs 16 kg (IQR: 10–18 kg; P < .001); median age, 1.7 years (IQR: 0.3–5.4 years) vs 3 years (IQR: 1.5–10 years; P < .001); use of the Broselow tape, 61 (59%) of 104 cases vs 36 (97%) of 37 cases (P < .001); intravenous route of administration, 47 (45%) of 104 cases vs 34 (92%) of 37 cases (P < .001). We also found that, when subjects in the pre-LA Kids cohort for whom the Broselow tape was used were compared with subjects for whom the tape was not used, the estimated odds ratio for obtaining an epinephrine dose within 20% of the correct dose was 1.0 (95% CI: 0.5–2.2), and the odds ratio for obtaining the exact dose was 4.9 (95% CI: 1.7–14.3).
Comparison of the Pre- and Post-LA Kids Program Cohorts
The differences in the weight and age distributions for the 2 cohorts were likely attributable to the fact that there were fewer subjects in the 2003 to 2004 cohort who could receive endotracheal epinephrine treatment (eg, those with preexisting tracheostomies or those who were large enough to be intubated with an adult-size endotracheal tube) and successful intravenous placement is easier for older and heavier children. Therefore, smaller children were less common in the later cohort, because many had no route available for epinephrine administration.
DISCUSSION
Between the pre-LA Kids (1994–1997) and post-LA Kids (2003–2004) time periods for data collection, LA County adopted a tool for drug dosing in pediatric resuscitations called LA Kids, aimed at simplifying and standardizing prehospital pediatric drug administration. Before the institution of LA Kids, the medication dosing error rate exceeded 70%, although the Broselow tape had been used to determine weights in more than one half of the cases. Merely providing paramedics with ready access to the Broselow tape seems to be insufficient to prevent a high rate of medication dosing errors in the prehospital setting. After implementation of the LA Kids program, paramedics used the Broselow tape to determine weights nearly all of the time. The correct dose was administered more than one half of the time, whereas the dose was incorrect by <20% in two thirds of cases. This reduction in medication error rates underscores the importance of implementing system changes such as mandating the use of the Broselow tape, instituting a team approach, and conducting quality review of performance within the system, to provide checks and balances to delivery of care.
It was apparent even in the pre-LA Kids cohort that there were fewer dosing errors in the subgroup in which the Broselow tape was used. Therefore, it is not surprising that, in the post-LA Kids cohort, in which a larger proportion of subjects were measured with the Broselow tape (because the LA Kids program mandated use of the Broselow tape), a reduction in the medication dosing error rate was seen.
On the basis of crew resource management behavioral principles developed in the aviation industry, teamwork is known to reduce error rates in nonmedical settings. Sequences of events that lead to errors can be interrupted by checks and balances by team members.17 Features of this teamwork approach that can be adapted to medical settings include assigning roles and responsibilities to team members and using check-back processes to verify communications and actions. For example, errors resulting from verbal orders or telephone reports of critical laboratory results may be decreased if the recipient must repeat the verbal order or result.18,19
Our study has several limitations. Because of the observational design, we cannot be certain that the change in error rates that we observed in the 2003 to 2004 cohort was attributable to the LA Kids program, because secular changes and increased vigilance in reducing iatrogenic error rates had been occurring concomitantly. In addition, because endotracheal intubation was removed from the scope of practice for LA County's paramedics in 1997, we observed low rates of use of the endotracheal route for epinephrine administration in the post-LA kids cohort. Error rates for intravenous and endotracheal dosing were different, but stratification according to route of administration and logistic regression modeling demonstrated that this difference alone did not account for the marked improvement in dosing accuracy after the system change. There were statistically significant differences between the 2 cohorts in median age and weight, with older and heavier children in the post-LA Kids cohort. However, after adjustment for these potentially confounding variables in the exploratory logistic models, the resulting odds ratios were similar to the results obtained from the stratified analysis.
The results of this study are from an unplanned interim analysis in which we evaluated the error rate for 37 subjects from 2003 to 2004 and compared it with the rate for the original intervention cohort. Although we found a statistical and clinical difference in error rates that far exceeded our expectations, such an analysis should be interpreted with care.
The retrospective nature of the chart review limited the analysis to data that had actually been recorded. From the written records, we were unable to discern whether the base station or the prehospital care provider was making the dose determination for the first dose of epinephrine. Although weight determinations and drug dosages were both required fields in the paramedic run sheet and base station contact form, inaccuracies in data recording are possible. There is no reason to suspect that there would be greater errors in recording or greater misclassification biases after LA Kids was implemented, however.
CONCLUSIONS
The LA Kids program seems to have resulted in a marked reduction in the rate of epinephrine dosing errors in the prehospital treatment of children in cardiopulmonary arrest in LA County. The rate of incorrect drug dosing is still far from ideal, however, and other measures should be taken to improve medication dosing for children.
Acknowledgments
We acknowledge grant support to M.G.-H. from the Agency for Healthcare Research and Quality and the Health Resources and Services Administration (grant 1R01HS9166), the State of California EMS Authority under special project grants (grants EMS 2031, 3036, 4051, and 5034), and the Maternal and Child Health Bureau in collaboration with the National Highway Traffic Safety Administration, Health Resources and Services Administration (grant MCH-064004). K.D.Y. was supported by a Mentored Patient-Oriented Career Research Award from the National Institutes of Health, National Center for Research Resources (grant K23-RR16180). This project was supported by a National Center for Research Resources grant awarded to the General Clinical Research Center at the LA Biomedical Research Institute, Harbor-University of California, Los Angeles, Medical Center (grant M01 RR00425). We acknowledge equipment donations from Laerdal Medical Corp, Medical Plastics Laboratory, Nellcor, STI Medical Products, Armstrong Medical, and Vital Signs.
We thank the LA County EMS Agency and Maureen Hasbrouk, the LA County coroners' offices, research nurses Pamela D. Poore and Suzanne M. Goodrich, and all of the paramedic provider agencies and field paramedics in LA County for their dedication to the Pediatric Airway Management Project.
Footnotes
- Accepted May 23, 2006.
- Address correspondence to Amy H. Kaji, MD, MPH, Department of Emergency Medicine, Harbor-UCLA Medical Center, 1000 W Carson St, Box 21, Torrance, CA 90509. E-mail: akaji{at}emedharbor.edu
The authors have indicated they have no financial relationships relevant to this article to disclose.
REFERENCES
- Copyright © 2006 by the American Academy of Pediatrics
In this issue
Jump to section
Related Articles
Cited By...
- Pediatric Readiness in Emergency Medical Services Systems
- Pediatric Readiness in Emergency Medical Services Systems
- Medic One Pediatric (MOPed) cards: standardising paramedic paediatric resuscitation
- Decreasing Prescribing Errors During Pediatric Emergencies: A Randomized Simulation Trial
- An assessment of the accuracy of a novel weight estimation device for children
- Interventions to Reduce Pediatric Medication Errors: A Systematic Review