Results of Survey on Pediatric Medication Safety (Part 1): More is Needed to Protect Hospitalized Children from Medication Errors
As many as 1 in 10 hospitalized children are impacted by a medication error.1 2 Up to 35% of these errors are serious or life threatening.3 The challenge is to learn from these events and to adopt effective strategies to prevent harmful errors from happening again. Based on the results of a recent ISMP survey, it appears we still have a long way to go to meet that challenge. The survey results make it clear that more needs to be done to protect pediatric patients from harmful medication errors.
During March and April 2015, 1,463 clinicians, mostly pharmacists (45%) and nurses (43%), completed our online Survey on Pediatric Medication Safety Practices. Respondents were asked to select the frequency with which they employed key error-prevention strategies. Most respondents worked in pediatric hospitals (43%) or general hospitals where pediatric patients are treated (41%). In Part 1 of our analysis of the survey, we discuss the aggregate findings from all respondents and also compare the 2015 results to a similar survey we conducted 15 years ago. In Part 2, to be published in a subsequent issue, we will compare subsets of data based on the respondents’ care setting, practice site, patient care unit, and professional designation.
Aggregate Findings from 2015 Survey
General strategies. The survey included five general error-prevention strategies involving all phases of the medication use process (Table 1). With four of the strategies, 90% or more respondents reported implementation at least 90% of the time. Three of these strategies included using metric units of measure to: 1) express the volume of liquid medications; 2) weigh patients; and 3) document the weight on medical records and prescriptions. The fourth strategy was to standardize and limit the concentrations and dosage strengths of pediatric high-alert medications.
The fifth strategy, and the one that scored lowest in this section, involved storing adult, pediatric, and neonatal medications in separate storage locations. Only about half of the respondents reported full compliance with this strategy. Five percent of respondents said that adult, pediatric, and neonatal medications were never separated or sequestered at their practice sites, and another 5% reported employing this precaution less than 20% of the time, leaving clinicians particularly vulnerable to product selection errors.
Strategies when prescribing medications. The survey included six error-prevention strategies associated with prescribing pediatric medications (Table 1). A large number of respondents (85%) reported that, at least 90% of the time, their organizations require: 1) the use of metric doses when ordering pediatric liquid medications; and 2) the entry/verification of the patient’s weight in the computerized prescriber order entry (CPOE) system before entering medication orders. The remaining respondents reported implementation of these practices less consistently, which could lead to serious dosing errors.
Dose range checking software was always available and enabled to provide alerts to prescribers about unsafe doses in only 61% of the respondents’ CPOE systems; 7% reported that dose range checking was never available and/or enabled with their CPOE systems. For the remaining 32% of respondents, the dose checking capabilities appear to be inconsistent. Two other prescribing strategies involved parenteral nutrition (PN) or other complex electrolyte solutions. On units where these products were prescribed, only 64% of respondents reported that prescribers always ordered each ingredient as weight/kg/day for younger children, and 53% reported that prescribers always ordered each ingredient per day for older children. Using variable units of measure and ways of expressing doses when prescribing PN or electrolyte ingredients could be a source of serious errors.
Surprisingly, the lowest scoring error-prevention strategy requires minimal prescriber effort and is one that ISMP has long endorsed: including the mg/kg, mg/m2, or other basis for the dose and the calculated amount per dose with pediatric drug orders. In the survey, we allowed exceptions for drugs that do not lend themselves to weight-based dosing. Despite this, only 37% of respondents reported full compliance with the strategy. Another 27% reported implementation of the strategy for 90-99% of applicable orders. The remaining 36% of respondents reported inconsistent practices, making it difficult for pharmacists and nurses to verify the patient’s dose and detect a prescribing error.
|Pediatric Medication Error-Prevention Strategies||Always (.99%)||Almost Always (90-99%)||Often (50-90%)||Some-times (20-49%)||Rarely (1-19%)||Never (<1%)|
|Metric units of measure are standard nomenclature for pediatric weights documented on medical records and prescriptions.||88||8||2||1||<1||1|
|The volume of liquid pediatric medication doses is expressed using metric units.||81||14||4||1||<1||<1|
|Pediatric patients are weighed using metric units of measure.||80||13||3||2||1||1|
|Concentrations/strengths of high-alert drugs are standardized and limited.||65||25||8||1||1||<1|
|Adult, pediatric, and neonatal medications are not stored near one another.||54||20||10||6||5||5|
|Strategies When Prescribing Medications|
|Patient’s weight in kg or g is entered in the computerized prescriber order entry (CPOE) system before orders are entered.||59||29||8||3||<1||1|
|Prescribers order pediatric liquid medications in metric doses.||51||34||10||4||1||<1|
|Prescribers order each ingredient of PN/ complex electrolyte solutions as weight/kg/day for younger children.||64||18||7||3||3||5|
|Dose range checking software is available and enabled in the CPOE system.||61||21||7||3||1||7|
|Prescribers order the total amount of each ingredient of PN/ complex electrolyte solutions per day for older children.||63||21||8||4||4||10|
|Prescribers include both the mg/kg or mg/m2 dose (or other basis for the dose) and the calculated amount per dose for pediatric drug orders.||37||27||21||8||6||1|
|Strategies When Dispensing Medications|
|Automated compounding devices are used to prepare PN/complex electrolyte solutions (or solutions are outsourced).||77||13||4||1||1||4|
|PN/complex electrolyte solutions are entered into compounding software exactly as each ingredient is prescribed (no unit conversions).||72||20||5||1||1||1|
|PN/complex electrolyte solutions are entered into the pharmacy computer exactly as each ingredient is prescribed (no unit conversions).||68||20||7||1||1||3|
|The pharmacy dispenses patient-specific doses of liquid oral/enteral medications in cups or specially designed oral/ enteral syringes.||72||15||6||2||2||3|
|The patient’s weight in kg or g is entered/verified in the pharmacy computer before entering/verifying
|Pharmacists verify the mg/kg or mg/m2 dose used (or other basis for the dose) to calculate the final dose of a drug before preparing/ dispensing medications.||64||24||5||3||2||2|
|Pharmacists recalculate the dose before preparing/ dispensing medications.||63||27||5||3||1||1|
|Dose range checking software is available and enabled in the pharmacy computer.||57||22||8||3||3||7|
|The pediatric patient’s age is entered/verified in the pharmacy computer before entering/verifying
|Pharmacists/technicians who prepare pediatric parenteral solutions have undergone specialized training and have demonstrated competency.||47||20||9||8||6||10|
|Preparation of IV/oral liquid doses includes barcode verification of ingredients.||41||15||7||5||5||27|
|Pharmacists verify components of pediatric/ neonatal sterile preparations prior to adding to the solution
(syringe pull-back method afterwards not acceptable).
|A clinical pharmacist is present on patient care units to participate in rounds and provide input when prescribing/ administering medications.||23||23||15||10||9||20|
|Pharmacists who prepare pediatric parenteral solutions spend time in the neonatal/pediatric units to observe prescribing and administration.||20||11||13||10||20||26|
|Strategies When Administering Medications|
|Oral syringes that do not connect to IV tubing are available in patient care units.||81||12||2||2||1||2|
|Doses for emergency drugs and common medications have been calculated for each pediatric patient and are available for reference.||72||16||4||3||1||4|
|Nurses who administer medications to pediatric patients have undergone specialized training and have demonstrated competency.||67||21||6||3||2||1|
|Before administering high-alert parenteral drugs, a second nurse independently verifies the patient, drug, dose, line attachment, pump settings, & infusion rate.||65||24||6||3||1||1|
|Bedside barcode scanning is used to verify patients & medications/ solutions before administration.||63||23||2||1||<1||11|
|A smart infusion pump with an activated drug library is used to administer pediatric parenteral solutions
that contain (or are) high-alert medications.
|Before administration of pediatric infusions, nurses trace the line from the medication/ solution to the patient (or vice versa) to verify attachment.||54||32||11||2||1||<1|
|Bedside barcode scanning is used to verify patients and breast milk before each feeding.||46||14||3||2||1||34|
Strategies when dispensing medications. The survey included fourteen error-prevention strategies encompassing the dispensing process (Table 1). It is within this category that both the highest and lowest scoring error-prevention strategies were found. For half of the strategies, at least 87% of respondents reported that they always (>99%) or almost always (90-99%): 1) use automated compounding devices to prepare PN/complex electrolyte solutions; 2) enter PN/electrolyte solution orders into the pharmacy system and compounding software exactly as each ingredient is prescribed without needing unit conversions; 3) enter/verify the patient’s weight in the pharmacy computer system before entering medication orders; 4) verify the mg/kg or mg/m2 dose (or other basis for the dose) before preparing the medication; 5) recalculate the patient’s dose before dispensing medications; and 6) dispense patient-specific doses of liquid oral/enteral medications in cups or oral syringes. Yet, full compliance with these strategies ranged between 63-77%, leaving serious gaps in practice and room for improvement even with the highest scoring dispensing strategies.
For the two lowest scoring strategies, less than one-quarter of respondents reported full compliance. These low-scoring strategies were associated with having a clinical pharmacist present on patient care units, and having pharmacists who prepare parenteral solutions spend time in neonatal and pediatric units to observe prescribing and administration procedures. About one-quarter of respondents told us that these two strategies are never implemented in their practice areas. Pharmacists who spend time in clinical areas may have a better understanding of how physicians and nurses prescribe and administer medications, and may subsequently dispense medications in a ready-to-administer form, reducing manipulation of the drug on the unit and the risk of contamination or an error.
For the next two lowest scoring strategies, only about half of respondents reported implementation at least 90% of the time, and only 40-41% reported full compliance. These strategies included: 1) requiring a pharmacist to verify components of pediatric and neonatal compounded sterile preparations prior to adding to an admixture (syringe pull-back method not acceptable); and 2) employing barcode verification of ingredients during preparation of IV and oral liquid doses. Deficiencies in these pharmacy practices make it nearly impossible to detect a drug preparation and selection error because the applied label may still list the correct prescribed ingredient(s). Numerous harmful and fatal pediatric events have been reported to ISMP over the years, each with similar practice deficiencies.
Similar to dose range checking capabilities with CPOE systems, only 57% of respondents said their pharmacy system consistently provided alerts about potentially incorrect doses; 7% reported that dose range checking was never available and/or enabled with their pharmacy systems. It is also concerning that all pharmacists/technicians who prepare pediatric parenteral solutions have not undergone specialized training and demonstrated competencies. Surprisingly, 10% of respondents reported that such training and competency validation never occurs, and another 6% indicated that it rarely takes place.
Strategies when administering medications. The survey included eight error-prevention strategies associated with the drug administration process (Table 1). For most of these strategies, more than 86% of respondents reported implementation at least 90% of the time. These strategies included: 1) calculating patient-specific doses of emergency drugs and common medications and making them available for reference for each patient during hospitalization; 2) providing nursing units with oral syringes that do not connect to IV tubing; 3) using a smart infusion pump with an activated library to administer pediatric parenteral solutions that contain (or are) high-alert medications; 4) requiring an independent double check before administering parenteral high-alert medications; 5) using bedside barcode scanning systems for medications; 6) tracing tubing lines from the solution to the patient (or vice versa) to verify line attachments before administration; and 7) requiring nurses to undergo specialized training and demonstrate competency associated with pediatric medication administration.
Only 3% of respondents do not use smart pumps in any locations across all care areas for all high-alert medications. However, 35% reported partial compliance, perhaps suggesting that smart pumps are not used in all locations or that the drug library is not activated, diminishing the safety benefits of this technology. Independent double checks prior to administration of high-alert medications occurred consistently in only 65% of respondents’ practice sites, making this an unreliable strategy in the remaining 35% of respondents’ practice sites. Eleven percent of respondents have not implemented bedside barcode scanning with pediatric drug administration.
The relatively simple strategy of tracing the line from the medication/solution source to the patient (or vice versa) to verify line attachment before drug administration only garnered full compliance by about half of the respondents, leaving patients at the remaining half of respondents’ practice sites exposed to the risk of life-threatening wrong route/wrong site errors and other types of errors. The lowest scoring strategy included the use of barcode scanning at the bedside to verify breast milk before each feeding. Despite the complexity associated with implementing this practice, almost half of all respondents (46%) for whom the strategy was applicable reported full compliance with this technology, and another 14% reported compliance 90-99% of the time.
Comparison Between 2000 and 2015 Survey Findings
In 2000, in cooperation with the Pediatric Pharmacy Advocacy Group (PPAG), ISMP distributed a survey to newsletter subscribers about pediatric medication safety practices. Nine of the 33 current strategies are the same as in the 2000 survey and can be compared to the 2015 survey (Table 2). The 2000 survey data are available in aggregate as well as by setting. Thus, we have compared the findings from the 2015 survey using the same setting categories, although Part 2 will cover these findings in more detail. For comparison, the 2015 categories of Almost Always and Often were combined to represent the 2000 category of Frequently, and the 2015 categories of Sometimes and Rarely were combined to represent the 2000 category of Sometimes. The categories of Always and Never remained unchanged.
When comparing the survey findings, we found that three of the strategies showed marked improvement, three of the strategies showed modest gains, and three of the strategies actually worsened in compliance over time (Table 2).
The strategies that showed marked improvement included: 1) listing the mg/kg or mg/m2 dose (or other basis for the dose) and the calculated dose with pediatric drug orders; 2) requiring an independent double check before administering parenteral high-alert medications; and 3) having a clinical pharmacist present in clinical areas. In 2000, only 4% of respondents reported full compliance with including the mg/kg or mg/m2 dose (or other basis for the dose) and the calculated dose with pediatric drug orders; another 21% reported that the strategy was implemented frequently. In 2015, 37% reported full compliance, and another 48% reported frequent implementation. In 2000, 30% of respondents reported full compliance with an independent double check prior to administering high-alert parenteral medications, and another 17% reported frequent implementation. In 2015, 65% reported full compliance, and another 30% reported frequent implementation. Having a clinical pharmacist present on the unit always or frequently increased from 40% in 2000 to 61% in 2015, with the most significant gains in general pediatric units (23 percentage points), PICUs (20 percentage points), and NICUs (16 percentage points).
The three strategies with modest gains included: 1) entering/verifying the patient’s weight in the pharmacy system before preparing and dispensing medications; 2) verifying the mg/kg or mg/m2 dose (or other basis for the dose); and 3) recalculating the patient’s actual dose before preparing and dispensing pediatric medications.
The three strategies that worsened in compliance included: 1) verifying the patient’s age in the pharmacy computer system before entering medication orders; 2) requiring pharmacists/technicians who prepare pediatric parenteral solutions to undergo specialized training and competency validation; and 3) requiring nurses to undergo specialized training and competency validation (Table 2).
Full compliance with entering/verifying the patient’s age in the pharmacy system before entering orders decreased from 85% in 2000 to 50% in 2015 in the aggregate data, with the smallest decrease in nurseries (91% to 82%), and the greatest decrease in PICU (95% to 43%). The decreases in full compliance related to specialized training and demonstrated competency for pharmacists/technicians who prepare and dispense pediatric parenteral solutions, and nurses who administer medications to pediatric patients, are less dramatic but still significant. There was an average decrease of 10 percentage points, with one exception: respondents who provided services for NICU reported a decrease of 23 percentage points in full compliance with pharmacists’/technicians’ training and competency verification between 2000 and 2015.
While compliance with several safety practices is high, and some improvements can be noted between 2000 and 2015, further efforts are needed to implement safety strategies in pediatric settings since pediatric patients are at risk of harm from errors because of their size, immature renal and hepatic function, and an inability to communicate signs of the adverse effects of drugs. Look for Part 2 of the survey analysis for additional insight into differences in recommended practices between settings and among various professional disciplines.
- 1Takata GS, Mason W, Taketomo C, Logsdon T, Sharek PJ. Development, testing, and findings of a pediatric-focused trigger tool to identify medication-related harm in US children’s hospitals. Pediatrics. 2008;121(4):927-35
- 2Takata GS, Taketomo CK, Waite S. California Pediatric Patient Safety Initiative. Characteristics of medication errors and adverse drug events in hospitals participating in the California Pediatric Patient Safety Initiative. Am J Health Syst Pharm. 2008;65(21):2036–44
- 3Tham E, Calmes HM, Poppy A, et al. Sustaining and spreading the reduction of adverse drug events in a multicenter collaborative. Pediatrics. 2011;128(2):438-45