A Weighty Mistake

SPOTLIGHT CASE

Commentary by Seth J. Bokser, MD, MPH

The Case

A 17-month-old toddler was brought to the emergency department (ED) by her parents with concerns about an infection around the eye. A triage nurse took the patient's vital signs, including her weight, and escorted her and the parents to a room for evaluation. The resident and attending physician diagnosed an uncomplicated periorbital cellulitis and prescribed clindamycin 225 mg orally three times daily based on the patient's weight. The patient received the first dose in the ED and was then discharged home.

The following evening, the patient's mother called the ED to report that the patient's discharge paperwork listed a weight of 25 kg rather than her true weight of 25 lb (11.3 kg). The mother, a medical student, realized the potential implications of such an error and asked the resident on duty to recalculate the appropriate dosage. The patient's dose was changed to 113 mg three times daily, and the patient continued the course with resolution of the cellulitis. Other than mild and self-limited diarrhea, there were no significant adverse effects from the four larger doses administered prior to the dose change. On further investigation, the initial triage nurse incorrectly entered the patient's weight (25 pounds) as a weight in kilograms in the electronic medical record (EMR), an error that was not caught by other providers throughout the ED visit.

The Commentary

Medication dosing for the pediatric patient is fraught with potential for error.(1,2) While it is the gift for every pediatric provider to be surrounded by healthy growth in their patients, pediatric growth adds significant variability and complexity to medication prescribing.(3) For each patient under 40 kg, the pediatric provider must calculate individualized medication doses based on the patient's weight, as a proxy for physiologic development and volume of drug distribution. This standard of pediatric practice is called "weight-based dosing." A 2006–2007 analysis of the United States Pharmacopeia's MEDMARX database illustrated the risk inherent in weight-based dosing by revealing that one-third of pediatric medication errors were the result of "improper dose/quantity" (significantly higher fraction than in adults [4]), and 2.5% of those pediatric dosing errors led to patient harm.(5) The case presented is a striking reminder of the risks associated with weight-based dosing errors.

Over the last decade, pediatric providers, especially those associated with children's hospitals, large academic centers, and integrated delivery networks, have made considerable improvements in pediatric medication safety by adopting electronic health record (EHR) software with computerized provider order entry (CPOE) and automated clinical decision support (CDS).(6-8) Although the use of these tools has unfortunately lagged among pediatricians as compared with other specialists (9), pediatric clinicians, like the clinicians in this case, are increasingly leveraging the flawless math of computers to perform weight-based dosing calculations. This functionality automates the following multistep process: (i) multiplying the medication's recommended weight-based daily dose (usually expressed in mg/kg) by the patient's weight (expressed in kg), (ii) dividing that daily dose into the desired interval for administration (in the described case, three times daily), and (iii) rounding that interval dose to match a measurable volume of pediatric suspension or breakable fraction of a tablet.

Importantly, the above calculations are all based on one foundational, patient-specific variable—weight expressed in kilograms. If a pediatric patient's weight is entered incorrectly into the EHR, the world's largest supercomputer with the best CPOE software will produce a dosing error every time with 100% precision. In the era of integrated EHRs, once "bad data" (inaccurate data) are recorded, especially data as foundational as weight, patient safety is at risk. In addition to medication prescribing, pediatric clinicians titrate many important medical interventions based on recorded patient weight. These interventions include fluid resuscitation, blood product delivery, and nutritional supplementation.

Although there is obvious safety benefit of taking these complex weight-based treatment calculations out of the hands of math-challenged human clinicians (10), there is also the real risk of automation complacency. This automation-complacency risk was realized in the described case when clinicians trusted the computer-assisted calculation and likely did not review the patient's weight as part of dosing process. If the clinician had to type, or write by hand, the weight of 25 kilograms, or 55 pounds, rather than relying on automation, would she have realized that this weight was incongruent with the 17-month-old child who was sitting before her? Perhaps so.

Automation complacency creates an environment where these data entry errors are difficult to catch. Therefore, these errors can be insidious, propagating through future pediatric weight-based interventions with little resistance. If the mother in this case were not so vigilant, and—instead of calling to report the weight-entry error—had called asking for a different antibiotic or antipyretic, there is little doubt that the new prescription(s) would have also included the twofold overdose medication error. Furthermore, because the pharmacy likely would have received the identical "bad data" (incorrect weight) sent through a third-party broker from the CPOE system to the pharmacy dispensing system, there would have been little to safeguard the child from receiving another erroneous and potentially dangerous prescription. Indeed, a 2010 publication produced by the Institute for Safe Medication Practices, the ECRI Institute, and the Pennsylvania Patient Safety Authority demonstrated that of the 479 medication error reports submitted from 2004–2008 related to obtaining and/or recording patient weight, 67.2% (322) of the events reached the patient.(11)

Given the importance of weight as a foundational piece of data in the integrated EHR, we must consider interventions to support accurate measurement, recording, and display of pediatric patient weights. The following recommendations span the domains of policy, biomedical equipment, clinical standards of practice, and information technology (IT).

Recommendation 1. In a joint policy statement entitled Guidelines for Care of Children in the Emergency Department, the American Academy of Pediatrics, American College of Emergency Physicians, and the Emergency Nurses Association have stated that every ED in the United States that cares for pediatric patients must have a weight scale "that displays in kilograms only (not pounds) for infants and children."(12) Just this simple equipment upgrade, though not free, would improve the accuracy of recorded weights in intervention-heavy environments like the ED. I can remember well during my pediatric residency training working in an ill-equipped ED and asking parents to hold their child on the adult scale, which measured in pounds. In some environments without an infant or pediatric scale, pediatric weights are documented based on parental history: "What did your child weigh at his/her last pediatric visit?" Of course, these parental reports are indirect, and the units of measurement are unreliable. Unfortunately, many environments that care for children continue to "make do" with these unacceptable workarounds because they do not have safe pediatric equipment. A 2002–2003 National Hospital Ambulatory Medical Care Survey reported that only 6% of US EDs had the recommended pediatric equipment as outlined in previously published national guidelines.(13) One would hope in the interceding decade that compliance has increased, but there is surely a need for many pediatric care environments—especially outside of children's hospitals—to upgrade scales and standard operating procedures for documenting pediatric patients' weights (Figures).

Recommendation 2. Because the most common type of medical error involving patient weight is confusion between pounds and kilograms (5), clinicians should measure, record, and display weight in kilograms in the EHR. This recommendation is also endorsed by the Institute for Safe Medication Practices.(14) The challenges to implementing such a standard of practice has been that we (in the US) live in a non–metric-based society, we wish to communicate with families in units of measure they understand, and we cater to relatives yearning to hear the "pounds and ounces" of familial offspring in stylized birth announcements. Still, I and the organizations listed above believe it is safest and should be standard of practice to measure, record, and display weights in kilograms only within the clinician-facing interfaces of the EHR. If we wish to convert the kilograms recorded in our clinician-facing EHR into pounds when facing our non–metric-speaking patients and families (in person or within patient portals), software and mobile applications can assist us with conversion. However, we should not muddy the clinician-facing EHR.

Recommendation 3. As pediatric providers continue to adopt EHRs with CPOE, they should consider strongly implementing automated CDS at the critical moment of data entry. This type of CDS consists of rule-based algorithms that compare entered patient weight to expected patient weight in real time. If the discrepancy between entered and expected weight is too large, they fire an onscreen alert to the data entry user at the moment of entry. The EHR software can calculate expected weight in a variety of ways, most commonly by placing a reasonable range (e.g., plus or minus 10%) around the most recent previously entered patient weight. Software can also be programmed to compare entered weight to age-adjusted norms according to standardized growth charts. If such data entry decision support were implemented in this case, the ED triage nurse certainly would have been warned that a weight of 25 kg is exceedingly uncommon for a 17-month-old patient and likely an error. World Health Organization standard growth charts indicate that the 98th percentile cut-off for a 17-month-old female is 12.9 kg (15)—nearly half the value that the triage nurse entered and that was used to calculate the erroneous medication dose.

Recommendation 4. Biomedical device integration (the ability to integrate data coming from digital medical devices, such as scales, directly into the EHR) holds the promise to limit human error inherent in observing and recording of biometric data. In the past few years, we have seen the growth of many such provider-targeted, and consumer-focused, solutions come to market, including integrated digital scales. The solutions vary but generally include (i) biomedical devices capable of outputting a standard digital signal (often RS232 standard), (ii) software that can accurately convert this message into a standard data format (often HL7 standard), and (iii) an EHR that can accept the message and display these data, usually allowing for manual user over-ride to correct for spurious misreads (e.g., the patient had only one foot on the scale).

In conclusion, this case describes an important and classic system error and, more importantly, provides us with tremendous quality-improvement opportunities. We should not continue to rely on humans to remember when to record pediatric weight in pounds and when in kilograms. This system has failed and will continue to fail—sometimes with disastrous consequences. The EHR can amplify this system failure because of automation complacency; once bad data is entered into the EHR it can perpetuate more bad data and erroneous decision support. I presented recommendations above that range from the simple purchase of a pediatric kilograms-only scale to more sophisticated technical forcing functions. One or more of these solutions should be implemented in every environment that cares for pediatric patients regardless of its current level of IT advancement.

Take-Home Points

In the era of integrated electronic medical records, where automated calculations and CDS algorithms drive patient care, bad foundational data (e.g., incorrectly entered patient weights) are dangerous.
CPOE systems have decreased dosing error rates associated with weight-based dosing. Yet, they have also introduced the risk of automation complacency. Medication errors resulting from obtaining or recording incorrect patient weight have a high likelihood of reaching patients.
Recommendations to constrain data entry error of important values such as patient weight include the following: (i) All providers that care for children should have access to infant and pediatric scales that weigh children in kilograms. (ii) Standard of practice should be to measure, record, and display weight in kilograms in the clinician-facing EHR. (iii) As pediatric providers continue to adopt EHR software data, they should consider implementing decision support at the point of data entry for key pieces of foundational data such as weight. (iv) EHR-integrated digital scale solutions can significantly constrain the potential for patient weight–entry errors.

Seth J. Bokser, MD, MPH

Associate Professor for Pediatrics, UCSF Medical School

Medical Director for IT, UCSF Benioff Children's Hospital

San Francisco, CA

Faculty Disclosure: Dr. Bokser has declared that neither he, nor any immediate member of his family, has a financial arrangement or other relationship with the manufacturers of any commercial products discussed in this continuing medical education activity. In addition, the commentary does not include information regarding investigational or off-label use of pharmaceutical products or medical devices.

References

1. Kaushal R, Bates DW, Landrigan C, et al. Medication errors and adverse drug events in pediatric inpatients. JAMA. 2001;285:2114-2120. [go to PubMed]

2. Wong IC, Ghaleb MA, Franklin BD, Barber N. Incidence and nature of dosing errors in paediatric medications: a systematic review. Drug Saf. 2004;27:661-670. [go to PubMed]

3. Sullivan JE, Buchino JJ. Medication errors in pediatrics—the octopus evading defeat. J Surg Oncol. 2004;88:182-188. [go to PubMed]

4. Bates DW, Boyle DL, Vander Vliet MB, Schneider J, Leape L. Relationship between medication errors and adverse drug events. J Gen Intern Med. 1995;10:199-205. [go to PubMed]

5. Sentinel Event Alert. Preventing pediatric medication errors. April 11, 2008;(39):1-5. [go to PubMed]

6. Potts AL, Barr FE, Gregory DF, Wright L, Patel NR. Computerized physician order entry and medication errors in a pediatric critical care unit. Pediatrics. 2004;113(1 Pt 1):59-63. [go to PubMed]

7. Kaushal R, Barker KN, Bates DW. How can information technology improve patient safety and reduce medication errors in children's health care? Arch Pediatr Adolesc Med. 2001;155:1002-1007. [go to PubMed]

8. Bates DW, Teich JM, Lee J, et al. The impact of computerized physician order entry on medication error prevention. J Am Med Inform Assoc. 1999;6:313-321. [go to PubMed]

9. Leu MG, O'Connor KG, Marshall R, Price DT, Klein JD. Pediatricians' use of health information technology: a national survey. Pediatrics. 2012;130:e1441-e1446. [go to PubMed]

10. Potts MJ, Phelan KW. Deficiencies in calculation and applied mathematics skills in pediatrics among primary care interns. Arch Pediatr Adolesc Med. 1996;150:748-752. [go to PubMed]

11. Medication errors: significance of accurate patient weights. PA-PSRS Patient Saf Advis. March 2009;6:10-15. [Available at]

12. American Academy of Pediatrics, Committee on Pediatric Emergency Medicine; American College of Emergency Physicians, Pediatric Committee; Emergency Nurses Association Pediatric Committee. Joint Policy Statement—Guidelines for Care of Children in the Emergency Department. Pediatrics. 2009;124:1233-1243. [go to PubMed]

13. Middleton KR, Burt CW. Availability of pediatric services and equipment in emergency departments: United States, 2002–2003. Adv Data. 2006;(367):1-16. [go to PubMed]

14. Emergency Nurses Association. Weighing pediatric patients in kilograms. Position statement. [Available at]

15. WHO Growth Standards Are Recommended for Use in the U.S. for Infants and Children 0 to 2 Years of Age. Atlanta, GA: Centers for Disease Control and Prevention; 2006. [Available at]