Abstract
BACKGROUND:
Although pediatric electronic prescribing systems are increasingly being used in pediatric care, many of these systems lack the clinical decision-support infrastructure needed to calculate a safe and effective rounded medication dose. This infrastructure is required to facilitate tailoring of established dosing guidance while maintaining the medication's therapeutic intent.
OBJECTIVE:
The goal of this project was to establish best practices for generating an appropriate medication dose and to create an interoperable rounding knowledge base combining best practices and dose-rounding information.
METHODS:
We interviewed 19 pediatric health care and pediatric pharmacy experts and conducted a literature review. After using these data to construct initial rounding tolerances, we used a Delphi process to achieve consensus about the rounding tolerance for each commonly prescribed medication.
RESULTS:
Three categories for medication-rounding philosophy emerged from our literature review: (1) medications for which rounding is used judiciously to retain the intended effect; (2) medications that are rounded with attention to potential unintended effects; and (3) medications that are rarely rounded because of the potential for toxicity. We assigned a small subset of medications to a fourth category—inadequate data—for which there was insufficient information to provide rounding recommendations. For all 102 medications, we were able to arrive at a consensus recommendation for rounding a given calculated dose.
CONCLUSIONS:
Results of this study provide the pediatric information technology community with a primary set of recommended rounding tolerances for commonly prescribed drugs. The interoperable knowledge base developed here can be integrated with existing and developing electronic prescribing systems, potentially improving prescribing safety and reducing cognitive workload.
Keywords: biomedical informatics, electronic prescribing, clinical practice, computer software, prescriptions, medical informatics, pharmacotherapy
WHAT'S KNOWN ON THIS SUBJECT:
Pediatric electronic-prescribing systems improve care by providing weight-based dose calculations. However, they often generate liquid medication doses that are difficult for families or caregivers to measure and administer accurately and do not consider dosing guidelines of the prescribed drug.
WHAT THIS STUDY ADDS:
This study provides evidence-based and expert-validated rounding recommendations. These data are usable by commercial vendors to improve the rounding capabilities of electronic-prescribing systems. This process should continue with infrequently prescribed medications to improve pediatric prescribing safety.
Pediatric prescribing is a complicated process that requires the prescriber to calculate a medication dose that is appropriate for the treatment goals and for the child's weight or body surface area.1 Although some medications are relatively tolerant of inaccurate dosing, others that have narrow therapeutic indices, such as digoxin or insulin, have a great potential for adverse consequences if dosed improperly.2–4 This process is sufficiently complicated that most pediatricians rely on prescribing guides in print or electronic form to practice safely.5,6 Emerging therapies with less well-published dosing guidelines also make safe and effective pediatric prescribing difficult.5,7
Electronic prescribing (e-prescribing) has become a national initiative supported by the Centers for Medicare & Medicaid Services, the Institute of Medicine, the American Academy of Pediatrics, and other professional societies and organizations. In part, this support has been facilitated by mounting evidence about prescribing errors that have been well summarized8–10 and that promise to improve the rate of medication errors in pediatrics.1,11,12
There are less data to support the use of e-prescribing in the ambulatory pediatric community,13 despite the challenges associated with pediatric patient medication management.14 A study by Kaushal et al15 noted the potential for e-prescribing to prevent up to 21% of adverse drug events in outpatient settings, including those related to drug frequency and weight/dose checks. Most e-prescribing systems do not generate easily administered (rounded) doses. Therefore, these systems often require clinicians to edit the computed dose, increasing the risk of introducing dosing errors.16 In addition, there is no validated resource containing knowledge about the appropriate rounding tolerances for different pediatric medications. There also is no established approach for generating a rounded dose that is easily administered by patients; therefore, it is difficult for e-prescribing systems to adopt a rounding algorithm with the currently available knowledge. The goal of this project was to establish best practices for generating an appropriate medication dose and to create an interoperable rounding knowledge base combining best practices and dose-rounding information.
METHODS
To address the project goals, we relied on information from interviews and a literature review. We also used a consensus-based process to validate proposed rounding tolerances.
Interviews
We conducted semi-structured interviews to assess the overall considerations used in rounding prescribed medications. We used a purposive sampling plan to recruit subjects who included local and national experts in general pediatrics, subspecialty pediatrics, pediatric pharmacy, biomedical informatics, and pharmacology. The sampling plan focused primarily on ambulatory locations, where medication dosing takes into account the needs of home caregivers who must measure and administer the medication.
A member of the project team provided a brief project overview to each domain expert before the interview. Each interview started with a recapitulation of the project goals, and then focused the interview around 3 framing questions: “Please describe how you create a prescription where the effects, both intentional and unintentional, are related to the dose”; “Please describe the process for prescribing a compound medication“; and “What is your process for rounding?” All interviews were transcribed and posted to a project Wiki site (www.dokuwiki.org) with restricted access for review and discussion among project team members. An iterative process based on discussion and revision of rounding techniques was used to focus and refine subsequent interviews. Results from the interviews were used to help define practical approaches to rounding and to develop a framework for categorizing medications according to the usual rounding philosophy as described by our experts.
Literature Review
We reviewed pharmacologic information for 120 medications that comprised >95% of the most commonly prescribed pediatric medications in 2 academic medical centers.17 We collected weight-based dosing guidelines, minimum and maximum dosing amounts, and drug toxicity or adverse effect information from a series of commonly cited articles and texts.18–20 We also consulted the gray literature, including Web resources and the US Food and Drug Administration Web site, for information about some medications. Using the framework developed in the interview phase, we evaluated dosing knowledge for each medication and assigned it to a category within this framework. We used the properties of the assigned rounding category, along with dosing information from domain experts and information about the drug's therapeutic intent and potential adverse reactions obtained from the literature review, to propose an initial rounding percentage.
Rounding Percentage Validation
We used a Delphi approach aimed at generating consensus about each rounding recommendation. This technique works well “to correlate informed judgments on a topic spanning a wide range of disciplines.”21 In our project, this model for consensus building was ideal for a primarily online group of pharmacists, primary care providers, and hospitalists to discuss the degree to which medications should be rounded automatically.
We began by entering proposed rounding percentages into an electronic survey instrument familiar to the expert reviewers called REDCap.22 Each question presented the recommended rounding percentage and allowed each group member to agree or disagree with the recommendation. A sample question is shown in Table 1. Each expert completed the survey anonymously. After experts completed the first round of surveys, a facilitator (Dr Johnson or Dr Helmke) tabulated all responses and provided feedback to the expert group as a whole. All medications for which there was <80% consensus about the proposed rounding percentage were discussed among the participants. On the basis of this discussion, we modified the rounding percentages for these medications, and completed a second round of the survey. We included 2 pediatric neurologists later in this process to review and agree on rounding tolerances for sets of medications typically prescribed by members of that specialty. Finally, there were 16 medications that the group unanimously agreed were out of scope for this effort, either because the group was unfamiliar with their use (eg, ziprasidone) or because it was determined that dosing in pediatric offices typically did not use existing weight-based guidelines (eg, ursodiol). Group consensus for all other medications was achieved after 4 rounds of discussion.
TABLE 1.
Pretend you have a patient who is being prescribed a dose of each medication below. You have calculated this dose using standard milligram per kilogram per day formulae and arrive at 5.82 mL/dose, and there is only 1 formulation to choose from. The e-prescribing system you use returns a rounded dose, based on ease of home administration. |
If the medication allows 0% rounding, the dose will be 5.8 mL. |
If the medication allows 1% rounding, the dose will be between 5.7 and 5.9 mL. You would likely pick 5.9 mL. |
If the medication allows 5% rounding, the dose will be between 5.5 and 5.9 mL. You would likely pick 5.9 mL. |
If the medication allows 10% rounding, the dose will be between 5.2 and 6.4 mL. You would likely pick 6 mL. |
If the medication allows 15% rounding, the dose will be between 4.9 and 6.7 mL. You would likely pick 6 mL. |
Give these choices, for each drug below, do you agree or disagree with the maximum amount we would suggest to allow rounding? |
Digoxin. Do not round further; okay? |
Must provide value |
Calcitriol. Do not round further; okay? |
Must provide value |
Ondansetron (Zofran). Do not round further; okay? |
Must provide value |
Morphine. Do not round further; okay? |
Must provide value |
Tylenol with codeine, round by 5%; okay? |
Must provide value |
Clarithromycin (Biaxin), round by 5%; okay? |
Must provide value |
Diphenhydramine, round by 5%; okay? |
Must provide value |
Griseofulvin, round by 10%; okay? |
Must provide value |
Penicillin, round by 15%; okay? |
Must provide value |
Guaifenesin/dextromethorphan, round by 15%; okay? |
Must provide value |
RESULTS
Rounding Framework Development
Table 2 describes the experience and the roles played by each member of the advisory group. Some members participated in interviews or rounding knowledge validation, or both, as designated below.
TABLE 2.
Pediatric Specialty | Total Participants, n | Median Time in Specialty, y | Interviewed, n | Validated Knowledge (Delphi Process), n |
---|---|---|---|---|
Generalist | 7 | 25 | 4 | 6 |
Hospitalist | 2 | 13 | 2 | 1 |
Cardiologist | 1 | 24 | 1 | 1 |
Nephrologist | 1 | 22 | 1 | 0 |
Pharmacist (PharmD) | 5 | 22 | 3 | 3 |
Hematology | 1 | 8 | 1 | 0 |
Neurologista | 2 | 15 | 0 | 2 |
Involved in second and third rounds of consensus.
On the basis of these interviews, we discovered that domain experts approach medication dosing and rounding by balancing the goals of therapy with the potential for adverse effects related to dosing. Three philosophical approaches emerged. We also created an additional category for medications with insufficient data available to assess the risk of automated rounding.
Dose-Dependent Intended Effects
The first approach was relevant when the intended effect was itself dose dependent. The iconic active ingredient for this approach was furosemide, which produces diuresis in rough proportion to the amount of the medication given per dose. In this case, domain experts typically start low and titrate the drug upward, typically in small (10%) increments. Medications in this group may be automatically rounded up or down in increments smaller than 10% to reach a more easily administered dose with the same intended effect.
Dose-Dependent Unintended Effects
The second approach had as a goal avoiding unintentional adverse effects. This approach is typically used for medications such as antibiotics or systemic steroids, in which dose-dependent adverse effects may be avoided by lowering the dose. For most medications in this group, dosing tends to begin at the highest well-tolerated dosage for the indication, and then rounded down to an easily administered dose, bearing in mind the maximum dosing recommendation guidelines.
Narrow Therapeutic Range
The third approach recognized the potential for drugs to have a narrow therapeutic index and a high risk for toxicity. Drugs such as digoxin and insulin are in this category and typically are not rounded, or are rounded to the nearest one-tenth of a milliliter from the originally calculated dose.
Insufficient Data Available
We included a fourth category for medications, such as mesalamine, in which insufficient data exist about the proper dosing model for children, and toxicity is likely with even a slight overdose based on adult data. In these cases, no rounding is typically performed. For many of these medications, there is no manufactured liquid formulation, further complicating the automated process by requiring the pharmacist to construct a compounded form of the medication. Because the final formulation may not be known, it is not possible to create an easy-to-administer dose during the e-prescribing process.
Recommendations for Rounding
On the basis of conversations with domain experts, we assigned tolerable rounding ranges to each category above. These data are listed in Table 3.
TABLE 3.
Category | Unintended Adverse Effects Dependent on Dose | Impact of Effect Dependent on Dose | Narrow Therapeutic Index | Rounding Tolerance, % |
---|---|---|---|---|
Avoiding unintentional adverse effects | Yes | No | No | 10–15 |
Controlling intended effects | No | Yes | No | 5–10 |
Avoiding toxicity | No | No | Yes | 1–5 |
Insufficient data | NA | NA | Usually | 0–1 |
NA indicates not available.
Table 4 lists the medications, rounding categories, and rounding percentages for each of the 120 medications in our database along with degree of agreement among all members of the advisory group. Medications such as amitryptyline and digoxin were sufficiently toxic that our team recommended rounding down to the nearest one-tenth of a milliliter.
TABLE 4.
Medication (Rounding Tolerance) | Category | Consensus Round |
---|---|---|
Amitriptyline (0%) | Unintended effects | 2 |
Acetaminophen (10%) | Toxicity | 1 |
Amlodipine (10%) | Unintended effects | 1 |
Amoxicillin/clavulanic acid (15%) | Unintended effects | 1 |
Amoxicillin (15%) | Unintended effects | 1 |
APAP/codeine (5%) | Unintended effects | 3 |
Aspirin (10%) | Unintended effects | 1 |
Atenolol (5%) | Excess intended effect | 2 |
Methotrexate (0%) | Toxicity | 1 |
Azathioprine (5%) | Unintended effects | 2 |
Azithromycin (15%) | Unintended effects | 2 |
Budesonide (10%) | Unintended effects | 1 |
Carbamazepine (2%) | Unintended effects | 4 |
Cefdinir (10%) | Unintended effects | 1 |
Cefixime (10%) | Unintended effects | 1 |
Cefuroxime (10%) | Unintended effects | 1 |
Cephalexin (15%) | Unintended effects | 1 |
Chlorothiazide (10%) | Excess intended effect | 1 |
Cimetidine (10%) | Unintended effects | 1 |
Ciprofloxacin (1%) | Unintended effects | 2 |
Clarithromycin (10%) | Unintended effects | 2 |
Clindamycin (10%) | Unintended effects | 2 |
Clonazepam (0%) | Toxicity | 4 |
Clonidine (0%) | Toxicity | 2 |
Cyproheptadine (5%) | Unintended effects | 3 |
Dexamethasone (5%) | Unintended effects | 2 |
Dextroamphetamine (10%) | Unintended effects | 1 |
Diazepam (2%) | Toxicity | 4 |
Digoxin (0%) | Toxicity | 2 |
Diphenhydramine (10%) | Excess intended effect | 3 |
Docusate sodium (10%) | Unintended effects | 2 |
Doxycycline (10%) | Unintended effects | 2 |
Enalapril maleate (10%) | Unintended effects | 1 |
Erythromycin (10%) | Unintended effects | 2 |
Ethosuximide (2%) | Toxicity | 3 |
Famotidine (15%) | Unintended effects | 1 |
Felbamate (0%) | Unintended effects | 3 |
Fluconazole (10%) | Unintended effects | 1 |
Fluoxetine (5%) | Unintended effects | 3 |
Folic acid (10%) | Unintended effects | 1 |
Furosemide (5%) | Excess intended effect | 2 |
Gabapentin (10%) | Unintended effects | 1 |
Glycopyrrolate (5%) | Unintended effects | 3 |
Griseofulvin (10%) | Unintended effects | 2 |
Guaifenesin (15%) | Unintended effects | 1 |
Hydrocodone/APAP (5%) | Unintended effects | 3 |
Hydroxychloroquine (0%) | Toxicity | 2 |
Hydroxyzine (10%) | Unintended effects | 1 |
Ibuprofen (15%) | Unintended effects | 1 |
Iron supplements (10%) | Unintended effects | 3 |
Ketoconazole (10%) | Unintended effects | 1 |
Lamotrigine (10%) | Unintended effects | 1 |
Levetiracetam (2%) | Toxicity | 2 |
Levofloxacin (1%) | Unintended effects | 2 |
Levothyroxine (0%) | Toxicity | 2 |
Lithium (0%) | Toxicity | 3 |
Lorazepam (2%) | Toxicity | 3 |
Mercaptopurine (0%) | Toxicity | 2 |
Mesalamine (0%) | Insufficient evidence | 4 |
Methadone (0%) | Toxicity | 2 |
Methimazole (5%) | Unintended effects | 1 |
Methylprednisolone (10%) | Unintended effects | 1 |
Metoclopramide (10%) | Unintended effects | 1 |
Metronidazole (5%) | Unintended effects | 2 |
Minocycline (10%) | Unintended effects | 1 |
Morphine (0%) | Toxicity | 2 |
Moxifloxacin (1%) | Unintended effects | 2 |
Mycophenolate (5%) | Unintended effects | 1 |
Naproxen (10%) | Unintended effects | 1 |
Nortriptyline (0%) | Insufficient evidence | 2 |
Omeprazole (10%) | Unintended effects | 2 |
Ondansetron (10%) | Unintended effects | 2 |
Oxcarbazapine (5%) | Unintended effects | 4 |
Oxycodone (5%) | Unintended effects | 2 |
Oxycodone/APAP (5%) | Unintended effects | 2 |
Pancreatin (15%) | Unintended effects | 3 |
Pancrelipase (15%) | Unintended effects | 3 |
PEG (15%) | Unintended effects | 3 |
Penicillin V (15%) | Unintended effects | 2 |
Permethrin (15%) | Unintended effects | 3 |
Phenobarbital (2%) | Toxicity | 4 |
Pimozide (15%) | Unintended effects | 4 |
Prednisolone (10%) | Unintended effects | 1 |
Prednisone (10%) | Unintended effects | 1 |
Pregabalin (2%) | Unintended effects | 3 |
Propranolol (5%) | Excess intended effect | 2 |
Pyridoxine (15%) | Unintended effects | 1 |
Ranitidine (15%) | Unintended effects | 1 |
Risperidone (5%) | Toxicity | 3 |
Rizatriptan (5%) | Toxicity | 3 |
Sertraline (15%) | Unintended effects | 3 |
Spironolactone (10%) | Excess intended effect | 1 |
Sucralfate (15%) | Unintended effects | 1 |
Sumatriptan (5%) | Toxicity | 1 |
Tizanidine (5%) | Toxicity | 3 |
Topiramate (2%) | Excess intended effect | 4 |
Trazodone (15%) | Excess intended effect | 4 |
Trimethoprim/SMX (10%) | Unintended effects | 2 |
Valproic acid (2%) | Toxicity | 4 |
Warfarin (0%) | Toxicity | 1 |
Zolmitriptan (15%) | Excess intended effect | 1 |
Zonisamide (2%) | Unintended effects | 4 |
Category refers to the general dosing philosophy used by practitioners for this medication; consensus round refers to the number of times this medication was discussed before consensus was achieved. APAP indicates acetaminophen; PEG, polyethylene glycol; SMX, sulfamethoxazole.
In most cases, expert review resulted in a widening, rather than a narrowing, of the rounding tolerance. For example, despite the risk of dose-dependent tardive dyskinesia associated with using metoclopramide, in practice this drug is often rounded more aggressively. Therefore, the rounding percentage was increased from 5% as initially proposed to 10%. We achieved unanimous consensus for 39% of the proposed medication-rounding tolerances in the first round of voting. There were medications that the group agreed were rarely used in practice and were therefore out of scope for this initial project. These drugs included antiretroviral medications (ritonavir, oseltamivir, and lamivudine) and some rarely used neurologic medications. After discussion at a face-to-face meeting, 2 subsequent rounds resulted in all but 7 medication-rounding percentages being acceptable to the group. This final group of medications, typically started by pediatric neurologists, required extensive discussion between the expert group and 2 guest neurologists before consensus was reached.
DISCUSSION
Using information obtained from pediatricians, pharmacists, and an extensive literature review, we proposed and validated a set of rounding ranges and then assigned each medication to a rounding category. These data form an important body of knowledge that can be used by e-prescribing systems to automatically calculate administrable and safe doses for young children. Moreover, by using an iterative consensus process, we obtained a high level of agreement across 99.3% of all medications.
Our validation process was limited in power by the relatively small number of pediatricians and pharmacies in our advisory group; however, the high level of agreement among pediatricians with an average of 20 years of experience suggests that our set of recommendations can serve as the critical first step to establishing a comprehensive knowledge base. Additional studies should evaluate the extent to which these recommendations are accepted or overridden by pediatric prescribers, as well as the incidence of adverse events that may arise related to rounding. We also were only able to create rounding tolerances for a subset of medications that are commonly prescribed. This process should be continued to address other medications prescribed using weight or body surface area to derive a dose. In addition, it will be important to reconvene an expert group to evaluate new medications that reach the consumer market. We estimate that this process would need to be done at least every 3 years, to address the ∼12 new medications that are added to the Harriet Lane Handbook every edition (C. K. K. L., unpublished data, 2011).
As noted here, this project is the first attempt at creating a knowledge base of pediatric rounding tolerances that can be incorporated into any e-prescribing system whose vendor has an interest in using this knowledge. However, this work is only the beginning of developing tools required to create a safe pediatric e-prescribing environment. In part, the challenge to this task is the lack of evidence supporting common dosing standards, such as those published in typical pediatric textbooks. However, even these “eminence-based” recommendations do not exist in a computable form. In an era in which there are mandates and incentives for adopting e-prescribing systems, this deficit is likely to negatively impact the otherwise highly usable systems recommended by regional extension centers or certifying bodies. Our study provides an initial expert-developed set of recommendations that can be readily incorporated into an evolving pediatric e-prescribing milieu. It is our hope that these recommendations will provide a starting point for all vendors to improve their automated processes.
It is also important to recognize the challenges of home medication administration, as noted by Kaushal et al,15 when creating a medication dose. These recommendations will need to be combined with models that facilitate the cognitive processes associated with selecting an appropriate formulation, taking into account variables such as the patient's age, the presence of a feeding tube, patient preferences for volume of medication, and other considerations.
ACKNOWLEDGMENTS
This research project was funded by the Agency for Healthcare Research and Quality under grant R18HS17216. We thank the agency for their support and thoughtful comments throughout this project, and we thank the members of the STEPSTools Working Advisory Group. They all played a key role in the development of the Web service tool, and their contributions to this article are greatly appreciated.
The following are members of the STEPSTools Working Advisory Group: John Canning, Physician's Computer Company; Coda L. Davison, MPA, BBA, Vanderbilt University School of Medicine; Cynthia S. Gadd, PhD, MBA, Vanderbilt University School of Medicine; Robert Grundmeier, MD, Children's Hospital of Philadelphia; Chip Hart, Physician's Computer Company; Jill S. Helmke, DPh, NPh, Vanderbilt University School of Medicine; Yun-Xian Ho, PhD, Vanderbilt University School of Medicine; Carlton K. K. Lee, PharmD, MPH, The Johns Hopkins University School of Medicine; Jennifer Mansour, MSW, CPHIT, CPEHR, American Academy of Pediatrics; Eugenia Marcus, MD, Pediatric Health Care; Beki Marshall, CPHIT, CPEHR, American Academy of Pediatrics; Jerome A. Osheroff, MD, Thomas Reuters; Marvin Palmer, Vanderbilt University School of Medicine; Mark M. Simonian, MD, American Academy of Pediatrics Liaison; S. Andrew Spooner, MD, MS, Cincinnati Children's Hospital; and Stuart T. Weinberg, MD, Vanderbilt University School of Medicine.
Dr Johnson drafted the majority of the manuscript, participated in the data acquisition, and conducted the analysis of the data and contributed to its interpretation; Dr Lee edited the manuscript and verified the rounding tolerance data; Dr Helmke developed the knowledge base for the data acquisition and participated in the review and open coding of expert interviews; Dr Weinberg edited the manuscript and reviewed the rounding tolerance data for accuracy; Dr Spooner participated in subject recruitment, data collection, and manuscript preparation; and Mr Davison drafted parts of the introduction and acknowledgments section and assisted with the literature review.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
Abbreviation:
- e-prescribing
- electronic prescribing
REFERENCES
- 1. Sullivan JE, Buchino JJ. Medication errors in pediatrics: the octopus evading defeat. J Surg Oncol. 2004;88:182–188 [DOI] [PubMed] [Google Scholar]
- 2. Lesar TS. Errors in the use of medication dosage equations. Arch Pediatr Adolesc Med. 1998;152(4):340–344 [DOI] [PubMed] [Google Scholar]
- 3. Lesar TS, Briceland LL, Delcoure K, Parmalee JC, Masta-Gornic V, Pohl H. Medication prescribing errors in a teaching hospital. JAMA. 1990;263(17):2329–2334 [PubMed] [Google Scholar]
- 4. Rowe C, Koren T, Koren G. Errors by paediatric residents in calculating drug doses. Arch Dis Child. 1998;79(1):56–58 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Kim GR, Bartlett EL, Jr, Lehmann HP. Information resource preferences by general pediatricians in office settings: a qualitative study. BMC Med Inform Decis Mak. 2005;5:34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Khan AN, Frank J, Geria R, Davidson S. Utilization of personal digital assistants (PDAS) by pediatric and emergency medicine residents. J Emerg Med. 2007;32(4):423–428 [DOI] [PubMed] [Google Scholar]
- 7. D'Alessandro DM, Kreiter CD, Peterson MW, Kingsley P, Johnson-West J. An analysis of patient care questions asked by pediatricians at an academic medical center. Ambul Pediatr. 2004;4(1):18–23 [DOI] [PubMed] [Google Scholar]
- 8. Aspden P, Corrigan JM, Wolcott J, Erickson SM. eds. Institute of Medicine. Patient Safety: Achieving a New Standard for Care. Committee on Data Standards for Patient Safety. Washington, DC: National Academies Press; 2004 [PubMed] [Google Scholar]
- 9. Fortescue EB, Kaushal R, Landrigan CP, et al. Prioritizing strategies for preventing medication errors and adverse drug events in pediatric inpatients. Pediatrics. 2003;111(4 pt 1):722–729 [DOI] [PubMed] [Google Scholar]
- 10. Aspden P, Wolcott JA, Bootman JL, Cronenwett LR. eds. Institute of Medicine, Committee on Identifying and Preventing Medication Errors. Preventing Medication Errors. Washington, DC: National Academies Press; 2007 [Google Scholar]
- 11. 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(9):1002–1007 [DOI] [PubMed] [Google Scholar]
- 12. 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 [DOI] [PubMed] [Google Scholar]
- 13. American Academy of Pediatrics Committee on Clinical Information Technology, Gerstle RS. Electronic prescribing systems in pediatrics: the rationale and functionality requirements. Pediatrics. 2007;119(6):1229–1231 [DOI] [PubMed] [Google Scholar]
- 14. Levine SR, Cohen MR, Blanchard NR, et al. Guidelines for preventing medication errors in pediatrics. J Ped Pharmacol Ther. 2001;6:426–442 [Google Scholar]
- 15. Kaushal R, Goldmann DA, Keohane CA, et al. Adverse drug events in pediatric outpatients. Ambul Pediatr. 2007;7(5):383–389 [DOI] [PubMed] [Google Scholar]
- 16. Killelea BK, Kaushal R, Cooper M, Kuperman GJ. To what extent do pediatricians accept computer-based dosing suggestions? Pediatrics. 2007;119(1). Available at: www.pediatrics.org/cgi/content/full/119/1/e69 [DOI] [PubMed] [Google Scholar]
- 17. Johnson KB, Weissenberger S, Lee CK, Walker A, Serwint JR. Myth Versus Reality: What and How Do General Pediatricians Prescribe? Paper presented at: 2007 Pediatric Academic Societies' Annual Meeting; May 5–8, 2007; Toronto, Ontario, Canada [Google Scholar]
- 18. Custer JW, Rau RE. eds. The Harriet Lane Handbook : A Manual for Pediatric House Officers. 18th ed Philadelphia, PA: Mosby/Elsevier; 2009 [Google Scholar]
- 19. Lexi-Comp, Inc, American Pharmaceutical Association Drug Information Handbook. Hudson, OH/Washington, DC: American Pharmaceutical Association; 2003 [Google Scholar]
- 20. Katzung BG. Basic and Clinical Pharmacology. 9th ed New York, NY: Lange Medical Books/McGraw Hill; 2004 [Google Scholar]
- 21. Delbecq AL, van de Ven AH, Gustafson DH. Group Techniques for Program Planning: A Guide to Nominal Group and Delphi Processes. Glenview, IL: Scott Foresman; 1975 [Google Scholar]
- 22. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap): a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42(2):377–381 [DOI] [PMC free article] [PubMed] [Google Scholar]