11
Errors Associated with the Errors Associated with the Use of E-PrescribingUse of E-Prescribing
Eric Poon MD, MPHEric Poon MD, MPHIS Director of Clinical Informatics, IS Director of Clinical Informatics,
Associate Physician, Associate Physician, Brigham and Women’s Hospital, Boston, MABrigham and Women’s Hospital, Boston, MA
Assistant Professor, Harvard Medical SchoolAssistant Professor, Harvard Medical School
Supported by a grant from the Risk Management Foundation
22
AgendaAgenda
Errors associated with e-prescribingErrors associated with e-prescribingTaxonomy with examplesTaxonomy with examplesContributing factorsContributing factors
E-Prescribing Comes of Age!
Strong evidence that electronic prescribing in the inpatient setting prevents serious medication errors Emerging evidence in the ambulatory setting
Significant interest at the local, state, and federal level 100% e-prescribing target for PCPs and specialists at
Partners Healthcare Use of e-prescribing as P4P metric National efforts
44
Unintended Consequences: The Underbelly Unintended Consequences: The Underbelly of Healthcare Information Technology (HIT)of Healthcare Information Technology (HIT)
Events or outcomes that are neither Events or outcomes that are neither anticipated noranticipated nor the specific goals of the the specific goals of the associated HIT associated HIT May be negative or positiveMay be negative or positive
Best studied more recently in the inpatient Best studied more recently in the inpatient settingsettingNew errors from use of inpatient CPOE has New errors from use of inpatient CPOE has
received significant attentionreceived significant attention
66
New Errors: New Errors: Revisiting an Old PhenomenonRevisiting an Old Phenomenon
New errors observed in the inpatient New errors observed in the inpatient setting (Koppel, Han, Ash)setting (Koppel, Han, Ash)Extend and severity not fully understood in Extend and severity not fully understood in
the outpatient settingthe outpatient settingStudy goalsStudy goals
Develop taxonomy to classify errorsDevelop taxonomy to classify errorsDescribe contributing factorsDescribe contributing factors
Range of Errors Associated with e-prescribing
88
Key Components of a prescriptionKey Components of a prescription
Drug productDrug product DoseDose FrequencyFrequency RouteRoute Dispensing quantityDispensing quantity Refill quantitiesRefill quantities InstructionsInstructions Identification and authentication of prescriberIdentification and authentication of prescriber
Methods Drafting of Taxonomy
2 sets of electronic prescriptions with errors reviewed Convenience sample at CRICO site 1, collected by outpatient
pharmacist and medical director of adult primary care practice (200+ prescriptions)
Electronic prescription reviewed at CRICO site 2 as part of AHRQ-funded project to evaluate new e-prescribing standards
Validation of Taxonomy 2 focus groups with community pharmacists
Contributing factors Recreation of prescriptions with errors in native system Consultation with informatics and IT professionals
1010
Drug Product Errors Drug Product Errors
Incomplete Drug NameIncomplete Drug NameStrength Omitted or Error in StrengthStrength Omitted or Error in Strength Incorrect Drug Chosen and Script Incorrect Drug Chosen and Script
Manually AlteredManually Altered
1111
Error in Product StrengthError in Product Strength
1212
Dose ErrorsDose Errors
Dose OmittedDose Omitted Incorrect DoseIncorrect DoseAmbiguity in Sig FieldAmbiguity in Sig FieldDose IncompleteDose IncompleteOverdoseOverdoseUnderdoseUnderdose
1313
Ambiguous DoseAmbiguous Dose11.25mg once
a day, or 7.5mg once a
day?
1414
Ambiguity in Sig FieldAmbiguity in Sig Field
Which do you trust? The
computer or the provider’s
Spanish?
1515
Route ErrorsRoute Errors
Incorrect RouteIncorrect RouteOmitted and Should Be POOmitted and Should Be POOmitted and should be Other RouteOmitted and should be Other Route
1616
Route IncorrectRoute Incorrect
Underlying cause: Inadequate synonym support in medication lookup ‘Nuva ring’ (not recognized) vs ‘Nuvaring’ (recognized)
1717
Frequency ErrorsFrequency Errors
Frequency OmittedFrequency OmittedFrequency ChangedFrequency Changed
1818
Ambiguous frequency with incorrect Ambiguous frequency with incorrect hand-written alterationhand-written alteration
1919
Special InstructionsSpecial Instructions Dose MismatchDose Mismatch Drug Product Strength MismatchDrug Product Strength Mismatch Dose Form MismatchDose Form Mismatch Frequency MismatchFrequency Mismatch Route Dosage Form MismatchRoute Dosage Form Mismatch PRN is Indicated in Special Instructions but not in the PRN is Indicated in Special Instructions but not in the
Sig. Sig. Duration of Therapy MismatchDuration of Therapy Mismatch Quantity MismatchQuantity Mismatch Dose and Frequency MismatchDose and Frequency Mismatch PRN Indication MismatchPRN Indication Mismatch Previous Special Instructions Carrying over on RenewalPrevious Special Instructions Carrying over on Renewal
2020
Directions that contradict rest of the Directions that contradict rest of the prescriptionprescription
Provider lack of familiarity with ‘alternate dosing’ prescribing method
Factors Contributing to Errors:A Preliminary Look
Categories of Contributing Factors
Technology Factors
Environmental FactorsUser Factors•Knowledge deficit about application features•Knowledge deficit about the prescription •Multi-tasking•Improper use of short-cuts•Improper use of free-text
•User-interface design•Medication dictionary•Data entry form factor
•Lack of time during visit•Placement of computer equipment
2727
Dissecting an ErrorDissecting an Error
What the prescribing module would have looked like…
7.5 mg
Free Text Entry Possible at the
time
Dose field relatively new to clinicians
History
1. Patient’s BP not well controlled on 5mg of Norvasc
qd. Increase to 7.5mg qd
3. Instead of specifying the higher dose in the dose field, provider
entered the dose as free text in the strength/form field
2. Norvasc only available in 2.5mg, 5 mg,and 10mg tablets
3030
Writing prescriptions – do we know how to?Writing prescriptions – do we know how to?
How much training have you received?Has anyone ever shown you this
relationship?
Dose
Strength Take
Dose Dose ≈ Strength x Take≈ Strength x Take
Lisinopril 40mg Lisinopril 40mg po once a daypo once a day
Lisinopril 20mg Lisinopril 20mg tabletstablets
Take one tablet Take one tablet a daya day
Other common contributing factors
Medication dictionary issuesWrong/not available product in dictionary
Multi-tasking/distractions/short-cutsNot seeing/checking what is actually been
typed inKeyboard vs. pen;
Propagation of previous errorsCarry forward of erroneous information
during renewals
Preliminary Recommendations
Possible User Interventions Focused training on use of e-prescribing module
Hands-on, real-life examples ‘Real life’ context Focus on common errors Ongoing effort
Education about the prescription Medical students & physicians in training
Develop mechanism so that prescription errors become teaching tools.
Develop and disseminate best practices for how to incorporate EMR into the visit
Possible Technology Interventions
Usability study/pilot before new features go-live
Explore different form factors for data entry
Monitor for use of free-text in structured fields
Formalize mechanism for using prescription errors as opportunities to improve user-interface design
Examples of recent improvements
Relationship between dose, strength and take
made more explicitWarning appears when
free text instructions are typed in
3737
Concluding RemarksConcluding Remarks
E-prescribing has significant potential to E-prescribing has significant potential to improve medication safetyimprove medication safety
Like all other technologies, e-prescribing Like all other technologies, e-prescribing opens doors to new errors. opens doors to new errors. Causation often multi-factorialCausation often multi-factorialFrustratingly common-placeFrustratingly common-place
Investment in time, energy and Investment in time, energy and mechanism for interventionsmechanism for interventions
Are we still in denial?
3939
AcknowledgementsAcknowledgements
Thomas Moniz, RPhThomas Moniz, RPhJeff Rothschild, MD MPHJeff Rothschild, MD MPHMatvey Palchuk, MD MS Matvey Palchuk, MD MS Matthew DitmoreMatthew DitmoreHans Kim, MDHans Kim, MD