Impact of the Institute for Healthcare Improvement Global Trigger Tool Compared to Current Adverse Event Monitoring Systems at a Large Teaching Hospital Tania Vila, PharmD Drug Information Specialty Resident Duke University Hospital Co-Investigators: Heidi Cozart, R.Ph.; Lynn Eschenbacher, PharmD, MBA; Diana Brown, RN; William Richardson, MD

Background: Adverse Event Detection Methods 44,000 to 98,000 deaths per year occur in U.S. hospitals as a result of errors Three-pronged approach to adverse event monitoring Voluntary Reporting Computerized Monitoring Global Trigger Tool – New methodology Kohn LT, Corrigan JM, Donaldson MS, editors. To err is human: building a safer health system. Washington, DC: National Academy Press; 2000.

Background: Adverse Event Detection Methods Voluntary Reporting Qualitative data informs and guides safety and quality (sentinel events, near misses) First-hand account of event from reporters Anonymous & accessible to all hospital employees (available online) Under-reporting

Background: Adverse Event Detection Methods Computerized Monitoring Automated surveillance for possible ADEs based on logic-based rules followed by chart review Scans demographic, laboratory, pharmacy system, and other clinical databases Quantitative data allows for trending Standardized scoring system for severity and causality (Naranjo Scale) Only high risk medication rules are evaluated Killbridge P, Classen. Surveillance for adverse drug events: history, methods and current issues. VHA’s 2002 Research series;3:1-44. 4

Example of Computerized Monitoring (ADE-S) Rules Antidotes or combinations of medications and laboratory values Naloxone IV Dextrose 50% IV AND low blood glucose (<50 mg/dL) 2 consecutive aPTT values > 100 OR one >150 seconds in patient receiving heparin

Background: Adverse Event Detection Methods IHI Global Trigger Tool™ New methodology Trigger-based manual chart review Review team minimum of 3 people 20 charts per month recommended 20 minutes/chart Less resource intensive Griffin FA, Resar RK. IHI Global Trigger Tool for Measuring Adverse Events. IHI Innovation Series white paper. Cambridge, MA: Institute for Healthcare Improvement; 2007.

7

Purpose: Primary Objective To determine the frequency and types of adverse events in urology and orthopedics services as identified by the IHI Global Trigger Tool™ Frequency defined as follows: Adverse events/1,000 patient days

Purpose: Secondary Objectives To compare the frequency and characteristics of AEs detected by the IHI Global Trigger Tool™, computerized monitoring, and the voluntary reporting system To identify potential quality improvement opportunities and to recommend enhancements to the computerized monitoring system based on results

Design Retrospective chart review of patients admitted between January 1, 2007 and June 30, 2007 Inclusion Criteria ≥18 years Patients admitted to urology or orthopedics Chart must be closed and complete Approved by the Duke University Institutional Review Board

Methods: IHI Global Trigger Tool™ 120 patients admitted to urology and orthopedic services were randomly selected (5/service) in 2 week blocks Training: Nurse, pharmacist, and physician completed IHI training process Phase I: IHI standardized practice charts with key Phase II: IHI trigger tool with Duke-specific charts Data Collection: Nurse and pharmacist reviewed each chart; MD reviewed discordant cases

IHI Chart Review Process Data Collection Process Discharge summary Laboratory results Physician orders/MAR Nursing notes/flow sheets and progress notes Documentation All triggers documented whether or not an adverse event occurred Event severity was scored according to an internal scale and a nationally validated scale

Methods: Severity Scores Severity Index, (internal scale; 0 through 6) Only 3 through 6 represent patient harm 3 Transient adverse patient effects occurred which required some corrective therapy, increased length of stay (LOS) 1-2 days or resulted in lab values, vital signs or medication effects outside the desirable parameters 4 Significant adverse patient effects occurred which required aggressive intervention such as code, intubation, transfer to ICU, antidote, interventional drug therapy or increased LOS > 2 days 5 Permanent adverse patient effects occurred such as paralysis, brain damage, disability or loss of limb, organ, or bodily function 6 Patient death

Methods: Severity Scores National Coordinating Council for Medication Error Reporting and Prevention (NCC MERP), (A through I) Only E through I represent patient harm E Temporary harm to the patient and required intervention F Temporary harm to the patient and required initial or prolonged hospitalization G Permanent patient harm H Intervention required to sustain life I Patient death

Patient Population Definitions IHI Global Trigger Tool™ Random sample of 120 patients from all orthopedics and urology admissions between Jan. 1 to June 30, 2007 Voluntary Reporting and Computerized Monitoring Unique admissions to orthopedics or urology services on the 2 patient care units most likely to care for these patients between Jan. 1 to June 30, 2007

Results: Demographics of Adverse Event Population Detection Method (admissions with AE*) IHI (n = 50) Computerized Monitoring (n = 17) Voluntary Reporting (n =11) Mean Age, years ± SD Range 18-64 years, n (%) ≥65 years, n (%) 59.9 ± 15.5 21-90 29 (58%) 21 (42%) 69.5 ± 14.2 37-90 5 (29%) 12 (71%) 51.2 ± 13.1 30-74 10 (91%) 1 (9%) Gender Male, n (%) 31 (62%) 7 (41%) 5 (46%) Admit Service Orthopedics, n (%) Urology, n (%) 24 (48%) 26 (52%) 9 (53%) 8 (47%) 9 (73%) 3 (27%) Median LOS †, Days 4 1-23 8 1-44 5 3-14 *unique admissions to orthopedics or urology with an adverse event identified †LOS = Length of stay

Results: Adverse Event Detection Methods 59 events 57 19 events IHI Trigger Tool 289 positive triggers IHI Trigger Tool Computerized Monitoring 222 rules fired Computerized Monitoring 1 16 1 2 Voluntary Reporting 85 reports Voluntary Reporting Random sample of 120 unique admissions 10 1294 unique admissions 13 events 17

Results: Adverse Event Frequencies IHI Global Trigger Tool Computerized Monitoring Voluntary Reporting High risk medication related adverse events/1000 patients Adverse events/1000 patients Primary endpoint 136 SRS compared to IHI global trigger tool and ADE-S – less than 2-fold difference; but IHI captures much more overall than ADE-S or SRS; HC: It is amazing how close these numbers are.. 4 3 4 2 23 18

Results: IHI Global Trigger Tool™ Adverse Event by Category Fall, 1.7% Miscellaneous†, 5.1% Narcotics/Benzodiazepines, 11.9% Care Module (47.5%) Medication Module (30.6%) Surgical (22%) Hospital-associated infection, 13.6% Chest pain, 3.4% Thrombosis/ Anticoagulation, 5.1% Adverse drug reaction, 13.6% Surgical complications, 22% 59 adverse events detected Transfusion/blood loss anemia, 23.7% †Miscellaneous includes chemo-induced neutropenia (n = 1), hyponatremia/delirium (n = 2)

Results: Computerized Monitoring Adverse Event by Category Compared to IHI… 11.9% captured by IHI Narcotics/ Benzodiazepines, 26.3% 19 adverse events detected Anticoagulants, 52.6% Hypoglycemia, 21.1% 5.1% captured by IHI 0% captured by IHI

Results: Voluntary Reporting Adverse Events by Category Cathartics/Laxatives, 7.7% Electrolyte Replacement Therapy (potassium), 7.7% Compared to IHI… More high risk medication adverse events No hypoglycemia events Anesthetics, 7.7% Narcotics/ Benzodiazepines, 38.5% Anticoagulants, 15.4% 13 adverse events detected Antibiotics, 23.1%

Results: Adverse Event Severity Internal Severity Index Nationally validated NCC MERP 84.7% 79% 66% Percentage (%) 67% 34% 21% 33% 15.3% <3 = did not reach patient; 3 = temporary harm, required corrective therapy OR resulted in labs/vitals outside desirable range; 4 = required aggressive treatment, prolonged LOS > 2 days OR resulted in initial hospitalization

Future Enhancements Expand computerized monitoring Discharge notes and nursing notes were valuable in adverse event identification in IHI As technology advances and nursing notes become electronic, free-text scanning may become a valuable implementation Example “oversedation” search versus IV naloxone rule

Limitations Retrospective chart review Not generalizable to other services Causality and preventability not assessed on IHI Global Trigger Tool™ Physician reviewer was an orthopedic surgeon which may introduce bias

Limitations Subjective judgment of adverse events (e.g. anemia due to blood loss from surgery) Not 1:1 comparisons with IHI Computerized monitoring and voluntary reporting data extracted by the units most likely to care for these orthopedics and urology patients IHI Trigger Tool™ used a random sample rather than all admissions to orthopedics and urology

Conclusions IHI Trigger Tool™ identifies the greatest number of events overall Computerized monitoring and voluntary reporting identify greater proportion of medication-related events Voluntary systems good at capturing near misses/rare events whereas surveillance (IHI and computerized monitoring) better at capturing harm All 3 adverse event detection methods are complementary 26

Acknowledgements Department of Pharmacy Residency Research Committee Computerized Patient Safety Initiatives, Duke Health Technology Solutions Julie Eckstrand, Pharm.D. Monica Horvath, Ph.D. Andrea Long, Pharm.D. Julie Whitehurst, Pharm.D.

Questions

ADE-S System Pharmacy Clinical Rules Engine Laboratory Demographic Data Pharmacy Immediate Intervention required? Health-system clinical pharmacists List of Triggers (daily) Computerized Patient Safety Initiative (CPSI) pharmacists Score and document ADE causality, severity, and narrative Investigate trigger (e.g. chart review) Kappa pharmacist No Yes Kappa pharmacist Intervene and document intervention

Adverse Drug Reaction Probability Scale Do Not Know Adverse Drug Reaction Probability Scale Yes No Are there previous conclusive reports on this reaction? +1 Did the adverse event appear after the drug was administered? +2 -1 Did the adverse reaction improve when the drug was discontinued or a specific antagonist was administered? Did the adverse reaction reappear when the drug was re-administered? Are there alternative causes (other than the drug) that could, on their own, have caused the reaction? -2 Did the reaction re-appear when a placebo was given? Was the drug detected in the blood (or other fluids) in concentrations known to be toxic? Was the reaction more severe when the dose was decreased? Did the patient have a similar reaction to the same or similar drugs in any previous exposure? Was the adverse event confirmed by an objective evidence? ≥9 Definite; 8 to 5 Probable; 1 to 4 Possible; ≤0 Doubtful Naranjo CA, Busto U, Sellers EM, et al. A method for estimating the probability of adverse drug reactions. Clin Pharmacol Ther 1981;30:239-245.

Sample Size Calculation 87% power at 120 charts to detect a 4-fold difference in rates between IHI Trigger Tool™ and ADE-S during a 6 month period 31

IHI Surgical Trigger Tool™ Surgical Trigger Tool™ contains 28 triggers Surgical triggers have corresponding code on global trigger tool (not limited to surgical module) 193 surgical triggers vs. 289 global triggers (67% correspondence) 34

Results: IHI Triggers by Frequency Return to surgery Admission to ICU post-op X-ray or doppler studies for emboli Any procedure complication SCr > 2 times baseline Consult requested in PACU C. diff positive culture Romazicon (flumazenil) use Time in ED > 6 h Narcan (naloxone) use Abrupt med stop 3.5% op-time >6 h 3.1% Fall 2.8% Other triggers 20% X-ray intra-op 5.6% Transfusion 10.1% Benadryl use 6.9% Abrupt drop in Hct of 25% 13.9% Oversedation/ hypotension 7.6% Anti-emetic use 18.4% Readmission 8% 289 triggers