Table 2 Nine triggers that most commonly represented adverse drug events (ADEs)
No. of triggers reviewedDuplicate triggers*False triggers†Qualifying triggers‡ADEsPositive predictive value (%)§PADEsNon-preventable ADEs
No. (percentage of all ADEs)Primary process at which event was preventable (percentage of PADEs)Effect on patient¶No. (percentage of all ADEs)Effect on patient¶
1. Medication stop59037950815526.347 (30.3%)23 prescribing (48.9%)8 none/minimal108 (69.7%)14 none/minimal70 mild6 moderate4 severe14 missing data
6 mild
4 moderate
5 severe
22 administration (46.8%)15 none/minimal
4 mild
1 moderate
1 severe
1 missing data
2 monitoring (4.3%)2 severe
2. Hospitalisation10102992221.816 (72.7%)9 prescribing (56.3%)9 severe6 (37.5%)6 severe
5 administration (31.25%)5 severe
2 monitoring (12.5%)2 severe
3. Emergency-department visits9414891414.98 (57.1%)2 prescribing (25%)2 moderate6 (42.9%)5 moderate1 severe
3 administration (37.5%)1 mild
2 moderate
2 monitoring (25%)2 severe
1 no stage given (12.5%)1 moderate
4. INR>580088100.08 (100.0%)6 prescribing (75%)4 none/minimal0 (0%)NA
1 mild
1 missing data
2 monitoring (25%)2 none/minimal
5. TSH<0.03 mIU/l on thyroxine10019990.03 (33.3%)2 prescribing (66.6%)2 none/minimal6 (67.7%)4 none/minimal
1 monitoring (33.3%)1 none/minimal2 missing data
6. Creat>2.5 mg/dl152112213.32 (100%)2 prescribing (100%)1 mild0 (0%)NA
1 moderate
7. BUN>60 mg/dl1592416.71 (100%)1 prescribing (100.0%)1 none/minimal0 (0%)NA
8. ALT>84 IU/l133010538.51 (20.0%)1 prescribing (100.0%)1 mild4 (80%)3 none/minimal
1 mild
9. AST>80 IU/l154011320.00NANA3 (100%)2 none/minimal
1 mild
Total861228975021925.486 (39.3%)46 prescribing33 none/minimal133 (60.7%)23 none/minimal
30 administration14 mild72 mild
11 moderate11 moderate
9 monitoring26 severe11 severe
1 no stage given2 missing data16 missing data
  • *Duplicate triggers are triggers that were deemed at the review stage to be repeat occurrences of an already reviewed trigger that was deemed to be due to a chronic disease and therefore was expected. For example, for a patient with elevated alanine aminotransferase (ALT), the first occurrence was reviewed, and it was determined that it was due to chronic liver disease. Further occurrences of this trigger were then labelled as duplicate triggers, since they did not represent a new event but merely an expected chronic abnormality

  • †False-positive triggers are those triggers that were determined, upon review, to be erroneously recorded. This happened most commonly with the “Medication stop” trigger

  • ‡Qualifying triggers are those that qualified for review, meaning that they were neither duplicate nor false-positive triggers.

  • §Positive predictive value is the probability that a positive trigger represents an ADE.

  • ¶Effect on patient was classified into one of four categories: (1) none/minimal, no change in symptoms, may have abnormal labs; (2) mild, mild reversible complications, not requiring hospitalisation—for example, brief symptoms lasting less than 1 day; (3) moderate, moderate reversible complications, not requiring hospitalisation—for example, emergency-department visit, prolonged symptoms lasting a day or longer); (4) severe, severe or irreversible complications—for example, hospitalisation, permanent disability, death.

  • AST, aspartate aminotransferase; BUN, blood urea nitrogen; INR, international normalized ratio; PADE, preventable adverse drug event; TSH, thyroid-stimulating hormone.