Table 3.
Predictive value of co-occurrence of a medical incident commonly captured by automated data systems and a suspect prescribing discrepancy.
| Occurrences | ADEs | Positive predictive value | |
|---|---|---|---|
| Incident/example of suspect discrepancy | (N)* | (N) | (N/N) (95% CI) |
| Pain**/analgesic discrepancy | 59 | 22 | .37 (.25–.51) |
| High blood pressure**/antihypertensive discrepancy | 32 | 8 | .33 (.16–.55) |
| Renal insufficiency**/angiotensin blocker discrepancy | 14 | 4 | .29 (.10–.58) |
| Low blood pressure**/antihypertensive discrepancy | 34 | 6 | .18 (.07–.35) |
| Fever**/antibiotic discrepancy | 28 | 2 | .07 (.01–.25) |
Episode types are included that had at least 10 occurrences.
Pain: new or worse pain; fever: temperature > 100.5 F (38 C); high blood pressure: new systolic blood pressure > 185, or diastolic blood pressure > 105, or an increase in systolic or diastolic blood pressure of 30mm; low blood pressure: new systolic blood pressure < 95, or a drop in systolic blood pressure of 30mm; renal insufficiency: new creatinine increase >0.5 mg/dl.