Question: If a patient reports an allergy to codeine but has tolerated Demerol in the past, can that information should be stored such that an alert is not fired when Demerol is ordered? Also, if you are prescribing a narcotic that fires an alert and you override it, can the information be saved for that patient so that you don't have to override it every time you prescribe that drug?
Medications are entered into the allergy profile for a variety of reasons. Physicians, RNs and other clinical staff will enter a particular drug into the allergy profile as allergies when, in fact, it would be more accurate to report intolerance, side effect, idiosyncratic reaction, or another more fitting reason to avoid the drug. While it is virtually unheard of that a patient is truly allergic to a narcotic/synthetic opioid, there are a variety of intolerances that warrant notation. Though it might seem gratuitous to fire an allergy alert when trying to prescribe Demerol in a patient with a documented codeine intolerance, we must remain mindful of potential issues with drugs within the same class.
Of course, there are instances when patients have had a reaction to a drug that should raise a red flag when a similar drug within the same class is being prescribed. Amoxicillin would never be prescribed if a penicillin allergy was documented. We also ought to be warned about the PCN allergy when prescribing a cephalosporin, e.g. cephalexin. This alert needs to be fired even if we know that, statistically, most reported PCN allergies are not true allergies and, even if they are, it is rarely anaphylaxis, making the 10% crossover risk of a cephalosporin allergy pretty near negligible. But the physician needs to be forewarned.
We gave serious thought to parsing out the alerts based on the categorization re: allergy, intolerance, side effect, etc. and realized a few things: first, most users do not enter the correct category due to misunderstanding, misapprehension or simple omission; second, there are certain side effects and idiosyncratic reactions that should give us pause to prescribe a different medication from the same class as the inciting agent, e.g. Stevens-Johnson syndrome, dystonic reaction, acute bronchospasm; and third, most practically speaking, our nursing colleagues are not comfortable being asked to interpret a patient’s allergy report so as to discriminate between a true vs. a benign event.
We have made all efforts to keep the alerts to a reasonable baseline and have chosen, for example, to avoid posting alerts for medications that might interact with certain foods, e.g. grapefruit juice. But the computer cannot be set to discriminate one class from another as to whether a same-class alert should or should not be fired. In other words, we can’t turn off drug class alerts for narcotics and not for all other classes or vice vesa. Furthermore, if we did and there was a true allergy to warrant pan-class alerting, there would be no way to prevent the potentially harmful prescription.
The question of why the alert fires time and again if the alert had been overridden in the past is simply a matter of safety. It's the "are you sure?" double-check given the enormous incidence of medication-related morbidity and mortality in the US. I agree that this seems an absolute nuisance and I would at least like to see the last reason selected for the override, e.g. MD will monitor, stick. I will bring this up with the system architect sometime soon.
We are continuing to refine and fine tune medication management as much as possible, directing energies towards drug-drug interactions and dose-range checking . Determining the “bandwidth” for alerting thresholds is a complex process and, while practice efficiency is crucial, patient safety is paramount. It is conceivable that the narcotics class could be further divided such that synthetics and non-synthetic agents would not cause cross-class alerting. But we are not there yet.
No comments:
Post a Comment