MyDrugAdvice

Drugs can be more harmful than beneficial if patient specific precautions be ignored. Therefore many applications have been developed for EHR/CDS that notify clinicians for different patient related alerts. However “Alert Fatigue” is one of the major problems with current EMR/CDS systems; one recent study showed that safety alerts being overridden by clinicians 49-96% of the time (http://www.informatics-review.com/wiki/index.php/Alert_Fatigue). This means high probability of prescribing inappropriate drug for patient. Our goal is to reduce the effect of alert fatigue by shifting some checks to patient side. MyDrugAdvice is an application that reduces patient risks caused by drug consumption by providing relevant advices based on the patients health records. Patients can search the name of the prescribed drug in MyDrugAdvice and it returns the related alerts to the patient considering his/her health history and the queried drug. For example, a person having the record of “Hypertension” who is prescribed “Vazotab tablets” for runny nose can be notified that “Vazotab” is harmful for patients with high blood pressure and he/she can notify the clinician accordingly. For each queried drug, MyDrugAdvice searches the list of all available advices for the drug and checks whether any of them is applicable to this patient based on the data in the patient health records. Drug related advices are available online and can be found in websites (such as MedlinePlus, Drugs.com and etc). Health conditions (like allergies and diseases) were extracted from online drug contents using natural language processing techniques. These extracted health conditions are matched against the data in patient records. MeSH ontology is used to map the same medical concept which may be mentioned with different vocabularies. MyDrugAdvice can change the way we search Internet and bring the most relevant contents to the query and the user health records.