If you mix the two drugs? Ask Dr. Google

22:29
If you mix the two drugs? Ask Dr. Google -

A risky mix? The exploration of search engine terms of data can reveal clues about what drugs combinations have side effects.

ParentingPatch / Creative Commons

Analysis of queries made to Google, Bing and other search engines can reveal potentially dangerous consequences of mixing requirements before they are known to the Food and Drug administration (FDA), according to a new study. This data mining could even expose the medical risks that slip through undetected clinical trials.

Pharmaceuticals often have side effects that go unnoticed until they are already available to the public. This is especially true of the side effects that occur when two drugs interact, largely because drug tests attempt to identify the effects of a drug at a time. Doctors have few ways to drive these hidden risks, such as reports to the FDA of doctors, nurses and patients. A study in 2011, the data extracted from these reports and FDA discovered a hidden drug interactions: When taken together, the antidepressant paroxetine and pravastatin cholesterol suppressor because hyperglycemia or high blood sugar. After checking that research experiments, the researchers behind the study asks what other sources of information have been left unexploited.

Enter the search engines. Just as Google Flu Trends reveals influenza outbreaks by tracking search terms related to influenza, search queries on combinations of drugs and possible side effects say, "Paroxetine", "pravastatin" and "hyperglycemia" -might enable researchers to identify unforeseen drawbacks to drugs, said researcher Nigam Shah bioinformatics from Stanford University in Palo Alto, California. "If a lot of people are concerned about a symptom, which in itself is valuable information."

Although many bad drug reactions are never reported to doctors, people talk about what bothers them all the time on an occasional basis to friends or online, says the biologist Nicholas Tatonetti calculation of Columbia University, who was also involved in the study. "They do not really know," he said. "They are just reports of their symptoms, which is just a normal thing that humans love to do."

Thus, researchers have turned to Microsoft, which gets permission many Internet Explorer users to collect Google, Bing and Yahoo! search queries for research. Microsoft has provided a database containing 82 million queries search engines from 6 million unique users as of 2010. The research team looked for users interviewed "paroxetine" and "pravastatin" or each drug alone, to determine if the same users also sought "hyperglycemia" or other terms describing hyperglycemic symptoms. for example, they may have looked "dry mouth" or the technical term "xerostomia."

Before anyone knew the side effects of the drug combination, one in 10 people looking for both drugs also looked for terms related to hyperglycemia, or about twice what people looking paroxetine or pravastatin alone, reports the team today in the Journal of the American Medical Informatics Association . The researchers then looked for another 62 pairs of drugs, half known to cause high blood sugar and half unknown. They found that the data mining procedure correctly predicted whether the combo drug did or did not cause hyperglycemia about 81% of the time.

The results add yet another resource for scientists to find clues about the risks of drugs, Shah said, and it can be followed in real time. Hope, he notes, is that companies search engine could analyze data and send the results to the FDA, which would then track down information and alert doctors to potential new risks. "This is information we have access. As a society, we're sitting on it," he said. "We could use it to help the FDA, which is currently based on sources reported."

The new study is "exciting" because the field is so new, said researcher Hojjat Salmasian Biomedical Informatics at Columbia University, who has not participated. "We do not know how much it has potential," he said. "But based on the results that have been published so far in this study and other similar studies, it seems that this is a very important area explore. "

Previous
Next Post »
0 Komentar