Facebook “Likes” Accurately Predict Hospital Mortality Rates

FOR EVERY 93 LIKES, MORTALITY DROPS 1%.

0 Comments

inShare

It can be tough to pick a hospital. And while it’s tempting to just use some online database--to Yelp it, so to speak--every problem with online ratings systems is only exacerbated by the scale of human mortality. If I can’t trust Yelpers’ opinion on what’s authentic Szechuan cuisine, how can I possibly depend on their opinions of coronary artery bypass surgery?

A new study published in The American Journal of Medical Quality points to an unlikely solution to finding a good hospital: Facebook. After analyzing the 30-day mortality rates across 40 New York hospitals and cross-referencing their Facebook page likes, they found a strong correlation between more likes and lower mortality rates. From the study:

The findings suggest that the number of Facebook “Likes” is associated with hospital quality as measured by the 30-day mortality rate. Based on this data set, a 1 percentage point decrease in 30-day mortality corresponds with almost 93 more Facebook “Likes.” This is probably because Facebook “Likes” tap into levels of patient satisfaction, as the positive relationship between “Likes” and the patient recommendation variable suggest. In other words, people are less apt to “Like” a hospital if that hospital has higher 30-day mortality rates and more prone to “Like” a hospital if they would recommend it. This somewhat intuitive finding suggests that Facebook offers an additional resource, beyond surveys, to gauge the attitudes of patient populations.

Incredibly, Facebook likes were found to be a more accurate predictor of mortality rates than traditional patient surveys. This fact leads the study’s authors to go so far as to conclude that “the number of ‘Likes’ on a hospital’s Facebook page can be used as a proxy for patient satisfaction and an indicator of hospital quality.” It just goes to show how big data is poised to rock the world of health care. Well, that, and the next time you randomly like something on Facebook, it could actually save someone’s life. Sort of. (Not really.)

read the study here