Skip to main content

Leveraging Twitter for Insights on Patient Experience

 |  By  
   November 04, 2015

Hospital marketers can use Twitter and other social media tools in conjunction with HCAHPS and other surveys to paint a clearer picture of the organization's quality and patient experience performance.

If you've ever read the comment section on a YouTube video, you know just how brutally honest social media users can be. The anonymity that the internet affords often leads to people posting harsher opinions than they would say aloud in their everyday life, and countless online trolls have made a hobby of being downright cruel.

But researchers at Boston Children's Hospital have found the candid opinions expressed on social media—specifically on Twitter—may provide hospitals with valuable insight when it comes to measuring quality and patient experience.

The study, "Measuring patient-perceived quality of care in US hospitals using Twitter," published in the October issue of BMJ Quality & Safety, broke down 400,000 public tweets directed at the Twitter handles of almost 2,400 hospitals in the U.S. between 2012 and 2013.

Though a variety of methods, 34,735 patient experience-related tweets directed at 1,726 hospital-owned Twitter accounts, were tagged. Researchers determined the sentiment of those tweets and sorted the tweets into topical categories.

"As healthcare is becoming more patient-centric, and outcome- and value-driven, it is increasingly important that we listen to the patient's voice," says Jared Hawkins, MMSc, PhD, faculty in Boston Children's Hospital's Computational Health Informatics Program (CHIP) and director of informatics for Boston Children's Innovation Acceleration Program.

"Our research group has used many novel data streams for public health research as part of our exploration of the digital phenotype. It became apparent to us that patients are using Twitter to discuss their experience receiving healthcare. We decided to conduct this study to see if we could capture this discussion and whether the content could be useful."  

Correlation with HCAHPS
The Children's Hospital research team hoped that the Twitter data would expand on the findings of HCAHPS and other quality surveys.

"Traditional quality surveys, while useful, have some major limitations: low response rates (and selection biases in those that do respond), targeted questions, and significant time lag before the results are available to the public," Hawkins says. "Twitter is an incredibly rich resource for unsolicited, real-time feedback that is accessible to anyone who wishes to have their voice heard?making it a perfect companion to traditional surveys."

Ideally, the research team would like to correlate Twitter data to outcome metrics that relate to quality of care, but they're not quite there, yet. When researchers compared their data with outcomes data on the Hospital Compare website, they found a weak negative correlation between tweet sentiment and hospitals' 30-day readmission rates. However, they didn't find a relationship between tweet sentiments and HCAHPS experience data.

Of course, this could be because the sample size they were working with was relatively small. There was much less Twitter volume in 2012, and of the 2,400 hospital Twitter accounts researchers looked at, only 300 received more than 50 inbound tweets in a year. Three years later, those numbers have likely increased significantly. As a result, the Children's Hospital team believes measuring patient experience via social media will become commonplace in the future.

"Social media is already widely used as a method for consumers to voice their opinion, and often facilitates the quick resolution of complaints. For example, see how consumers have used this to communicate with airlines," says John Brownstein, PhD, chief innovation officer at Boston Children's Hospital and director of its Computational Epidemiology Group.

In the meantime, there are several steps hospitals can take to prepare to incorporate social media data into their quality reporting process.

Guidelines for Hospitals
"First, and we know many hospitals are doing this, it is important to have a dedicated Twitter account that is regularly monitored," Hawkings says. "Patient comments should quickly be responded to. While Twitter is often the first point of contact, the majority of the discussion will take place offline. As we have shown, the majority of tweets directed towards a hospital (~91%) do not discuss patient experience. Those tweets that are not about patient experience may still be valuable to the hospital, but will need to be filtered out for any patient experience analysis."

Once irrelevant tweets are filtered out, hospital marketers will be able to use Twitter and other social measures in conjunction with HCHAPS and other surveys to paint a clearer picture of the organization's quality and patient experience performance.

"Traditional and nontraditional data streams should be analyzed in concert, as they both have their own strengths and weaknesses," Brownstein says. "Interested stakeholders should strive for the most complete information, and social media provides valuable content that should not be ignored."


Marianne Aiello is a contributing writer at HealthLeaders Media.

Tagged Under:

Get the latest on healthcare leadership in your inbox.