Using Social Media to Change Health Behavior/Service Utilization

What is the headline saying or claiming?
Link to article: Using Social Media for Public Health, Patient Behavior Change

Social media campaigns can be useful for sparking conversation about public health issues and driving patient behavior change and education

What is the research article saying?

Social media can be an effective tool for disseminating public health messages and support better patient access to mental healthcare… (example given is the “Bell Let’s Talk” campaign which introduced Twitter as the main platform in 2012)

More awareness about mental health treatment and reducing the stigma often associated with mental health treatment access may help encourage some patients to utilize treatments when they otherwise would not have done so…

There were temporal increases in care access during the Bell Let’s Talk Twitter campaigns.

What are the implications of this headline?
Social media campaigns can drive behavior change when it comes to health issues

Are there similar and/or opposing headlines from other outlets?
N/A. Social media campaigns are often used to raise awareness about an issue.

What are the data sources?
Research study that assessed the Bell Let’s Talk Campaign to see if the social media campaign impacted youth outpatient mental health services in the province of Ontario, Canada. Researchers studied the impacts of the campaign on rates of monthly outpatient mental health visits between 2006 and 2015: Youth Mental Health Services Utilization Rates After a Large-Scale Social Media Campaign: Population-Based Interrupted Time-Series Analysis

What is the study design?
The researchers used a cross sectional time series analysis of youth that accessed outpatient mental health services during the time period mentioned previously.

Additional data source that I referred to:
Interrupted time series regression for the evaluation of public health interventions: a tutorial

Interrupted time series (ITS) analysis is a valuable study design for evaluating the effectiveness of population-level health interventions that have been implemented at a clearly defined point in time…

It is particularly suited to interventions introduced at a population level over a clearly define time period and that target population-level health outcomes…

A time series is a continuous series of observations on a population, taken repeatedly (normally at equal intervals) over time. In an ITS study, a time series of a particular outcome of interest is used to establish an underlying trend, which is ‘interrupted’ by an intervention at a known point in time…

There is an expected or counterfactual trend/scenario established for comparison purposes (includes data collected prior to the intervention)…

A priori information/key information needed for study design (based on the article above):

Yes. The headline supports claims made by and summarizes the key points of the article.

Appropriate design?

1. Clear differentiation between pre-intervention and post-intervention periods

2. Outcome should be short-term, with the possibility of changing quickly after an intervention has been implemented

Appropriate data?

1. There are no fixed limits regarding data points (amount needed); inspect pre-intervention data points using descriptive statistics (visualize)

2. Routine data (usually administrative), gathered over a long period of time/long time series

3. Understand potential bias in results related to changes in recording or data collection methods

Where is change expected?

1. Gradient of the Trend

2. Change in the Level

3. Both

When should change occur?

1. Immediately after

2. After some lag

What should be taken into account?

1. Time-varying confounders

Control for seasonality (leads to autocorrelation and over-dispersion)

Adjust for residual autocorrelation using ARIMA (autoregressive integrated moving average modeling)

Control for infectious diseases that are prone to outbreaks (use sensitivity analysis)

Are these data sources credible when applied to the article?
Yes. The source is credible since the study design was followed/implemented as intended.

What are the data sources saying?

There was an increase between 2006 and 2015 in the rates (monthly mental health visit rates) of outpatient mental health (primary healthcare and psychiatric visits) use by youth aged 10 to 24 years old in the province of Ontario for males and females. The 2012 Bell Let’s Talk campaign was temporarily (temporally?) associated with increases in the trends of outpatient mental health visits, especially within the adolescent female cohort. Although no discernible difference in the immediate change in the rate of mental health visits (magnitude/level change) was observed among the adolescent groups, young adults exhibited a slight drop in most outpatient mental health visits, followed by a moderate increase or plateauing of rates…

Results broken down:

1. Over 10-year period (2 time points, 2006 & 2015)

Adolescents (10-17) saw an increase in the monthly mental health visit rate for primary care and psychiatric services.
Young Adults (18-24) saw an increase in monthly mental health visit rates for primary care and psychiatric services.

2. Immediate change associated with intervention:

Adolescents (10-17)

There was no discernible difference in the immediate change in the rate of mental health visits observed that could be attributed to the campaign.

Young Adults (18-24)

There was an immediate drop in rates of mental health visits after the campaign- this group experienced a decrease and plateau in the slope of all psychiatric service visits after 2012.

3. Both female age cohorts saw increases in accessing primary health care for mental health services after the 2012 intervention.

Are the data sources being interpreted correctly?
The article makes the claim that “each year during which the campaign ran, mental healthcare access saw a spike amongst adolescent and young adult patients.” However, since only two data points were compared (the 2006 and 2015 data points) the statement about seeing a spike “each year” does not appear to be accurate. This statement also appears to contradict the one made right after it: “Following the month-long campaigns, visit rates decreased or plateaued, researchers found.” The article also advocates for a more targeted campaign with specific calls to action, to see if this may lead to more health behavior change.

Overall, the researchers discuss how the, “lack of substantive step change in health care utilization from normal levels is not surprising,” since the goal of the campaign was to increase awareness of mental health and stigma. At most, the data from this study may suggest that the campaign contributed to a gradual rather that immediate change in behavior as it relates to youth in Ontario, Canada accessing mental health services. The researchers call for further exploration of the increase in female mental health service utilization over the 10-year period (possibly with an “emphasis on gender and sex within health sciences research”) and further research on “more precise modeling techniques to measure the effect of social media on population and public health.”

Are limitations provided?
The research study provides the following limitations:

1. Administrative data was used, so illness severity could not be measured. The study also could not analyze individual presentations/usage of mental health services.

2. Emergency department visits for mental health services were not included in the study (this was so that the study could focus on planned mental health activities that could possibly be attributed to the campaign).

3. Specific sub-populations could not be studied to see how the campaign may have impacted homogeneous populations/smaller groups.

4. Although there was a temporal change associated with the campaign, other factors could have contributed to this change.

5. The cumulative effect of the campaign on people over time was not explored.

I would add that caution should be taken when trying to generalize the results from this ecological study of youth in Ontario, Canada to other populations.


What does this mean for the general public and public health professionals?
Although mental health awareness can be increased using social media outlets and campaigns, more research needs to be done to see if these campaigns can also influence behavior change that leads to an increase in the utilization of mental health services on a population-level (or in specific sub-populations).

The impact of social media campaigns on population health should be evaluated using an appropriate study design.


Scrutinizer Product




Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s