Public sector, enterprises need mental health analytics to improve decision-making
The intersection between mental health and technology has gotten a lot closer in recent times. With the pandemic altering lifestyles and creating new behavior patterns, comprehending mental health is now becoming a prerogative not just for enterprises trying to understand the pressures on their employees, but governments too.
Yet, the general understanding of mental health is normally associated with the negative behaviors and outcomes that poor mental hygiene can lead to. And despite the high amount of behavioral data at hand, using that data to improve society is still not fully realized.
Today, data comes from almost anywhere and everywhere. Be it mobile devices or sensors on buildings, wearables, the data collected is used to discern patterns in understanding the behavior of consumers, products, health, and more. The impact of data can be so powerful as to change perceptions and influence big decisions — cybercriminals know how valuable such insights have become, which is why they’re also after it.
In understanding consumers, data is recorded from their past behavior, preferences, and activities. But what about in understanding their mental health? Can data be used to develop mental health analytic models that can help understand people better? Can data be used for the study of employees’ mental health or even understanding what goes through their minds by the actions they take?
Understanding the data
According to Josh Morgan, the National Director of Behavioural Health and Whole Person Care at SAS, the story of data and mental health is all about the need for its treatment, as well as the impact its services can bring. It also about proving how mental health services can be provided for, and improved via data analytics.
Most data available mostly focuses on the negative impacts of mental health. For example, hospitalization, suicide, crime, homelessness, and such. This is the only part of the data that is always being served.
“We need to understand what mental health is. 50% of the population may be diagnosed with mental health in the future, but we need to combat the stigma and discrimination associated with that. Mental health should not be defined by that. That’s the core values of what data and mental health can do,” said Morgan.
Echoing Morgan’s views is Antonio De Castro, Senior Industry Consultant for Global Health & Life Sciences Practice, SAS. He explains that we need to look at the data in multiple ways as there is a deeper story there. For example, the pandemic caused a big impact on mental health. But not everyone would have the same mental health impact. And this is where the data is vital in measuring it properly.
Mental health analytics is all about bringing behavioral health data together and being able to tell the impact on services, especially in understanding the impact of mental health treatment. Enterprises, public sectors understand that mental health goes beyond negativity only.
“The quality of data has changed. It is different in how we capture it. There are vast volumes of data available in different channels. Understanding the new levels data, what precautions are needed, the limits and benefits of differentiation are key,” said De Castro.
Implementing mental health analytics
In the public sector, the data can be used to understand how funds for community development and such, as well as social problems, can be dealt with. For example, government agencies or NGOs can rely on mental health data to design programs or understand their community better in a certain demographic. The advantage of this is that they only need to leverage existing data and analyze it.
For enterprises, mental health is becoming increasingly important, especially with the pandemic changing the way people work. Companies want to understand their employees and find ways they can improve their working lifestyle. And the best way of doing this is by understanding their mind and behavior.
Some of the use cases on how this analysis is improving society can be seen being implemented in the US, Canada, and Europe. In Canada, the Centre for Addiction and Mental Health (CAMH) is working to remove the stigma of mental illness and addiction while providing world-class care to those in need. As Canada’s largest mental health teaching hospital, the Toronto-based institution treats more than 34,000 patients each year.
The team there used SAS to optimize care for alternate level of care (ALC) patients – people who occupy acute care hospital beds but no longer require hospital care. By predicting which patients are ALC upon admission, CAMH can ensure these patients are seamlessly moved into the right care setting at the appropriate time. Using social determinant data captured at admission to perform the analysis, they tested several predictive models including univariate and multivariate analysis. In the end, they landed on a predictive model that was 80% accurate, a major step forward in streamlining treatment for ALC patients and optimizing bed space.
To reduce suicide risk among Canadian youth, Canada Health Infoway collected 2.3 million tweets and used text mining software to identify 1.1 million of them as likely to have been authored by 13- to 17-year-olds in Canada by building a machine learning model to predict age, based on the open-source PAN author profiling dataset. Their analysis made use of natural language processing, predictive modeling, text mining, and data visualization. The team looked at the percentage of people in the group who were talking about depression or suicide, and what they were talking about.
Mental health and privacy
With access to data, there are also concerns on the privacy implications that can arise from it. Interestingly, De Castro pointed out that when they use data from Twitter for example, they do not have access to such personal information. In fact, they need to understand the patterns and from the characters used in Tweets to understand user behavior.
“The goal is not about identifying individuals. It’s about preventing suicide from a public health perspective. That’s a piece sometimes we forget about in the use of data including for mental health. For this case, what geographic areas needed to be targeted for suicide prevention as better use of resources. It is a great use of public funds and upholding privacy,” said Morgan.
Another use case is in San Bernardino County. Using data management and advanced analytics, California’s San Bernardino County Department of Behavioral Health improves care and addresses misconceptions. This is done via SAS Data Management paired with sophisticated SAS Analytics to improve the quality of data and find insight within all this information. Now, the department can find better ways to connect people to the right type of behavioral health services – and enable them to live healthier, more satisfying lives.
In Southern California, Riverside County engaged with SAS to build a whole person care solution based on SAS Data Preparation and SAS Visual Analytics on SAS Viya. The solution enables the county to integrate health and non-health data from its public hospital, behavioral health system, county jail, social services systems, and homelessness systems.
By connecting these databases, Riverside County can now see how individuals interact with different services across its health system. This is enhanced by entity resolution, a feature of the technology that enables the county to identify unique entities, even if a person’s name, address, or other personal identifiers don’t match up across different databases.
“I have seen more interest and procurement in behavioral health data this year. Things are changing. People recognize that we can’t not look at behavioral health data. This is the requirement to do good work, to do quality improvement, and to tell outcomes. You can’t not have it in 2021. Behavioral health, mental health, data, and SAS is a natural evolution of what we have done and continue to do, especially in using data for good to improve our communities,” concluded Morgan.
25 September 2023
21 September 2023
20 September 2023