Data-Driven Therapy: The Future of Mental Health Care
Technology changes our lives. As innovators produce new applications for existing technology, they continue to reshape daily norms. Not long ago, students and researchers were directed to libraries to search texts and other published materials, such as newspapers and scientific journals. Now, thanks to information technologies, students can sit in a classroom and access research sources through smartphones, tablets, and computers.
Technology has transformed healthcare. It allows the storage of an almost endless amount of electronic health records and medical data for the efficient tracking of treatment effectiveness and coordinated care for many health conditions. Applications have helped businesses improve customer service by enhancing the analytics of their experiences and providing real-time data that can be used to help resolve customer issues. This same focus on patient experience is found in healthcare, too. Electronic health resources allow people to be more aware, to be better educated, and to experience less shame about mental health and substance use struggles.
While it might not immediately jump to mind, even highly personal services such as therapy for individuals and families have significantly benefited from technology. Data is revolutionizing the way people seek and receive mental healthcare. Using a data-driven approach enables patients to benefit from more effective therapy and leads individuals to experience higher success rates. This is one of the reasons why individuals considering therapy may want to seek data-driven therapy.
There are several reasons why individuals choose to pursue therapy. They may have experienced a traumatic event or a significant loss and need professional help to process and cope with the impact of the trauma or grief. Another individual may seek guidance to help address factors related to their sexuality, gender, sexual orientation, or cultural identity. Others may be struggling with depression, bipolar disorder, an anxiety disorder, or adjusting to life in a relationship, a new city, a new job, or as a new parent.
The varied reasons that people seek therapy make it impossible for psychologists to develop a one-size-fits-all behavioral health care approach that is effective with all patients. Nor would that even be a goal, given the complexity and diversity of people’s life stories, experiences, and predispositions. A virtually limitless range of dynamics makes each individual — and their situation — distinct. One person may be dealing with trauma from an abusive relationship that is compounded by active substance abuse and years of mistreatment by a parent. Another may be dealing with obsessive-compulsive disorder, as well as a history of neglect and feeling like an outsider. A person suffering a loss due to death may struggle because the grief intensifies unresolved feelings they have about a turbulent and dysfunctional relationship.
Everyone’s mental health is a highly complex web of struggles and elements of extraordinary resilience. Feelings such as guilt, doubt, fear, and anxiety can be developed and triggered by a wide range of circumstances.
This means that, when it comes to the treatment of substance abuse and mental health, psychotherapists should continually seek to expand what they know and how they unpack significant experiences and relevant historical information with their patients.
Therapy is a highly effective form of treatment for a range of mental health, behavioral health, and substance use disorders, yet research suggests that some therapists are more effective than others. Furthermore, it seems that the relapse rates of various approaches to therapy differ. For example, it’s common for more than 50 percent of those presenting to Cognitive Behavioral Therapy (CBT) for major depression to struggle with similar themes within a year of completing treatment.
Part of the reason for variations in research findings is that limited real-world emphasis is placed on on-going performance measures and tracking of patients’ progress and their reactions and feelings to the therapy process. That is, if psychotherapists aren’t assessing people’s change throughout and after therapy, then important information is missing. As a profession, without this kind of data and feedback, we’re unable to fully assess and understand why some patients accelerate towards their treatment goals while others flounder and drop out of therapy too soon. Relatedly, if therapists aren’t able to follow-up and check-in with patients who have ended therapy, then there is a striking absence of data about the long-term effects of therapy and therapists.
As the complexity of people’s lives increases and technology breaks further barriers, therapists are presented with an inspiring opportunity. We’re even better able to adapt our approach to treatment to account for an increasing number of variables that can be significant when working with a patient. It’s far from simple, and it’s almost always easier to leave things as they are now. However, to be increasingly effective, forward-looking, and maximally responsive to a variety of struggles and mental health concerns, therapists can employ novel uses of technology to develop personalized therapy, thus improving quality of care.
At its best, therapy recognizes the needs and unique attributes of the individual instead of attempting to fit patients into an artificially created one-size-fits-all framework. When developing scientific theories, making sense of what factors contribute to or detract from mental health is a group-level exercise. Well-planned studies with sophisticated statistical analyses reveal much about what can help decrease anxiety, or how to deal with symptoms of depression more effectively. However, findings like this apply to an average person of the group of people who are being studied. While exceedingly useful in many ways, there is no such thing as an average person in the real world, especially when someone is sitting in a therapist’s office. So when moving from the psychology lab to the therapy room, skilled therapists learn to “hold theories lightly” and recognize that the best therapy is the therapy that is tailored to each individual.
Similar to how personalized medicine is a powerful new frontier in providing essential health benefits to patients, new uses of technology health systems and increasing computational power means that personalized therapy is likely to be the future norm — and we certainly believe it should be. For instance, an emerging approach to therapy involves consideration of patient needs as well as personal preferences that are incorporated into the process of selecting a therapist.
Future versions of data-driven therapy combine the idea of personalized therapy with understanding fluctuations in how well patients report they are feeling and their impressions of the therapy process, allowing for therapeutic adjustments that would have otherwise not been made in the absence of data. An excellent example of this is Feedback-Informed Treatment (FIT), which allows therapists to invite their patients to be even more active participants in their mental health treatments.
Feedback-Informed Treatment (FIT)
One of the most basic applications of personalized therapy is to evaluate information based on an individual’s sessions to identify patterns of behavior and establish personalized care plans. FIT involves gathering and utilizing the patient’s perception of the relationship with their therapist, along with reports and health screenings of their current week-by-week wellbeing, to check in with patients and evaluate the progress of the specific type of mental health treatment they are receiving. Frequent and consistent collaboration with patients about their experience of therapy and the therapeutic alliance can have a significant impact on what leads to positive, lasting change in therapy. The process may also involve reviewing data related to other patients who have a similar type and/or level of symptoms or concerns for which they sought therapy.
With FIT, therapists and primary care providers have access to statistical data related to the patterns of their patient’s responses and to the mental health care provided. This can help to identify potential problems that patients may be hesitant to address directly, such as when a patient is more likely to end therapy prematurely because they feel that their treatment is off-track. It can also highlight the most effective course of treatment for that particular individual to help the therapist adjust their approach to treatment in real-time to determine the best strategy for each patient and improve the likelihood of success and improved emotional health. FIT utilizes a plan based on both the expertise of trained therapists and data-driven technology. It provides regular and consistent patient feedback that creates discussions in session to confirm or disprove a therapist’s clinical hypotheses and instincts and ultimately help them make informed decisions about their patient’s treatment.
Benefits of FIT
Enhancing traditional therapy with FIT has some specific advantages that can improve mental health treatment outcomes. It encourages more transparency around feedback and the exploration of the therapeutic alliance. The therapeutic alliance— how well someone clicks with their therapist — is particularly vital to ensuring positive outcomes in therapy.
Specifically, we believe the significant benefits of FIT are it:
- provides multiple sources of data,
- allows for more efficient assessments, and
- can help therapists create more personalized and highly effective treatment plans that result in improved mental health.
Multiple Input Sources of Data
Data-driven therapy allows therapists to receive health information from more than one source. For instance, care providers can have their patients complete questionnaires that address current mental health conditions and symptoms, treatment goals and progress, and how well they think therapy is going. This can be done on tablets or computers before and/or following each of their therapy sessions. The data gathered can then be reviewed and discussed together with the therapist in session, which would likely better inform the treatment and hopefully foster more openness about progress and patient feedback.
This type of information may help therapists identify potential issues before they even meet with a patient. For example, a therapist may be able to identify inherent relationship struggles in a patient’s history just from data concerning a divorce. An individual who indicates several moves within a relatively short period may struggle with commitment or attachment that keeps them from putting down roots.
Post-session surveys can also be used to evaluate the effectiveness of the therapist’s approach while underscoring potential indicators that the patient may prematurely end therapy, without discussing what type of therapy or what kind of therapist might be a better match for them.
In typical psychotherapy, a therapist relies on their training, clinical experience, and finely-tuned empathic skills to identify patterns and form hypotheses related to each individual patient. Much of the time, this results in positive outcomes. However, any given therapist may be limited by the information they pick up from sessions and use it to provide feedback and help the patient identify patterns. FIT systems are designed to supplement this process by collecting feedback and data directly from patients so that therapists have more information. This means that the FIT approach enables therapists to have additional, concrete data they can use, in conjunction with traditional therapy methods, to provide individual-specific plans of action.
Individualized and Highly Effective Treatment Plans
The FIT approach is designed to consider specific information from individual patients so that their treatment can be customized to meet their needs. This eliminates many of the limitations of the one-size-fits-all approach to therapy.
FIT guides a treatment approach for a patient based on trends in their data. This means that the therapist will be better informed to approach their work with each individual patient to increase the likelihood of achieving a positive treatment outcome. As mentioned, therapists can also use ongoing data collection from each session to determine how well their therapeutic approach with a given patient is working and use that information to customize their treatment further.
Deep learning is a computational process that is designed to mimic the way our brain pathways work. For example, let’s suppose you have $100 to spend on personal items for the month, and you learn about a concert that costs $85 per ticket. To attend the show, you also need several things, such as a bus pass and food, but $15 will not be enough to cover those items. You want to go to this concert, so what do you do?
When individuals think through decisions like this, they may factor in a wide range of data to make their choice. Two different people with a similar opportunity to buy that ticket, and with the same financial resources, may consider distinct variables.
One might ultimately decide to pass on the ticket and spend their money on their other monthly needs. Another may rationalize how much money they can save or search for additional resources for funding so that they can afford the ticket. Whatever their circumstances and factors, each individual draws on different information to help them reach a decision.
Deep learning artificially replicates this thought process. As reported by Forbes, “Deep learning is a subset of machine learning where artificial neural networks, algorithms inspired by the human brain, learn from large amounts of data. Similarly to how we learn from experience, the deep learning algorithm would perform a task repeatedly, each time tweaking it a little to improve the outcome.”
Deep Learning Applied to Psychotherapy
A deep learning program for therapy can be designed to review medical records and other points of mental health data collected over time to determine distinct patterns of symptoms and responses to therapy. Data like this would be analyzed to assess treatment progress, the quality of the therapeutic alliance between mental health providers and patients, and whether treatment goals are being addressed. The deep learning program can be used to make predictions that enhance the likelihood of positive outcomes in therapy and allow a therapist to develop a highly personalized form of therapy.
Why We Believe Data-Driven Therapy is Here to Stay
Data-driven therapy techniques have been proven to reinforce more openness between patient and therapist, as feedback from each is welcomed and encouraged. Recent research has demonstrated that monitoring ongoing client progress and providing mental health professionals with real-time feedback resulted in behavioral health benefits and improved treatment outcomes by nearly 65 percent. Data-driven therapy practices like FIT are successful in part because patients receive more customized treatment approaches that work for them as individuals. The increased likelihood that therapy will be effective helps motivate individuals to commit to the process. The success of FIT is already changing views towards therapy and can be expected to continue to attract patients who might otherwise be hesitant to participate in therapy as well as contribute to the de-stigmatization of therapy and mental illness. From our vantage point, new technology and increasing openness to using technology to enhance therapy are just beginning to show its true potential for transforming mental healthcare.