How useful is it to know a client is high risk for bad outcomes?
The Myndplan prototype project is well underway. So far we have the Scans working seamlessly and are preparing to launch our innovative symptom monitoring component of the app. In addition to this systematic roll-out of features, we’ve been busy responding to requests from clients and clinicians using the app. Some of this feedback has concerned the usual bugs and fixes that new software always requires. But in other cases the questions and suggestions we get revolve around new features that we may already be working on, but haven’t launched. Recently we have completed preliminary work on an example of this sort of challenge, which relates to profiles with a pattern that includes an unusually high number of elevated scale scores.
The Symptom Checklist & the “Ceiling Profile”
To understand the feature I’m about to describe, it will help to review some of the history of our scan development process. For many years as part of clinicians’ usual assessment and intake process, we provided clients with a well-known measure of acute distress, the Symptom Checklist-90-Revised (1), along with the Myndplan scans. Because the scans were under development, we needed additional tools like the SCL-90-R to help ensure that everything was working as intended. This meant that each time a clinician looked at scan scores for a client, they also had a Symptom Checklist profile for comparison. This kind of validity check is pretty common when questionnaires are being built, and was a critical part of the scan development.
Over the years, clinicians began to notice one particular pattern with the SCL-90-R that we knew was important, but could not decipher much further. It involved exceptionally high scores on all but a few of the scales making up the score profile, which for the SCL-90-R was charted on a single piece of paper in the form of a line graph connecting the various scale scores across the page. The shape of a typical SCL-90-R profile was of a jagged line, with a variety of peaks and troughs representing high and low scores. However, in some cases, scores would line up to form a distinctively flat line across the upper margin of the page – the ceiling of the profile form. At this level of distress no single score can be distinguished from another.
What did the SCL-90-R manual say about this “ceiling profile”? Only that the presence of extreme scores on several scales indicated a client’s overall level of psychological distress, or the extent to which they were “augmenting” their scores. The SCL-90-R manual describes the latter as either “dramatizing” or “faking bad”.
Many clinicians using the SCL-90-R had noticed that there seemed to be more than this going on with profiles that displayed a distinct “ceiling”. Furthermore, they were not seeing such a clear pattern on the profile of Myndplan scan scores. While it was true that the clients with these ceiling profiles on the SCL-90-R were frequently in some sort of crisis, there were often other features that suggested the solution was more complex, such as vague but disabling somatic complaints, overuse of opioid medications, and issues with employment.
Beyond the Ceiling
Recently, I brought this problem to Dr. Rod Martin, Myndplan’s research director and a crucial advisor on the prototype project. Dr. Martin has a career of academic teaching and research in clinical psychology under his belt, and along with all the accolades, articles, and books, he has accumulated a treasure trove of knowledge when it comes to data analysis. He was intrigued, and suggested that we have a closer look. First, he developed a small program to extract all the ceiling profiles we could find in our database. We then identified the strongest examples, and created a simple rule set that defined them.
In the next step, we performed some analytics using the Myndplan results for the clients who had classic versions of the SCL-90-R ceiling profile. It took a while to follow this process to its conclusion, but in the end we were able to confirm a solution using the Scans that predicts the “ceiling” effect with surprising accuracy. This means that we can use the scans to predict which clients will obtain a ceiling profile, without them having to actually complete the SCL-90-R. More importantly, the solution tells us a lot more about this group of people than the SCL-90-R ever did.
Into the Abyss
This is where our discussion gets to the crux of what a psychologist like myself – or any therapist for that matter – desires most from a good assessment tool: accurate predictions that help identify more clearly not just who the client is, but what some of their major issues might be, and how they will influence the therapeutic relationship and its goals over the course of therapy. In the case of the present investigation, we have discovered that some very important characteristics are associated with many ceiling effect clients:
> Co-morbid Problems: In terms of the scales making up the various scans, several are involved, including suspicion, intrusive thoughts, drug use, depression, and what we call Factor I. Factor I is a combination of scales that describe people who are preoccupied with their health, struggling with their sleep, and extremely fatigued. It is frequently associated with chronic pain disorders and depression. We have tentatively labelled it “Somatic”, given that it contains symptoms that people with a lot of medical problems often complain about.
> No Work: A close look at the problems that “ceiling profile” clients experience helps to fill out the picture. Most have symptoms that interfere with their ability to function – so much so that their job is at stake, if they are even able to work. The rest don’t have any work, and are desperate to secure some means of keeping food on the table and a roof over their head.
> Addictions are Common: Two thirds have a history of drug problems, or current issues with drug use, most notably prescription opioids, but also benzodiazepines and alcohol.
> Soma, Soma, Soma: Sixty percent would meet the criteria for somatic symptom disorder or illness anxiety disorder in the new DSM5, or the older somatization disorder that continues to exist in ICD10. This is a group of people with a long history of medical complaints, many of which have proven frustrating to treat. The distress caused by these problems has often been severe in comparison to other clients with the same symptoms and has resulted in frequent admissions to the local emergency department and visits to the doctor.
> Secondary Gain: Close to 25% are probably malingering, though it is not difficult to understand why. For example, most do not have jobs and rely on very meagre disability benefit payments of one sort or another to survive. This segment of the group are desperately in need of money to survive and deathly fearful that their benefits will be cut off or reduced if they don’t take some sort of action.
> Lots of Legal Problems: Another 20% are in trouble with the law, usually around domestic violence that erupted amongst all the stress.
> Elevated Suicide Risk: Suicide is a prominent risk for more than 10% of this group, and the odds of someone actually dying are 150x greater than would be expected with non-clients. In other words, whether the distress is real or imagined, the consequences for people caught under circumstances that promote this pattern of responding can be dire, if not deadly. Essentially, this is a pattern that seems to signal a personality type that gets into frequent difficulties that have very real, negative consequences.
As a clinician, this is not a hard group to recognize. To say that it is made up of people who are struggling and in crisis is an understatement of epic proportions. For the most part, they present with an exceptionally difficult set of problems to work with. They experience frequent setbacks and their prognosis for recovery is exceptionally poor. For a sizeable minority, poverty is a significant barrier to progress in treatment. So is drug abuse. Unless these problems are addressed their symptoms will not improve. Personal crises and somatic complaints often dominate sessions and derail efforts to apply more helpful and effective treatment methods.
These are all things that a clinician wants to know before starting counselling. They suggest many issues that need to be explored early in therapy as the treatment relationship is established. They confirm that the client is feeling desperate, and as with all of us, desperate times often call for desperate measures. This is a problem set that genuinely puts a person at risk of harm, and may put others at risk too. After all, when someone is in crisis they are more likely to make bad decisions, and bad decisions often affect everyone.
In the end, we’ve identified some important characteristics that often go with a common pattern in extreme scores that may be helpful for both clinicians and clients to know. This provides a tool that flags a very specific set of problems indicating that a client may be stuck “between a rock and a hard place”. This often involves the kind of challenges that blur the line between psychology and sociology, or more precisely psychiatric illness and the desperation that goes with disadvantage and crisis. It turns out this is not so much a ceiling of distress as it is a sign of how circumstances have taken a person to “rock bottom”, and reflects a facet of some people’s desperate circumstances that cannot be labeled or fit neatly into a single diagnosis. Just the sort of situation any clinician wants to know before they embark on a course of therapy with a client, as it can have a huge impact on the options for change.
Derogatis, L.R. (1975). The Symptom Checklist 90 Revised: Administration Scoring and Procedures Manual I. Baltimore: Clinical Psychometric Press.
Leave a Reply