Off-roading with Science: Disciplined Improvisation (Part 5)

Here's the last in the science-in-practice series. These are posts I've used to organize my thinking about the challenges we face doing evidence-based practice. To recap

lego jeep
  • It's often difficult to find the evidence you need to guide clinical decisions (see post 1 & post 2).
  • Without great evidence, we rely on clinical judgment. But that's problematic (post 2 & post 3). Biases outside of our awareness function like an automatic spotlight.  Unconsciously perception, attention, and problem solving are caught by a few elements right in front of us.

To remedy this, we need kind environments, work routines and tools that help to see and learn the relationship between clinical judgment, intervention, and outcome. That way true expertise in clinical intuition and judgment can develop.

To make a kind environment, start with progress monitoring and use of a standardized protocol for the client's main problem (if available). It's not that this is guaranteed to produce good outcomes if delivered with high fidelity-it's that human brains need guidance (post 4).

But what if there is no evidence-based protocol or the protocol doesn't produce the desired change? Then what's needed is disciplined improvisation. Despite it limitations (1), case formulation is our best tool to harness the scientific method for clinical work (2). The basic idea is to spell out what you are trying to change (the dependent variables) and what you will do to change it (the independent variable(s)). You do it systematically. You observe what happens (progress monitoring) and base decisions on the data.

Use A Target Hierarchy Informed By Science (dependent variables)

A target hierarchy gives you if-then guidelines about what-to-treat-when. It works like a manual spotlight so that the most important problems are addressed first (helping offset the ways unconscious bias might otherwise drag us off track).

Target hierarchies provide a process to organize session time. That way we think consistently and coherently, sort the relevant from irrelevant, and better manage the immense cognitive load of doing therapy.

They also help because when the target hierarchy is informed by the best science, the target hierarchy tells you the right content to target. For example, research shows that directly targeting suicidal behavior as a problem in itself gives better outcomes (rather than seeing it as a sign or symptom that will naturally resolve when the underlying disorder is treated) (3).

Targets in a hierarchy can be disorder specific. For example, if you summarize research on insomnia and its treatment research, you'd target shifting and regularizing the sleep-wake window, reducing sleep interfering behaviors and cognitions, and improving sleep efficiency and quality. For substance abuse, you'd include a target to treat physical and psychological discomfort of withdrawal symptoms and urges to use, since these predict relapse. Targets can also be transdiagnostic, in other words, fundamental processes that contribute to or maintain problems across what our current diagnostic labels consider distinct.  Check out Mansell et al's fantastic paper (4) if you're interested in more on transdiagnostic processes.
I think treatment target hierarchies are one of the practice-friendliest ways to consolidate scientific findings.

Treat Targets with Robust Change Processes

When a client's problems don't match well with one established protocol, or have failed to respond to an established protocol, improvise in a disciplined way by using the components of evidence-based protocols. Scientific knowledge is most often packaged into a manual for a disorder. Manuals are treated as if they are distinct, but most component strategies are not unique-they are common and shared across manuals.

Important efforts are underway to disentangle the component strategies or modules from the package/protocol/manual (5,6). The science here is early days, but it looks like modular approaches work and practitioners may prefer them and find them easier to learn (7).

So when you move to disciplined improvisation, assemble a protocol of the most robust modular components of evidence-based protocols relevant to your client's problems.

To match modular components to treatment targets, use the general means-ends problem solving heuristic common to all cognitive behavioral approaches.

  • Assess whether the desired behavior is not happening due a capability deficit (the client doesn't know how to do the needed behavior).  If so, then use skills training procedures.
  • If the client has the skills but emotions, contingencies, or cognitive processes or content interfere with the ability to behave skillfully, then use the procedures and principles from exposure, contingency management, and cognitive modification so that skillful behavior can happen as needed.
  • When you need them, pull disorder specific procedures and principles from relevant protocols.

Here's an example of what I mean. Say you have a late 20s professionally employed Latina woman who seeks treatment for depression. Based on the evidence, Behavioral Activation would be a good starting standard protocol (8,9). In addition to depression, she has other problems that she ranked in order of importance: #2 problem-drinking, and insomnia, marital conflict, and behavior problems with her children all about equal importance. There are protocols for each of these separately, but not a lot about how to sequence or combine treatment (10,11).

Disciplined improvisation might look like this as a rough sketch PICO table. ("PICO" is a way to frame questions for literature search that I like for shared decision-making. P stands for patient, problem or population, I for intervention, C for comparison, control or comparator, and O for outcomes (12).) Behavioral Activation is the starting template. Because it's based on the idea that depression results from a transdiagnostic process, lack of reinforcement, it's easy to adapt for other problems, too. Targets to do with problematic drinking, insomnia, parenting strategies, and even steps the client can take to improve her couple relationship can each be targeted through the robust common procedure of activation assignments to reduce avoidance that interferes with reinforcing contingencies and improve mastery and satisfaction to improve reinforcement. Then I can pull disorder specific principles and strategies from any disorder specific evidence-based protocols (e.g., for insomnia, problem drinking, or parent training).

Make sense? This way of going off-roading by assembling a modular treatment when doing EBP gives me guidance to offset the weaknesses of clinical judgment. When combined with progress monitoring, it's a disciplined way to improvise. Just like using a standard protocol, it has "If, then" tests: "if we use this protocol, we should see this change." This feedback loop makes a more learnable (kinder) environment where I can learn the relationship between clinical judgment, intervention, and outcome.

What do you think?  As always, appreciate hearing from you in the comments or by email.



1. Kuyken W. (2006), Evidence-based case formulation: Is the emperor clothed? In Tarrier, N, & Johnson, J. (Eds), Case Formulation in Cognitive Behaviour Therapy: The Treatment of Challenging and Complex Cases, pp. 12 - 35.

2. Persons, J. B. (2008). The Case Formulation Approach to Cognitive-Behavior Therapy Guildford Press: New York.

3. Comtois, K.A., & Linehan, M.M. (2006). Psychosocial treatments of suicidal behaviors: a practice-friendly review. Journal of Clinical Psychology, 62, 161-70.

4. Mansell, W., Harvey, A., Watkins, E., & Shafran, R. (2009). Conceptual foundations of the transdiagnostic approach to CBT. Journal of Cognitive Psychotherapy, 23(1), 6-19.

5 Chorpita, B. F., Daleiden, E. L., & Weisz, J. R. (2005). Modularity in the design and application of therapeutic interventions. Applied and Preventive Psychology, 11(3), 141e156.

6 Roth, A., & Pilling, S. (2008). Using an evidence-based methodology to identify the competences required to deliver effective cognitive and behavioral therapy for depression and anxiety disorders. Behavioral and Cognitive Psychotherapy, 36, 129-147. doi:10.1017/S1352465808004141

7 Weisz, J. R., Chorpita, B. F., Palinkas, L. A., Schoenwald, S. K., Miranda, J., Bearman, S. K., et al.Mayberg, S. (2012). Testing standard and modular designs for psychotherapy treating depression, anxiety, and conduct problems in youth: a randomized effectiveness trial. Archives of General Psychiatry, 69(3), 274-282.

8 Collado, A., Calderón, M., MacPherson, L., & Lejuez, C. (2016). The Efficacy of Behavioral Activation Treatment Among Depressed Spanish-Speaking Latinos.

9 Kanter, J. W., Santiago-Rivera, A. L., Santos, M. M., Nagy, G., López, M., Hurtado, G. D., & West, P. (2015). A randomized hybrid efficacy and effectiveness trial of behavioral activation for Latinos with depression.Behavior therapy, 46(2), 177-192.

10 Manber, R., Edinger, J. D., Gress, J. L., San Pedro-Salcedo, M. G., Kuo, T. F., & Kalista, T. (2008). Cognitive behavioral therapy for insomnia enhances depression outcome in patients with comorbid major depressive disorder and insomnia. Sleep, 31(4), 489-495.

11 Jacobson, N. S., Dobson, K, Fruzzetti, A. E., Schmaling, K. B., & Salusky, S. (1991) Marital therapy as a treatment for depression. Journal of Consulting and Clinical Psychology, 59, 547-557. .

12 Huang X, Lin J, Demner-Fushman D (2006). "Evaluation of PICO as a knowledge representation for clinical questions". AMIA Annu Symp Proc: 359-63. PMC 1839740. PMID 17238363.