Research on Clinical Intervention

Question 1

Therapy-Outcome: History

Answer 1

1950’s – Eysenck argued “spontaneous remission” (improvement of symptoms without any treatment) - said recovery rate of patents with therapy was about the same as those who had none..
—-> pessimistic about traditional therapy - but thought of all behavioral was the best.

—–made ppl ask - is psychotherapy effective? ——

1970’s: Gordon Paul “ultimate question” – WHAT treatment, by WHO, is most effective for THIS person with THAT specific problem, under WHICH set of circumstances, and HOW does it come about?

• • so people start to try to…
• determine efficacy of treatment and compare to alternatives
• assess what components are responsible for the change

then. ..
- assess durability of changes
- any negatives?
- how acceptable is treatment to different clients?
- how expensive? cost effective?
- how do they lead to changes in behavior?

Early Research : determine if therapy effective in the lab/community

Later Research : which forms are superior, and superior for what (closer to answering ‘ultimate question’)

Question 2

Outcome Research Designs: Past and Present

Answer 2

Designed so results can be interpreted w/o ambiguity

–> most powerful design = controlled experiment (independent = therapy given, dependent = observed changes)

• -> most = either within subjects or between subjects
• —>within = client gets single treatment but its altered at various points and changes are observed.
• —>between = diff groups have different treatments - ant and type of changes are observed and compared.
• randomized = more statistically significant

Question 3

Outcome Research Designs: Within Subject

Answer 3

Dependent variables measured on several occasions.

• pretreatment/ baseline
• then intervention phase begins

.. usually conducted over time with just 1 or a few patients..
allows for intensive study of treatment process
(popular for eval of new treatments or when rare disorder)

–Case Study Model - eval services by developing specific treatment for each client. then assessing effects using sing-subject like research design.

Question 4

Outcome Research Designs: Between Subjects

Answer 4

Experimental and control groups - compare changes between different manipulations.
– measure dependent before (pretest) shortly after termination (posttest) and maybe later at various times (follow up)

Radom assignment important (avoid sampling error)

Can have…

• treatement/no treatment
• treatment/alternative treatment (motivational interviewing and motivational interviewing + CBT in alcoholism)
• “Dismantling” - which aspects of therapy most associated with positive outcomes – groups receive combinations of different components of one therapy type to see which are most important.

+ = because can maimpulate several independent variables at the same time rather than sequentially.. (which is how within subjects does it)
BUT v expensive.. and a lot of effort.

Question 5

Outcome Research Designs: Randomized Clinical Trials **

Answer 5

Can use either within or between subject designs.

–>require homogenous samples, random assignment, and carefully monitored treatment regimens

Question 6

Results of Outcome Research on Treatments - Past and Present: Box Score Reviews

Answer 6

The narrative - Traditional/earliest approach to summarizing outcome research

• -make categorical judgments about whtehr each study yielded positive or negative results
• —> then tally the # of each (pos/neg outcomes)

X = subjective, unsystematic, huge number of outcome studies makes it hard to weigh the importance of each, disagreements over results,

Question 7

Results of Outcome Research on Treatments - Past and Present: Meta-Analytic Studies

Answer 7

Standardizes the outcomes of a large number of studies (so that they can be compared or combined)
-Treatment effect size calculated (avg difference in outcome between treatment and non treatment groups)
effect size small = .2 - .5
effect size large = .8 +

First meta-analysis = 1977 - found medium effect size .69 of therapy - thought therapy v effective.
–> 2nd done by same authors = larger – found rager effect size of .85 .

– Today - in general they confirm that psycho treatment is effective intervention for many psych disorders

Question 8

Results of Outcome Research on Treatments - Past and Present: Client Satisfaction Surveys

Answer 8

1990’s Cosumer reports public survey –>

• 90% felt better after treatment,
• no difference in psychotherapy alone vs plus medication,
• no approach rated more highly than others,
• family physicians and marriage counselors showed less improve nets than the rest of professionals.

Question 9

Empirically Supported Treatments and Evidence Based Treatments

Answer 9

Movement to understand which therapies work best for which problems, AND to ID which practices/through which treatment maximum benefit for clients will be produced..

EMPIRICALLY SUPPORTED:
“Task Force on Promotion and Dissemination of Psychological Procedures” - set criteria for research deigns that can reach reliable and valid conclusions about efficacy.
—> then reined them and identified list of treatments from high quality research that qualify as ‘efficacious’ (later changed to ‘empirically validated’ then changed to ‘empirically supported’)
—->2001 - 108 treatments and 37 child therapies as either:
1) well established/ efficacious and specific (2+ rigorous randomized control studies showing benefits or large # single case experiments)

2) probably/possibly efficacious (1+ “ “)
3) Promising (supported by studies whose signs are less convincing than the first two categories)

–>60-90% of Empirically Supported Treatments (ESTs) list are CBT in nature

Evidenced Based Practice:
also classify as:
-strong research support
-modest research support
-controversial/conflicting research support
—-> bulk with strong support still fall into CBT.. why?
- some say because outcome studies often need treatment to be conducted with standardized treatment manuals.. not on the basis of the therapists own preference and experience (this could bias towards CBT)

Question 10

Evidence Based Practice (EBP)

Answer 10

APA Task Force on EBP to find how treatment can be shown to be effective in the real world.

Question 11

Common / Nonspecific Factors of Therapy

Answer 11

1) Therapist Variables
2) Client Variables
3) Relationship Variables

Question 12

Nonspecific Factors of Therapy: Therapist Variables

Answer 12

What therapist characteristics predict positive outcomes?

• empathy
• collecting client feedback

also probably effective =

• positive regard towards client
• goal consensus with client
• collaboration with client

So all associated with better outcomes regardless of technique or orientation.

Question 13

Nonspecific Factors of Therapy: Client Variables

Answer 13

Some client attitudes might be more helpful in benefiting from treatment than others..

• clients who are open and offerer higher disclosure and lower resistance from the start have better outcomes.
• higher depression show greater gains than lower levels
• strong expectations of successful treatment have better outcomes how don’t expect it

Question 14

Nonspecific Factors of Therapy: Relationship Variables

Answer 14

Therapeutic Alliance - the better it is the better the outcome.

Also - evidence that customizing therapy based on client character is beneficial:
ex) non resistant patients tend to do better with structured treatment from directive therapists. More resistant do better with less directive therapists..

Question 15

Findings on Group Therapy

Answer 15

Can be effective - esp with strong group cohesion & therapeutic alliance.

• • esp - for schizo (supportive group) and for depression (CBT)
• • cost effective too

Question 16

Findings on Couples Therapy

Answer 16

Show improved relationship satisfaction and psycho adjustment.

• -> especially strong support for behavioral couples therapy
  (esp. for alcoholism)
• 2 key success factors of couples therapy shown to be:
  1) communication training
  2) teaching problem solving skills

Question 17

Findings on Family Therapy

Answer 17

Behavioral and Structural FT’s show strongest empirical support.
–> but usually all show improvements in communication patterns, and in the ID’d fam member whose problems prompted therapy.

– strongest fam therapy procedures include behavioral and psychoeducational components (ex - parent mgmt training, fam psychoeducation for schizos)

Question 18

Findings on Preventive Interventions

Answer 18

A number of effective prevention programs have been identified…
EX)
- preventing HIV by educating and addressing motivation / behavioral skills competence related to safe sex in inner city hs.
- preventing substance abuse by increasing parenting skills and strengthening healthy connections in families

Question 19

Findings on Self Help Resources/Groups

Answer 19

..Internet resources and groups are hard to evaluate empirically..

…more research on books - suggests can be effective for mild depression, eating disorders, gambling, anxiety, and mild alcohol abuse.

Question 20

Findings on Combo of Medication and Psychotherapy

Answer 20

Combining medication and psychotherapy can be helpful and in some cases more effective than either approach alone.
– studies show that for anxiety and depression (at least) combined treatments don’t greatly improve on what therapy does on its own..
AND therapy may result in benefits that are greater and more enduring than medication..

-some surveys show clients would even rather psychological therapy over pharmacological - but might not be provided for by insurance.

Question 21

Internal Validity

Answer 21

Based on ability to assert that observed changes were caused by manipulated variable not by a confounding factor.

High internal validity associated with:

• random group assignment
• defining independent variable carefully
• fidelity checks (treatment done properly?)
• careful selection of samples (client characteristics like diagnosis are screened)
• multiple dependents (all evidence based)

Question 22

External Validity

Answer 22

Based on how applicable results are to clients / problems / situations beyond the experimentally controlled environment.

-do the results apply to clients of different ages? genders? races? different types or severities of diagnosis?

Doesn’t always follow internal validity - normally a give and take.

Question 23

Therapy Research: Concerns and Compromises

Answer 23

• to have high internal validity you sacrifice external validity.
  (so well controlled experimental results can only make limited conclusions about therapy in the clinical environment)
  –> need to focus on efficacy (how well it works in tightly controlled conditions) as well as effectiveness (how well it works in the real world)
• RCT’s only focus on comparing different treatments, not common factors
• -> effects of common factors are often responsible for improvements, rather than specific ‘active ingredients’
• Treatment manuals in research designs - many don’t normally use them or feel restricted by the structure imposed.
• Dissemination and implementation is necessary to keep clinicians up to date on current practices *