Creating a Professional Resumé

Psychopathology_History_Diagnosis_and_Empirical_Fo..._----_Pages_73_to_108.pdf

Home >Psychology homework help >Creating a Professional Resumé

THE SECOND P: PRESCRIPTION 57

to indicate the new, posterior probability; Jenkins, Youngstrom, Washburn, et al. (2011) provide a worked example.

The starting probability on the left-hand line can be chosen to reflect the base rate of the diagnosis in a particular setting based either on local data (such as would be obtained from a record review) or from published estimates drawn from similar settings, such as epidemiological or clinical-epidemiological studies drawn from clinical or forensic settings (e.g., Teplin, Abram, McClelland, Dulcan, & Mericle, 2002). Alternately, the entry point for the nomogram could be an estimate based on the particular risk factors present in the given case, such as family history of the disorder. Most flexibly, clinicians could begin by quantifying their own clinical impression as a probability (assigning a number from .00 to 1.00) and then use the nomogram to see how the test result is changing their impression. This can include sensitivity analyses, where the clinician examines the impact of what-if scenarios by bracketing initial estimates with more liberal and conservative values.

Where does the practitioner find the likelihood ratios to use in the nomogram? Occasionally these are published directly in a table in a test manual or research report (e.g., Youngstrom et al., 2004), but this is not yet common practice. More often, articles will include the sensitivity and specificity values, which are sufficient information to calculate the likelihood ratios. The likelihood ratio associated with a positive test result is the sensitivity divided by (1—specificity). The likelihood ratio for a negative test result is the false negative rate (1—sensitivity) divided by the true negative rate (i.e., the specificity). It also is possible to estimate likelihood ratios if normative data are given for people with the condition and without the condition (Frazier & Youngstrom, 2006). Once the practitioner identifies the most valid estimates of sensitivity and specificity, it is necessary to calculate the likelihood ratio only once, and then the clinician can use the nomogram with all subsequent evaluations that include the measure in question.

The advantages of using a nomogram to facilitate interpretation of test results are considerable. Foremost among them is that the nomogram makes it possible to apply test results to a specific individual, directly estimating the posterior probability. The posterior probability is what all parties to an assessment find most relevant, but it varies depending on the base rate, which will change across settings. Additional advantages of the nomogram include: (a) substantial improvements in the accuracy and precision of test interpretation (Jenkins, Youngstrom, Washburn, et al., 2011), (b) reducing the influence of cognitive heuristics on the interpretation of test results (Galanter & Patel, 2005), (c) flexibility in choice of starting point (i.e., base rate, clinical impression, or the combination of either of those with other quantitative information), (d) elimination of computation, and (e) facilitation of discussion between the practitioner and the family to determine when to initiate treatment versus continuing assessment. The principal drawbacks of the nomogram approach are that it is unfamiliar to most clinicians and training programs, and that few researchers are publishing the likelihood ratios directly in their research reports. More technical limitations include that the nomogram involves a loss of precision compared to using Bayes’ theorem (which handheld devices or web applications could support), and that applications combining multiple sources of information assume that the sources are independent. In practice, as long as the data come from different sources (e.g., parent versus teacher report, or family history versus youth self-report), then the correlations are likely to be only small to moderate, and the effect on the nomogram results will be minor compared to the increases in accuracy and precision afforded by the nomogram versus typical intuitive test interpretation.

Psychopathology : History, Diagnosis, and Empirical Foundations, edited by Linda W. Craighead, et al., John Wiley & Sons, Incorporated, 2013. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/ashford-ebooks/detail.action?docID=1380177. Created from ashford-ebooks on 2021-10-31 19:43:21.

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

58 STRATEGIES FOR EVIDENCE-BASED ASSESSMENT OF CHILDREN AND ADOLESCENTS

ASSESSMENT AS AID IN TREATMENT SELECTION

A second way that assessment can help guide the prescription of interventions is by providing a mechanism for comparing treatments. Within the evidence-based medicine literature, a variety of different metrics have been developed. These include the absolute risk reduction (ARR), which subtracts the rate of a negative categorical event (suicide, relapse, etc.) in the treatment group from the rate in the comparison group, and the number needed to treat (NNT), which is equal to 1 divided by the ARR. The NNT has limitations from a statistical standpoint, but it has advantages as a commonsense metric—it indicates the number of people who would need to receive a particular treatment in order for one more case to achieve a desirable outcome (Straus et al., 2011). For example, if a particular therapy achieved a response rate of 70%, versus treatment as usual achieving a response rate of 35%, then the ARR would be 35%, and the NNT would be about 3, meaning that for every three people receiving the new treatment instead of treatment as usual, one more patient would achieve the desirable outcome.

There is a corollary to the NNT, focused on risk of iatrogenesis—the number needed to harm (NNH). The NNH is calculated as 1 divided by the rate of adverse events in the group receiving the treatment versus the rate of the same adverse events in the comparison group. Until recently, there has not been much attention to the fact that psychosocial interventions can produce harmful effects (Lilienfeld, 2007). Estimation of the NNH would remind therapists that some treatments pose significant risks, and it also would provide a common metric that could be used to help families compare the risks associated with psychosocial versus pharmacological or other interventions. If both the NNT and the NNH are available, then it is also possible to compute the likelihood of help versus harm (LHH) (Straus et al., 2011). The LHH is the ratio of the NNT to NNH, with both being adjusted for the clinician’s judgment of the patient’s risk compared to the risk of the average patient in the studies providing the values of NNT and NNH. Straus et al. (2011) present detailed description and examples, including methods for weighting outcomes according to patient preferences. The take-home message from this chapter is that there is a framework for comparing risks and benefits of treatment that can be used to compare treatment options, and that also can incorporate clinical impressions and patient values (see also Kraemer, 1992).

Although the NNT, NNH, and LHH derive from outcome data, they are most helpful in terms of guiding treatment selection for an individual (and hence are included in the “Prescription” section instead of the “Process” section of the chapter). A limitation of these parameters is that they assume dichotomous outcomes. In practice, decisions about treatment choice are also typically categorical, and continuous measures can be dichotomized if the practical advantages outweigh the statistical consequences of lost precision and power (Kraemer et al., 1999).

Clinical Implications

The NNT statistics for most mental health interventions tend to be large, underscoring the need for improved interventions. The NNH statistics are a helpful reminder that treatments almost always involve some risk, offering a nudge toward Hippocratic caution (i.e., “First, do no harm”) in prescribing interventions. The LHH provides a promising framework for thinking about the trade-off between benefits and risks for competing treatment options. Evaluation of psychosocial interventions has not focused in depth on cost-benefit analysis (cf. Lilienfeld, 2007), but the advantages of this view include a

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

THE THIRD P: PROCESS 59

clear emphasis on the lower risk of harm associated with most psychosocial interventions compared with pharmacotherapies, and also the promotion of active negotiation about choice of treatment with consumers.

The Third P: Process

A third major way that assessment tools can be informative in a developmental psy- chopathology framework is by measuring process. One major subcategory would be clinical outcomes, quantifying the degree of change occurring as the result of treatment; but the concept of processes can be used more broadly, to also encompass measurement of variables that are informative about the mechanisms for growth and change. In statistical terminology, the variables that account for change processes have been described as “mediators” (Baron & Kenny, 1986; Kraemer, Wilson, Fairburn, & Agras, 2002). In a clinical context, markers of adherence to treatment or fidelity of delivery represent another class of important process variables to consider. Finally, assessment can be used to detect variables that will change the prognosis or response to treatment. Such variables have been called “moderators,” and their effect on other variables is sometimes termed “interaction” (Cohen et al., 2003). Treatment moderators help identify which sub- groups may show the best treatment response, optimizing treatment matching to specific issues or clinical groups (Ebesutani, Bernstein, Chorpita, & Weisz, 2012). The following paragraphs elaborate on each of these uses of assessments to inform about process.

TREATMENT OUTCOME EVALUATION

Outcomes most often have been defined as reductions in the severity of symptoms, or alternately, as no longer meeting criteria for a diagnosis (Jacobson & Truax, 1991). Assessment scales have most commonly been given at the end of treatment, and then either compared to pretest baseline measurement or compared to the mean outcome in a comparison group. There has been movement toward repeating outcome assessments, sometimes using long-term follow-ups to look at the maintenance of gains, or the prevention of relapse or disease progression (e.g., Findling et al., 2007; MTA Cooperative Group, 1999). Systems that give therapists access to feedback about change during the course of treatment have produced lower rates of dropout, more efficient resource allocation across caseloads, and better outcomes (Howard, Moras, Brill, Martinovich, & Lutz, 1996; Lambert, Hansen, & Finch, 2001).

Outcome assessment has an inherent tension between the goals of psychometric accuracy versus feasibility. Reliable assessment requires the use of either longer scales or shorter scales that are repeated many times. Either of these strategies increases the respondent’s burden. If people rush to fill out a questionnaire, they may respond less thoughtfully and accurately, lowering the scale’s reliability and validity. They also are likely to skip items, creating missing data. In clinical contexts, reliance on too intensive an assessment strategy will add to the burden for the practicing clinician, who may not have the time, training, resources, or motivation to pursue such an intensive schedule of evaluation. A practical consequence of this has been a push to develop shorter outcome scales (e.g., Ware, Kosinski, & Keller, 1996). One logical end point of this process of adaptation would be to have single-item outcome measures that are repeated multiple times over the course of treatment. Alternately, a focal construct could be used as a

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

60 STRATEGIES FOR EVIDENCE-BASED ASSESSMENT OF CHILDREN AND ADOLESCENTS

primary outcome measure, or else included as part of a suite of specialized instruments capturing a broader sense of the person’s functioning.

Repeated, Brief Measures

The simplest form of a brief, repeated measure would be when a therapist asks a patient, “How are you doing now?” For research, brief assessments need to be standardized, written down, and quantitative. Using a simple scaling question, such as, “How has your anxiety been this week, on a scale from 1 to 10?” would be an example of the next level of refinement. The Longitudinal Interview Follow-up Evaluation (LIFE; Keller et al., 1987) is an example of making relatively simple ratings on a fixed scale and then repeating them each week. The reliability of a single-item rating for any given week is likely to be low. The approach’s strength comes from amassing ratings about lots of weeks, allowing a picture of the dynamic course to emerge. Other versions of brief, repeated assessments include “life charting” of mood and energy levels (Denicoff et al., 1997), or daily report cards used as a way of monitoring changes in high-frequency behaviors in settings such as the classroom or home (Evans & Youngstrom, 2006).

The principal advantages of these sorts of assessment strategies include the relatively low burden placed on the respondent, and their sensitivity to trends of change over the course of treatment. If the provider monitors these regularly, then it can lead to more responsive treatment and better engagement and outcome. A major disadvantage of these tools from a research perspective is that the resulting data are challenging to analyze using conventional statistical methods. Another major drawback is that compliance rates often are low in spite of efforts to minimize the burden, with patients often skipping days or stopping their tracking entirely.

Behavioral Checklists and Questionnaires as Outcome Measures

Behavior checklists and questionnaires, or interview-based rating scales, have been the most widely used method for evaluating outcomes in both psychotherapy and pharmacotherapy research. Ironically, the proliferation of potential outcome measures has clouded the fundamental questions of how effective a treatment is, or whether one treatment performs significantly better than another. A 10-point reduction in the Internalizing Problems T-score does not indicate the same degree of efficacy as a 10-point reduction in the Child Depression Inventory, due to differences in scaling and also differences in content validity. Converting outcomes to effect sizes, such as Cohen’s d (the mean difference between the treatment and comparison group, divided by either the pooled standard deviation or the standard deviation of the comparison group), provides only a partial solution. Effect sizes convert the outcomes of different studies (or different measures) into a common metric, but they do not address the possibility of differential validity changing the magnitude of the outcomes (De Los Reyes, Alfano, & Beidel, 2011). Meta-analytic methods testing the homogeneity of effects provide an indication of when the effect sizes are more discrepant than could be attributed to simple sampling variations (Lipsey & Wilson, 2001); and with enough studies, it becomes possible to estimate how much variation might be due to choice of outcome measure instead of other sample or treatment characteristics.

Two other limitations of effect sizes such as d or r include that they do not yield response rates, and they also are group statistics versus measures of individual outcome status. Rates are more intuitive for many consumers and also policy makers, and many of the summary statistics favored in evidence-based medicine (e.g., the NNT, NNH, LHH)

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

THE THIRD P: PROCESS 61

are derived from rates. This is not a major barrier to using effect sizes, however, as effect sizes can be converted into each other (Lipsey & Wilson, 2001). It is possible to have the best of both approaches. Keeping outcome measures as continuous scores preserves more information and increases statistical power to detect treatment effects (Cohen et al., 2003), but outcomes can be expressed in dichotomous rates when doing so would be helpful.

The more complicated challenge comes from the goal of defining successful treatment response at the individual level. Many ad hoc definitions have been used, including percentage reductions from the initial symptom level (e.g., 30% reduction in severity of depression counts as a response, etc.). Percentage reduction definitions have been used frequently in pharmacotherapy trials and older psychosocial treatment studies. However, they are generally acknowledged to be problematic for various reasons, including that the choice of a threshold percentage is usually arbitrary. Such definitions also ignore the varying degrees of retest stability afforded by different instruments, conflating instability due to poor reliability with treatment response. This artifact has probably inflated the size of the placebo response in many clinical trials (Youngstrom et al., in press).

Clinically Significant Change

One response has been to develop more complicated definitions of treatment response and remission, including algorithms that combine different sources of information or different aspects of functioning, as well as sometimes including durational requirements (Findling, et al., 2003; Frank et al., 1991). In the psychosocial literature, the need for a psychometrically sophisticated definition of clinical response that also could be meaningfully applied to individual cases led to the development of the “clinically significant change” model (Jacobson & Truax, 1991). This model focuses on two key components: reliable change and change that moves the patient’s score below a predefined normative threshold.

The “reliable change” portion of the definition focuses on whether the amount of change shown by an individual case is large enough to reflect true change (in the classical test theory sense) rather than possibly being attributable to the unreliability of the measure. Jacobson proposed that the raw change score should be divided by the standard error of the difference for the measure, which takes into account the retest stability of the measure. The reliable change index (RCI) is expressed as a z-score with the standard error of the difference acting as the denominator. The RCI makes it possible to compare individual change across multiple outcome measures, because RCIs for each will be in the same z-metric, and all will be adjusted for the amount of stability inherent to each measure. Jacobson and colleagues suggested interpreting RCIs of 1.96 or larger as reflecting reliable change (based on the 95% threshold for the normal distribution, two-tailed), or 1.65 reflecting reliable change with 90% confidence. Jacobson argued that interpretations of reliable change should always be two-tailed, as there was the possibility that treatment might actually harm the individual (Lilienfeld, 2007).

A suitably large RCI is a necessary, but not sufficient, condition for clinically significant change in Jacobson’s model. The second key element is demonstrating that the patient’s outcome score had moved past at least one of three potential benchmarks based on normative data. These could be labeled the “ABCs of change”: moving away from the clinical range (operationally defined as the threshold two standard deviations below the clinical average score on the measure), moving back into the normal range (operationally defined as being within two standard deviations of the mean for a nonclinical standardization sample), or crossing closer to the nonclinical than the clinical mean

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

62 STRATEGIES FOR EVIDENCE-BASED ASSESSMENT OF CHILDREN AND ADOLESCENTS

Nonclinical Average

Clinical Average

Away from the Clinical Distribution (>2 SD from Clinical Mean)

Back into the Nonclinical Range (<2 SD from Nonclinical Mean)

Crossing Closer to the Nonclinical Mean (below weighted average of two group means)

FIGURE 2.3 Thresholds for Clinically Significant Change Based on Clinical and Nonclinical Distributions of Scores

(defined by calculating a weighted average of the two means that adjusts for differences in the standard deviations of the clinical versus nonclinical groups). Figure 2.3 presents these three thresholds for a hypothetical measure with a very high degree of separation between the clinical and nonclinical means (d > 4). Table 2.2 presents these benchmarks for the broadband scales for the parent, teacher, and youth report versions of the Child Behavior Checklist, calculated via data in the technical manual (Achenbach & Rescorla, 2001). In the neuropsychology literature, further refinements include calculation methods that adjust for practice effects, with the preferred method being a regression-based RCI that corrects for both practice effects and regression to the mean (Sawrie et al., 1996).

Multivariate Outcomes

One of the limitations of the models just discussed is that they are univariate. Each outcome measure is considered separately. In clinical trials, this leads to concerns about Type I errors and spurious findings. It also creates the opportunity for ambiguous situations where there is adequate response on one measure but not others (such as a reduction in hyperactivity but no improvement in attention, social skills, or grades). Clustering methods might offer an attractive, multivariate definition of treatment outcomes. Rather than focusing on changes in a single scale or area of functioning, these approaches would make it possible to evaluate whether multivariate changes in functioning and problems were sufficient to change the person’s “cluster membership” from a more severe or impaired group to a less severe or better-functioning profile. An interesting start in this direction was supported by the Ohio Department for Mental Health, which sponsored the development of a cluster-based typology of adult consumers using statistical methods, with supplemental qualitative methods being used to refine and validate the cluster descriptions in collaboration with providers and consumers (Rubin & Panzano, 2002). The cluster typologies developed for different behavior checklists (as discussed previously) could similarly be used to define profiles of behavior at treatment end points as well as initial assessment.

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

THE THIRD P: PROCESS 63

TABLE 2.2 Clinically Significant Change Benchmarks Calculated for the Child Behavior Checklists (Achenbach & Rescorla, 2001)

Cut Scores∗ Critical Change

(Unstandardized Scores)

Measure A B C 95% 90% SEdifference

CBCL T-Scores

Total 49 70 58 5 4 2.4

Externalizing 49 70 58 7 6 3.4

Internalizing n/a 70 56 9 7 4.5

Attention Problems n/a 66 58 8 7 4.2

TRF T-Scores

Total n/a 70 57 5 4 2.3

Externalizing n/a 70 56 6 5 3.0

Internalizing n/a 70 55 9 7 4.4

Attention Problems n/a 66 57 5 4 2.3

YSR T-Scores

Total n/a 70 54 7 6 3.3

Externalizing n/a 70 54 9 8 4.6

Internalizing n/a 70 54 9 8 4.8

∗A = Away from the clinical range, B = Back into the nonclinical range, C = Closer to the nonclinical than clinical mean.

Note: Benchmarks for YSR Attention Problems are not presented because of consistent findings that self-report is not an effective modality for assessing attention problems.

Research and Clinical Implications

There is a great deal of work to do in order to realize the potential for evidence- based assessment of outcomes in developmental psychopathology. The research agenda includes (a) establishing relevant normative data for clinical and nonclinical groups, (b) publishing standard errors of the measure and the difference, (c) using item response theory and other methods to enhance the precision of estimates of individual functioning and to better calibrate comparisons across samples and settings (Embretson, 1996), and (d) comparing different measures and measurement strategies in terms of their sensitivity to treatment effects and their criterion validity in terms of associations with client satisfaction and functioning. Additional contributions could be made by exploring multivariate approaches to defining functioning and outcomes, such as cluster-based methods. The creative tension will be in developing more sophisticated methods that inform treatment without requiring cumbersome methods that dissuade clinicians or consumers from adopting the methods.

MEDIATORS

Mediators are the intermediate variables that act as vehicles carrying cause to effect, providing mechanisms of change (Kraemer et al., 2002). There has been extensive

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

64 STRATEGIES FOR EVIDENCE-BASED ASSESSMENT OF CHILDREN AND ADOLESCENTS

debate about the necessary requirements for mediation, including temporal sequencing or theoretical explanatory models, along with the statistical models used to test and demonstrate mediation (Baron & Kenny, 1986; Kraemer et al., 2002; MacKinnon, 2008). A regression-based approach presented by Baron and Kenny has been highly influential, but it was published before the widespread availability of covariance modeling statistical software that makes it possible to correct for measurement error and to directly estimate indirect efforts through mediational pathways. Measurement error, potential suppression effects (where two or more variables have opposite effects on the outcome), and low statistical power all complicate the assessment of mediational models (MacKinnon, 2008).

Despite these difficulties, mediation is a core theme in developmental psychopathol- ogy. Mediation tells the story of processes, rather than merely cataloging correlations. Mediational models explicate routes of development. Within the treatment context, mediation can isolate the active ingredients and how they produce therapeutic change. Better understanding of mechanisms offers strong confirmation of the theoretical models underpinning therapies, as would occur if cognitive behavior therapies produced mea- surable changes in cognitions (the purported mediator) that accounted for the changes in the outcome variables, or if family therapy produced quantifiable improvements in communication and problem solving that in turn explained the reductions in symptoms or gains in functioning. Careful examination of biological variables over the coming decades will also help clarify whether shifts in neurophysiological activity or morphology reflect a mediator or a peripheral outcome of change processes in psychopathology and treatment. Although there is great enthusiasm for using imaging techniques with so-called brain diseases, for example, there are many examples in medicine of supposed mechanisms or proxy variables having either no relationship or the opposite of what was believed prior to rigorous trials (Silverman, 1998).

Mediational studies also promote conceptual consolidation. It is unlikely that all of the various purported mechanisms of therapeutic change contribute equally to outcome. Mediational analyses can identify the more powerful drivers of change. Conversely, there has been a proliferation of psychotherapies, with reviews finding more than 300 psychosocial interventions that are at least nominally different (Kazdin, 2011). It is unlikely that this plethora of treatments involves a similar number of unique mechanisms. In fact, it is more likely that the majority share only nonspecific factors that permeate most interventions and account for a surprisingly large proportion of the outcome variance (Wampold, 2001).

A third potential benefit of elucidating change mechanisms would be that techniques could then be refined to produce larger effects on the intervening variables, and thus potentially better outcomes. For example, if improved emotion recognition were identified as a mediating agent (Izard et al., 2001), then other therapies could add emotion recognition components to the treatment package (Greenberg, Kusche, Cook, & Quamma, 1995), or more powerful methods for improving emotion recognition could be developed (Duke, Nowicki, & Martin, 1996).

MODERATORS

Moderators change the relationship between another pair of variables. Statistical mod- eration has also been termed an “interaction effect.” Moderator variables qualify the main effects that variables have on development and outcome. The diathesis-stress model of developmental psychopathology is an example of an interaction effect: Neither the

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

THE THIRD P: PROCESS 65

predisposition (diathesis) nor the stressor is sufficient in itself to produce the poor out- come, whereas the combination is. The 5HTTP short allele of the serotonin transporter gene offers an example of this type of interaction: Individuals with this genetic poly- morphism are more susceptible to the effects of stress (although they still show good outcomes in the absence of acute stressors) (Caspi et al., 2003).

The range of possible moderator variables is broad. Genetic differences will receive a great deal of attention in the next decade, now that the human genome is mapped (e.g., WTCCC, 2007). Genotypic differences will interact with the metabolism of medications, producing changes in efficacy as well as side effects (Reiss, 2010). However, genes will also be linked with responses to changes in diet, social environment, and other environmental factors. All of these factors will often be moderators of psychological processes in their own right.

Cultural variables are another level of analysis likely to yield fruitful moderator variables. Again drawing on psychotherapy for examples, cognitive therapy may be particularly well matched to families from European backgrounds, whereas Asians tend to focus more on somatic and less on cognitive features of illness, and Latino families may be especially receptive to interpersonal approaches to therapy (Mufson, Dorta, Olfson, Weissman, & Hoagwood, 2004). Cultural differences are likely to influence rates of treatment seeking, degree of engagement with different modalities of therapy, extent of available social support, and other factors likely to play key roles in development and outcome.

Patient and therapist characteristics also may serve crucial roles in determining response course. For example, higher cognitive ability appears to be a protective factor that moderates the deleterious effects of many other adverse events and factors (Gottfredson, 1997), is generally associated with more positive response to psychotherapy (Neisser et al., 1996), and probably is linked with the ability to benefit from cognitive behavior therapy in particular at younger ages (Garfield, 1994). Personality and temperament are also likely to be important predictors of engagement in therapy, chronicity, and proneness to relapse (Barnett et al., 2011; Harkness & Lilienfeld, 1997). Therapist characteristics potentially moderate treatment, too (Garfield, 1997; Wampold, 2001). The dearth of studies in this area more likely reflects our hesitance as psychologists to examine ourselves publicly rather than the incremental contribution of these variables to the outcome.

Adherence and Fidelity

Both client adherence and therapist fidelity influence the effectiveness of an intervention by changing the degree of engagement (and thus the dosing of the therapy) for the two main parties in the treatment endeavor. Some have argued that moderators should be variables that are fixed prior to the beginning of treatment (Kraemer et al., 2002). However, fidelity and adherence seem to be important counterexamples. They are not immutable, and they can change over the course of treatment; and they certainly can have a substantial impact on the outcome. Measures of adherence can include things such as rate of completion of homework assignments, attendance of scheduled sessions, pill counts, and other markers of activities integral to treatment. Measures of fidelity include therapist progress notes, checklists or rating scales completed by the therapist (or less frequently by the client) after each session, or ratings of audio- or videotape of sessions.

Some psychotherapies appear to produce their positive effects by means of improved adherence to other components of a comprehensive treatment package. For example, psychoeducation about mood disorders produces better outcomes in part through improved

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

66 STRATEGIES FOR EVIDENCE-BASED ASSESSMENT OF CHILDREN AND ADOLESCENTS

adherence to adjunctive pharmacotherapy (Fristad, Verducci, Walters, & Young, 2009). Cultural factors may influence adherence by promoting enthusiasm or opposition to treatment components.

Similarly, fidelity is also likely to influence outcome. Lower fidelity implementations would dilute the dosage of any active ingredients specific to a therapy (Crits-Christoph et al., 1991). More controversially, high degrees of fidelity to manualized treatments might undercut the flexibility and interpersonal factors that might be major, nonspecific contributors to the effectiveness of therapy (Wampold, 2001). There are competing hypotheses with regard to fidelity: Some predict that greater fidelity explains much of the gap between the magnitude of outcomes in efficacy versus effectiveness studies, and others argue that too much fidelity might actually constrain the therapist in a way that handicaps outcome.

Priority Areas for Research, Training, and Practice

The goal of this chapter has been to review the assessment of developmental psychology from the perspective of both practicing clinicians as well as researchers. Evidence- based medicine contributes a lens for evaluating the psychological literature in terms of methodological rigor as well as clinical relevance. Consistent with this spirit, the chapter closes with a summary that outlines actions for change.

CRITICAL REAPPRAISAL AGAINST THE THREE PS OF ASSESSMENT

Although literally scores of books, hundreds of instruments, and thousands of articles have been written for psychological assessment, only a meager number have clear significance for clinical work with developmental psychopathology. The Three Ps of prediction, prescription, and process are not intended to be exhaustive, but instead provide a core set of heuristics that promote reevaluation of existing tools against clear criteria tied to clinical utility. As alluded to earlier, there are more than 300 measures of depression available to clinicians. They cannot all be equally good for each distinct clinical purpose.

Three strategies will yield swift rewards by winnowing the field. One is for the practitioner critically to read the literature and look for new champion measures to dethrone incumbents. The assessment tools in one’s cabinet should be sized up in terms of their ability to predict criteria of interest (e.g., validity correlations); prescribe treatments (by means of diagnostic efficiency or by criterion validity without relying on diagnosis as a proxy, as could happen with predictors of suicide, aggression, recidivism, or other important clinical conditions); or measure process (as quantified by sensitivity to treatment effects, or else providing guidance about treatment course and response via measuring factors that moderate treatment) (Youngstrom, 2013).

A second strategy would be to conduct studies that directly compare assessment methods on these key dimensions. Investigations that simultaneously compare multiple measures are extremely informative because they level the playing field. All participating instruments are performing in the same theater, equally susceptible to the vagaries of study design characteristics. Clinicians would be well served by finding and digesting such articles.

Meta-analysis offers a third approach for comparing tests. More recent meta-analyses have become much more sophisticated in terms of testing whether parameters are

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

PRIORITY AREAS FOR RESEARCH, TRAINING, AND PRACTICE 67

converging on a single general value versus being moderated by sample or design characteristics, and they are starting to provide useful guidance about patient features that might change the validity of different assessments or treatments (Liberati et al., 2009). Recent attention to the preparation of meta-analyses and systematic reviews has also led to the promulgation of guidelines to critically evaluate the design characteristics of assessment studies (Bossuyt et al., 2003). Although there are ways that these tools can be formally applied in meta-analyses (Whiting, Rutjes, Reitsma, Bossuyt, & Kleijnen, 2003), the primary intent is for clinicians to have a convenient yet systematic way of critically appraising assessment tools and determining the appropriateness of their use with any given individual client (Straus et al., 2011).

At present, scant articles and manuals present such clinically relevant information while also drawing on good research designs. Evidence-based medicine authorities estimate that less than 2% of articles indexed in MedLine currently fulfill all of these criteria (Straus et al., 2011). They suggest referring to outlets of systematic reviews, such as Evidence- Based Mental Health and the Cochrane Collaborative Systematic Reviews, as ways of finding syntheses of rigorous studies with a clinical focus. They also provide suggestions for search strategies in databases such as MedLine and PsycINFO to increase the chances of finding relevant information, such as incorporating MeSH terms in searches. To identify studies relevant to prediction, consider using MeSH terms such as “prognosis” and “criterion validity.” To identify studies of prescription, use “sensitivity or specificity” and “prescription.” To identify studies of “process,” use “outcome measurement or outcome assessment,” “clinical evaluation,” or perhaps “adherence” or “fidelity” and “assessment or measurement”—depending on the purposes of the search.

Once relevant articles are found, it is straightforward to compare studies on the basis of sample composition, design features, and the magnitude of effects. There are helpful checklists to facilitate evaluation of studies against methodological criteria (Whiting et al., 2003), as well as recommendations about determining the match between features of the patient in question versus the sample composition of the study. If two or more measures appear to perform similarly, then considerations of cost effectiveness and burden break ties. All other things being equal, then less expensive and less invasive procedures will clearly be preferable (Youngstrom, 2013). This will be important to keep in mind during the coming wave of enthusiasm for neuron-imaging and biomarkers of brain disease: Unless these methods demonstrate substantially improved performance over less expensive methods, there is the risk that their adoption will increase costs without providing commensurate improvements in outcomes. Assessment tools are analogous to athletes—few will excel across the full range of demands for assessment, much as it is difficult for an athlete to perform exceptionally well at every event in a decathlon. Specialized assessment tools are more likely to perform well in the niches for which they were optimized, but they should not be applied to all cases any more than a single athlete should be entered into all events in a tournament.

DEVELOP NEW MEASURES

Other commentators have stated that the world has enough assessment tools and does not new ones (Kazdin, 2005). Although we agree that there is a surfeit of assessment tools that have failed to distinguish themselves from one another, most instruments have not yet been evaluated against the metrics attached to the Three Ps of assessment. As these metrics are applied to the portfolio of tools available, not only is it likely that a much

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

68 STRATEGIES FOR EVIDENCE-BASED ASSESSMENT OF CHILDREN AND ADOLESCENTS

smaller number of competitive tools will emerge, but the process will reveal areas where it will be beneficial to refine or develop new measures. The application of more conservative statistical methods will expose many examples of mediocre measurement, where tools fail to adequately measure constructs from a rigorous statistical perspective, let alone from the vantages of development, prediction, or high-stakes decisions about individuals. There are considerable costs involved in using poorly validated scales. Among the psychometric deficiencies are poor content coverage, measurement invariance across target groups, lower reliability (often due to insufficient length), lower validity (limited by reliability and poor content coverage), unnecessary or unsupportable increase in complexity of the overall battery (including Type I errors due to the proliferation of scales, as well as Type II errors due to reduction of power via low reliability or validity) (Silverstein, 1993), and increased burden and expense (Kraemer, 1992). Not only does each test add to the burden and expense for the clinician and participant, but the addition of lower-quality or irrelevant tests can actually lead to degradation of the quality of clinical decision making (Kraemer, 1992).

Another specific need is for tools that produce large effect sizes. Most existing measures do not deliver adequate effect sizes to support the intended clinical or research applications. A large effect size is d ∼ .8, based on Cohen’s reviews of typical effect sizes found in top psychology journals (Cohen, 1988, 1994). However, this translates into an AUC of .71, meaning that what the field has accepted as a large difference in group distributions is not adequate when trying to use the instrument to classify individuals. Even d of .8 renders many of the outcome benchmarks implied by the Jacobson and Truax clinical significance model silly or impossible (such as requiring negative raw scores in order to achieve a statistical definition of clinical significance).

Adaptive testing, using item response theory to quantify item properties, will make it possible to obtain increased precision and a better range of scores without increasing participant burden (but not without increasing expenses, as these approaches typically will require computer-assisted testing). The increases in validity would be worth some added investment, particularly as computing costs continue to fall. Priorities for new measures should include demonstration of more robust factor structures (and more indicators per factor), effect sizes larger than incumbent measures using clinically relevant compar- isons, and more evidence of clinical validity (ideally quantified in terms linked to the Three Ps).

FACILITATE EVIDENCE-BASED PRACTICE

There is a great deal that could be done to improve the clinical utility of existing assessment practices, even without developing new instruments. Much good could be accomplished by presenting additional psychometric information about instruments, especially by adopting the statistics espoused by evidence-based medicine. Table 2.3 presents a ladder of improvements that could be climbed for the validation of assessment tools, organized around the Three Ps. Incorporation of these techniques into the evaluation and packaging of assessment tools would markedly increase their utility for prediction, prescription, and monitoring the processes of development and treatment. Another refinement that could be woven into the developmental psychopathology literature would be to more consistently use MeSH search terms as keywords to index articles, making them easier for users to locate in the future.

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

PRIORITY AREAS FOR RESEARCH, TRAINING, AND PRACTICE 69

TABLE 2.3 Ladder of Improvements for Developmental Psychopathology Assessments

Prediction

Most Refined Present number needed to treat (NNT), number needed to harm (NNH), likelihood of help versus harm (LHH)

Multivariate regression models with interaction effects

Unstandardized regressions weights (for complete prediction model)

Confidence intervals

Effect sizes

Statistical significance (p < .05)

Entry Level Intuitive, impressionistic interpretation

Prescription

Most Refined Multistage assessment evaluations

Differential item functioning analyses and other evidence of moderators

Quality receiver operating characteristic (ROC) and cost-benefit analyses

Logistic regressions—testing incremental validity and combinations of variables

Multilevel likelihood ratios

Receiver operating characteristic (ROC) analyses

Sensitivity and specificity for one threshold

Effect size

Statistical significance

Entry Level Intuitive, impressionistic formulation

Process

Most Refined Moderators, such as adherence, fidelity, patient characteristics (including gene X environment interactions), and therapist characteristics

Mediators of treatment (although these may be more useful in treatment

development than in daily clinical practice)

Social validation

Clinical significance definitions

Multivariate models (clusters of functioning)

Elaborated tests of significance (e.g., considering positive functioning as well as

symptom reduction)

Statistical significance

Entry Level Intuitive

DO THESE RESULTS APPLY TO MY CLIENT?

Practitioners can read Table 2.3 as a treasure hunt list. Each rung further up the ladder provides more directly clinically relevant information, or more nuanced applications of assessment results to an individual. The ideal end point would be to have an evidence base that supports the choice of instrument and the customization of interpretation to meet

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

70 STRATEGIES FOR EVIDENCE-BASED ASSESSMENT OF CHILDREN AND ADOLESCENTS

the individual circumstances of the client. The choice of measure is always provisional, to the degree that a better instrument could be published at any time. “Better” within this context is primarily determined either by having demonstrated greater validity (under equal or more challenging conditions) or else by showing more validity for a subgroup to which the patient also belongs.

When the library of assessment tools is rated against these standards, there are many gaps in coverage. The gaps grow geometrically larger when trying to consider potential differences in needs for persons from culturally diverse backgrounds or with distinct clinical profiles. Does a measure of attention problems still work when translated into Spanish? Does it provide meaningful information when used with a patient who also has a pervasive developmental disorder?

The recommended strategy is to proceed with the best available tools based on the evidence, but with caution. The gains inherent to shifting to an evidence-based strategy of assessment are profound. Moderators of test performance in a particular subgroup can be empirically tested, and they have proven to be fairly rare to date. Some of this scarcity is probably due to a lack of systematic investigation, and some is due to the challenges of demonstrating an interaction effect. But from a clinician’s or client’s perspective, this is still good news: It means that the full armamentarium of assessment tools is available, unless there is a published example that contradicts the validity for a particular characteristic of the patient (Jenkins, Youngstrom, Youngstrom, Feeny, & Findling, 2011). Just as we would not withhold an influenza vaccine from Asian families simply because it had been developed on a European participant pool, we should not refrain from using tests with validity data that imperfectly match the patient. This perspective does not undercut the importance of research to examine potential moderators. Quite the opposite: Until the appropriate large group studies can be performed, the clinician should adopt a “single subject study” mentality and cautiously proceed with the best available methods, while carefully attending to the possibility that the individual outcome might diverge from the published tendency. There are definitely risks in using assessment tools developed on one group and applying them to patients from other populations. The attached risks are often greater for denying the use of an assessment instrument until it can be validated via a rigorous large group design.

IMPLICATIONS FOR TRAINING: USING THE EVIDENCE BASE AS OCCAM’S RAZOR

This chapter began with a review of the content of assessment training in both graduate programs and predoctoral internships. It concludes with recommendations about training, but it extends to include continuing education and other forms of self-improvement.

The first challenge is to reevaluate the core of one’s clinical battery, whether in private practice or in teaching an assessment course. We should not keep using or teaching the same instruments out of convention (and often the standard of practice is difficult to distinguish from habit, unless there is a fresh infusion of evidence; Youngstrom, 2013). Teaching and practice could be improved by shifting resources from measures that lack evidence (including old standards) to those that have it (Spring, 2007). A similarly radical but helpful change could be achieved by including more clinically oriented statistical methods in the core curriculum of clinical, counseling, developmental psychopathology, and school psychology programs. Many programs could accomplish this without

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

news:It

PRIORITY AREAS FOR RESEARCH, TRAINING, AND PRACTICE 71

increasing the number of required courses, by deemphasizing hand computation and crafting exercises that involve doing statistics in assessment and therapy.

Continuing education historically has been a weak method of imparting new information and skills in a way that influences practice. The situation might be improved by combining (a) clear demonstrations of the clinical utility of new methods (Jenkins, Youngstrom, Washburn, et al., 2011); (b) teaching the skills for locating new improvements (such as searching Evidence-Based Mental Health or Cochrane databases, or using MeSH terms to search PubMed); (c) teaching the Standards for the Reporting of Diagnostic accuracy studies (STARD) (Bossuyt et al., 2003) and other guidelines for critically evaluating methods; and (d) giving practitioners opportunities to practice applying the new methods to existing cases, with corrective feedback and discussion. In the evidence-based medicine literature, there has been discussion of the “leaky pipeline” connecting research innovation to the improvement of care for a specific case. The pipeline has multiple leaks, each of which is an opportunity for improved connection between research and practice (Glasziou & Haynes, 2005).

Perhaps the most productive way to lead to improvement in the quality of assessment will be to teach an attitude and behaviors that continuously check for updates and improvements. Advocates of evidence-based medicine have captured this concept in a dramatic exhortation to “burn your textbooks.” The point they want to convey is that the ideas and data available when practitioners were completing training will become outdated, and the only solution is to regularly update knowledge and practices against the current evidence base. Computer software provides a convenient metaphor: Many programs now provide automated reminders to check for upgrades. Assessment upgrades will also become available, and it is hoped that practitioners and educators will get into the habit of applying the updates.

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

References

Achenbach, T. M. (1993). Empirically based taxonomy: How to use syndromes and profile types derived from the CBCL/4-18, TRF, and YSR. Burlington: Univer- sity of Vermont Department of Psychiatry.

Achenbach, T. M. (1999). The child behavior checklist

and related instruments. In M. E. Maruish (Ed.), The use of psychological testing for treatment planning and outcomes assessment (2nd ed., pp. 429–466). Mahwah, NJ: Erlbaum.

Achenbach, T. M., McConaughy, S. H., & How-

ell, C. T. (1987). Child/adolescent behavioral and

emotional problems: Implication of cross-informant

correlations for situational specificity. Psychological Bulletin, 101, 213–232. doi:10.1037/0033-2909.101 .2.213

Achenbach, T. M., & Rescorla, L. A. (2001). Manual for the ASEBA school-age forms & profiles. Burlington: University of Vermont.

Aldenderfer, M. S., & Blashfield, R. K. (1984). Cluster analysis. Beverly Hills, CA: Sage.

Angold, A., Costello, E. J., & Erkanli, A. (1999).

Comorbidity. Journal of Child Psychology and Psy- chiatry, 40, 57–87.

Archer, R. P., & Newsom, C. R. (2000). Psychological

test usage with adolescent clients: Survey update.

Assessment, 7, 227–235. Barkley, R. A. (2002). Major life activity and health out-

comes associated with attention-deficit/hyperactivity

disorder. Journal of Clinical Psychiatry, 63, 10–15. Barnett, J. H., Huang, J., Perlis, R. H., Young, M. M.,

Rosenbaum, J. F., Nierenberg, A. A., . . . Smoller,

J. W. (2011). Personality and bipolar disorder:

Dissecting state and trait associations between

mood and personality. Psychological Medicine, 41, 1593–1604. doi:10.1017/s0033291710002333

Baron, R. M., & Kenny, D. A. (1986). The moderator-

mediator variable distinction in social psycholog-

ical research: Conceptual, strategic, and statistical

considerations. Journal of Personality and Social Psychology, 51, 1173–1182.

Beauchaine, T. P., & Beauchaine, R. J. (2002). A com-

parison of maximum covariance and k-means cluster

analysis in classifying cases into known taxon groups.

Psychological Methods, 7, 245–261. Beck, A. T., & Steer, R. A. (1987). Beck depression

inventory manual. San Antonio, TX: Psychological Corporation.

Beck, A. T., Steer, R. A., & Garbin, M. G. (1988).

Psychometric properties of the Beck Depression

Inventory: Twenty-five years of evaluation. Clinical Psychology Review, 8, 77–100.

Bird, H. R. (1996). Epidemiology of childhood dis-

orders in a cross-cultural context. Journal of Child Psychology and Psychiatry, 37, 3549.

Bird, H. R., Yager, T. J., Staghezza, B., Gould, M. S.,

Canino, G., & Rubio-Stipec, M. (1990). Impairment

in the epidemiological measurement of childhood

psychopathology in the community. Journal of the American Academy of Child & Adolescent Psychia- try, 29, 796–803.

Blanchard, J. J., Gangestad, S. W., Brown, S. A., &

Horan, W. P. (2000). Hedonic capacity and schizo-

typy revisited: A taxometric analysis of social anhe-

donia. Journal of Abnormal Psychology, 109, 87–95. Bollen, K. A. (1989). Structural equations with latent

variables. New York, NY: Wiley. Borsboom, D., Romeijn, J. W., & Wicherts, J. M.

(2008). Measurement invariance versus selection

invariance: Is fair selection possible? Psychologi- cal Methods, 13, 75–98. doi:10.1037/1082-989X.13 .2.75

Bossuyt, P. M., Reitsma, J. B., Bruns, D. E., Gatso-

nis, C. A., Glasziou, P. P., Irwig, L. M., . . . de Vet,

H. C. W. (2003). Towards complete and accurate

reporting of studies of diagnostic accuracy: The

STARD initiative. British Medical Journal, 326, 41–44. doi:10.1136/bmj.326.7379.41

Brooks, S. J., & Kutcher, S. (2001). Diagnosis and

measurement of adolescent depression: A review of

commonly utilized instruments. Journal of Child and Adolescent Psychopharmacology, 11, 341–376.

Brown, T. A. (2006). Confirmatory factor analysis for applied research. New York, NY: Guilford Press.

Buros, O. K. (1965). Foreword. In O. K. Buros

(Ed.), The mental measurements yearbooks (6th ed., p. xxii). Lincoln: University of Nebraska.

Camara, W., Nathan, J., & Puente, A. (1998). Psy- chological test usage in professional psychology: Report of the APA practice and science directorates (p. 51). Washington, DC: American Psychological

Association.

Cantwell, D. P. (1996). Classification of child and

adolescent psychopathology. Journal of Child Psy- chology and Psychiatry, 37, 3–12.

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

REFERENCES 73

Carey, W. B. (1998). Temperament and behavior prob-

lems in the classroom. School Psychology Review, 27, 522–533.

Carson, R. C. (1997). Costly compromises: A critique

of the Diagnostic and Statistical Manual of Mental Disorders. In S. Fisher & R. P. Greenberg (Eds.), From placebo to panacea (pp. 98–112). New York, NY: Wiley.

Cashel, M. L. (2002). Child and adolescent psycholog-

ical assessment: Current clinical practices and the

impact of managed care. Professional Psychology: Research and Practice, 33, 446–453. doi:10.1037// 0735-7028.33.5.446

Caspi, A., Sugden, K., Moffitt, T., Taylor, A., Craig,

I., Harrington, H., . . . Poulton, R. (2003). Influence

of life stress on depression: Moderation by polymor-

phism in the 5-HTT gene. Science, 301, 386–398. Childs, R. A., & Eyde, L. D. (2002). Assessment train-

ing in clinical psychology doctoral programs: What

should we teach? What do we teach? Journal of Personality Assessment, 78, 130–144. doi:10.1207/ S15327752JPA7801_08

Cicchetti, D., & Cohen, D. J. (1995). Developmental psychopathology: Vol. 1. Theory and methods. New York, NY: Wiley.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erl- baum.

Cohen, J. (1994). The earth is round (p < .05). American Psychologist, 49, 997–1003.

Cohen, J., & Cohen, P. (1983). Applied multiple regression/correlation analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.

Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003).

Applied multiple regression/correlation analysis for the behavioral sciences (3rd ed.). Hillsdale, NJ: Erlbaum.

Crawford, J. R., Garthwaite, P. H., Denham, A. K., &

Chelune, G. J. (2012). Using regression equations

built from summary data in the psychological assess-

ment of the individual case: Extension to multiple

regression. Psychological Assessment, 24, 801–814. doi:10.1037/a0027699

Crits-Christoph, P., Baranackie, K., Kurcias, J. S., Beck,

A. T., Carroll, K., Perry, K., . . . Zitrin, C. (1991).

Meta-analysis of therapist effects in psychotherapy

outcome studies. Psychotherapy Research, 1, 81–91. Cuthbert, B. N. (2005). Dimensional models of psy-

chopathology: Research agenda and clinical utility.

Journal of Abnormal Psychology, 114, 565–569. De Los Reyes, A., Alfano, C. A., & Beidel, D. C.

(2011). Are clinicians’ assessments of improvements

in children’s functioning “global”? Journal of Clin- ical Child & Adolescent Psychology, 40, 281–294. doi:10.1080/15374416.2011.546043

De Los Reyes, A., Henry, D. B., Tolan, P. H., &

Wakschlag, L. S. (2009). Linking informant discrep-

ancies to observed variations in young children’s

disruptive behavior. Journal of Abnormal Child Psychology, 37, 637–652. doi:10.1007/s10802-009- 9307-3

De Los Reyes, A., Youngstrom, E. A., Swan, A. J.,

Youngstrom, J. K., Feeny, N. C., & Findling, R. L.

(2011). Informant discrepancies in clinical reports of

youths and interviewers’ impressions of the reliabil-

ity of informants. Journal of Child and Adolescent Psychopharmacology, 21, 417–424. doi:10.1089/ cap.2011.0011

Denicoff, K. D., Smith-Jackson, E. E., Disney, E. R.,

Suddath, R. L., Leverich, G. S., & Post, R. M. (1997).

Preliminary evidence of the reliability and validity

of the prospective life-chart methodology (LCM-p).

Journal of Psychiatric Research, 31, 593–603. Depue, R. A., & Lenzenweger, M. F. (2001). A neu-

robehavioral dimensional model. In W. J. Livesley

(Ed.), Handbook of personality disorders: Theory, research, and treatment (pp. 136–176). New York, NY: Guilford Press.

Depue, R. A., Luciana, M., Arbisi, P., Collins, P., &

Leon, A. (1994). Dopamine and the structure of

personality: Relation of agonist-induced dopamine

activity to positive emotionality. Journal of Person- ality and Social Psychology, 67, 485–498.

Derryberry, D., & Rothbart, M. K. (1997). Reac-

tive and effortful processes in the organization of

temperament. Development and Psychopathology, 9, 633–652.

Duke, M. P., Nowicki, S., & Martin, E. A. (1996).

Teaching your child the language of social success. Atlanta, GA: Peachtree.

Duncan, T. E., Duncan, S. C., Strycker, L. A., Li, F., &

Alpert, A. (1999). An introduction to latent vari- able growth curve modeling: Concepts, issues, and applications. Mahwah, NJ: Erlbaum.

Ebesutani, C., Bernstein, A., Chorpita, B. F., & Weisz,

J. R. (2012). A transportable assessment proto-

col for prescribing youth psychosocial treatments

in real-world settings: Reducing assessment burden

via self-report scales. Psychological Assessment, 24, 141–155. doi:10.1037/a0025176

Eisman, E. J., Dies, R. R., Finn, S. E., Eyde, L.

D., Kay, G. G., Kubiszyn, T. W., . . . Moreland, K.

L. (1998). Problems and limitations in the use of psychological assessment in contemporary health

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

74 STRATEGIES FOR EVIDENCE-BASED ASSESSMENT OF CHILDREN AND ADOLESCENTS

care delivery: Report of the Board of Professional Affairs Psychological Assessment Workgroup, Part II (p. 22). Washington, DC: American Psychological

Association.

Embretson, S. E. (1996). The new rules of measurement.

Psychological Assessment, 8, 341–349. Evans, S. W., & Youngstrom, E. A. (2006). Evidence

based assessment of attention-deficit hyperactivity

disorder: Measuring outcomes. Journal of the Amer- ican Academy of Child & Adolescent Psychiatry, 45, 1132–1137.

Fabrigar, L. R., Wegener, D. T., MacCallum, R.

C., & Strahan, E. J. (1999). Evaluating the use of

exploratory factor analysis in psychological research.

Psychological Methods, 4, 272–299. Findling, R. L., Frazier, T. W., Youngstrom, E. A.,

McNamara, N. K., Stansbrey, R. J., Gracious, B. L.,

. . . Calabrese, J. R. (2007). Double-blind, placebo-

controlled trial of divalproex monotherapy in the

treatment of symptomatic youth at high risk for

developing bipolar disorder. Journal of Clinical Psy- chiatry, 68, 781–788.

Findling, R. L., Jo, B., Frazier, T. W., Youngstrom,

E. A., Demeter, C. A., Fristad, M. A., . . . Horwitz, S.

M. (under review). The 24-month course of manic

symptoms in children. Bipolar Disorders. Findling, R. L., McNamara, N. K., Gracious, B. L.,

Youngstrom, E. A., Stansbrey, R. J., Reed, M.

D., . . . Calabrese, J. R. (2003). Combination lithium

and divalproex in pediatric bipolarity. Journal of the American Academy of Child & Adolescent Psychia- try, 42, 895–901.

Flanagan, D. J., McGrew, K. S., & Ortiz, S. O. (2000).

The Wechsler intelligence scales and Gf-Gc theory. Needham Heights, MA: Allyn & Bacon.

Frank, E., Prien, R. F., Jarrett, R. B., Keller, M. B.,

Kupfer, D. J., Lavori, P. W., . . . Weissman, M. M.

(1991). Conceptualization and rationale for consen-

sus definitions of terms in major depressive disor-

der: Remission, recovery, relapse, and recurrence.

Archives of General Psychiatry, 48, 851–855. Frazier, T. W., & Youngstrom, E. A. (2006). Evidence

based assessment of attention-deficit/hyperactivity

disorder: Using multiple sources of information.

Journal of the American Academy of Child & Ado- lescent Psychiatry, 45, 614–620.

Frazier, T. W., Youngstrom, E. A., & Naugle, R. I.

(2007). The latent structure of attention-deficit/

hyperactivity disorder in a clinic-referred sample.

Neuropsychology, 21, 45–64. Frazier, T. W., Youngstrom, E. A., Speer, L., Embacher,

R., Law, P., Constantino, J., . . . Eng, C. (2012).

Validation of proposed DSM-5 criteria for autism spectrum disorder. Journal of the American Academy of Child & Adolescent Psychiatry, 51, 28–40 e23. doi:10.1016/j.jaac.2011.09.021

Fristad, M. A., Verducci, J. S., Walters, K., & Young,

M. E. (2009). Impact of multifamily psychoedu-

cational psychotherapy in treating children aged

8 to 12 years with mood disorders. Archives of General Psychiatry, 66, 1013–1021. doi:10.1001/ archgenpsychiatry.2009.112

Galanter, C., & Patel, V. (2005). Medical decision

making: A selective review for child psychiatrists

and psychologists. Journal of Child Psychology & Psychiatry, 46, 675–689.

Galanter, C. A., & Leibenluft, E. (2008). Frontiers

between attention deficit hyperactivity disorder and

bipolar disorder. Child and Adolescent Psychiatric Clinics of North America, 17, 325–346. doi:10.1016/ j.chc.2007.11.001

Garb, H. N. (1998). Studying the clinician: Judgment research and psychological assessment. Washington, DC: American Psychological Association.

Garfield, S. L. (1994). Research on client variables in

psychotherapy. In S. L. Garfield & A. E. Bergin

(Eds.), Handbook of psychotherapy and behavior change (4th ed., pp. 190–228). New York, NY: Wiley.

Garfield, S. L. (1997). The therapist as a neglected vari-

able in psychotherapy research. Clinical Psychology: Science and Practice, 4, 40–43.

Glasziou, P. P., & Haynes, B. (2005). The paths from

research to improved health outcomes. ACP Journal Club, 142, A8–A10. doi:ACPJC-2005-142-2-A08 [pii]

Glorfeld, L. W. (1995). An improvement on Horn’s

parallel analysis methodology for selecting the cor-

rect number of factors to retain. Educational and Psychological Measurement, 55, 377–393.

Glutting, J. J., Youngstrom, E. A., Oakland, T., &

Watkins, M. (1996). Situational specificity and gen-

erality of test behaviors for samples of normal and

referred children. School Psychology Review, 25, 94–107.

Gottesman, I. I., & Gould, T. D. (2003). The endopheno-

type concept in psychiatry: Etymology and strategic

intentions. American Journal of Psychiatry, 160, 636–645.

Gottfredson, L. S. (1997). Why g matters: The com- plexity of everyday life. Intelligence, 24, 79–132.

Gray, J. A., & McNaughton, N. (1996). The neuropsy-

chology of anxiety: Reprise. In D. A. Hope (Ed.),

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

REFERENCES 75

Perspectives in anxiety, panic and fear (Vol. 43, pp. 61–134). Lincoln: University of Nebraska Press.

Greenberg, M. T., Kusche, C. A., Cook, E. T., &

Quamma, J. P. (1995). Promoting emotional com-

petence in school-aged children: The effects of

the PATHS curriculum. Special Issue: Emotions in

developmental psychopathology. Development and Psychopathology, 7, 117–136.

Grimm, L. G., & Yarnold, P. R. (1995). Reading and understanding multivariate statistics. Washington, DC: American Psychological Association.

Grove, W. M., Zald, D. H., Lebow, B. S., Snitz, B. E., &

Nelson, C. (2000). Clinical versus mechanical pre-

diction: A meta-analysis. Psychological Assessment, 12, 19–30.

Guion, R. M. (1998). Assessment, measurement, and prediction for personnel decisions. Hillsdale, NJ: Erlbaum.

Guyatt, G. H., & Rennie, D. (Eds.). (2002). Users’ guides to the medical literature. Chicago, IL: AMA Press.

Hanley, J. A., & McNeil, B. J. (1983). A method

of comparing the areas under receiver operating

characteristic curves derived from the same cases.

Radiology, 148, 839–843. Harkness, A. R., & Lilienfeld, S. O. (1997). Individual

differences science for treatment planning: Personal-

ity traits. Psychological Assessment, 9, 349–360. Haslam, N., Holland, E., & Kuppens, P. (2012).

Categories versus dimensions in personality and

psychopathology: A quantitative review of taxomet-

ric research. Psychological Medicine, 42, 903–920. doi:10.1017/S0033291711001966

Hasler, G., Drevets, W. C., Gould, T. D., Gottesman,

I. I., & Manji, H. K. (2006). Toward construct-

ing an endophenotype strategy for bipolar disorders.

Biological Psychiatry, 60, 93–105. Haynes, S. N., & Lench, H. C. (2003). Incremen-

tal validity of new clinical assessment measures.

Psychological Assessment, 15, 456–466. Horn, J. L. (1965). A rationale and test for the num-

ber of factors in factor analysis. Psychometrika, 30, 179–185.

Howard, K. I., Moras, K., Brill, P. L., Martinovich,

Z., & Lutz, W. (1996). Evaluation of psychotherapy:

Efficacy, effectiveness, and patient progress. Ameri- can Psychologist, 51, 1059–1064. doi:10.1037/0003- 066X.51.10.1059

Hudziak, J. J., Althoff, R. R., Derks, E. M., Faraone, S.

V., & Boomsma, D. I. (2005). Prevalence and genetic

architecture of Child Behavior Checklist—juvenile

bipolar disorder. Biological Psychiatry, 58, 562–568.

Hudziak, J. J., Heath, A. C., Madden, P. F., Reich,

W., Bucholz, K. K., Slutske, W., . . . Todd, R. D.

(1998). Latent class and factor analysis of DSM-IV ADHD. Journal of the American Academy of Child & Adolescent Psychiatry, 37, 848–857.

Insel, T., Cuthbert, B., Garvey, M., Heinssen, R., Pine,

D. S., Quinn, K., . . . Wang, P. (2010). Research

Domain Criteria (RDoC): Toward a new classifi-

cation framework for research on mental disorders.

American Journal of Psychiatry, 167, 748–751. doi:10.1176/appi.ajp.2010.09091379

Izard, C. E., Fine, S., Schultz, D., Mostow, A., Ack-

erman, B., & Youngstrom, E. A. (2001). Emotion

knowledge as a predictor of social behavior and aca-

demic competence in children at risk. Psychological Science, 12, 18–23.

Jacobson, N. S., & Truax, P. (1991). Clinical sig-

nificance: A statistical approach to defining mean-

ingful change in psychotherapy research. Journal of Consulting and Clinical Psychology, 59, 12–19. doi:10.1037/0022-006X.59.1.12

Jaeschke, R., Guyatt, G. H., & Sackett, D. L. (1994).

Users’ guides to the medical literature: III. How to

use an article about a diagnostic test: B. What are

the results and will they help me in caring for my

patients? Journal of the American Medical Associa- tion, 271, 703–707.

Jenkins, M. M., Youngstrom, E. A., Washburn,

J. J., & Youngstrom, J. K. (2011). Evidence-based

strategies improve assessment of pediatric bipolar

disorder by community practitioners. Professional Psychology: Research and Practice, 42, 121–129. doi:10.1037/a0022506

Jenkins, M. M., Youngstrom, E. A., Youngstrom, J. K.,

Feeny, N. C., & Findling, R. L. (2011). Generalizabil-

ity of evidence-based assessment recommendations

for pediatric bipolar disorder. Psychological Assess- ment. doi:10.1037/a0025775

Jensen-Doss, A., & Weisz, J. R. (2008). Diagnos-

tic agreement predicts treatment process and out-

comes in youth mental health clinics. Journal of Consulting and Clinical Psychology, 76, 711–722. doi:10.1037/0022-006x.76.5.711

Kagan, J. (1997a). Conceptualizing psychopathology:

The importance of development profiles. Develop- ment and Psychopathology, 9, 321–334.

Kagan, J. (1997b). Temperament and the reactions to

unfamiliarity. Child Development, 68, 139–143. Kamphaus, R. W., Petoskey, M. D., Cody, A. H., Rowe,

E. W., & Huberty, C. J. (1999). A typology of par-

ent rated child behavior for a national U.S. sample.

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

76 STRATEGIES FOR EVIDENCE-BASED ASSESSMENT OF CHILDREN AND ADOLESCENTS

Journal of Child Psychology and Psychiatry, 40, 607–616.

Kaplan, D., & Elliott, P. R. (1997). A didactic example

of multilevel structural equation modeling applicable

to the study of organizations. Structural Equation Modeling, 4, 1–24.

Kazdin, A. E. (2005). Evidence-based assessment

for children and adolescents: Issues in measure-

ment development and clinical application. Journal of Clinical Child & Adolescent Psychology, 34, 548–558.

Kazdin, A. E. (2011). Evidence-based treatment

research: Advances, limitations, and next steps.

American Psychologist, 66, 685–698. doi:10.1037/ a0024975

Kazdin, A. E., & Kagan, J. (1994). Models of dys-

function in developmental psychopathology. Clinical Psychology: Science and Practice, 1, 35–52.

Keller, M. B., Lavori, P. W., Friedman, B., Nielsen,

E., Endicott, J., McDonald-Scott, P., & Andreasen,

N. C. (1987). The Longitudinal Interval Follow-up

Evaluation: A comprehensive method for assessing

outcome in prospective longitudinal studies. Archives of General Psychiatry, 44, 540–548.

Kihlstrom, J. F. (2001). To honor Kraepelin: From

symptoms to pathology in the diagnosis of mental ill-

ness. In L. E. Beutler & M. Malik (Eds.), Rethinking the DSM: Psychological perspectives. Washington, DC: American Psychological Association.

Konold, T. R., Glutting, J. J., & McDermott, P. A.

(1997). The development and applied utility of a

normative aptitude-achievement taxonomy for the

Woodcock-Johnson Psycho-Educational Battery—

Revised. Journal of Special Education, 31, 212–232. Kraemer, H. C. (1992). Evaluating medical tests: Objec-

tive and quantitative guidelines. Newbury Park, CA: Sage.

Kraemer, H. C., Kazdin, A. E., Offord, D. R., Kessler,

R. C., Jensen, P. S., & Kupfer, D. J. (1999). Measur-

ing the potency of risk factors for clinical or policy

significance. Psychological Methods, 4, 257–271. Kraemer, H. C., Wilson, G. T., Fairburn, C. G., &

Agras, W. S. (2002). Mediators and moderators

of treatment effects in randomized clinical tri-

als. Archives of General Psychiatry, 59, 877–883. doi:10.1001/archpsyc.59.10.877

Lambert, M. J., Hansen, N. B., & Finch, A. E. (2001).

Patient-focused research: Using patient outcome data

to enhance treatment effects. Journal of Consulting & Clinical Psychology, 69, 159172. doi:10.1037/0022- 006X.69.2.159

Lazarus, R. S. (1991). Emotion and adaptation. New York, NY: Oxford University Press.

Liberati, A., Altman, D. G., Tetzlaff, J., Mulrow, C.,

Gotzsche, P. C., Ioannidis, J. P., . . . Moher, D. (2009).

The PRISMA statement for reporting systematic

reviews and meta-analyses of studies that evaluate

healthcare interventions: Explanation and elabora-

tion. BMJ, 339, b2700. doi:10.1136/bmj.b2700 Lilienfeld, S. O. (2007). Psychological treatments that

cause harm. Perspectives on Psychological Science, 2, 53–70.

Lipsey, M. W., & Wilson, D. B. (2001). Practical meta-analysis (Vol. 49). Thousand Oaks, CA: Sage.

Lonigan, C. J., Vasey, M. W., Phillips, B. M., & Hazen,

R. A. (2004). Temperament, anxiety, and the pro-

cessing of threat-relevant stimuli. Journal of Clinical Child & Adolescent Psychology, 33, 8–20.

Lubke, G. H., Muthén, B., Moilanen, I. K., McGough,

J. J., Loo, S. K., Swanson, J. M., . . . Smalley, S.

L. (2007). Subtypes versus severity differences in

attention-deficit/hyperactivity disorder in the north-

ern Finnish birth cohort. Journal of the American Academy of Child & Adolescent Psychiatry, 46, 1584–1593. doi:10.1097/chi.0b013e31815750dd

MacKinnon, D. P. (2008). Introduction to statistical

mediation analysis. New York, NY: Erlbaum.

Markon, K. E., & Krueger, R. F. (2005). Categorical and

continuous models of liability to externalizing disor-

ders: A direct comparison in NESARC. Archives of General Psychiatry, 62, 1352–1359.

McCutcheon, A. L. (1987). Latent class analysis (Vol. 64). Thousand Oaks, CA: Sage.

McDermott, P. A. (1994). National profiles in youth psychopathology: Manual of Adjustment Scales for Children and Adolescents. Philadelphia, PA: Edu- metric and Clinical Science.

McDermott, P. A. (1998). MEG: Megacluster ana-

lytic strategy for multistage hierarchical grouping

with relocations and replications. Educational and Psychological Measurement, 58, 677–686.

McDermott, P. A., & Weiss, R. V. (1995). A normative

typology of healthy, subclinical, and clinical behav-

ior styles among American children and adolescents.

Psychological Assessment, 7, 162–170. McFall, R. M., & Treat, T. A. (1999). Quantifying the

information value of clinical assessment with signal

detection theory. Annual Review of Psychology, 50, 215–241. doi:10.1146/annurev.psych.50.1.215

Mechanic, D. (1989). Mental health and social policy. Englewood Cliffs, NJ: Prentice-Hall.

Meehl, P. E. (1954). Clinical versus statistical pre- diction: A theoretical analysis and a review of

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

REFERENCES 77

the evidence. Minneapolis: University of Minnesota Press.

Meyer, G. J., & Handler, L. (1997). The ability of

the Rorschach to predict subsequent outcome: A

meta-analysis of the Rorschach Prognostic Rating

Scale. Journal of Personality Assessment, 69, 1–38. doi:10.1207/s15327752jpa6901_1

Mineka, S., Watson, D., & Clark, L. A. (1998).

Comorbidity of anxiety and unipolar mood disor-

ders. Annual Review of Psychology, 49, 377–412. doi:10.1146/annurev.psych.49.1.377

MTA Cooperative Group. (1999). A 14-month ran-

domized clinical trial of treatment strategies for

attention-deficit/hyperactivity disorder. Archives of General Psychiatry, 56, 1073–1086.

Mufson, L. H., Dorta, K. P., Olfson, M., Weiss-

man, M. M., & Hoagwood, K. (2004). Effectiveness

research: Transporting interpersonal psychotherapy

for depressed adolescents (IPT-A) from the lab to

school-based health clinics. Clinical Child and Fam- ily Psychology Review, 7, 251–261.

Muthen, L., & Muthen, B. (2004). M-Plus user’s guide (3rd ed.). Los Angeles, CA: Muthen & Muthen.

Neisser, U., Boodoo, G., Bouchard, T. J., Jr., Boykin,

A. W., Brody, N., Ceci, S. J., . . . Urbina, S. (1996).

Intelligence: Knowns and unknowns. American Psy- chologist, 51, 77–101.

Nezu, A. M., Ronan, G. F., Meadows, E. A., &

McClure, K. S. (2000). Practitioner’s guide to empir- ically based measures of depression. New York, NY: Kluwer.

Nylund, K. L., Asparouhov, T., & Muthén, B. O. (2007).

Deciding on the number of classes in latent class

analysis and growth mixture modeling: A Monte

Carlo simulation study. Structural Equation Mod- eling: A Multidisciplinary Journal, 14, 535–569. doi:10.1080/10705510701575396

O’Connor, B. P. (2000). SPSS and SAS programs for

determining the number of components using parallel

analysis and Velicer’s MAP test. Behavior Research Methods, Instruments & Computers, 32, 396–402.

Oh, H.-J., Glutting, J. J., Watkins, M. W., Youngstrom,

E. A., & McDermott, P. A. (2004). Correct interpreta-

tion of latent versus observed abilities: Implications

from structural equation modeling applied to the

WISC-III and WIAT linking sample. Journal of Spe- cial Education, 38, 159–173.

Panksepp, J. (2000). Emotions as natural kinds within

the mammalian brain. In M. Lewis & J. M. Haviland-

Jones (Eds.), Handbook of emotions (2nd ed., pp. 137–156). New York, NY: Guilford Press.

Perloff, J. M., & Persons, J. B. (1988). Biases resulting

from the use of indexes: An application to attribu-

tional style and depression. Psychological Bulletin, 103, 95–104.

Piotrowski, C. (1999). Assessment practices in the

era of managed care: Current status and future

directions. Journal of Clinical Psychology, 55, 787–796.

The Psychological Corporation. (1992). Wechsler Indi- vidual Achievement Test manual. San Antonio, TX: Author.

Quay, H. C. (1993). The psychobiology of underso-

cialized aggressive conduct disorder: A theoretical

perspective. Special Issue: Toward a developmental

perspective on conduct disorder. Development and Psychopathology, 5, 165–180.

Quay, H. C. (1997). Inhibition and attention deficit

hyperactivity disorder. Journal of Abnormal Child Psychology, 25, 7–13.

Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchi- cal linear models: Applications and data analysis methods (2nd ed.). Newbury Park, CA: Sage.

Reiss, D. (2010). Introduction to the special issue:

Genetics, personalized medicine, and behavioral

intervention—Can this combination improve patient

care? Perspectives on Psychological Science, 5, 499–501. doi:10.1177/1745691610383514

Rettew, D. C., Lynch, A. D., Achenbach, T. M.,

Dumenci, L., & Ivanova, M. Y. (2009). Meta-

analyses of agreement between diagnoses made

from clinical evaluations and standardized diag-

nostic interviews. International Journal of Methods in Psychiatric Research, 18, 169–184. doi:10.1002/ mpr.289

Richards, H. J., Casey, J. O., & Lucente, S. W. (2003).

Psychopathy and treatment response in incarcerated

female substance abusers. Criminal Justice & Behav- ior, 30, 251–276.

Robins, E., & Guze, S. B. (1970). Establishment of diag-

nostic validity in psychiatric illness: Its application

to schizophrenia. American Journal of Psychiatry, 126, 983–986.

Rubin, W. V., & Panzano, P. C. (2002). Identifying

meaningful subgroups of adults with severe mental

illness. Psychiatric Services, 53, 452–457. Ruscio, A. M., & Ruscio, J. (2002). The latent structure

of analogue depression: Should the Beck Depression

Inventory be used to classify groups? Psychological Assessment, 14, 135–145.

Sattler, J. (2001). Assessment of children: Cognitive applications (4th ed.). San Diego, CA: Author.

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

78 STRATEGIES FOR EVIDENCE-BASED ASSESSMENT OF CHILDREN AND ADOLESCENTS

Sattler, J. M. (2002). Assessment of children: Behav- ioral and clinical applications (4th ed.). La Mesa, CA: Author.

Sawrie, S. M., Chelune, G. J., Naugle, R. I., & Luders,

H. O. (1996). Empirical methods for assessing mean-

ingful neuropsychological change following epilepsy

surgery. Journal of the International Neuropsycho- logical Society, 2, 556–564.

Schmidt, N. B., Kotov, R., & Joiner, T. (2004). Tax- ometrics: Toward a new diagnostic scheme for psychopathology. Washington, DC: American Psy- chological Association.

Shiner, R. L. (1998). How shall we speak of children’s

personalities in middle childhood? A preliminary

taxonomy. Psychological Bulletin, 124, 308–332. Silverman, W. A. (1998). Where’s the evidence?

Debates in modern medicine. New York, NY: Oxford University Press.

Silverman, W. K., & Ollendick, T. H. (2005). Evidence-

based assessment of anxiety and its disorders in

children and adolescents. Journal of Clinical Child & Adolescent Psychology, 34, 380–411.

Silverstein, A. B. (1993). Type I, Type II, and other

types of errors in pattern analysis. Psychological Assessment, 5, 72–74.

Solomon, A., Haaga, D. A. F., & Arnow, B. A. (2001). Is

clinical depression distinct from subthreshold depres-

sive symptoms? A review of the continuity issue in

depression research. Journal of Nervous and Mental Disease, 189, 498–506.

Spring, B. (2007). Evidence-based practice in clinical

psychology: What it is, why it matters, what you

need to know. Journal of Clinical Psychology, 63, 611–631.

Stedman, J. M., Hatch, J. P., & Schoenfeld, L. S.

(2001). The current status of psychological assess-

ment training in graduate and professional schools.

Journal of Personality Assessment, 77, 398–407. doi:10.1207/S15327752JPA7703_02

Steer, R. A., Cavalieri, T. A., Leonard, D. M., & Beck,

A. T. (1999). Use of the Beck Depression Inven-

tory for primary care to screen for major depression

disorders. General Hospital Psychiatry, 21, 106–111. Straus, S. E., Glasziou, P., Richardson, W. S., & Haynes,

R. B. (2011). Evidence-based medicine: How to practice and teach EBM (4th ed.). New York, NY: Churchill Livingstone.

Tabachnick, B. G., & Fidell, L. S. (2007). Using mul- tivariate statistics (5th ed.). Boston, MA: Allyn & Bacon.

Tengstroem, A., Grann, M., Langstroem, N., & Kull-

gren, G. (2000). Psychopathy (PCL-R) as a predictor

of violent recidivism among criminal offenders with

schizophrenia. Law and Human Behavior, 24, 45–58. Teplin, L. A., Abram, K. M., McClelland, G. M., Dul-

can, M. K., & Mericle, A. A. (2002). Psychiatric

disorders in youth in juvenile detention. Archives of General Psychiatry, 59, 1133–1143.

Thompson, B. (2000). Q-technique factor analysis: One

variation on the two-mode factor analysis of vari-

ables. In L. G. Grimm & P. R. Yarnold (Eds.),

Reading and understanding more multivariate statis- tics (pp. 207–226). Washington, DC: American Psychological Association.

Tillman, R., & Geller, B. (2005). A brief screening

tool for a prepubertal and early adolescent bipolar

disorder phenotype. American Journal of Psychiatry, 162, 1214–1216.

Velicer, W. F., Eaton, C. A., & Fava, J. L. (2000).

Construct explication through factor or component

analysis: A review and evaluation of alternative

procedures for determining the number of factors

or components. In R. D. Goffin & E. Helmes

(Eds.), Problems and solutions in human assessment (pp. 41–71). Hillsdale, NJ: Erlbaum.

Wakefield, J. C. (1997). When is development dis-

ordered? Developmental psychopathology and the

harmful dysfunction analysis of mental disorder.

Development and Psychopathology, 9, 269–290. Waller, N. G., & Meehl, P. E. (1998). Multivariate

taxometric procedures: Distinguishing types from continua (Vol. 9). Thousand Oaks, CA: Sage.

Wampold, B. E. (2001). The great psychotherapy debate: Models, methods, and findings. Mahwah, NJ: Erlbaum.

Ware, J. E., Kosinski, M., & Keller, S. D. (1996).

A 12-item short-form health survey: Construction of

scales and preliminary tests of reliability and validity.

Medical Care, 34, 220–223. Watkins, M. W., Glutting, J. J., & Youngstrom, E. A.

(2005). Issues in subtest profile analysis. In D. P.

Flanagan (Ed.), Contemporary intellectual assess- ment: Theories, tests, and issues (pp. 251–268). New York, NY: Guilford Press.

Wechsler, D. (2003). Wechsler Intelligence Scale for Children—Fourth Edition: Technical and interpre- tive manual. San Antonio, TX: Psychological Cor- poration.

Whiting, P., Rutjes, A. W., Reitsma, J. B., Bossuyt,

P. M., & Kleijnen, J. (2003). The development of

QUADAS: A tool for the quality assessment of

studies of diagnostic accuracy included in system-

atic reviews. BMC Medical Research Methodology, 3, 25.

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

REFERENCES 79

Wiggins, J. S. (1973). Personality and prediction: Principles of personality assessment. Reading, MA: Addison-Wesley.

Willett, J. B., & Singer, J. D. (1993). Investigating

onset, cessation, relapse, and recovery: Why you

should, and how you can, use discrete-time survival

analysis to examine event occurrence. Journal of Consulting and Clinical Psychology, 61, 952–965.

Wood, J. M., Nezworski, M. T., & Stejskal, W. J. (1996).

The comprehensive system for the Rorschach: A crit-

ical examination. Psychological Science, 7, 3–10. doi:10.1111/j.1467-9280.1996.tb00658.x

WTCCC [Wellcome Trust Case Control Consortium].

(2007). Genome-wide association study of 14,000

cases of seven common diseases and 3,000 shared

controls. Nature, 447, 661–678. Youngstrom, E. A. (2013). Future directions in psy-

chological assessment: Combining Evidence-Based

Medicine innovations with psychology’s historical

strengths to enhance utility. Journal of Clinical Child & Adolescent Psychology, 42, 139–159. doi: 10.1080/15374416.2012.736358

Youngstrom, E. A., Arnold, L. E., & Frazier, T. W.

(2010). Bipolar and ADHD comorbidity: Both arti-

fact and outgrowth of shared mechanisms. Clinical Psychology: Science & Practice, 17, 350–359.

Youngstrom, E. A., Findling, R. L., Calabrese, J. R.,

Gracious, B. L., Demeter, C., DelPorto Bedoya, D., &

Price, M. (2004). Comparing the diagnostic accuracy

of six potential screening instruments for bipolar dis-

order in youths aged 5 to 17 years. Journal of the American Academy of Child & Adolescent Psychi- atry, 43, 847–858. doi:10.1097/01.chi.0000125091 .35109.1e

Youngstrom, E. A., Meyers, O. I., Youngstrom, J. K.,

Calabrese, J. R., & Findling, R. L. (2006a). Compar-

ing the effects of sampling designs on the diagnostic

accuracy of eight promising screening algorithms for

pediatric bipolar disorder. Biological Psychiatry, 60, 1013–1019. doi:10.1016/j.biopsych.2006.06.023

Youngstrom, E. A., Meyers, O. I., Youngstrom,

J. K., Calabrese, J. R., & Findling, R. L. (2006b).

Diagnostic and measurement issues in the assess-

ment of pediatric bipolar disorder: Implications for

understanding mood disorder across the life cycle.

Development and Psychopathology, 18, 989–1021. doi:10.1017/S0954579406060494

Youngstrom, E. A., Zhao, J., Mankoski, R., Forbes,

R. A., Marcus, R. M., Carson, W., . . . Findling, R. L.

(in press). Clinical significance of treatment effects

with aripiprazole versus placebo in a study of manic

or mixed episodes associated with pediatric bipolar I

disorder. Journal of Child & Adolescent Psychophar- macology.

Zumbo, B. D. (2007). Three generations of DIF anal-

yses: Considering where it has been, where it is

now, and where it is going. Language Assessment Quarterly, 4, 223–233.

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

Chapter 3

Obsessive-Compulsive Disorder

JONATHAN S. ABRAMOWITZ, LAURA E. FABRICANT, AND RYAN J. JACOBY

Introduction

Obsessive-compulsive disorder (OCD) is one of the most devastating psychological dis- orders. Its symptoms often interfere with work or school, with interpersonal relationships, and with activities of daily living (e.g., using the bathroom, going to bed). Moreover, the psychopathology of OCD is among the most complex of the psychological disorders. Sufferers appear to struggle against seemingly ubiquitous unwanted thoughts, doubts, and urges that, while senseless on the one hand, are perceived as signs of danger on the other. The wide array and intricate associations between behavioral and cognitive symptoms can perplex even the most experienced of clinicians. This chapter describes the nature of OCD symptoms, the leading explanatory theories, and empirically supported approaches to assessment and treatment.

The Nature of OCD

DIAGNOSTIC CRITERIA

Obsessive-compulsive disorder (OCD) is defined by the presence of obsessions or compulsions that produce significant distress and cause noticeable interference with various aspects of role functioning (e.g., academic, occupational) (American Psychiatric Association [APA], 2013). The DSM-5 diagnostic criteria appear in Table 3.1. Obsessions are intrusive thoughts, ideas, images, impulses, or doubts that the person experiences as senseless and that evoke anxiety. Examples include unwanted ideas of germs and contamination, unwanted doubts that one has been negligent, and unacceptable thoughts of a violent, sexual, or blasphemous nature. Compulsions are urges to perform overt (e.g., checking, washing) or mental rituals (e.g., praying) in response to obsessions or to reduce anxiety or distress. The person typically perceives compulsive rituals as senseless or excessive.

Traditionally, OCD has been considered an anxiety disorder because its cardinal features are anxiety and fear, and efforts to control and escape from anxiety and fear. In DSM-5, however, OCD is now removed from the anxiety disorders and is

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

SIGNS AND SYMPTOMS OF OCD 81

TABLE 3.1 Clinical Features of Obsessive-Compulsive Disorder

Diagnosis requires either obsessions or compulsions or both:

Obsessions are recurrent, persistent thoughts, urges, or images that are intrusive and unwanted, which

the individual tries to ignore, suppress, or neutralize.

Compulsions are repetitive behaviors or mental acts that are intended to prevent or reduce anxiety

or distress or to prevent a dreaded event that the person feels driven to do, either to neutralize

an obsession or prevent an event, even though the behaviors or mental acts are not reasonably

related to the obsession or are excessive.

In addition, the obsessions and compulsions must be time-consuming (more than an hour a day) and

cause distress or impairment, not be due to other substances or medical conditions, and not better

explained by other mental conditions.

Specifiers:

Note if good/fair insight (recognition that beliefs probably not true), poor insight (believed to be

probably true) or absent insight/delusional (believed to be definitely true).

Note if tic-related (current or past history of a tic disorder).

Source: Based on DSM-5.

now the flagship diagnosis for a new diagnostic category: obsessive-compulsive and related disorders (OCRDs). Other conditions in this proposed category include body dysmorphic disorder, hair pulling disorder (aka trichotillomania), skin picking disorder, and hoarding disorder.

Signs and Symptoms of OCD

OBSESSIONS

As mentioned previously, obsessions are thoughts, images, impulses, doubts, and ideas that are experienced as unwanted, persistent, and intrusive; anxiety- or guilt-provoking; or repugnant and senseless (APA, 2013). Although highly individualized, the general themes of obsessions can be organized into categories such as contamination; guilt and responsibility for harm (to self or others); uncertainty; taboo thoughts about sex, violence, and blasphemy; and the need for order and symmetry. Most patients evince multiple obsessional themes and forms, and sometimes there are shifts in the content of these phenomena.

Unlike other types of repetitive thoughts, obsessions are experienced as unwanted or uncontrollable in that they intrude into consciousness (often triggered by something in the environment). The content of obsessions is also incongruent with the individual’s belief system and is not the type of thought the person would expect of himself or herself. Finally, obsessions are resisted; that is, they are accompanied by the sense that they must be dealt with, neutralized, or altogether avoided. The motivation to resist is activated by

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

82 OBSESSIVE-COMPULSIVE DISORDER

TABLE 3.2 Examples of Obsessions Reported by Clinic Patients With OCD

Category Example

Contamination What if I get rabies from driving over a dead animal on the street?

I used a public bathroom; what if I have someone else’s germs on me?

Responsibility for harm or

mistakes

By mistake, I might have kissed someone other than my spouse without

realizing it.

What if I left the door unlocked and someone will break into my home?

What if I called my friend a racial slur without realizing it?

What if I hit someone with my car without realizing it?

Symmetry/order Odd numbers are incorrect.

The books must be evenly placed on the shelf or else I will have bad luck.

Unacceptable thoughts

with immoral, sexual,

or violent content

Image of Jesus with an erection on the Cross.

Image of my grandparents having sex.

Thought about stabbing my husband in his sleep.

the fear that if action is not taken, disastrous consequences may occur. Table 3.2 presents examples of common obsessions observed in people with OCD.

COMPULSIONS

Compulsive rituals are the most conspicuous features of OCD and often the most func- tionally impairing. They typically belong to the following categories: decontamination (washing/cleaning), checking (including asking others for reassurance), repeating routine activities (e.g., going back and forth through a doorway), ordering and arranging, and mental rituals. Compulsions are senseless and excessive, and often need to be performed according to rules. They are also intentional, in contrast to mechanical or robotic repeti- tive behaviors such as tics. Rituals in OCD are performed to reduce distress, in contrast to repetitive behaviors in addictive or impulse-control disorders (e.g., sexual addiction, trichotillomania), which are carried out because they produce pleasure, distraction, or gratification (APA, 2013).

It is usually clear when compulsive rituals are performed to reduce obsessional anxiety about particular feared consequences. Examples include excessively checking appliances to reduce fears of electrical fires and excessive cleaning to avoid a feared sickness. In other cases, patients have difficulty articulating the presence of obsessionally feared consequences, and instead perform rituals to reduce general anxiety or to achieve a feeling of completeness. Table 3.3 presents examples of common compulsive rituals observed in people with OCD.

As Table 3.3 shows, compulsions can be overt or covert. Additional examples of covert (mental) rituals include repetition of special “safe” phrases or prayers in a specific manner, and mentally checking (or analyzing) one’s previous conversations to be sure one has not said anything offensive. Most people with OCD also deploy strategies that do not meet DSM criteria for compulsions in response to their obsessional fears (i.e., the strategies are not rule-bound or repeated over and over). Examples include purposeful

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

SIGNS AND SYMPTOMS OF OCD 83

TABLE 3.3 Examples of Compulsive Rituals Reported by Clinic Patients With OCD

Category Example

Decontamination Hand washing for 45 minutes in response to using the bathroom.

Wiping down all objects brought into the house for fear of germs from recently

applied pesticides on an adjacent lawn.

Checking Driving back to recheck that no accidents were caused at the intersection.

Returning home after seeing a fire engine to make sure the house wasn’t on fire.

Repeating routine

activities

Going through a doorway over and over to prevent bad luck.

Retracing one’s steps to make sure that no mistakes were made.

Ordering/arranging Saying the word “left” whenever one hears the word “right.”

Rearranging the books on the bookshelf until they are “just right.”

Mental rituals Canceling a bad thought by thinking of a good thought.

Excessive praying to prevent feared disastrous consequences.

distraction and thought suppression. Such rituals can take infinitely diverse forms, and some may be remarkably subtle. Functionally, however, all of these behaviors serve to neutralize obsessional thoughts or fears. The following examples illustrate neutralization strategies.

• One man gripped the steering wheel tightly when he had distressing thoughts of driving his car into opposing traffic.

• A woman with obsessional thoughts of her child drowning tried to suppress and dismiss such images when they came to mind (thought suppression).

• A woman with obsessions about harming her husband confessed these thoughts to him whenever they came to her mind. She explained, “If I tell my husband that I’m thinking about hurting him, he’ll be ready to stop me if I start to act.”

Many individuals with OCD also engage in repeated attempts to gain ultimate certainty that obsessional doubts are invalid. Such attempts to gain assurances might be overt (e.g., asking questions) or covert (checking one’s body for signs of sexual arousal in response to inappropriate stimuli), although the most straightforward style is asking similar questions over and over.

AVOIDANCE

Avoidance behavior is present in most people with OCD and is intended to prevent obsessional fears and compulsive urges altogether. The aim of avoidance might be to prevent specific consequences such as contamination or illness, whereas in other instances avoidance is focused on preventing obsessional thoughts from occurring in the first place. For example, one woman avoided using public staircases because they evoked thoughts and fears of impulsively pushing unsuspecting people down the steps. Other patients engage in avoidance so that they do not have to carry out tedious compulsive rituals.

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

84 OBSESSIVE-COMPULSIVE DISORDER

For instance, a young man with obsessional fears of contamination from his family’s home computer (because it had been used to view pornography) engaged in elaborate and time-consuming compulsive cleaning and showering rituals. During the morning and afternoon he avoided the computer room so that he would not have to perform these rituals during the day. In the evening, however, he relaxed his avoidance and allowed himself to enter the room and become contaminated knowing that he could “work in” his ritualistic showering before bedtime.

SUBTYPES AND DIMENSIONS OF OCD

Although there are grounds for conceptualizing OCD as a homogeneous disorder, research has identified reliable and valid OCD symptom dimensions (Abramowitz et al., 2010; McKay et al., 2004). These include (a) contamination (contamination obsessions and decontamination rituals), (b) responsibility for harm and mistakes (aggressive obsessions and checking rituals), (c) incompleteness (obsessions about order or exactness and arranging rituals), and (d) unacceptable taboo violent, sexual, or blasphemous thoughts with mental rituals.

POOR INSIGHT

As is shown in Table 3.1, the DSM-5 criteria for OCD include the specifiers “good or fair insight,” “poor insight,” and “absent insight” to denote the degree to which the person views his or her obsessional fears and compulsive behavior as reasonable. Although most people with OCD recognize that their obsessions and compulsions are senseless and excessive, there is a continuum of insight, with 4% of patients convinced that their symptoms are realistic (i.e., poor or absent insight; Foa & Kozak, 1995). Poorer insight appears to be associated with religious obsessions, fears of mistakes, and aggressive obsessional impulses (Tolin, Abramowitz, Kozak, & Foa, 2001).

TIC-RELATED OCD

The DSM-5 criteria also denote a subtype of OCD in which the individual has a history of tic disorders such as Tourette’s syndrome. While the data are not conclusive, this putative variant of OCD appears to run in families and involve an early onset and male predominance. Obsessions typically concern symmetry and exactness, and compulsions often involve ordering and arranging.

INTERPERSONAL ASPECTS OF OCD

OCD frequently has a negative impact on the sufferer’s interpersonal relationships—such as that with a romantic partner, spouse, or other family members. In turn, dysfunctional relationship patterns can promote the maintenance of OCD symptoms so that a vicious cycle develops. For example, a partner or spouse might inadvertently behave in ways that maintain OCD symptoms by helping with compulsive rituals and avoidance behavior

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

SIGNS AND SYMPTOMS OF OCD 85

out of love for the sufferer (i.e., symptom accommodation). OCD symptoms also create relationship distress and conflict, which exacerbate the anxiety and obsessional symptoms. In this section we discuss these patterns and their effects on OCD symptoms.

Symptom Accommodation

Accommodation occurs when a friend or relative participates in the loved one’s rituals, facilitates avoidance strategies, assumes daily responsibilities for the sufferer, or helps to resolve problems that have resulted from the patient’s obsessional fears and compulsive urges. The accommodation might occur at the request (or demand) of the individual with OCD, who deliberately tries to involve loved ones to help with controlling his or her anxiety. In other instances, loved ones voluntarily accommodate because they feel the need to show care and concern for their suffering partners and do not wish to see them become highly anxious. Table 3.4 shows examples of accommodation behaviors we have observed in our work with couples in which one partner has OCD.

Conceptually, since avoidance and compulsive rituals prevent the natural extinction of obsessional fear and ritualistic urges, accommodation to these symptoms by a relative or close friend perpetuates OCD symptoms. For instance, consider a woman with obsessional fears of acting on unwanted impulses to molest her newborn infant. By accommodating his wife’s avoidance of changing or bathing their newborn child by doing it himself, her husband prevents his wife from learning that she is unlikely to act on her unwanted obsessional thoughts and that she can manage the temporary anxiety that accompanies these thoughts. Indeed, researchers have found that family accommodation predicts an attenuated response to cognitive-behavioral treatment for OCD (Steketee & Van Noppen, 2003).

TABLE 3.4 Examples of Family Accommodation Behaviors in OCD

OCD Symptom Partner Accommodation Behavior

Contamination and washing

symptoms

Washing or cleaning for the patient

Doing extra laundry

Avoiding contaminated stimuli

Obsessional doubting and

compulsive checking

Assisting with checking rituals

Providing reassurance

Helping the patient avoid ambiguous situations that might

trigger doubts

Violent, sexual, and religious

obsessions

Providing reassurance

Helping with avoidance of stimuli that trigger obsessional

thoughts

Helping with praying or interpreting Bible passages or

religious doubts

Ordering and symmetry (“not just

right”) obsessions and

compulsions

Checking to make sure things are in order or arranged

properly

Repeating answers until they are just right

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

86 OBSESSIVE-COMPULSIVE DISORDER

Relationship Conflict

Relationship stress and conflict also play an important role in the maintenance of OCD. Couples in which one partner suffers with OCD often report problems with interde- pendency, unassertiveness, and avoidant communication patterns that foster stress and conflict. In all likelihood, OCD symptoms and relationship distress influence each other, rather than one exclusively leading to the other. For example, a husband’s contentious relationship with his wife might contribute to overall anxiety and uncertainty that develops into his obsessional doubting. His excessive checking, reassurance seeking, and overly cautious actions could also lead to frequent disagreements and relationship conflict. Particular aspects of a relationship that might increase distress and contribute to OCD maintenance include poor problem-solving skills, hostility, and criticism. Moreover, crit- icism, hostility, and emotional overinvolvement are associated with premature treatment discontinuation and symptom relapse (Chambless & Steketee, 2000).

OBSESSIVE-COMPULSIVE RELATED DISORDERS

In DSM-5, OCD has been removed from its traditional categorization among the anxiety disorders, and placed in a new category of putatively similar obsessive-compulsive and related disorders (OCRDs), which includes body dysmorphic disorder, hoarding disorder, skin picking disorder, and hair pulling disorder (trichotillomania). This change has been criticized on conceptual, practical, and empirical grounds as we briefly describe next (for a more thorough discussion see Storch, Abramowitz, & Goodman, 2008).

One impetus for the creation of the OCD-related disorders category in DSM-5 is the fact that the disorders of this group all involve repetitive thoughts or behaviors. Repetitive behaviors are indeed present in both OCD and the proposed related disorders—for example, trichotillomania (now called hair pulling disorder in DSM-5) and excessive skin picking. Yet whereas compulsive rituals in OCD are performed in response to obsessional fear and they function as an escape from distress, hair pulling and skin picking are not triggered by obsessions or fear, and do not function to reduce fears of negative consequences (e.g., Stanley, Swann, Bowers, & Davis, 1992). Although individuals with these problems may experience guilt, shame, and anxiety associated with their repetitive behaviors, this is not the same as obsessional fear.

Body Dysmorphic Disorder (BDD)

Both BDD and OCD can involve intrusive, distressing thoughts concerning one’s appearance, and repeated checking might be observed in both disorders. However, the focus of BDD symptoms is limited to one’s appearance, whereas people with OCD also tend to have other obsessions. Nevertheless, similar psychological treatments can be effective for both conditions.

Hoarding

Once considered to be a symptom of OCD, hoarding is now understood as a separate problem. Indeed, many individuals with hoarding do not meet diagnostic criteria for OCD (Pertusa et al., 2010), and hoarding symptoms are no more prevalent in patients with OCD than in patients with other psychological disorders (Abramowitz, Wheaton, & Storch, 2008). Hoarding also differs from other OCD symptom domains and does

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

PREVALENCE AND EPIDEMIOLOGY OF OCD 87

not fit into conceptual frameworks for understanding OCD (Rachman, Elliot, Shafran, & Radomsky, 2009). Hoarding typically involves thoughts about acquiring and maintaining possessions, thoughts that are not particularly intrusive or unwanted; in fact, these thoughts are generally emotionally positive or neutral and thus do not meet criteria for obsessions (Rachman et al., 2009). It is also difficult to conceptualize excessive saving as compulsive or ritualistic, and this behavior does not result in an escape from (or neutralization of) obsessional anxiety in the way that checking or washing compulsions do (Rachman et al., 2009).

OBSESSIVE-COMPULSIVE PERSONALITY DISORDER (OCPD)

OCPD involves the presence of personality traits such as excessive perfectionism, inflexibility, and need for control that negatively impact interpersonal relationships, occupational functioning, or other important domains of an individual’s life. Individuals with this condition often maintain strict principles and are intolerant of others who do not conform to their standards. Historical clinical opinion has proposed a special relationship between OCPD and OCD, which can be traced back to Freud’s anecdotal description of the “Rat Man,” who was described as having both conditions. Similarities between the two conditions can sometimes be observed, such as excessive list making and arranging. However, the functional roles of these symptoms are notably distinct in each syndrome. The experience of individuals with OCPD is ego-syntonic in that they consider their behaviors and urges as rational and appropriate. In contrast, the obsessive thoughts experienced by individuals with OCD are ego-dystonic in that they are experienced as unwanted, upsetting, and personally repugnant. People with OCD harming obsessions may feel compelled to write down everything they have done during a day in order to reassure themselves that they have not caused a catastrophe, whereas an individual with OCPD may believe that making lists of daily activities maximizes efficiency and ensures that no details are overlooked.

Although distinct, OCD and OCPD may co-occur. Comorbidity estimates have sug- gested that between 23% and 32% of OCD patients also display one or more symptoms of OCPD, and some studies have suggested that comorbid OCPD is associated with poorer treatment outcome for OCD (Eisen, Mancebo, Chiappone, Pinto, & Rasmussen, 2008). However, OCPD has also been found to co-occur with a variety of other anxiety disorders as well as depression (Dowson & Grounds, 1995). In addition, other personality disorders, such as avoidant and dependent personality disorder, have been estimated to co-occur with OCD at least as frequently as OCPD, if not more so (Pfohl & Blum, 1991), suggesting the lack of a unique relationship between OCD and OCPD.

Prevalence and Epidemiology of OCD

PREVALENCE

The lifetime prevalence of OCD has been estimated at between 0.7% and 2.9% (e.g., Kessler et al., 2005) and there is a slight preponderance of females (Rasmussen & Eisen, 1992a). The disorder typically begins by age 25, although childhood or adolescence onset is not rare. Mean onset age is earlier in males (about 21 years) than in females (22 to 24 years) (Rasmussen & Eisen, 1992b).

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

88 OBSESSIVE-COMPULSIVE DISORDER

COURSE

OCD is a chronic condition with a low rate of spontaneous remission. Left untreated, symptoms fluctuate, with worsening during periods of increased life stress. Fortunately, more patients now receive effective treatments than ever before, leading to increased rates of symptom remission. Full recovery, however, is the exception rather than the rule.

Psychological Models

LEARNING MODELS

Behavioral (conditioning) models of OCD are based on Mowrer’s (1960) two-stage theory of fear acquisition and maintenance. In the first stage (classical conditioning), a previously neutral stimulus (the conditioned stimulus, or CS) is paired with an aversive stimulus (the unconditioned stimulus, or UCS; e.g., a traumatic experience), so that the CS comes to elicit a conditioned fear response, or CR. As a result, situations (e.g., driving, using the bathroom); objects (e.g., door handles, knives); and thoughts, images, doubts, or impulses (e.g., thoughts of harm) that pose no objective threat come to evoke obsessional fear.

In the second stage (operant conditioning), avoidance behaviors develop as a means of reducing anxiety; avoidance is negatively reinforced by the immediate (albeit temporary) reduction in distress it engenders. Compulsive rituals, which develop as an escape behavior from obsessional fear when avoidance is impossible, are also negatively reinforced in this way. Avoidance and escape behaviors, however, prevent the natural extinction of obsessional fears, and thereby maintain such fear.

Although the conditioning explanation has fallen out of favor as an explanation for the development of OCD symptoms, operant conditioning (negative reinforcement) does appear to play a role in the maintenance of OCD symptoms. Experimental research, for example, has repeatedly found that compulsive behavior leads to anxiety reduction (e.g., Hodgson & Rachman, 1972). Overall, then, negative reinforcement provides an empirically valid explanation for the persistence of compulsive rituals and avoidance behavior in OCD.

COGNITIVE DEFICIT MODELS

Memory

Some theorists have proposed that OCD symptoms arise from abnormally functioning cognitive processes, such as memory. Compulsive checking, for example, could develop as a consequence of not being able to remember whether one has locked the door, and so on. Research, however, has found no evidence of a memory deficit in OCD (e.g., Woods, Vevea, Chambless, & Bayen, 2002). In fact, patients appear to have a selectively better memory for OCD-related information relative to non-OCD-relevant stimuli (Radomsky, Rachman, & Hammond, 2001).

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

PSYCHOLOGICAL MODELS 89

Reality Monitoring

It has also been proposed that OCD is related to problems with reality monitoring—the ability to discriminate between memories of actual versus imagined events. Compulsive checking, for example, could be prompted by difficulties discerning whether an action (e.g., locking the door) was really carried out or merely imagined. Across a series of studies addressing this issue, however, no differences in reality monitoring between OCD patients and control groups were found (Woods et al., 2002).

Inhibitory Deficits

The intrusive and repetitious quality of obsessions has led some researchers to hypothesize that OCD is characterized by deficits in cognitive inhibition—the ability to dismiss extraneous mental stimuli. Studies examining recall and recognition suggest that people with OCD have more difficulty forgetting negative material and material related to their obsessional fears relative to other sorts of material.

Synthesis

There are a number of limitations of cognitive deficit models of OCD. First, they do not account for the heterogeneity of OCD symptoms (e.g., why do some people have washing compulsions while others have checking rituals?). Second, they do not account for the fact that similar mild cognitive deficits have been found in many psychological disorders. Thus, if cognitive deficits play a causal role in OCD, it is most likely to be a nonspecific vulnerability factor, as opposed to a specific cause.

COGNITIVE-BEHAVIORAL MODELS

The most promising psychological model of OCD is the cognitive-behavioral approach, which is based on Beck’s (1976) cognitive theory that emotional disturbance is brought about not by situations and stimuli themselves, but by how one makes sense out of such situations or stimuli. Accordingly, obsessions and compulsions are thought to arise from specific sorts of dysfunctional beliefs, with the strength of these beliefs influencing the person’s degree of insight into his or her OCD symptoms. In particular, the cognitive- behavioral model of OCD is based on the finding that unwanted intrusive thoughts (i.e., thoughts, images, and impulses that intrude into consciousness) are a normal experience (e.g., Rachman & de Silva, 1978). These normal intrusions are postulated to develop into clinical obsessions when they are appraised as significant and harmful. To illustrate, consider an intrusive doubt about one’s home burning down while one is away on vacation. Most people experiencing such an intrusion would regard it as an insignificant cognitive event. Such an intrusion, however, could develop into a clinical obsession if the person appraises it as having serious consequences; for example: “If I think about it, it might happen,” or “I must take extra precautions to ensure that it doesn’t happen.” Such appraisals evoke distress and motivate attempts to suppress or remove the unwanted intrusion (e.g., by replacing it with a “good” thought), or prevent any harmful events associated with it (e.g., compulsive checking of appliances).

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

90 OBSESSIVE-COMPULSIVE DISORDER

TABLE 3.5 Domains of Dysfunctional Beliefs in OCD

Belief Description

Inflated responsibility/

overestimation of threat

Belief that one has the power to cause and/or the duty to prevent

negative outcomes

Belief that negative events are likely and would be unmanageable

Exaggeration of the importance

of thoughts and need to

control thoughts

Belief that the mere presence of a thought indicates that the

thought is significant

Belief that complete control over one’s thoughts is both

necessary and possible

Perfectionism/intolerance for

uncertainty

Belief that mistakes and imperfection are intolerable

Belief that it is necessary and possible to be 100% certain that

negative outcomes will not occur

According to this approach, compulsive rituals and avoidance represent efforts to remove intrusions and prevent feared consequences. Salkovskis (1996) advanced two reasons that compulsions and avoidance become persistent and excessive. First, they are negatively reinforced by their ability to reduce distress (as in the learning model). Second, they prevent people from learning that their appraisals of intrusions are exaggerated and unrealistic. That is, performing the ritual robs the person of the opportunity to discover that the anticipated negative outcome would most likely not have occurred in the first place. If the individual avoids obsessional triggers, there is no opportunity to learn that distressing obsessional thoughts do not pose danger.

The cognitive-behavioral model has a good deal of empirical support (e.g., Clark, 2004). Psychometric research indicates that there are three principal domains of dys- functional beliefs (shown in Table 3.5) associated with OCD symptoms (e.g., Wheaton, Abramowitz, Berman, Riemann, & Hale, 2010), and laboratory experiments have demon- strated that inducing such beliefs influences dysfunctional appraisals and exacerbates obsessional symptoms (e.g., Rassin, Merckelbach, Muris, & Spaan, 1999). Longitudinal prospective research has also found that these types of beliefs confer vulnerability to the onset or worsening of obsessive-compulsive symptoms under certain conditions (e.g., Abramowitz, Khandher, Nelson, Deacon, & Rygwall, 2006).

Implications

The cognitive-behavioral approach implies that a successful treatment for OCD symptoms must accomplish two things: (1) the correction of maladaptive beliefs and appraisals that lead to obsessional fear and (2) the termination of avoidance and compulsive rituals that prevent the self-correction of maladaptive beliefs and extinction of anxiety. In short, the task of cognitive behavior therapy (CBT) is to foster an evaluation of obsessional stimuli as nonthreatening and therefore not demanding of further action. Patients must come to understand their problem not in terms of the risk of feared consequences, but in terms of how they are thinking and behaving in response to stimuli that objectively pose a low risk of harm.

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

NEUROBIOLOGICAL MODELS OF OCD 91

Neurobiological Models of OCD

SEROTONIN HYPOTHESIS

The serotonin hypothesis proposes that obsessions and compulsions arise from abnormal- ities in this neurotransmitter system, specifically a hypersensitivity of the postsynaptic serotonergic receptors (Zohar & Insel, 1987). Three lines of evidence are cited to support the serotonin hypothesis: medication outcome studies, biological marker studies, and bio- logical challenge studies in which OCD symptoms are evoked using serotonin agonists and antagonists. The most consistent findings come from the pharmacotherapy litera- ture, which suggests that selective serotonin reuptake inhibitor (SSRI) medications (e.g., fluoxetine) are more effective than medications with other mechanisms of action (e.g., imipramine) in reducing OCD symptoms. In contrast, studies of biological markers—such as blood and cerebrospinal fluid levels of serotonin metabolites—have provided incon- clusive results regarding a relationship between serotonin and OCD (e.g., Insel, Mueller, Alterman, Linnoila, & Murphy, 1985). Similarly, results from studies using the phar- macological challenge paradigm are largely incompatible with the serotonin hypothesis (Hollander et al., 1992).

STRUCTURAL MODELS

Structural models hypothesize that OCD is caused by neuroanatomical and functional abnormalities in particular areas of the brain, specifically the orbitofrontal-subcortical circuits, which are thought to connect brain regions involved in processing information with those involved in the initiation of behavioral responses. The classical conceptual- ization of this circuitry consists of a direct and an indirect pathway. The direct pathway projects from the cerebral cortex to the striatum to the internal segment of the globus pallidus/substantia nigra, pars reticulata complex, then to the thalamus and back to the cortex. The indirect pathway is similar, but projects from the striatum to the external segment of the globus pallidus to the subthalamic nucleus before returning to the common pathway. Overactivity of the direct circuit is thought to give rise to OCD symptoms.

Structural models of OCD are derived from neuroimaging studies in which activity levels in specific brain areas are compared between people with and without the condition. Investigations using positron emission tomography (PET) have found increased glucose utilization in the orbitofrontal cortex (OFC), caudate, thalamus, prefrontal cortex, and anterior cingulate among patients with OCD compared to nonpatients (e.g., Baxter et al., 1988). Studies using single photon emission computed tomography (SPECT) have reported decreased blood flow to the OFC, caudate, various areas of the cortex, and thalamus in OCD patients compared to nonpatients (for a review see Whiteside, Port, & Abramowitz, 2004). Finally, studies comparing individuals with OCD to healthy controls using magnetic resonance spectroscopy (MRS) have reported decreased levels of various markers of neuronal viability in the left and right striatum and in the medial thalamus (e.g., Fitzgerald, Paulson, Stewart, & Rosenberg, 2000). Although findings vary across studies, a meta-analysis of 10 PET and SPECT studies found that relative to healthy individuals, those with OCD evince more activity in the orbital gyrus and the head of the caudate nucleus (Whiteside et al., 2004).

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .

92 OBSESSIVE-COMPULSIVE DISORDER

EVALUATION OF BIOLOGICAL MODELS

One limitation of biological models is that no explanation has been offered for how neurotransmitter or neuroanatomical abnormalities translate into OCD symptoms (e.g., why does hypersensitivity of postsynaptic receptors cause obsessional thoughts or compulsive rituals?). In addition, biological models are unable to explain (a) the fact that OCD symptoms are generally constrained to the particular themes discussed earlier, and (b) why someone would experience one type of obsession (e.g., contamination), but not another (e.g., sexual). A third problem with biological models is their logical (as opposed to empirical) basis. Since the serotonin hypothesis originated from the findings of preferential efficacy of serononergic medication (SSRIs) over nonserotonergic antidepressants (e.g., imipramine), the assertion that the effectiveness of SSRIs supports the serotonin hypothesis is circular. Further still, there is a logical fallacy in deriving etiological models from treatment results. This fallacy is best illustrated with the following example: “When I use steroid cream, my rash goes away. Therefore, the reason I got the rash in the first place was that my steroid level was too low.” Evidence from controlled studies of differences in serotonergic functioning between individuals with and without OCD is especially inconsistent, so there is actually little convincing evidence that OCD is caused by an abnormally functioning serotonin system. A final problem with biological models is that they are based on correlational studies, which cannot address (a) whether true abnormalities exist and (b) whether the observed relationships are causal.

Assessment

DIAGNOSTIC INTERVIEWS

The Structured Clinical Interview for DSM Disorders (SCID; First, Spitzer, Gibbon, & Williams, 2002); the Mini International Neuropsychiatric Interview (MINI; Sheehan et al., 1998); and the Anxiety Disorders Interview Schedule for DSM-IV (ADIS-IV; Di Nardo, Brown, & Barlow, 1994) all assess the cardinal features of OCD, although these will likely be updated for DSM-5. The ADIS provides the most detail about OCD symptoms, assessing their severity using dimensional rating scales.

CLINICIAN-RATED SEVERITY SCALES

The Yale-Brown Obsessive Compulsive Scale (Y-BOCS; Goodman et al., 1989a, 1989b) is the most widely used clinician-rated measure of OCD. It contains three parts: First, the interviewer provides definitions of obsessions and compulsions to help in identifying these symptoms. Second, using a symptom checklist of over 50 common obsessions and compulsions, the interviewer asks the patient to indicate whether each symptom is currently present, is absent, or was present only in the past. The clinician and patient then generate a list of the three most severe obsessions, compulsions, and OCD-related avoidance behaviors. The third section is a 10-item severity scale that assesses the (a) time spent with, (b) interference from, (c) distress associated with, (d) efforts to resist, and (e) ability to control obsessions (items 1–5) and compulsions (items 6–10). Each item is rated on a scale from 0 (no symptoms) to 4 (extremely severe), and scores on the 10 items

C o p yr

ig h t ©

2 0 1 3 . Jo

h n W

ile y

& S

o n s,

I n co

rp o ra

te d . A

ll ri g h ts

r e se

rv e d .