Lynn F. Bufka and Nicholas Camp
In addition to the foregoing functional considerations, properties related to the psychometric soundness and development of the tool must also be evaluated. The approach must be reliable and valid. It also needs to have been evaluated and determined to be appropriate for use with the given population or setting. Information on norms should be relevant to the given patient population in terms of gender, age, and ethnicity, at the very least.
In addition to being reliable and valid,
screening tools should also possess high levels of sensitivity and specificity.
“Sensitivity” and “specificity” refer to how accurately an assessment approach
identifies target “cases.” The greater the sensitivity of the instrument, the greater the likelihood that the
assessment will identify an individual who actually has the particular problem
the approach is designed to identify.
Conversely, the greater the specificity of an
assessment approach, the greater the likelihood of not identifying an individual who does not have the
particular problem the approach is designed to
assess (Andrykowski, Cordova, Studts, & Miller, 1998). Clearly, tools
with high specificity and sensitivity will identify the
most individuals with the problem in question, while yielding the fewest false
positives.
However, few screening tools are both highly sensitive and highly specific. Rather,
selecting a screening approach that is highly
sensitive (that is, with a low false-negative rate) and has moderate
specificity is useful. Such a tool will identify patients whose symptoms might
possibly meet diagnostic criteria, while still
excluding a fair number of those who clearly do not have the problem in question (Baldessarini, Finkelstein, &
Arana, 1983).
Sensitivity and specificity are less important
criteria for measures of treatment outcome. More critical in the
evaluation of treatment outcome is identifying measures that actually capture
the factors of interest. Mental health treatment can target numerous factors, including symptom reduction, improved
functioning, and changes in relationships and
quality of life. To gauge treatment success, it is important to select measures that assess what the treatment is designed to
impact.
Often, measures of symptom
reduction and changes in psychopathology are used to track outcome, but sometimes other measures are necessary. Of the screening
tools, those that provide more information about psychological status may be
suitable for use as outcome measures as well. Those that are brief screens or
that provide a cutoff score indicating presence
or absence of distress are likely not sufficient for use in tracking outcomes.
Another important criterion in the selection of an approach is related to appropriate use in the given setting; that is, before choosing an assessment instrument for use in daily practice, one should first consider the utility and feasibility of the approach for the specific setting. For instance, to ensure sound psychometric properties, an approach often must be conducted in a standardized fashion, yet standardized administration is usually difficult in typical clinical settings.
Therefore, selection of an assessment approach should take into consideration
factors such as ease of administration, ability
to train staff to administer the assessment appropriately, and the degree of administration variability that still
yields accurate results. If the tool is easy to
administer and interpret, more staff members will willingly use it as part
of their regular patient care. A single screening tool
is usually preferable within a setting, so that
staff can easily compare patients and evaluate overall functioning of those who obtain care at the site, but a few different measures
to assess outcomes may be selected on the basis
of treatment target.
Once psychological measurement is a routine part of patient care, staff will have greater
information about all patients. And, as staff
members find routine psychological screening and outcomes assessment useful in their daily functioning, they will likely become
more committed to widespread, accurate use of the selected measures.
The language of the tool is also an important feature worthy of consideration. English is not the first language of many patients. Obviously, for those who do not speak or read English, the assessment must be conducted in another language, so determining that the tool’s psychometric properties are acceptable across languages is important. However, not all tools have acceptable translations, and translators might instead be employed, thus losing some reliability and validity but ensuring the screening of all patients.
The language of the tool is also an important feature worthy of consideration. English is not the first language of many patients. Obviously, for those who do not speak or read English, the assessment must be conducted in another language, so determining that the tool’s psychometric properties are acceptable across languages is important. However, not all tools have acceptable translations, and translators might instead be employed, thus losing some reliability and validity but ensuring the screening of all patients.
A second language consideration occurs when
individuals with less than fluent mastery of English
complete paper-and-pencil or computer-based measures
in English. Little is currently known about how language fluency impacts
assessment accuracy, but it may have an impact on the
validity of assessment results.
A final language consideration is the level of
literacy necessary for accurate completion of paper-and-pencil assessments.
Regular surveys of adult literacy suggest that approximately 20%
of Americans read at or below fifth-grade level (National Center for Education Statistics, 2003); therefore, any assessment
tool that requires reading must be written so
that the majority of adults can understand the material. Tools for children must be written in language appropriate for their
age.
The last important practical consideration is
whether the assessment is culturally appropriate. Manifestations of
psychopathology can vary across cultures, and various
assessment strategies may be more or less culturally sensitive. In settings
with highly diverse patient populations, this is
particularly important. Determining acceptable
adaptations to the assessment tool that still result in meaningful scores is
often necessary when selecting an instrument
(Geisinger, 1998). Tools that have been evaluated
cross-culturally should be selected for use in settings with more diverse
populations.
References
Martin
m. Antony & David H. Barlow, 2010, Handbook of assessment and treatment
planning for Psychological disorders,
Second
edition, structured and semi-structured diagnostic interviews, The Guilford
Press
No comments:
Post a Comment