Types of Assessment Measures

Structured Diagnostic Interviews

Clinician-administered structured diagnostic interviews are considered extremely valuable tools for assessing PTSD symptomatology (Keane et al., 1995). Although it is standard practice in clinical research settings to employ structured diagnostic interviews, the use of these types of interviews in the clinical setting is less common, with perhaps the single exception of clinical forensic practice (Keane, 1995; Keane, Buckley, & Miller, 2003). In general, the infrequency of use may be due to time and cost burdens, as well as the need for specialized training to master the administration of many of these interviews. Nonetheless, it has been suggested that increased use of structured diagnostic interviews for PTsD in clinical settings may well improve diagnostic accuracy and aid in treatment planning (Litz & Weathers, 1994).

several structured interviews are available that were developed for the assessment of PTsD either as modules of comprehensive diagnostic assessment tools or as independent PTSD measures. These are described below.


The Structured Clinical Interview for the DSM-IV (SCID-IV; First et al., 2000) is designed to assess a broad range of psychiatric conditions on Axes I and II. It is divided into separate modules corresponding to DSM-IV diagnostic criteria (American Psychiatric Association, 1994); each module provides the interviewer with specific prompts and follow-up inquiries intended to be read verbatim to respondents. Symptom presence is rated on a 3-point confidence scale based on the interviewer's interpretation of the individual's responses to the questions. To assess PTSD, respondents are asked to frame symptoms in terms of their "worst trauma experience." The SCID is intended for use only by clinicians and highly trained interviewers.

Although the administration of the full SCID-IV can be time consuming, the modular structure allows clinicians to limit their assessment to conditions that are frequently comorbid with PTSD. Within the context of a trauma clinic, it is recommended that modules for anxiety disorders, mood disorders, and substance use disorders be administered. Administration of the psychotic screen will also help to rule out conditions that require a different set of interventions (Keane & Barlow, 2002).

The SCID-PTSD module is considered psychometrically sound. Keane et al. (1998) examined the interrater reliability of the SCID by asking a second interviewer to listen to audiotapes of an initial interview. They found a kappa of .68 and agreement across lifetime, current, and never PTSD of 78%. Similarly, in a sample of patients who were reinterviewed within a week by a different clinician, they found a kappa of .66 and diagnostic agreement of 78%. McFall, Smith, Roszell, Tarver, and Malas (1990b) reported evidence of convergent validity, finding significant correlations between the number of SCID-PTSD items endorsed and other measures of PTSD (e.g., Mississippi Scale [.65; Keane, Caddell, & Taylor, 1988] and MMPI-PTSD Scale [.46; Keane, Malloy, & Fairbank, 1984]). The SCID-PTSD module also yielded substantial sensitivity (.81) and specificity (.98) and a robust kappa (.82) in one clinical sample against a composite PTSD diagnosis (Kulka et al., 1988), indicating good diagnostic utility.

Disadvantages to the SCID have been described previously. First, the scoring algorithm of the SCID permits only a dichotomous rating (e.g., presence or absence of a PTSD diagnosis), which has limitations. Most clinicians agree that psychological symptoms occur in a dimensional rather than dichotomous fashion (Keane, et al., 2000). Another disadvantage of the SCID is that it does not assess for the frequency or severity of symptoms. Third, by assessing symptoms in response to the "worst event" experienced, important information may be lost regarding the effects of other traumatic events (Cusack, Falsetti, & de Arellano, 2002). Finally, the trauma screen of the SCID may miss significant traumatic events (Falsetti et al., 1996).


Developed by the National Center for PTSD (Blake et al., 1990), the Clinician Administered PTSD Scale (CAPS) is currently one of the most widely used structured interviews for diagnosing and measuring the severity of PTSD (Weathers, Keane, & Davidson, 2001). The CAPS assesses all DSM-IV diagnostic criteria for PTSD, including criteria A (exposure), B-D (core symptom clusters), E (chronology), and F (functional impairment), as well as associated symptoms of guilt and dissociation. An important feature of the CAPS is that it contains separate ratings for the frequency and intensity of each symptom, which can be summed to create a severity score for each symptom. This permits flexibility in scoring and analyses. The CAPS also promotes uniform administration and scoring through carefully phrased prompt questions and explicit rating scale anchors with clear behavioral referents. Once trained, interviewers are able to ask their own follow-up questions and use their clinical judgment in arriving at the best ratings.

Similar to the SCID, flexibility is built into the administration of the CAPS. Interviewers can administer only subscales for the 17 core symptoms, all DSM-IV criteria (American Psychiatric Association, 1994), and/or add subscales for the associated symptoms. If administered in its entirety, the CAPS takes approximately 1 hour, but the time for administration is cut in half if only the 17 core symptoms are assessed.

Weathers et al. (2001) extensively reviewed the psychometric studies conducted on the CAPS; Weathers, Ruscio, and Keane (1999) also examined the reliability and validity data of the CAPS across five samples of male Vietnam veterans, collected at the National Center for PTSD. Robust estimates were found for interrater reliability over a 2- to 3-day interval for each of the three symptom clusters (.86-.87 for frequency, .86-.92 for intensity, and .88-.91 for severity) and all 17 symptoms (.91 for total frequency, .91 for total intensity, and .92 for total severity). Test-retest reliability for a CAPS-based PTSD diagnosis was also high (kappa = .89 in one sample and 1.00 in a second sample). Thus the data indicate that trained and calibrated raters can achieve a high degree of consistency in using the CAPS to rate PTSD symptom severity and diagnose PTSD. Weathers et al. (1999) also found high internal consistency across all 17 items in a research sample (alphas of .93 for frequency and .94 for intensity and severity) and a clinical sample (alphas of .85 for frequency, .86 for intensity, and .87 for severity), which supports the use of the CAPS in both research and clinical settings.

Strong evidence for validity of the CAPS was also provided by Weathers et al. (1999), who found that the CAPS total severity score correlated highly with other measures of PTSD (Mississippi Scale =.91; MMPI-PTSD Scale = .77; the number of PTSD symptoms endorsed on the SCID = .89; and the PTSD Checklist [PCL] = .94). As expected, correlations with measures of antisocial personality disorder were low (.14-.33). Weathers et al. (1999) also found strong evidence for the diagnostic utility of the CAPS, using three CAPS scoring rules for predicting a SCID-based PTSD diagnosis. The rule having the closest correspondence to the SCID yielded a sensitivity of .91, specificity of .84, and efficiency of .88, with a kappa of .75, indicating good diagnostic utility (see Weathers et al., 1999, for a detailed discussion of different scoring rules and their implications).

The CAPS has now been used successfully with a wide variety of trauma populations (e.g., combat veterans, victims of rape, crime, motor vehicle accidents, incest, the Holocaust, torture, and cancer), has served as the primary diagnostic or outcome measure in more than 200 empirical studies on PTSD, and has been translated into at least 12 languages (Weathers et al., 2001). Thus the existing data strongly support its continued use in both clinical and research settings.


Developed by Foa, Riggs, Dancu, and Rothbaum (1993), the PTSD Symptom Scale Interview (PSS-I) is a structured interview designed to assess symptoms of PTSD in individuals with a known trauma history. Using a Likert scale, interviewers rate the severity of 17 symptoms corresponding to the DSM-III-R (American Psychiatric Association, 1987) criteria for PTSD. One limitation of the PSS-I is that it measures symptoms over the past 2 weeks, rather than 1 month, which the DSM criteria specify as necessary for a diagnosis of PTSD (Cusack et al., 2002). According to the authors, the PSS-I can be administered by lay interviewers who are trained to recognize the clinical picture presented by traumatized individuals.

The PSS-I was originally tested in a sample of women with a history of rape and nonsexual assault (Foa et al., 1993) and found to have strong psychometric properties. Foa et al. reported high internal consistency (Cronbach alphas = .85 for full scale, .65-.71 for subscales), test-retest reliability over a 1-month period (.80), and interrater agreement for a PTSD diagnosis (kappa = .91, 95% agreement). With respect to validity, the PSS-I was significantly correlated with other measures of traumatic stress (e.g., .69, Impact of Event Scale [IES] Intrusion score [Horowitz, Wilner, & Alvarez, 1979]; .67, Rape Aftermath Symptom Test total score [RAST; Kilpatrick, 1988]) and demonstrated good diagnostic utility when compared to a SCID-PTSD diagnosis (sensitivity = .88; specificity = .96). The PSS-I appears to possess many strong features that warrant its consideration for clinical and research use, especially with sexual assault survivors.


Originally developed by Davidson, Smith, and Kudler (1989), the Structured Interview for PTSD (SIP) is designed to diagnose PTSD and measure symptom severity. It includes 17 items focused on the DSM-IV (1994) criteria for PTSD as well as two items focused on survivor and behavior guilt. Each item is rated by the interviewer on a Likert scale. There are initial probe questions and follow-up questions to promote a more thorough understanding of the respondent's symptom experiences. It can be administered by clinicians or appropriately trained paraprofessionals. The SIP takes 10-30 minutes to administer, depending upon the level of symptomatology present.

Psychometric data for the SIP is good. In a sample of combat veterans, Davidson et al. (1989) reported high interrater reliability (.97-.99) on total SIP scores and perfect agreement on the presence or absence of PTSD across raters. High alpha coefficients have also been reported (.94 for the veteran sample [Davidson et al., 1989] and .80 for PTSD patients enrolled in a clinical trial [Davidson, Malik, & Travers, 1997]). In the veteran sample, test-retest reliability for the total SIP score was .71 over a 2-week period. With respect to validity, the SIP was significantly correlated with other measures of PTSD, but not with measures of combat exposure (.49-.67;

Davidson et al., 1989, 1997, as cited in Orsillo, 2001, p. 291). Davidson et al. (1989) compared the SIP scores of current and remitted SCID-defined PTSD cases and reported good sensitivity (.96) and specificity (.80) against the SCID. At a cutoff score of 25, the SIP correctly classified 94% of cases relative to a structured clinical interview (Davidson et al., 1997). Overall, the SIP appears to be a sound instrument


Derived from the 19-item SIP, the Treatment Outcome PTSD Scale (TOP-8) was designed as a brief interview to assess core symptoms of PTSD in treatment outcome studies (Davidson & Colket, 1997). Item selection was drawn from a sample of patients with chronic PTSD who were taking part in a clinical trial for pharmacological treatment. The interview includes eight items reflecting symptoms that are thought to occur most frequently in PTSD and demonstrate the most change in response to treatment. Using a Likert scale, interviewers rate how much each symptom has "troubled the person" during the past week. It takes 5-10 minutes to administer.

Initial data on the TOP-8 scale indicate that it has good psychometric properties, including high internal consistency (Cronbach alpha = .73), test-retest reliability (.88; Connor & Davidson, 1999), and interrater reliability (.96; Davidson & Colket, 1997). Evidence for convergent validity has also been provided (e.g., correlations of .91 with the Davidson Trauma Scale [Davidson, Book et al., 1997], .89 with the IES, and .98 with the SIP).

According to Davidson and Colket (1997), the advantages of the TOP-8 are that it takes less time than many other structured interviews, eliminates items reflective of symptoms that are rare or unlikely to change, and may reduce counter- or therapeutic effects of lengthy interviews. They acknowledge its disadvantages of eliminating clinically important or distressing symptoms and offering less ability to explore properties of treatment.

Unfortunately, the methodological strategies used to construct and validate the scale are not commonly accepted in the psychometric literature. The results of the preliminary studies might be a function of these idiosyncratic methods, the specific nature of the small samples employed, or the fact that the scale was derived entirely retrospectively rather than prospec-tively. Thus replications using additional methods are necessary before endorsing the shortened version of the SIP. The authors appropriately recommend that clinicians and researchers may want to use this interview only in conjunction with other clinical measures.


Originally developed by DiNardo, O'Brien, Barlow, Waddell, and Blanchard (1983), the Anxiety Disorders Interview Schedule (ADIS) was designed to permit differential diagnoses among the DSM-III (American Psychiatric Association, 1980) anxiety disorder categories and to provide detailed symp tom ratings. The interview was revised to fit DSM-III-R (American Psychiatric Association, 1987) (ADIS-R; DiNardo & Barlow, 1988) and more recently DSM-IV (American Psychiatric Association, 1994) criteria (ADIS-IV; DiNardo, Brown, & Barlow, 1994). The ADIS-IV also includes an assessment of mood disorders, substance use disorders, and selected somatoform disorders, a diagnostic time line, and a dimensional assessment of the key and associated features of the disorders. The provision of a dimensional as well as a categorical assessment allows the clinician to describe subthreshold manifestations of each disorder and offers more possibilities for analyses. The ADIS has been translated into numerous languages and used in over 150 clinical and research settings around the world. It is recommended only for trained, experienced interviewers.

Psychometric studies on the ADIS-PTSD module provide mixed results. Originally tested in a small sample of Vietnam combat veterans, the ADIS-PTSD module yielded strong sensitivity (1.0) and specificity (.91), and 93% agreement with interview-determined diagnoses (Blanchard, Gerardi, Kolb, & Barlow, 1986). DiNardo, Moras, Barlow, Rapee, and Brown (1993) tested the reliability of the ADIS-R in a community sample recruited from an anxiety disorders clinic and found only fair agreement between two independent raters when PTSD was the principal diagnosis or an additional diagnosis (kappa = .55). In a test of the ADIS-IV, the interrater reliability across two interviews given 10 days apart was also fair for current diagnoses (kappa = .59; Brown, DiNardo, Lehman, & Campbell, 2001) but slightly improved for lifetime diagnoses (kappa = .61). Additional reliability and validity data on the ADIS-IV are needed to ensure its continued use in clinical and research settings.


Developed by Watson, Juba, Manifold, Kucala, and Anderson (1991), the PTSD-Interview (PTSD-I) is a diagnostic interview based on the DSM-III-R (American Psychiatric Association, 1987) that differs in administrative format from most other interviews. Patients are given a copy of the scale to read along with the interviewer and are asked to provide a rating (based on a Likert scale) for each of the symptoms. This format shares much in common with self-report questionnaires yet deviates from the other diagnostic scales in that it does not allow clinicians to make ratings of their own, based on their expertise and experience. The PTSD-I yields both dichotomous and continuous information.

Psychometric data on the PTSD-I is excellent. Watson et al. (1991) administered the PTSD-I to a sample of veteran outpatients and found high test-retest reliability (.95) for the PTSD-I total score over a 1-week interval and high interrater reliability for a PTSD diagnosis (kappa = .61, 87% agreement). A high alpha coefficient (.92) indicated good internal consistency. With regard to validity, the total score of the PTSD-I has been shown to cor relate highly with other measures of PTSD (e.g., PTSD section of the Diagnostic Interview Schedule [DIS; Robins & Helzer, 1985] = .94, Watson et al., 1991; and the IES =.85, Wilson, Tinker, Becker, & Gillette, 1994). Compared to the DIS-PTSD scale, Watson et al. found that the PTSD-I yielded a sensitivity of .89, specificity of .94, and overall efficiency of .92, indicating good diagnostic utility.

Self-Report PTSD Questionnaires

Numerous self-report measures have been developed as a method of obtaining information on PTSD. For the most part, self-report measures are used as continuous measures of PTsD to reflect symptom severity; in several cases, however, specific cutoff scores have been developed to provide a diagnosis of PTSD. These measures are generally more time and cost efficient than structured interviews and can be especially valuable when used as screens for PTSD or in conjunction with structured interviews. The data also support the use of self-report questionnaires alone in clinical and research settings when administering a structured interview is not feasible or practical. Many of the measures can be used interchangeably, because the findings appear to be robust for the minor variations in methods and approaches involved. In selecting a particular instrument, the clinician is encouraged to examine the data for the population on which that instrument is to be employed. In so doing, the clinician is likely to maximize the accuracy and efficiency of the test employed (Keane & Barlow, 2002).


Developed by Horowitz et al. (1979), the IES is one of the most widely used self-report measures to assess psychological responses to a traumatic stressor. The initial 15-item questionnaire, which focused only on intrusion and avoidance symptoms, was derived from a model of traumatic stress developed by Horowitz (1976). Since the publication of the DSM-IV (American Psychiatric Association, 1994), a revised 22-item version of the scale (IES-R; Weiss & Marmar, 1997) was developed that includes items on hyperarousal symptoms and flashback experiences. Thus the IES-R more closely parallels DSM-IV criteria for PTSD. To complete the measure, respondents rate (on a Likert scale) "how distressed or bothered" they were by each symptom during the past week since a traumatic event. The IES has been translated into several languages and has been used with many different trauma populations. It takes approximately 10 minutes to complete.

Data on the psychometric properties of the revised IES-R are preliminary. In two studies that incorporated four samples of emergency workers and earthquake survivors, Weiss and Marmar (1997) reported satisfactory internal consistency for each of the subscales (alphas = .87-.92 for Intrusion, .84-.86 for Avoidance, and .79-.90 for Hyperarousal). Test-retest reliability data from two samples yielded a range of reliability coefficients for the sub-scales (Intrusion = .57-.94, Avoidance = .51-.89, and Hyperarousal = .59.92). Weiss and Marmar suggest that the shorter interval between assessments and the greater recency of the traumatic event contributed to higher coefficients of stability for one sample.

Convergent and discriminant validity data are not yet available for the IES-R. There were many questions raised about the validity of the original scale, in part because it did not assess all DSM criteria for PTSD (see Weathers et al., 1996; Joseph, 2000). Although it now more closely parallels the DSM-IV (1994), items measuring numbing are considered limited by some investigators (Foa, Cashman, Jaycox, & Perry, 1997). In a review of psychometric studies on the IES, Sundin and Horowitz (2002) report a range of correlations between the IES subscales and other self-report measures (e.g., .31-.46 on SCL-90 [Symptom Checklist] PTSD items; Arata, Saunders, & Kil-patrick, 1991) and diagnostic interviews (e.g., .32-.49 for SCID—McFall et al., 1990b; .75-.79 for CAPS—Neal et al., 1994). Neal et al. (1994) reported high sensitivity (.89) and specificity (.88) for the original scale when compared to a CAPS diagnosis. Additional studies with the revised instrument are clearly needed to establish its reliability and validity and ensure its continued use in clinics and research settings.


Developed by Keane et al. (1988), the 35-item Mississippi Scale is widely used to assess combat-related PTSD symptoms. The scale items were selected from an initial pool of 200 items generated by experts to closely match the DSM-III (American Psychiatric Association, 1980) criteria for the disorder. The Mississippi Scale has been updated and now assesses the presence of symptoms reflecting the DSM-IV (1994) criteria for PTSD and several associated features. Respondents are asked to rate, on a Likert scale, the severity of symptoms over the time period occurring "since the event." The Mississippi Scale yields a continuous score of symptom severity as well as diagnostic information. It is available in several languages and takes 10-15 minutes to administer.

The Mississippi Scale has excellent psychometric properties. In Vietnam veterans seeking treatment, Keane et al. (1988) reported high internal consistency (alpha = .94) and test-retest reliability (.97) over a 1-week time interval. In a subsequent validation study, the authors found substantial sensitivity (.93) and specificity (.89) with a cutoff of 107, and an overall hit rate of 90% when the scale was used to differentiate between a PTSD group and two non-PTSD comparison groups.

McFall, Smith, Mackay, and Tarver (1990a) replicated these findings and further demonstrated that PTSD patients with and without substance use disorders did not differ on the Mississippi Scale. Given the high comorbidity between PTSD and substance use disorders, the authors felt it was important to demonstrate that the test assesses PTSD symptoms rather than effects associated with alcohol and drug use. McFall et al. (1990a) also obtained information on convergent validity, finding significant correlations between the Mississippi Scale and other measures of PTSD, including the total number of SCID-PTSD symptoms (.57), total IES score (.46), and degree of traumatic combat exposure (.40; Vietnam-Era Stress Inventory, Wilson & Krauss, 1984). These findings suggest that the Mississippi Scale is a valuable self-report tool in settings where assessment of combat-related PTSD is needed.


Originally derived from the MMPI Form R (Keane et al., 1984), the Keane PTSD Scale (PK) now consists of 46 items empirically drawn from the MMPI-2 (Lyons & Keane, 1992). The items are answered in a true/false format. The scale is typically administered as part of the full MMPI-2 but can be useful as a stand-alone scale. The embedded and stand-alone versions are highly correlated (.90; Herman, Weathers, Litz, & Keane, 1996). The PK Scale yields a total score that reflects the presence or absence of PTSD. The standalone scale takes 15 minutes to administer.

Psychometric data on the embedded and stand-alone versions of the PK Scale are excellent. Herman et al. (1996) reported evidence from a veteran sample of strong internal consistency of the embedded and stand-alone versions of the MMPI-2 PTSD Scale (alphas ranging from .95 to .96), and high test-retest reliability coefficients for the stand-alone version over 2 to 3 days (.95). With regard to validity, the embedded and stand-alone versions of the MMPI-2 PTSD scale were correlated with other self-report measures of PTSD, including the Mississippi Scale (.81-.85), IES (.65-.71), and PCL (.77.83), and a diagnostic interview (CAPS; .77 to .80). The embedded and standalone versions differed slightly in their optimally efficient cutoff score (26 vs. 24, respectively), but both demonstrated good sensitivity (.72 for embedded, .82 for stand-alone), specificity (.82 for embedded, .76 for stand-alone), and efficiency (.76 for embedded, .80 for stand-alone) compared to a CAPS diagnosis.

More research is needed to determine the generalizability of the findings on veterans with other populations, as well as the optimal cutoff scores (Foa et al., 1997; Watson, Kucala, & Manifold, 1986). Although only a few studies have been conducted on the PK in nonveteran populations, the data presented appear to be promising (Koretzky & Peck, 1990; Neal et al., 1994). The PK may be particularly useful in the area of forensic psychology, where the MMPI-2 is frequently employed because of its validity indexes.


Developed by Foa et al. (1997), the Posttraumatic Stress Diagnostic Scale (PDS) is a 49-item scale designed to measure DSM-IV (American Psychiatric Association, 1994) PTSD criteria and symptom severity. The PDS is a revised version of an earlier self-report scale based on the DSM-III-R (American Psychiatric Association, 1987), referred to as the PTSD Symptom Scale—Self-Report Version (PSS-SR; Foa et al., 1993). The PDS reviews trauma exposure and identifies the most distressing trauma. It also assesses criterion A-2 (physical threat or helplessness), criteria B-D (intensity and frequency of all 17 symptoms), and Criterion F (functional impairment). This scale has been used with several populations, including combat veterans, accident victims, and sexual and nonsexual assault survivors. The PDS can be administered in 10-15 minutes.

The psychometric properties of the PDS were evaluated among 264 volunteers recruited from several PTSD treatment centers as well as from nontreatment-seeking populations at high risk for trauma (Foa, et al., 1997). Investigators reported high internal consistency for the PTSD total score (alpha = .92) and subscales (alphas = .78-.84) and satisfactory test-retest reliability coefficients for the total PDS score and for the symptom cluster scores (.77-.85). With regard to validity, the PDS total score correlated highly with other scales that measure traumatic responses (IES Intrusion = .80 and Avoidance = .66; RAST = .81). In addition, the measure yielded substantial sensitivity (.89), specificity (.75), and high levels of diagnostic agreement with a SCID diagnosis (kappa = .65, 82% agreement). Based on these data, the authors have recommended the PDS as an effective and efficient screening tool for PTSD.


Developed by researchers at the National Center for PTSD (Weathers et al., 1993), the PTSD Checklist (PCL) is a 17-item self-report measure of PTSD symptomatology. Different scoring procedures may be used to yield either a continuous measure of PTSD symptom severity or a dichotomous indicator of diagnostic status. Furthermore, dichotomous scoring methods include either an overall cutoff score or a symptom cluster scoring approach. The original scale was based on the DSM-III-R criteria for PTSD; the PCL has been updated to assess the 17 diagnostic criteria outlined in the DSM-IV (American Psychiatric Association, 1994). Respondents are asked to rate, on a Likert scale, "how much each problem has bothered them" during the past month. The time frame can be adjusted, as needed, to suit the goals of the assessment. There is a civilian (PCL-C) and a military version (PCL-M) of the measure. On the civilian version reexperiencing and avoidance symptoms apply to any lifetime stressful event, whereas on the PCL-M, reexperiencing and avoidance symptoms apply only to stressful events that are military-related. The PCL has been used extensively in both research and clinical settings and takes 5-10 minutes to administer. If needed, a 17-item Life Events Checklist, developed as a companion to the CAPS and aimed at identifying exposure to potentially traumatic experiences (thereby establishing criteria A for the diagnosis), can be used with the PCL.

The PCL was originally validated in a sample of Vietnam and Persian Gulf War veterans and found to have strong psychometric properties (Weathers, Litz, Herman, Huska, & Keane, 1993). Keen, Kutter, Niles, and Krinsley (2004) examined the psychometric properties of the updated PCL in veterans with both combat and noncombat traumas and found evidence for high internal consistency (alpha = .96 for all 17 symptoms). Test-retest reliability was not examined, but the original study suggested that reliability was robust (.96) over a 2- to 3-day interval, and other investigators have documented adequate test-retest reliability of this measure over a 2-week time frame (Ruggiero, Del Ben, Scotti, & Rabalais, 2003).

With respect to validity, Keen et al. (2004) found that the scale was highly correlated with other measures of PTSD, including the Mississippi Scale (.90) and CAPS (total symptom severity = .79). Using a slightly higher cutoff score (i.e., 60) than Weathers et al. (1993) used, Keen et al. also found that the PCL had a sensitivity of .56, a specificity of .92, and overall efficiency of .84 when compared to the CAPS, indicating good diagnostic power.

Several studies now offer evidence for the reliability and validity of the PCL in nonveteran samples, although there are discrepancies reported in the optimal cutoff score to obtain the highest level of diagnostic efficiency. The possible reasons for these discrepancies (e.g., gender, recency of trauma, severity of trauma, and treatment-seeking status; Manne, DuHamel, Gallelli, Sorgen, & Redd, 1998) warrant further investigation. In addition, there is evidence that different scoring options for the PCL (e.g., an absolute cutoff score vs. symptom cluster scoring) yield differences in sensitivity, specificity, and diagnostic efficiency. Keen et al. (2004) suggest that the selection of a scoring routine may depend on the goal of the assessment; for example, symptom cluster scoring was associated with higher sensitivity and may be preferable when the goal is to identify all possible cases of PTSD, whereas the cutoff method was associated with higher specificity and may be preferable for research or when clinical resources are limited.


Developed by King, King, Leskin, and Foy (1995), the Los Angeles Symptom Checklist (LASC) is a 43-item scale that diagnoses PTSD and describes symptom severity. The original scale, referred to as the PTSD Symptom Checklist (Foy, Sipprelle, Rueger, & Carroll, 1984), was designed to closely adhere to DSM-III (American Psychiatric Association, 1980) criteria and has now been updated to correspond to the DSM-IV (American Psychiatric Association, 1994). The LASC includes an assessment of B, C, and D criteria (17 items). Respondents are asked to rate, on a Likert scale, "how much of a problem" each symptom is for them. There is also a global assessment of distress and adjustment problems related to trauma exposure. No time frame is established for rating symptoms. Originally validated on a veteran sample, the LACS has now been used to assess trauma symptoms in several other trau matized groups (see King et al., 1995 for details). It takes approximately 15 minutes to administer.

Psychometric data on the LASC are strong. King et al. (1995) combined data from 10 studies that used the LACS with clinical samples derived from a diverse set of populations (i.e., Vietnam veterans, battered women, adult survivors of childhood abuse, maritally distressed women, psychiatric outpatients, and high-risk adolescents). Evidence was provided for high internal consistency among veterans (alpha = .94 for 17 items measuring PTSD and .94 for the 43-item index) and women (alpha = .89 for 17-item index and .93 for 43-item index). With regard to test-retest reliability over a 2-week period, the 17-item set yielded a coefficient of .94, and the 43-item set yielded a coefficient of .90, suggesting stability in responses over time. With respect to validity, LASC scores correlate to varying degrees with measures of combat exposure (.51 for Combat Exposure Scale; Foy et al., 1984) and traumatic stress (e.g., 38 for Intrusion and .48 for Avoidance subscales on the IES; Astin, Lawrence, & Foy, 1993). Using only the 17-item PTSD severity index, King et al. found a sensitivity of .74, specificity of .77, and an overall hit rate of 76% compared to a SCID diagnosis. Using a diagnostic categorization scheme, the sensitivity was .78, specificity was .82, and the overall hit rate was 80% compared to the SCID. Thus the results obtained by using either scoring scheme provide an acceptable level of precision in classifying patients with and without PTSD.


Developed by Kubany, Leisen, Kaplan, and Kelly (2000), the Distressing Event Questionnaire (DEQ) provides dichotomous and continuous information. It does not assess criterion A-1 (occurrence of the traumatic event) but has three items that assess criterion A-2 (presence of intense fear, helplessness, and horror at the time of the event) and 17 items that assess the DSM-IV (American Psychiatric Association, 1994) diagnostic symptoms of PTSD (criteria B-D). Respondents are asked to indicate, on a Likert scale, "the degree to which they experienced each of the symptoms" within the last month. Additional items focus on chronology (criterion E), distress and functional impairment (criterion F), and associated features of guilt, anger, and unresolved grief. The DEQ takes 5-7 minutes to complete.

Kubany et al. (2000) conducted a series of studies to evaluate the psychometric properties of the DEQ, and the results are excellent. Samples included male Vietnam combat veterans and women with mixed trauma histories (including incest, rape, partner abuse, prostitution, and sexual abuse). In the initial study they found high internal consistency (alpha = .93 for total score and .88-.91 across symptom clusters). In a second study they reported high test-retest reliability (.83-.94) over an average of 10 days using a variety of scoring methods. The third and largest study provided evidence for construct validity. The DEQ total scale score was highly correlated with the

CAPS (.82-.90) and Modified PTSD Symptom Scale (.86-.94; Falsetti, Resnick, Resick, & Kilpatrick, 1993). Furthermore, using the CAPS as a criterion measure, an optimal cutoff score of 26 yielded a sensitivity of .87, specificity of .85, and diagnostic efficiency of .86 in the veteran sample. For women, a cutoff score of 18 yielded a sensitivity of .98, specificity of .58, and overall efficiency of .90. A particular strength of this scale is that it was able to correctly classify PTSD in a high percentage of men and women despite differences in trauma exposure and ethnicities.

Psychophysiological Measures

Over the past 10 years, research on biologically based measures of PTSD has established a foundation for a psychobiological description of PTSD (Orr, Metzger, Miller, & Kaloupek, 2004). Much of the work found that PTSD alters a wide range of physiological functions (Yehuda, 1997), and some researchers assert that PTSD may affect structural components of the brain (particularly the hippocampus; Bremner et al., 1995). Overall, the most consistent finding in this area is that psychophysiological reactivity to trauma-specific cues is elevated in individuals with PTSD, but not in trauma-exposed individuals without PTSD (for reviews see Orr et al., 2004; Prins, Kaloupek, & Keane, 1995).

As an extension of these findings, a number of studies has attempted to identify and classify cases of PTSD on the basis of psychophysiological reactivity to trauma-related cues (Blanchard, Kolb, & Prins, 1991; Malloy, Fairbank, & Keane, 1983; Pitman, Orr, Forgue, de Jong, & Claiborne, 1987). A psychophysiological assessment usually involves presenting individuals with standardized stimuli (e.g., combat photos, noises, odors) or personalized cues (e.g., taped scripts of traumatic experiences) related to their trauma. Measurements are taken of one or more physiological indices (e.g., blood pressure, heart rate, muscle tension, or skin conductance level), subjective responses (e.g., arousal and distress), and behavior (e.g., startle response, averting gaze, crying). Because no one psychophysiological index is error-free, convergent measures are recommended.

The capacity of psychophysiological indices to identify and classify cases of PTSD on the basis of reactivity to trauma cues has been documented, with sensitivity values ranging from 60 to 90% and specificity values falling between 80 and 100% (Keane et al., 1998; Orr et al., 2004). Furthermore, these findings were replicated in individuals exposed to a range of traumatic events (e.g., motor vehicle accidents, combat, sexual assault, and terrorism).

Although psychophysiological assessment can provide unique and accurate information, widespread use of this approach is not anticipated. Psychophysiological assessment can be expensive in terms of time and patient burden, and requires specialized training on the part of the clinician. In many situations, more time- and cost-efficient methods of assessment, such as the diagnostic interview or self-report measures, are adequate. Nonetheless, in cases where diagnostic accuracy is of the utmost importance (e.g., in forensic evaluations), it may be wise to employ this assessment strategy (cf. Keane, 1995; Prins et al., 1995). The availability of portable systems to conduct this type of measurement makes this technique increasingly feasible.

Another biological system of great interest for the assessment of PTSD is the hypothalamic-pituitary-adrenocortical axis. In particular, indices of cortisol and norepinephrine, and their ratio, appear important in terms of their ability to improve assessment of PTSD, above and beyond the use of diagnostic interviews and self-report measures (Yehuda, Giller, Levengood, Southwick, & Siever, 1995). Further research will help to determine how much benefit is achieved by adding these biological measures.

Conquering Fear In The 21th Century

Conquering Fear In The 21th Century

The Ultimate Guide To Overcoming Fear And Getting Breakthroughs. Fear is without doubt among the strongest and most influential emotional responses we have, and it may act as both a protective and destructive force depending upon the situation.

Get My Free Ebook

Post a comment