Hertlein, Ricci / PLATINUM TRAUMA, VIOLENCE, & STANDARD TO ALUATE EMDR 10.

1177/1524838004264340ABUSE / July 2004 EV

ARTICLE

A SYSTEMATIC RESEARCH SYNTHESIS OF EMDR STUDIES

Implementation of the Platinum Standard
KATHERINE M. HERTLEIN RONALD J. RICCI

Virginia Polytechnic Institute and State University

Eye movement desensitization and reprocessing (EMDR) is a psychological treatment method used primarily for people who have experienced trauma. This article provides a systematic research synthesis of EMDR studies targeting trauma symptomatology published between 1997 and 2003. This synthesis builds on the Revised Gold Standard (RGS) as a guide to evaluate empirical EMDR studies. Modifications and additions to the RGS criteria are proposed. The resulting standard is referred to as the Platinum Standard (PS). Sixteen EMDR studies are reviewed and critiqued using the PS criteria. None of the studies reviewed met full PS criteria. The mean score for the studies on the PS was 8.28, with 9 of the studies exceeding the mean. The review calls for researchers to employ more rigorous research designs for EMDR effectiveness using PS criteria. Implications for practice, policy, and research are discussed.

Key words:

EMDR, empirical studies, trauma, Platinum Standard, research clients to focus on a traumatic or disturbing m e m o ry a s w e ll a s t h e a cco m pa n y in g cognitions and emotions. The therapist provides bilateral stimulation in the form of visual tracking, auditory stimulus, or tactile stimulation. Once the memory has been desensitized, as indicated by client report, the therapist g uide s t h e clie n t in re proce s s in g t h e accompanying negative cognitions, replacing them with client-generated positive cognitions. EMDR has received much attention and mixed reaction by the scientific and professional community. Initial results of EMDR effec-

EYE MOVEMENT DESENSITIZATION and reprocessing (EMDR) is a psychological treatment method developed by Francine Shapiro (Shapiro, 1989a, 1989b, 1995). Its initial and primary use is treatment for individuals who have experienced trauma. This eight-phase treatment protocol uses bilateral stimulation to allow clients to work through traumatic events with the goal of desensitizing and reprocessing memories to reduce posttraumatic stress disorder (PTSD) symptomatology (Shapiro, 1995, 2001). During the desensitization and reprocessing phases of treatment, the therapist asks

AUTHORS NOTE: The authors would like to thank Howard Protinsky, Ph.D., for the conceptualization and Lenore McWey, Ph.D., for her guidance in its completion. Please address correspondence to the first author at khertlein@yahoo.com. TRAUMA, VIOLENCE, & ABUSE, Vol. 5, No. 3, July 2004 285-300 DOI: 10.1177/1524838004264340 2004 Sage Publications 285

286

TRAUMA, VIOLENCE, & ABUSE / July 2004

tiveness studies appeared promising, and EMDR was given an effective treatment rating by the Treatment Guidelines Committee of the International Society of Traumatic Stress Studies (ISTSS) (cited in Chemtob, Tolin, van der Kolk, & Pitman, 2000). This designation came just 2 years after Chambless and Hollon (1998) reviewed studies of empirically validated treatments and reported that there were no well-established treatments for PTSD. The intent of this article is to review the latest empirical studies on EMDR as a treatment for trauma symptoms. This article does not intend to draw conclusions about the effectiveness of EMDR treatment. Its purpose, rather, is to hold the research studies up to criteria thought to assess the rigor and thoroughness of research designs intended to add to the literature on EMDR effectiveness. The analysis begins with a review of the Gold Standard (GS) (Foa & Meadows, 1997) and the Revised Gold Standard (RGS) created by Maxfield and Hyer (2002). The RGS is the currently accepted standard by which to evaluate EMDR effectiveness studies. In our application, we encountered facets of several of the research designs in empirical EMDR studies not addressed by the current model. We modified the RGS to create what we felt was a more complete tool for conducting our research review, and labeled this the Platinum Standard (PS). The PS was employed to critique empirically based EMDR effectiveness studies for trauma symptoms published between January 1997 and April 2003, commencing where DeBell and Jones (1997) review left off. Of the seven studies reviewed by DeBell and Jones (1997), the number of participants ranged from 6 to 62. Each study attempted to use EMDR on PTSD diagnoses. Two studies reported one session of EMDR treatment, two reported two sessions of EMDR, one reported four sessions, one reported six sessions, and another was not specific. Six of seven studies reported a reduction in SUDs (subjective units of disturbance), whereas the seventh did not report SUDs in the results. DeBell and Jones have suggested that more controlled research is needed to understand the impact of EMDR on PTSD symptoms. A synthesized analysis of the empirical literature follows individual study re-

KEY POINTS OF THE RESEARCH REVIEW

Standardized research designs would clarify much of the controversy surrounding outcomes from EMDR effectiveness studies. APA 12 standards offer specific guidelines for research designs to inform empirical studies, including EMDR studies. The Platinum Standard offers a more rigorous framework through which to design and/or evaluate EMDR empirical studies.

views for the purpose of assessing the legitimacy of conclusions drawn from recent EMDR research. Finally, implications for policy, research, and practice are considered. EVOLUTION OF GUIDELINES TO DETERMINE EMDR EFFECTIVENESS

Gold Standard (GS)

Several authors have identified criteria to guide evaluation of effectiveness in EMDR research. Foa and Meadows (1997) presented a GS of evaluating EMDR effectiveness studies. These criteria include (a) clearly defined target symptoms; (b) reliable and valid measures; (c) use of blind evaluators; (d) information about an assessor s training; (e) manualized, replicable, specific treatment; (f) unbiased treatment; and (g) treatment adherence. Lohr, Tolin, and Lilienfeld (1998) added to Foas and Meadowss (1997) work, expanding the list to include the criterion of comparison. Their list of criteria includes (a) clearly defined target symptoms, (b) reliable and valid measures of symptoms, (c) unbiased assignment to conditions, (d) specific treatment procedures, (e) blind assessors in applying measures, (f) adherence to treatment protocol, (g) comparing EMDR with treatment conditions controlling for nonspecific effects of treatment. Lohr et al. hold that comparing EMDR to other treatment categories is a means of giving researchers and clin icia n s s o m e in dica t ion of EMD Rs comparative effectiveness and efficiency.

Hertlein, Ricci / PLATINUM STANDARD TO EVALUATE EMDR

287

Revised Gold Standard (RGS)

Maxfield and Hyer (2002) offered three additional criterion guidelines to the Foa and Meadows (1997) GS. These additions are: (a) no conf o u n d e d co n d i t i o n s ( i . e . , co n c u r re n t psychotherapy), (b) use of multi-modal measures, and (c) length of treatment for participants. They did a preliminary analysis of the same data reviewed under the GS and found that the RGS appeared to provide a more comprehensive evaluation of the methodological strength and better differentiation among the studies (Maxfield & Hyer, 2002, p. 37), with a significant relationship between outcome and methodology. They note that confounded treatment conditions (concurrent psychotherapy) diminish construct validity, increase noise, and increase the chance of a Type II error. Multimodal measures are assumed to provide a more accurate evaluation than self-report instruments, although their analysis showed no relationship between this criterion and effect size. Finally, the addition of an adequate course of treatment was based on their own observation that multiple-trauma survivors typically require more extensive treatment. They hypothesized that an insufficient course of treatment would likely interfere with good assessment of efficacy. This criterion appeared to have a strong relationship with treatment outcome in their preliminary analysis, thereby justifying its inclusion.

The Platinum Standard

Our scale expands on the revised guidelines of Foa and Meadows (1997), Lohr et al. (1998), and Maxfield and Hyer (2002). Maxfield and Hyers three additions to the GS in conjunction with our revisions and additions result in 13 criteria for assessing research designs looking at EMDR efficacy or effectiveness. We label this the Platinum Standard (PS) and offer it as a more comprehensive evaluation standard for EMDR effectiveness studies. The PS incorporates modified ideas from Foa and Meadows (1997), Lohr et al. (1998), and Maxfield and Hyer (2002). Furthermore, it adds 3 criteria based on the American Psychological

Associations 12 guidelines and ideas forwarded by Isaac and Michael (1995). Foa and Meadows GS #5 was expanded to include a midrange score wherein the treatment is specific but does not necessarily follow standard EMDR protocol (Shapiro, 1995). We found this additional category applicable given that many studies did not fit into either of the two existing criterion originally detailed under this measure. We also revised GS #8 offered by Maxfield and Hyer (2002) concerning the assessment for confounding variables. We found differences between studies wherein confounding conditions may have been present yet were tightly controlled. This option was not accounted for in the prior version, and we felt that acknowledging and controlling for any number of confounds, as opposed to merely excluding those participants with the confound of extra-therapeutic factors (e.g., concurrent treatment), should be given additional weight in assessing the research design. Finally, we collapsed two separate ratings in Maxfield and Hyers (2002) GS #10 regarding length of treatment. The RGS assessed treatment length separately depending on singletrauma or multiple-trauma participants. What we found in the literature, however, was that researchers often failed to gather adequate history to assess for severity and frequency of traumatic events, thereby rendering this criterion unwieldy. Although there is not definitive agreement on the number of EMDR treatment sessions required (Shapiro, 1995), we chose to accept the theoretical rationale proposed by Carlson, Chemtob, Rusnak, Hedlund, and Muraoka (1998) whereby 10 sessions is in keeping with accepted protocol for traditional brief psychotherapy as well as PTSD treatment with exposure-based therapy (e.g., Keane, Fairbank, Caddell, & Zimmerman, 1989). Although extant literature supports EMDR effectiveness in fewer than 10 sessions (e.g., Vaughan et al., 1994), we felt that this number of treatment sessions would allow adequate time to test comparative treatments as well. Again, the exposure that clients had to EMDR differs significantly and should be consistent across studies for a true experimental methodology.

288

TRAUMA, VIOLENCE, & ABUSE / July 2004 TABLE 1: Platinum Standard

We added the GS #11 criterion that accounts for therapist training level, which we view as an important consideration in evaluating EMDR effectiveness. Though Davidson and Parker (2001) concluded that therapists who were trained by the EMDR Institute do not differ from therapists not trained through the institute, controlling this variable of treatment is important from a research perspective. In addition, therapist training should be reported for both the control and treatment conditions, not just restricted to the EMDR group. When the level of therapist training is inconsistent, researchers cannot make the statement that it was the treatment condition that makes a difference between two groups. Another addition to the standards is evidence of a control or comparison group. Most of the studies in EMDR have examined whether or not EMDR produces a treatment change in the PTSD symptoms of traumatized individuals. When conducting this type of research, known as treatment package (Kazdin, 1994), a basic requirement is the inclusion of a no-treatment or control group (Chambless & Hollon, 1998; Kazdin, 1994). This measure is not part of the current RGS scale but was included in Lohr et al.s (1998) criteria. We believe it is an important component and have adopted this criterion for the PS (PS #12). Finally, we added the criteria of inclusion of index of effect size in accordance with recommendations published in the Publication Manual of the American Psychological Association (American Psychological Association, 2001). This not only allows a reader to gauge the importance of findings but also facilitates meta-analytic review (PS #13). In addition, inclusion of effect size data helps a reader to make judgments about test power and sample size. Table 1 shows the first 10 items on the PS. Those items are adapted from the RGS (Maxfield & Hyer, 2002). Table 2 lists the three new additions to the RGS. As with the RGS, scores on the PS should be regarded as descriptive rather than quantitative measures (Maxfield & Hyer, 2002). Also, studies that did

PS Item
PS #1

Description

Clearly defined target symptoms 0: no clear diagnosis or symptom definition 0.5: not all participants meet target symptom criteria 1.0: all participants meet target symptom criteria PS #2 Reliable and valid measures 0: did not use reliable and valid measures 0.5: measures used inadequate to measure change 1.0: reliable, valid, and adequate measures used PS #3 Use of blind, independent assessor 0: assessor was therapist 0.5: assessor was not blind 1.0: assessor was blind and independent PS #4 Assessor reliability 0: no training in administration of instruments used in the study 0.5: training in administration of instruments used in the study 1.0: training with performance supervision, or reliability checks PS #5 Manualized, replicable, specific treatment 0: treatment was not replicable or specific 0.5: treatment replicable and specific but not standard EMDR protocol (Shapiro, 1995) 1.0: treatment followed EMDR training manual (Shapiro, 1995) PS #6 Unbiased assignment to treatment 0: assignment not randomized 0.5: only one therapist or other semi-randomized designs 1.0: unbiased assignment to treatment PS #7 Treatment adherence 0: treatment fidelity poor 0.5: treatment fidelity variable or self-monitored by therapist only 1.0: treatment fidelity independently checked and adequate PS #8 Nonconfounded conditions (e.g., concurrent psycho therapy or psychopharmacology, violent household, etc.) 0: most participants exposed to confounds with no control for variables 0.5: few participants exposed to confounds with no control for variables 1.0: confounds nonexistent or controlled for (e.g., exclusion, matched assignment, etc.) PS #9 Use of multimodal measures 0: self-report measures only 0.5: self-report plus interview of physiological or behavioral measures 1.0: self-report plus two or more other types of measures PS #10 Length of treatment 0: 1-6 sessions 0.5: 7-10 sessions 1.0: 11+ sessions SOURCE: Adapted from Maxfield and Hyers (2002) adaptation of Foa and Meadows (1997).

Hertlein, Ricci / PLATINUM STANDARD TO EVALUATE EMDR TABLE 2: Platinum Standard (PS) Scale: Additional Items Added to the RGS to Create the PS

289

PS Item
PS #11

Description

Reported level of therapist(s) training 0: no qualifications for treating clinicians provided 0.5: qualifications for treatment group, clinicians provided 1.0: qualifications for treatment and comparative group, clinicians provided PS #12 Use of control or comparison group 0: No use of a wait control/comparison group 0.5: Use of a comparison group but no control 1.0: Use of a no-treatment control group PS # 13 Effect size reporting 0: No effect size reported 1.0: Effect size reported

not report on a criterion (e.g., a number of studies provided no information on assessor training) received a score of zero, as it must be assumed that this criterion was not considered in the research design. EMPIRICAL LITERATURE Inclusion criteria for our review were all empirical studies of EMDR targeting PTSD symptoms in adults. Trauma studies were chosen given EMDRs origin as a treatment of PTSD (Shapiro, 1989b). Since DeBell and Jones (1997) published a review in 1997, we limited our review to those studies published between 1997 and April 2003. PsychINFO, MEDLINE, ScienceDIRECT.com, and Family and Society Studies Worldwide were the databases used to locate empirical studies for this research synthesis. However, we were aware of other articles that were not listed in the databases. As a result, we augmented our search with David Baldwins Trauma Pages (http://www.trauma-pages.com/). This site is periodically updated and provides an exhaustive EMDR bibliography by the year of publication. It is our belief that we have located all of the published EMDR studies within these parameters. Our search located 16 controlled studies of EMDR outcomes appearing in the English literature between 1997 and March 2003. We excluded the Macklin et al. (2000) study, as it was a

follow-up study inconsistent with our inclusion criteria. In a study by Marcus, Marquis, and Sakai (1997), EMDR was compared to standard care (SC) for PTSD. SC, however, included a variety of treatment models and was not clearly defined. The 67 participants were randomly assigned to each treatment condition. Measures include several reliable and valid measures such as the IES, MPTSD, BDI, STAI-I, STAI-S, SUD, GAF, SCL-90, GSI, SCL-90, and PSD (see the appendix for a list of inventories). Assessment was not blind, evidenced by the statement, It was not possible to keep the indep e n d e n t e va lua t o r blin d t o t re a t m e n t condition (p. 309). The authors describe adherence to Shapiros (1995) eight steps in the protocol, although this was not independently monitored via videotape or live supervision. EMDR performed significantly better than the SC group for the specific scales for PTSD (i.e., IES, MPTSD, STAI-T, portions of the STAI-S, and SUD). The authors also reported that EMDR participants required statistically fewer sessions, but the mean number of total sessions for each is unclear. For example, after three sessions, half of the participants in the EMDR group did not meet the criteria for PTSD, which was true for only 20% of the participants in the SC group. At posttreatment, about three fourths of the EMDR group (77%) no longer met the criteria for PTSD, true for only half (50%) of the SC group. Lazrove, Triffleman, Kite, McGlashan, and Rounsaville (1998) conducted a prepilot AB design study of 8 participants diagnosed with chronic PTSD. Strengths of this study include treatment fidelity by monitored, Level-II EMDR-trained therapists. Reliable and valid self-report and interview measures were used to assess the target symptomatology of chronic PTSD. Despite the power limitations of the small sample size (N = 8), results were clinically significant in that none of the 7 participants completing the three-session treatment met criteria for current PTSD based on a posttreatment structured interview. Depressive symptoms also improved significantly, which authors la-

290

TRAUMA, VIOLENCE, & ABUSE / July 2004

beled a secondary effect. A robust treatment effect was observed in only three sessions, leading the authors to assert that EMDR is a unique form of therapy rather than an alternative to exposure therapies. Effect sizes were reportedly consistent with findings in controlled studies of EMDR (Lazrove et al., 1998) Devilly, Spence, and Rapee (1998) compared EMDR to two groups: EMDR minus the eye movement component group (REDDR) and a psychiatric support control group (SPS). The purpose of the study was to determine the necessity of the eye movement component of EMDR. In a mixed-group design, 51 veterans with PTSD symptomatology completed several self-report measures. This study accounts for physiological measures such as heart rate and blood pressure. The scales used were reliable and valid. Another strength of this study was that treating therapists were all Level II trained. Though 51 participants appears to be a large sample, it was divided over three conditions. In this sense, the sample sizes were small in each condition. Participants were assigned by stratified randomization technique; however, random assignment was not complete (Devilly et al., 1998, p. 439). There were no fidelity checks reported in this study, so it is difficult to determine the extent to which the standard EMDR protocol was followed. In addition, the number of sessions was limited and might not provide an adequate judgment of effectiveness. The researchers found that all groups reported a decrease in symptoms, but at follow-up there were no lasting effects for the treatment groups. The authors concluded that eye movements were not likely a mechanism for change. Scheck, Schaeffer, and Gillette (1998) studied the efficacy of EMDR with a sample of 60 traumatized women aged 16 to 25. Though a PTSD diagnosis was not a prerequisite for inclusion, participants met criteria of at least two of eight dysfunctional behavior patterns (e.g., truancy, sexual promiscuity). Participants were randomly assigned to an EMDR treatment group or an active listening (AL) treatment group. This was a well-designed study meeting full PS criteria on 7 of 13 measures. Assessors were trained, supervised, and blind to conditions. Therapists were Level II trained and followed

specific treatment protocol, although this was determined by self-monitoring only. Therapist experience levels were provided for both treatment and control groups. The major weakness in this design was the arguably inadequate twosession treatment regime. There was also no mention of control for extra-therapeutic factors (e.g., concurrent treatment) that must be considered when interpreting the results. At posttreatment assessment, the EMDR participants were within one standard deviation of normal group means. The AL group was within one standard deviation on only one measure (STATE). Three-month follow-up was hampered both by unavailability of many participants and subsequent treatment by a number of participants; therefore, statistical comparisons were not made. Overall, the general pattern of outcome measures showed improvement for both groups, although EMDR showed pre- and posttest effect sizes approximately double those of the AL group. The authors recognized and reported a significant limitation of study with this unstable (p. 37), untrusting population. The participants might represent a more stable, more trusting, less resistant subpopulation of the high-risk population we targeted (Scheck et al., 1998, p. 37). However, their caveat speaks not to interpretation of a between-group comparison but rather to the generalization of results to those in crisis situations who have not yet been stabilized. Carlson et al. (1998) used a sample of 35 male veterans diagnosed with PTSD to explore EMDR effectiveness. The participants were randomly assigned to one of three conditions: EMDR, biofeedback-assisted relaxation, or routine clinical care. Reliable and valid measures used included the CAPS-1, CMS, IES, PTSD-SR, BDI, STAI, clinical scales of the MMPI-2, and the MAC-R subscale (see the appendix for a list of inventories). Participants were assessed at pretreatment, treatment, posttreatment, and follow-up (3- and 9-month). The assessors were independent but not blind to the interview process. One design strength was the inclusion of psychophysiological measures in the assessment. Another strength was the length of treatment (12 sessions). Although the treating therapists were formerly trained in the EMDR

Hertlein, Ricci / PLATINUM STANDARD TO EVALUATE EMDR

291

method (p. 8), the training level (Level I or Level II) is unclear. Also unclear was the potential for confounding conditions. At follow-up, EMDR was effective when compared to the other groups. Participants in the EMDR group had statistically significantly lower scores on self-report measures of PTSD symptoms than did the biofeedback or routine care group. Devilly and Spence (1999) performed a stratified random comparison of cognitivebehavioral treatment with EMDR targeting PTSD symptoms. The researchers used the STAI-Y2, SCL-90R, BDI, PPD, CMS, PSS-SR, CEQ, DEVS-T, SUD, and IES, all reliable and valid measures of change but all self-report (see the appendix for a list of inventories). The assignment of participants to two treatment conditions was random. The symptoms for treatment were specific. The treating therapists were Level II trained, and treatment adherence was independently assessed by videotape and reported as adequate. Authors claimed the EMDR procedure followed the typical protocol outlined by Shapiro (1995). However, there were several deviations, including inaccurate instructions and repeating the negative cognition during treatment. In addition, an independent rater was present to assure that the research was following the EMDR protocol, but assessment was not blind. Inclusion/exclusion criteria dictated that participants were not currently receiving psychotherapy for PTSD, but confounding conditions were present, with half of their sample (n = 23) receiving concurrent psychopharmacological treatment. It is not reported if and how the researchers controlled for this. The researchers concluded that not only was TTP more effective initially in the treatment of PTSD but also was more durable with the effects of EMDR dissipating over time. Rogers et al. (1999) examined the effectiveness of EMDR with a small sample (N = 12) of Vietnam veterans diagnosed with PTSD. Dependent measures were subjective (IES and SUD), but the study was unique in including physiological measures such as heart rate and blood pressure. Participants were randomly assigned to conditions (EMDR vs. relaxation condition), and treatment providers were blind to the assessment data. In addition, the authors

described the treatment protocol in specific detail, which strictly adheres to standard EMDR protocol (Shapiro, 1995). Sessions were audiotaped to check for treatment fidelity, but fidelity checks were not included in the present study. The length of treatment was limited to one session. There was a significant difference in the SUD from pre- to postassessment. Again, the study relied on a small sample of 12 participants, generating limited test power. No indication of the level of therapist training was provided. Also, all but 1 participant was taking antidepressant medication, which may confound results with controls unclear. EMDR was more effective than the exposure group for reducing the severity of the intrusive memory. Cusack and Spates (1999) compared EMDR and eye-movement desensitization (EMD) as treatment methods for PTSD symptoms (N = 27). EMD is described as being similar to EMDR but without the cognitive aspects that define EMDR. The strengths of this study were the random assignment to conditions as well as use of reliable and valid measures. Another unique feature was that this study employed a more comprehensive assessment using the BASA (a behavioral assessment) and an interview (SIPTSD) in addition to self-report measures. Blind observation assessed treatment fidelity, but typical treatment protocol was not followed. Length of treatment was one to three 90-minute sessions. The study did not report control of confounding variables, nor was it a requirement for inclusion to meet criteria for PTSD. The results indicated both EMDR and EMD were effective in reducing PTSD symptoms. Edmond, Rubin, and Wambach (1999) conducted a randomized experimental evaluation with 59 women recruited from newspaper advertisements and flyers to agencies and clinicians in Texas. Volunteers with reported history of childhood sexual abuse were screened for inclusion; however, neither PTSD nor any other diagnosis was used as a prerequisite. Pretest scores did reveal symptomatology consistent with a clinical population. Participants were randomly assigned to EMDR treatment (n = 20), routine individual treatment (n = 20), and waitlist control (n = 19). The random assignment to treatment, control, and comparison groups

292

TRAUMA, VIOLENCE, & ABUSE / July 2004

were strengths of this study, as was the training level and verified treatment protocol of treating clinicians. Although the standardized outcome measures used (i.e., STAI, IES, BDI) demonstrate adequate reliability and validity, they are all self-report measures that were administered by the primary researcher. Other outcomes (i.e., SUD, VOC) are subjective measures that are part of EMDR treatment protocol. This procedure warrants caution in interpreting results given the possibility of researcher bias. The sample size of 59 may be perceived favorably relative to other studies reviewed; however, it still leaves a fairly large risk of a Type II error. The researchers report large composite effect sizes between the EMDR and routine treatment groups of 1.46 at posttest and 1.08 at 3-month follow-up. Furthermore, EMDR-treated individuals did not show clinically significant levels of trauma-specific anxiety or depression, whereas the comparison group exhibited above-average symptomatology. Researchers concluded that EMDR was more effective than routine individual treatment at maintaining therapeutic gains. Stable to larger effect sizes at follow-up is consistent with the findings of a meta-analysis conducted by Van Etten and Taylor (1998). Korn and Leeds (2002) conducted an AB design using the self-report TSI and SCL-90 as outcome measures (for a list of assessment inventories and abbreviations, see the appendix). Assessment was not blind, as the treating therapist was the assessor. The target symptoms were clearly defined, with both participants meeting PTSD criteria. Treatment procedure employed resource development and installation (RDI) as opposed to standard EMDR. RDI was described as similar to EMDR, although it uses fewer eye movements per set. Test power is considered limited by almost any standard by the small sample size (N = 2). RDI was effective, and researchers suggest that RDI has promise in increasing stability for clients with a history of trauma, but the results must be interpreted with caution due to research design limitations. AB designs are historically problematic because of the single administration of each condition, the lack of a control or comparison group, and

because they do not account for any factors related to time that might affect results. Lytle, Hazlett-Stevens, and Borkovec (2002) examined the effectiveness of EMDR using past stressful life experiences as the identified target of treatment. Their controlled study used a randomized design to compare EMDR to a fixed eye condition (ED) and a nondirective therapy condition (ND). Researchers employed a series of reliable and valid instruments, although all instruments relied on self-report and were administered by a treating therapist. Authors reported that two of the therapists underwent Level I EMDR training, and they trained the third therapist in the procedure. Therapists training and/or experience with the comparative treatment modalities were not provided. The participants (N = 48) were from a sample of college students. Diagnosis was unclear, with only partial criteria for PTSD met. Treatment was one session. The findings included effect sizes for all treatments. Their results indicated EMDR was not significantly better over the comparative treatment methods. ED showed the best performance overall. The Penn State Worry Questionnaire (PSWQ) was significantly associated with poorer outcomes on both EMD and ED but not with ND Lee, Gavriel, Drummond, Richards, and Greenwald (2002) randomly assigned 24 participants (12 in each group) to one of two conditions: EMDR or stress inoculation training with prolonged exposure (SITPE). Participants were recruited from hospital psychology clinics and sexual assault referral centers, and each participant served as his own control. Measures to assess change included SI-PTSD, MMPI-K, IES, and BDI, all reliable and valid but all self-report measures (see the appendix for a list of inventories). Participants received seven 60-minute sessions of treatment and were assessed at pretreatment, posttreatment, and 3-month followup. The researchers used behaviorally trained assessors and Level-II EMDR-trained clinicians in the study, both strengths. Another design strength was the monitoring of treatment fidelity via videotape. However, the study did not provide for a blind or independent assessor. There was also evidence of extra-therapeutic factors (e.g., potential for concurrent psycho-

Hertlein, Ricci / PLATINUM STANDARD TO EVALUATE EMDR

293

therapy and/or medications). Although both EMDR and SITPE were effective, EMDR was superior to SITPE in the measurement of intrusion. The authors noted that participants in the EMDR condition required less homework than the SITPE condition. Soberman, Greenwald, and Rule (2002) conducted a study based on their hypothesis that the development and persistence of conduct problems may originate, in part, from a history of trauma. They selected EMDR as a means of testing this hypothesis due to its promise as a considerably more efficient (p. 219) trauma treatment. Twenty-nine boys aged 10 to 16 who exhibited acting-out behaviors severe enough to result in residential or day treatment placement participated. Data sources included six self- or other-report measures of behaviors, symptoms, and distress (i.e., SUD, IES-8, CROPS, PROPS, PRS, and BRS). The BRS was an existing, system-wide rating of behaviors from which clients earn privileges. Participants were randomly assigned to treatment or control groups, with both groups continuing in standard milieu treatment. Standard treatment included individual, group, and family therapy; psychopharmacology; behavior modification; and special education services. Random assignment to a therapist was not accomplished as only one Level-II-trained clinician provided all treatment, thus weakening external validity. Furthermore, although treatment in this study did follow standardized EMDR protocol (Shapiro, 1995), the three sessions provided are considered inadequate by our review standards, nor were the procedures independently monitored. Results showed statistical significance at posttreatment on two measures (i.e., SUDS, PRS); however, 2-month follow-up results favored EMDR on all measures except the IES, which showed a change from clinical to nonclinical levels for both groups. Other results included trends favoring EMDR over control groups. Strengths of this study were the random and controlled design with a homogeneous group of boys. However, the authors recognized a significant weakness of the centralized treatment center from which participants were drawn created a system-wide interest in trauma, possibly influencing the milieu and

thus affecting nonparticipants as well. Authors also discussed the small sample size and truncated treatment design as limitations to yielding statistically significant changes across all measures. The authors concluded that the treatment holds promise, particularly in view of the continuing improvements made by the treatment group, which echoes findings in similar research studies (e.g., Greenwald, 1994; Scheck et al., 1998). They suspected that the treatment reduced reactivity in the EMDR group, thereby allowing the boys to take advantage of their ongoing treatment milieu. Power et al. (2002) conducted one of the most methodologically sound studies in this review. Patients who met criteria for PTSD (N = 105) were randomly assigned to EMDR (n = 39), exposure plus cognitive restructuring (E + CR) (n = 37), and wait list (n = 29) groups. A maximum of 10 treatment sessions were allowed, although treating clinicians felt results had been achieved in fewer sessions. In that regard, the EMDR group showed subjectively measured efficiency requiring a mean of 4.2 sessions (SD = 2.5) compared to a mean of 6.4 (SD = 3.2) sessions for the E + CR group. This study met full score criteria for 9 of 13 platinum standards. Areas ranked in the midrange were lack of blind assessment and use of multimodal assessment measures. Assessor reliability was not reported, thereby registering no score on that criterion. EMDR and E + CR performed equally well across all measures, with both showing statistical significance compared to the wait-list group. EMDR performed better on one measure of depression (HADS), with 81% of EMDR patients achieving clinically significant reductions in symptoms compared to 43% of the E + CR participants (p < .05). The other measure revealing significance was the SDS, with 70% versus 38%, respectively. However, gains in both of these measures were lost at a 15-month follow-up. The authors concluded that although each experimental treatment outperformed the wait list, the majority of patients do not achieve clinically significant long-term gains without additional treatment. The researchers concluded that both EMDR and E + CR are effective in treatment of PTSD; however, EMDR required

294

TRAUMA, VIOLENCE, & ABUSE / July 2004

fewer treatment sessions to achieve these results. Ironson, Freund, Strauss, and Williams (2002) conducted a pilot study with 22 individuals with some history of trauma, meeting Diagnostic and Statistical Manual of Mental Disorders (DSM III-R) criterion for trauma. Participants agreed to three preliminary sessions and one active treatment session of either EMDR or prolonged exposure (PE). PE treatment followed instructions for Foa and Rothbaums (1998, as cited in Ironson et al., 2002) imaginal exposure. EMDR followed the eight-phase protocol per Shapiro (1995). The therapists were doctoral students who had received at least Level I training for EMDR and were supervised by clinicians with Level II training. Therapists were trained in the PE technique by Freund. Both treatments were manualized, and fidelity checks and supervision were applied throughout. Measures were self-report (SUDS, BDI, DES, PSS-SR). Results suggest that EMDR and PE were equally effective in reducing PTSD symptoms. Results were maintained at 3month follow-up with 12 participants who could be located. Taylor et al. (2003) compared the three treatment methods (prolonged exposure therapy, relaxation training, and EMDR) for efficacy, speed, and incidence of symptom worsening for PTSD. Their study design intended to avoid the methodological limitations of prior studies examining this issue. Consequently, they followed and met all of Foa and Meadows (1997) GS for sound treatment outcome research. Inclusion criteria were a diagnosis of PTSD according to DSM-IV-TR standards, legal ability to consent to treatment, and willingness to suspend any concomitant psychological treatment and maintain entry levels of psychotropic medication over the course of the study. Qualifying participants (N = 60) were randomly assigned to one of the three treatment groups, each providing eight 90-minute individual sessions. Treatment protocols followed detailed manuals. Two female clinicians were randomly assigned to participants. Clinician qualifications and experience are clearly outlined, as are the three treatment protocols. Blind assessors administered

interviews (SCID-IV and CAPS). Interview outcomes and symptom assessment were independently rated and showed satisfactory levels of interrater reliability (i.e., .80 to .93). Participants completed self-report questionnaires before, during, and after treatment. The authors conclude,
Compared with EMDR and relaxation training, exposure therapy (a) produced significantly larger reductions in avoidance and reexperiencing symptoms, (b) tended to be faster at reducing avoidance, and (c) tended to yield a greater proportion of participants who no longer met criteria for PTSD after treatment. EMDR and relaxation did not differ from one another in speed or efficacy. (Taylor et al., 2003, p. 330)

SYNTHESIS OF THE EMPIRICAL LITERATURE We performed a systematic research synthesis of this literature to assess methodological limitations and trends in the current studies. Table 3 provides individual and total scores on the PS criteria for each of the empirical studies. Figure 1 illustrates the total PS score with studies ranked from highest to lowest.

Sample Size
Sample size is a common weakness in clinical studies. Of the 16 studies examined, sample sizes ranged from 2 to 105. In many studies, there were two to three conditions among which this sample was divided. For example, although the Power et al. (2002) study reports 105 participants, they were divided among three groups, resulting in per-cell sample sizes of between 29 and 39. Devilly et al. (1998) divided 51 participants among three conditions, resulting in 16 to 19 participants per cell. Kazdin (1994) stated that power is commensurately reduced as the comparisons entail subgroups with smaller group ns (p. 47). Researchers, therefore, must be cautious to consider test power when designing studies.

PS #1: Treatment Target

Eighty-one percent of the studies reviewed (13 of 16) met full criterion for clearly defined

TABLE 3: Platinum Standard (PS) Scores for EMDR Studies, 1997-2003 N 1.0 1.0 1.0 1.0 0.5 1.0 1.0 1.0 1.0 0.5 1.0 1.0 1.0 0.5 1.0 1.0 1.0 1.0 0.5 1.0 0 1.0 1.0 1.0 0.5 1.0 1.0 1.0 1.0 1.0 1.0 0.5 0.5 1.0 0 0 1.0 1.0 1.0 1.0 0.5 0.5 1.0 0 0 0 1.0 1.0 1.0 0.5 1.0 1.0 1.0 1.0 0.5 1.0 1.0 0 1.0 0 0 1.0 1.0 0 0 0 0.5 0.5 0.5 0.5 1.0 1.0 0 1.0 1.0 0 0 0 0.5 1.0 1.0 0.5 1.0 1.0 0.5 0.5 0 1.0 0.5 0 0 0 0 0 1.0 1.0 0.5 1.0 1.0 1.0 1.0 0 0 1.0 1.0 1.0 0.5 1.0 1.0 1.0 0 1.0 0.5 0 0 0 0 0.5 0.5 0 0 0.5 0 1.0 1.0 0 0 1.0 0.5 0.5 0 0.5 0 0.5 1.0 1.0 0 0.5 1.0 1.0 0.5 1.0 0 1.0 0.5 1.0 1.0 1.0 1.0 1.0 1.0 0.5 1.0 0.5 0 1.0 0 1.0 0.5 1.0 0.5 0.5 0.5 1.0 0 0 0.5 1.0 1.0 0 1.0 0 0.5 1.0 1.0 1.0 0.5 0.5 0.5 1.0 1.0 0.5 0 0 0.5 1.0 0.5 1.0 0.5 0.5 0.5 0.5

References

PS #1

PS #2

PS #3

PS #4

PS #5

PS #6

PS #7

PS #8

PS #9

PS #10

PS #11

PS #12

PS #13
1.0 1.0 1.0 1.0 1.0 1.0 0 1.0 1.0 1.0 0 1.0 0 1.0 0 1.0

TOTAL PS score
10.5 8.5 8.5 7.0 9.0 8.0 4.0 7.5 9.5 6.5 8.5 11.0 6.5 9.0 6.5 12.0

Carlson et al. (1998) 35 Cusack & Spates (1999) 27 Devilly & Spence (1999) 23 Devilly, Spence, & Rapee (1999) 51 Edmond, Rubin, & Wambach (1999) 59 Ironson et al. (2002) 22 Korn & Leeds (2002) 2 Lazrove et al. (1998) 8 Lee et al. (2002) 24 Lytle, Hazlett-Stevens, & Borkovec (2002) 48 Marcus, Marquis, & Sakai (1997) 67 Power et al. (2002) 105 Rogers et al. (1999) 12 Scheck, Schaeffer, & Gillette 1998) 60 Soberman, Greenwald, & Rule (2002) 29 Taylor et al. (2003) 60

295

296

TRAUMA, VIOLENCE, & ABUSE / July 2004

Platinum Scale Scores

Taylor, Thordarson & Maxfield et al. (2003) Carlson, et al. (1998) Lee et al. (2002) Power et al. (2002) Devilly & Spence (1999) Scheck, Schaeffer, & Gillette (1998) Cusack & Spates (1999) Edmond, Rubin, & Wambach (1999) Marcus, Marquis, & Sakai (1997) Devilly, Spence, & Rapee (1999) Lazrove et al. (1998) Lytle, Hazlett-Stevens, & Borkovec (2002) Rogers et al. (1999) Soberman, Greenwald, & Rule (2002) Korn & Leeds (2002)

PS #3: Assessment
10.5 10.5 10 12

6.5 6.5 6.5

8 8 7.5 7.5

9 9 8.5

0 1 2 3 4 5 6 7 8 9 10 11 12 13 Platinum Standard Score

Maxfield and Hyer (2002) found that blind assessors had no significant effect on the outcome of EMDR effectiveness, though they still correlate with outcome (r = .35). Only 6 of the studies met criteria for blind assessment (see Table 3). Despite the limited correlation found by Maxfield and Hyer, blind assessment is a valid part of any good research design (Isaac & Michael, 1995).

Empirical Studies

Figure 1: Rank Ordering of the 16 Studies According to PS Scores NOTE: The numbers to the right of each bar represent PS score totals for each study.

PS #4: Assessor Reliability

Eleven of the 16 studies failed to report any information about the characteristics of the assessors (see Table 3). Maxfield and Hyer (2002) reported that assessor reliability was one of the highest correlated factors to outcome. Authors should be more conscientious in reporting assessor qualifications.

treatment target symptomatology (see Table 3). In the 3 studies that did not, the majority of participants met most criteria. Interestingly, however, Maxfield and Hyer (2002) found a negative bivariate correlation between diagnosis and treatment outcome (r = .20). In other words, those meeting full criteria for PTSD treatment may present a more difficult symptom picture, which negatively impacts effectiveness of the treatment. Further investigation of the value of this standard is thereby warranted.

PS #5: Manualized Treatment

There is wide variability in the EMDR treatment procedures used in these studies. Lee, Gavriel, and Richards (1996) indicated that some of the studies attempting to assess EMDR effectiveness have omitted critical components of the procedure. Consistency is an important component when testing a specific treatment modality (Rosen, 1999). Fourteen of the 16 studies under review met the full criteria for specific treatment procedures according to the EMDR training manual (Shapiro, 1995). Cusack and Spates (1999) were not clear in reporting their treatment protocol. As a result, it was difficult to determine whether the study followed Shapiros (1995) protocol. Korn and Leeds (2002) used a very specific treatment (RDI); however, RDI is not standard EMDR procedure.

PS #2: Reliable and Valid Measures

Reliable and valid measures of symptoms are important in outcome research (Lambert & Hill, 1994). All of the studies reviewed used at least one reliable and valid measurement instrument with good psychometric properties (see Table 3). Though all of the studies met full PS criteria, Maxfield and Hyer (2002) noted that the instruments need to be accurate and sensitive enough to detect the amount of change that would be anticipated from the treatment (p. 34). Rogers et al. (1999) used SUD and IES. Korn and Leeds (2002) employed the TSI and the SLC-90. All of these are self-report measures and are perhaps inadequate according to Maxfield and Hyers criterion.

PS #6: Random Assignment

Random assignment to conditions is considered strong research design (Isaac & Michael, 1995). Unbiased and random assignment to conditions helps to strengthen external validity

Hertlein, Ricci / PLATINUM STANDARD TO EVALUATE EMDR

297

of the outcome. Within the studies reviewed, nine used an unbiased random assignment to conditions (see Table 3). Random assignment, however, was found to not correlate significantly with effect size (Maxfield & Hyer, 2002). Others have questioned the necessity of randomized clinical trials, as compared with correlational designs, for estimated treatment effects (Benson & Hartz, 2000; Concato, Shah, & Horwitz, 2000). Much of this debate has to do with efficacy versus effectiveness and thereby remains a critical component of research design.

PTSD (e.g., Cusack & Spates, 1999; Devilly et al., 1998; Rogers et al., 1999), particularly because many of them were referred from psychology clinics or trauma centers.

PS #9: Multimodal Measures

Maxfield and Hyer (2002) found no correlation between multimodal measurement and treatment effect, even after controlling for blind and independent assessment. It can be difficult for researchers to obtain the maximum PS score on this item because researchers rarely incorporate physiological or interview measures into their assessment procedures. However, using multiple data points helps ensure internal validity and objectivity (Anfara, Brown, & Mangione, 2002). The reviewed studies commonly relied exclusively on self-report measures such as SUD, IES, and SLC-90, which are vulnerable to relational artifacts. Four of the reviewed studies included interviews, physiological or behavioral measures, and to self-report measures (see Table 3). One study under review that did not report favorable outcome for EMDR received the lowest PS score on this item because they only used self-report assessments in determining EMDR effectiveness (Devilly & Spence, 1999).

PS #7: Treatment Fidelity

Insufficient treatment fidelity has been cited as a criticism of EMDR research studies (Shapiro, 1999). Manualized treatment studies are often plagued by therapist drift, which can be mitigated by independent monitoring. Treatment adherence was determined to meet full criteria only if it was deemed adequate by an independent monitor. Although a difficult standard to achieve, 8 of the studies reviewed met this standard (see Table 3). Of the original seven GS scale items analyzed (Maxfield & Hyer, 2002), treatment fidelity correlated most significantly with effect size (r = .79, p < .05).

PS #8: Confounding Conditions

Maxfield and Hyer (2002) defined GS #8 as no confounding conditions and found a significant one-tailed correlation to treatment outcome (p < .01). This is a difficult standard to achieve in a sample consisting of trauma survivors, and recommendations to participants to cease concurrent treatment may have ethical ramifications. Had this review held to the original RGS parameters, none of the 16 studies would have met this criterion. To correct for this, GS #8 was modified to account for studies that attempted to control for any extra-therapeutic factors, whether they be concurrent medication, concurrent psychotherapy, or external life events. Even after allowing for the adjustment, only 6 of the studies met full criteria (see Table 3). Of the 3 studies that only received partial credit, it was unclear whether participants were receiving concurrent psychotherapy for

PS #10: Length of Treatment

Applying rationale proposed by Carlson et al. (1998), 4 of the studies we reviewed provided adequate treatment length. Those studies that did (Carlson et al., 1998; Lee et al., 2002; Marcus et al., 1997; Power et al., 2002) reported positive clinical results from the application of EMDR to PTSD symptomatology. Three studies were unclear in the number of EMDR sessions administered (see Table 3). Again, Maxfield and Hyer (2002) found a significant one-tailed correlation between adequate course of treatment and effect.

PS #11: Level of Therapist Training Reported

Researchers should be cautious to provide training and/or experience levels of therapists providing treatment. In the present review, 10 of

298

TRAUMA, VIOLENCE, & ABUSE / July 2004

the 16 studies met this criterion (see Table 3). The assignment of therapist to condition might also be problematic. For example, a few studies used therapists familiar with EMDR to provide treatment to both the EMDR treatment and other treatment condition (i.e., Carlson et al., 1999). Although this decision allows the researcher to control for differences between therapists, it is unclear whether any bias exists in how the therapists treated their clients, favoring EMDR or another comparative treatment.

PS #12: Control Group

Kazdin (1994) stressed the importance of using a no-treatment control or wait-list control group as a means of understanding whether treatment produced therapeutic change. Twelve of the studies we reviewed either neglected to use a control group (e.g., Korn & Leeds, 2002) or used only a comparison treatment group (see Table 3). Comparison treatment groups are particularly problematic when participants are drawn from a treatment milieu (e.g., Soberman, Greenwald, & Rule, 2002), as there is the possibility of system-wide influence. This also might be problematic if participants view the study as a competition (e.g., classroom studies), which can affect outcome. Other pitfalls of neglecting to use a control group include hypothesis guessing, evaluation apprehension, and experimenter expectancies (Isaac & Michael, 1995).

searchers find it useful in designing future research, we recognize that it may not be adaptable to studies unlike those in this review. We invite researchers to use the scale and to provide us with feedback that might improve its adaptability. We recommend that researchers adopt this, or a similarly comprehensive standard, as a systematic means to research uniformity for purposes of forwarding knowledge about treatment effectiveness. It is our belief that continued systematic efforts to critically review EMDR effectiveness studies in a format such as the PS would carry both science and clinical practice forward. IMPLICATIONS FOR PRACTICE, POLICY, AND RESEARCH
Practitioners should remain aware of emerging research in EMDR. Researchers should consider the PS Scale criteria to shape their EMDR studies. Researchers should clearly spell out symptom target, treatment procedure, and protocol so that the clinical community can extrapolate practice techniques found to be effective. Research should follow American Psychological Association Publication Manual (5th ed.) recommendations of reporting effect size to allow the reader to fully understand the importance of findings and/or facilitate meta-analyses that will continue to contribute to the scientific research base. Authors should report the training level of all therapists involved in research studies, not only for the target treatment (e.g., EMDR) but also for the comparative treatment (e.g., relaxation training).

PS #13: Effect Size

Effect size is one way that researchers can compare the strength of treatment from one study to another (Kazdin, 1994). Of the 16 studies under investigation, 12 reported effect sizes for their analyses (see Table 3). Within these effect sizes, researchers can make comparisons across studies about the effectiveness of a particular intervention. DISCUSSION The PS is designed as a comprehensive assessment tool specifically tailored to empirical studies of EMDR. Although we hope that re-

APPENDIX
BASA = Behavioral Assessment of Speech Anxiety BDI = Beck Depression Inventory BRS = Behavior Rewards Scale CAPS = Clinician Administered PTSD Scale CEQ = Credibility/Expectancy Questionnaire CMS = Civilian Mississippi Scale for PTSD COT = Credibility of Therapy Questionnaire CROPS = Child Report of Posttraumatic Symptoms DAST = Drug Abuse Screening Test DES = Dissociative Experience Scale DEVS-T = Distress Evaluation Scale for Treatment GAF = Global Assessment Functioning GSI = Global Severity Index HADS = Hospital Anxiety and Depression Scale IES = Impact Event Scale MACR-1 = McAndrew Alcoholism Revised

Hertlein, Ricci / PLATINUM STANDARD TO EVALUATE EMDR

299

MMPI-K = Keanes PTSD Scale from the Minnesota Multiphasic Personality Inventory MPTSD = Modified PTSD Scale PPD = Personal Problem Definition Questionnaire PROPS = Parent Report of Posttraumatic Symptoms PSD = Positive Symptom Distress Scale PSS-SR = PTSD Symptom ScaleSelf-Report PSWQ = Penn State Worry Questionnaire SI-PTSD = Structured Interview for PTSD SCL-90(R) = Symptom Checklist (Revised) STAI-T/S = State-Trait Anxiety Inventory SUD(S) = Subjective Units of Disturbance TSI = Trauma Symptom Inventory VOC = Validity of Cognition

REFERENCES
American Psychological Association. (2001). Publication manual of the American Psychological Association (5th ed.). Washington, DC: Author. Anfara, V. A., Brown, K. M., & Mangione, T. L. (2002). Qualitative research: Making the research process more public. Educational Researcher, 31(8), 28-38. Benson, K., & Hartz, A. J. (2000). A comparison of observational studies and randomized, controlled trials. New England Journal of Medicine, 342(25), 1878-1886. Carlson, J., Chemtob, C., Rusnak, K., Hedlund, N., & Muraoka, M. (1998). Eye movement desensitization and reprocessing (EMDR) treatment for combat-related posttraumatic stress disorder. Journal of Traumatic Stress, 11(1), 3-24. Chambless, D. L., & Hollon, S. D. (1998). Defining empirically supported therapies. Journal of Consulting and Clinical Psychology, 66(1), 7-18. Chemtob, C. M., Tolin, D. F., van der Kolk, B. A., & Pitman, R. K. (2000). Eye movement desensitization and reprocessing. In E. B. Foa, T. M. Keane, & M. Friedman (Eds.), Effective treatments for PTSD: Practice guidelines from the International Society for Traumatic Stress studies (pp. 333335). New York: International Society for Traumatic Stress. Concato, J., Shah, N., & Horwitz, R. I. (2000). Randomized, controlled trials, observational studies, and the hierarchy of research designs. New England Journal of Medicine, 342(25), 1887-1892. Cusack, K., & Spates, R. (1999). The cognitive dismantling of eye movement desensitization and reprocessing (EMDR) treatment of posttraumatic stress disorder (PTSD). Journal of Anxiety Disorders, 13(1-2), 87-99. Davidson, P. R., & Parker, K. C. H. (2001). Eye movement desensitization and reprocessing (EMDR): A metaanalysis. Journal of Consulting and Clinical Psychology, 69(2), 305-316. DeBell, C., & Jones, D. (1997). As good as it seems? A review of EMDR experimental research. Professional Psychology: Research and Practice, 28(2), 153-163.

Devilly, G., & Spence, S. H. (1999). The relative efficacy and treatment distress of EMDR and a cognitive-behavior trauma treatment protocol in the amelioration of posttraumatic stress disorder. Journal of Anxiety Disorders, 13(1-2), 131-157. Devilly, G., Spence, S. H., & Rapee, R. (1998). Statistical and reliable change with eye movement desensitization and reprocessing: Treating trauma within a veteran population. Behavior Therapy, 29(3), 435-455. Edmond, T., Rubin, A., & Wambach, K. (1999). The effectiveness of EMDR with adult female survivors of childhood sexual abuse. Social Work Research, 23(2), 103-116. Foa, E. B., & Meadows, E. A. (1997). Psychosocial treatments for posttraumatic stress disorder: A critical review. Annual Review of Psychology, 48, 449-480. Greenwald, R. (1994). Applying eye movement desensitization and reprocessing (EMDR) to the treatment of traumatized children: Five case studies. Anxiety Disorders Practice Journal, 1(2), 83-97. Ironson, G., Freund, B., Strauss, J. L., & Williams, J. (2002). Comparison of two treatments for traumatic stress: A community-based study of EMDR and prolonged exposure. Journal of Clinical Psychology, 58(1), 113-128. Isaac, S., & Michael, W. B. (1995). Handbook in research and evaluation (3rd ed.). San Diego, CA: Educational and Industrial Testing Services. Kazdin, A. E. (1994). Methodology, design, and evaluation in psychotherapy research. In A. E. Bergin & S. L. Garfield (Eds.), Handbook of psychotherapy and behavior change (4th ed., pp. 19-71). New York: John Wiley. Keane, T. M., Fairbank, J. A., Caddell, J. M., & Zimmerman, R. T. (1989). Implosive (flooding) therapy reduces symptoms of PTSD in Vietnam combat veterans. Behaviour Therapy, 20,245-260. Korn, D., & Leeds, A. (2002). Preliminary evidence of efficacy for EMDR resource development and installation in the stabilization phase of treatment of complex posttraumatic stress disorder. Journal of Clinical Psychology, 58(12), 1465-1487. Lambert, M. J., & Hill, C. E. (1994). Assessing psychotherapy outcomes and processes. In A. E. Bergin & S. L. Garfield (Eds.), Handbook of psychotherapy and behavior change (4th ed., pp. 72-113). New York: John Wiley. Lazrove, S., Triffleman, E., Kite, L., McGlashan, T., & Rounsaville, B. (1998). An open trial for EMDR as treatment for chronic PTSD. American Journal of Orthopsychiatry, 68(4), 601-608. Lee, C., Gavriel, H., & Richards, J. (1996). Eye movement desensitization: Past research, complexities, and future direction. Australian Psychologist, 31, 168-173. Lee, C., Gavriel, H., Drummond, P., Richards, J., & Greenwald, R. (2002). Treatment of PTSD: Stress inoculation training with prolonged exposure compared to EMDR. Journal of Clinical Psychology, 58(9), 1071-1089. Lohr, J., Tolin, D., & Lilienfeld, S. (1998). Efficacy of eye movement desensitization and reprocessing: Implications for behavior therapy. Behavior Therapy, 29(1), 123156.

300

TRAUMA, VIOLENCE, & ABUSE / July 2004

Lytle, R. A., Hazlett-Stevens, H., & Borkovec, T. D. (2002). Efficacy of eye movement desensitization in the treatment of cognitive intrusions related to a past cognitive event. Journal of Anxiety Disorders, 16, 273-288. Macklin, M., Metzger, L., Lasko, N., Berry, N., Orr, S., & Pitman, R. (2000). Five-year follow-up study of eye movement desensitization and reprocessing therapy for combat-related posttraumatic stress disorder. Comprehensive Psychiatry, 41(1), 24-27. Marcus, S., Marquis, P., & Sakai, C. (1997). Controlled study of treatment of PTSD using EMDR in an HMO setting. Psychotherapy, 34(3), 307-315. Maxfield, L., & Hyer, L. (2002). The relationship between efficacy and methodology in studies investigating EMDR treatment of PTSD. Journal of Clinical Psychology, 58(1), 23-41. Power, K., McGoldrick, T., Brown, K., Buchanan, R., Sharp, D., Swanson, V., et al. (2002). A controlled comparison of eye movement desensitization and reprocessing versus exposure plus cognitive restructuring versus waiting list in the treatment of post-traumatic stress disorder. Clinical Psychology and Psychotherapy, 9(5), 299-318. Rogers, S., Silver, S., Goss, J., Obenchain, J., Willis, A., & Whitney, R. (1999). A single session, group study of exposure and eye movement desensitization and reprocessing in treating posttraumatic stress disorder among Vietnam War veterans: Preliminary data. Journal of Anxiety Disorders, 13(1-2), 119-130. Rosen, G. (1999). Treatment fidelity and research on eye movement desensitization and reprocessing (EMDR). Journal of Anxiety Disorders, 13(1-2), 173-184. Scheck, M., Schaeffer, J., & Gillette, C. (1998). Brief psychological intervention with traumatized young women: The efficacy of eye movement desensitization and reprocessing. Journal of Traumatic Stress, 11(1), 25-44. Shapiro, F. (1989a). Eye movement desensitization: A new treatment for post-traumatic stress disorder. Journal of Behavior Therapy and Experimental Psychiatry, 20, 211217. Shapiro, F. (1989b). Efficacy of the eye movement desensitization procedure in the treatment of traumatic memories. Journal of Traumatic Stress, 2, 199-223. Shapiro, F. (1995). EMDR: Basic principles, protocols, and procedures. New York: Guilford. Shapiro, F. (1999). Eye movement desensitization and reprocessing (EMDR) and the anxiety disorders: Clinical and research implications of an integrated psychotherapy treatment. Journal of Anxiety Disorders, 13(1-2), 35-67. Shapiro, F. (2001). EMDR: Basic principles, protocols, and procedures (2nd ed.). New York: Guilford. Soberman, G., Greenwald, R., & Rule, D. (2002). A controlled study of eye movement desensitization and reprocessing (EMDR) for boys with conduct problems. Journal of Aggression, Maltreatment and Trauma, 6(1), 217236. Taylor, S., Thordarson, D., Maxfield, L., Federoff, I., Lovell, K., & Ogrodniczuk, J. (2003). Comparative efficacy, speed, and adverse effects of three PTSD treatments:

Exposure therapy, EMDR, and relaxation training. Journal of Consulting and Clinical Psychology, 71(2), 330-338. Van Etten, M., & Taylor, S. (1998). Comparative efficacy of treatments for posttraumatic stress disorder: A metaanalysis. Clinical Psychology and Psychotherapy, 5, 126144. Vaughan, K., Armstrong, M., Gold, R., OConnor, N., Jenneke, W., & Tarrier, N. (1994). A trial of eye movement desensitization compared to image habituation training and applied muscle relaxation in post-traumatic stress disorder. Journal of Behavior Therapy and Experimental Psychiatry, 25(4), 283-291.

SUGGESTED FUTURE READINGS

Chambless, D. L., & Hollon, S. D. (1998). Defining empirically supported therapies. Journal of Consulting and Clinical Psychology, 66(1), 7-18. Concato, J., Shah, N., & Horwitz, R. I. (2000). Randomized, controlled trials, observational studies, and the hierarchy of research designs. New England Journal of Medicine, 342(25), 1887-1892. DeBell, C., & Jones, D. (1997). As good as it seems? A review of EMDR experimental research. Professional Psychology: Research and Practice, 28(2), 153-163. Maxfield, L., & Hyer, L. (2002). The relationship between efficacy and methodology in studies investigating EMDR treatment of PTSD. Journal of Clinical Psychology, 58(1), 23-41.

Katherine M. Hertlein, M.S., is a doctoral candidate in the Marriage and Family Therapy (MFT) Program at Virginia Tech. She received her masters in MFT from Purdue University Calumet. She has published in several journals including Journal of Couple and Relationship Therapy, American Journal of Family Therapy, and Contemporary Family Therapy. She is currently co-editing a book on clinical interventions for couple and family therapists and another on infidelity treatment. Her areas of interest include research methodology and measurement, training in marriage and family therapy, infidelity treatment, and Bowen family systems theory. Ronald J. Ricci, M.S., M.A., is a doctoral candidate in the Marriage and Family Therapy (MFT) Program at Virginia Tech. He is a licensed marriage and family therapist in Virginia and a clinical member of AAMFT. He is a Level-II trained EMDR practitioner and has used EMDR in his clinical practice. His most recent research focus has been using EMDR for trauma resolution with sex offenders. He has published in Contemporary Family Therapy and The Qualitative Report.