Adoption of evidence-based practices (EBP) is an essential step in improving the quality of mental health services (McHugh and Barlow 2010). Unfortunately, such treatments are not widely available in many service settings. Multiple factors are involved in the successful delivery and adoption of EBPs (Aarons et al. 2012; Sanders and Turner 2005; Simpson 2002), including clinicians’ characteristics, such as demographics (e.g., age, education, years of experience) and EBP attitudes, which can affect implementation effectiveness (Beidas et al. 2014; Carpenter et al. 2012; Henggeler et al. 2008). Organizational factors including pressure for change, organizational climate, infrastructure support, and perceived fit between organizational practices and an EBP, can influence implementation (Asgary-Eden and Lee 2012; Garner et al. 2012; Simpson 2002; Zazzali et al. 2008). Thus, implementation success may be dependent on a combination of clinician and organizational factors (Asgary-Eden and Lee 2012; Moore 2002).
Clinician fidelity (the level of adherence and competence in delivering a protocol), is a key implementation outcome (Proctor et al. 2011) and EBPs delivered with a higher degree of fidelity are associated with better treatment outcomes (Farmer et al. 2017; Holder et al. 2017; Jerrell and Ridgely 1999; McHugo et al. 1999; Strunk et al. 2010a, b). Moreover, clinician fidelity measured in early treatment sessions is predictive of patient outcomes, highlighting the importance of early fidelity (Forsberg et al. 2015; Henggeler et al. 2008; Thijssen et al. 2017). Therefore, elucidating novel organizational and clinician factors that predict early treatment fidelity may inform efforts to identify clinicians in need of more or less post-workshop support to efficiently allocate limited resources. This paper explored clinician factors and clinician perceptions of organizational factors related to early fidelity (fidelity measured in one of the first three sessions in one of the first cases clinicians treated) to Cognitive Processing Therapy (CPT; Resick et al. 2016), an EBP for posttraumatic stress disorder (PTSD), in a randomized controlled implementation trial.
Anzeige
Clinician Factors
Prior studies of clinician demographics as predictors of implementation outcomes have yielded inconsistent results (Campbell et al. 2013; Carpenter et al. 2012; Henggeler et al. 2008; Meier et al. 2015). Level of education has been found to be positively related to fidelity in some studies (Campbell et al. 2013; Carpenter et al. 2012), negatively related to fidelity in other studies (Meier et al. 2015) and mixed directionality has also been found in examining years of clinical experience and fidelity (Campbell et al. 2013; Carpenter et al. 2012). Several studies have found non-significant associations with these variables and fidelity (Garner et al. 2012; Henggeler et al. 2008; James et al. 2001; Schoenwald et al. 2005; Whitaker et al. 2012). Notably, only a small subset of participants (1–4%) in most of these study samples had doctoral degrees (Garner et al. 2012; Henggeler et al. 2008; Schoenwald et al. 2005). Interestingly, a study by James et al. (2001) demonstrated more experience using cognitive therapy predicted higher competence, despite no association with general clinical experience. The association between clinician age and fidelity has also been mixed with most research demonstrating no significant relationship (James et al., 2001; Rodriguez 2016; Schoenwald et al. 2005; Whitaker et al., 2012), while some studies have found age to be negatively associated with fidelity (Henggeler et al. 2008; Whitaker et al. 2012). In light of contradictory findings, it is important to examine the relationship among multiple clinician factors and fidelity.
Clinician attitudes toward EBPs is conceptualized by Aarons et al. (2012) as a multifactorial process. Clinicians with open attitudes toward EBPs are more likely to implement EBPs (Aarons 2005; Beidas et al. 2015). However, divergent attitudes (e.g., the extent to which clinicians perceive EBPs to differ from the way they typically practice) are negatively associated with fidelity (Beidas et al. 2014). Furthermore, clinician willingness to adopt EBPs if required to by their employer has been positively associated with fidelity (Beidas et al. 2014). Fidelity monitoring is also linked to higher fidelity (Aarons et al. 2009; Swain et al. 2010). Similar to monitoring, receiving session feedback from a supervisor is related to adherence, and may also effectively increase clinician competence (Carpenter et al. 2012; Swain et al. 2010; Lu et al. 2014). The perceived fit between the prescribed EBP and the clinician needs is also a determinant of EBP adoption (Zazzali et al. 2008) and may influence the level of EBP adherence. EBP limitations, such as perceiving the treatment to be unsuitable for particular clients, may be a barrier to implementation (Lewis and Simons 2011; Godley et al. 2001) and perceived EBP limitations may impact fidelity. Attitudes toward limitations of EBPs and their fit with clinicians’ practice context appear to be important for EBP adherence, but these factors have not been examined with clinician competence. Overall, few studies have simultaneously examined diverse attitudinal facets with fidelity.
Organizational Factors
Clinicians are situated within organizations, and organizational factors also influence clinicians’ EBP fidelity. Organizational factors include an organization’s motivational needs and pressures for change (e.g., staff perception of needing more training using technology or new equipment) and these motivational needs are associated with greater clinician fidelity (Henggeler et al. 2008). Organizational climate includes clarity of mission and goals, staff cohesion, clinical autonomy, methods of communication, staff stress, and openness to change. More positive climates have been shown to be associated with higher adherence to treatment protocols (Asgary-Eden and Lee 2012; Beidas et al. 2014). Availability of resources (i.e., staffing, internet, equipment availability) has been identified as an important factor in the implementation of various EBPs (Asgary-Eden and Lee 2012; Tooley 2012; Zazzali et al. 2008). Higher clinician-reported resources significantly predicted MI skill immediately following training (Tooley 2012). Staff attributes within organizations such as high job satisfaction and room for professional growth were associated with greater adherence (Schoenwald et al. 2009). Lastly, organizational fit with the prescribed EBP is another key variable. If there is a lack of fit, clinicians might make changes to the EBP or the organization may need to make a change (Zazzali et al. 2008) to prevent organizational challenges (e.g., Stirman et al. 2017). As most studies have only examined either adherence or competence, it is important to examine these organizational factors in relation to multiple components of fidelity to better understand these associations.
Present Study
Research on trauma-focused treatments has mainly focused on how variables of clinician attitudes and organizational factors relate to EBP usage. However, associations between clinician and organizational factors and EBP fidelity have only been examined with a limited number of EBPs and rarely focus on observational measures of fidelity, or early fidelity after a workshop and prior to substantial consultation, which may indicate whether and how specific factors are associated with clinicians’ ability to take in information during initial training and apply it to their own clients. No prior studies have examined these variables in clinicians delivering CPT for PTSD. The present study examined potential clinician factors and clinician perceptions of organizational factors that may predict early fidelity in clinicians delivering CPT (Resick et al. 2016), in a randomized controlled implementation trial. It was hypothesized that both clinician factors and their perceptions of organizational factors would be associated with early fidelity to CPT. Given the scant and sometimes inconsistent findings in the literature in this area, no further specific hypotheses were made.
Anzeige
Method
Participants and Procedure
The study was conducted in accordance with approved procedures from the institutional ethics boards of all institutions involved in the research. Clinicians were recruited at the individual level, rather than the organizational clinic level. Clinicians came from 14 different clinic settings (e.g., hospitals, VA Canada Operational Stress Injury Clinic) and there were a number of clinicians from various private practice settings. However, some clinicians were from the same clinic setting (the number of clinicians per clinic ranged from 1 to 12). Participants were 78 mental health clinicians recruited from across Canada, who submitted baseline survey data and at least one audio-recorded CPT session at the beginning of a larger randomized controlled implementation trial examining different post-workshop training strategies (Monson et al. 2018). All clinicians received a 2-day standard CPT training workshop provided by the fourth author, the CPT manual with related materials, and access to resources through the online CPT-web training program (https://cpt.musc.edu). The workshop entailed session-by-session review of the treatment, demonstrations of cases using video, role-plays, and discussion of how to resolve common clinical problems in CPT. Although some clinicians responded to advertisements for the workshops, others attended at the request of their employers. Post-workshop, clinicians were randomly assigned to one of three consultation conditions: (1) standard consultation involving discussion and conceptualization of cases without session audio review; (2) consultation including audio review, which included a review of segments of audio-recorded CPT sessions; and, (3) no consultation with delayed feedback on CPT fidelity.
Clinicians were eligible to participate in the study if they: (1) attended a CPT training workshop; (2) were currently providing or were interested in providing CPT to clients with PTSD; (3) consented to be randomized to one of three post-workshop consultation study conditions; and, (4) were willing to solicit client participation. At baseline, clinicians completed self-report measures to assess individual and perceived organizational factors. Clinicians enrolled variable numbers of patients, ranging from one to five. Clinicians were required to audio-record sessions for at least two patients who agreed to participate in the study, for the purpose of fidelity ratings. Available recorded sessions from patients were randomly selected to be rated for fidelity; therefore, all 12 sessions of CPT were not rated for fidelity. Because the goal of the present study was to examine predictors of early fidelity, the earliest fidelity rating available within sessions one to three was selected for each clinician, and this session occurred prior to them having received extensive consultation on their cases. Table 1 provides demographic information on the participants. Clinicians were mostly female, and approximately half had a Master’s degree or equivalent, while nearly half had a doctoral degree. Nearly half of the clinicians had received some form of supervised training in CBT prior to study participation.
Table 1
Clinician participant demographics
Characteristic | n (%)a |
|---|---|
Age (M [SD]) | 47.8 (9.8) |
Sex | |
Female | 58 (74.4) |
Male | 20 (25.6) |
Educational degree type | |
BA, BSW MA/MS, MSW, RPN | 42 (53.8) |
MD, PhD, PsyD, eDd | 36 (46.2) |
Years of experience as mental health professional | |
1–9 years | 25 (32.1) |
10 + years | 53 (67.9) |
CBT experience | |
Low experience (i.e., little to no knowledge, training, or supervision) | 36 (46.2) |
High experience (i.e., several courses and/or undergone supervised training or consultation) | 42 (53.8) |
Practice setting type | |
VA Canada Operational Stress Injury Clinic | 15 (19.2) |
Canadian Forces Operational Trauma and Stress Support Clinic | 10 (12.8) |
Other federal service | 4 (5.1) |
Provincial health agency/facility | 21 (26.9) |
Private practice | 22 (28.2) |
Other | 6 (7.7) |
All clinicians provided written informed consent to participate in the study. Clinicians recruited patients from their routine practice settings and obtained informed consent verbally from them in order to help protect privacy and confidentiality. Clinicians attested to having sought verbal consent. Patients were eligible to participate if they: (1) had a current PTSD diagnosis; (2) were willing to complete symptom outcome measures and have their CPT sessions audio-recorded and reviewed by study personnel; and (3) be at least 18 years old. Exclusion criteria included: (1) an uncontrolled psychotic or bipolar disorder; (2) substance dependence requiring daily use or medical detoxification; (3) imminent suicide or homicide risk; (4) and cognitive impairment that prevented therapy engagement. Descriptive statistics were calculated for the 64 patients that completed demographic questionnaire measures. On average, patients were 41 years old, had 13 years of education, and a PTSD Checklist session one score of 62. Additionally, 50% of the patient sample was male, 91% was white.
Measures
The measures used in the present study included the Evidence-Based Practice Attitudes Scale-50 (EBPAS-50; Aarons et al. 2012) to measure individual clinician attitudes toward EBPs, the Dimensions of Organizational Readiness-Revised (DOOR-R; Hoagwood et al. 2003) and the Organizational Readiness to Change Questionnaire (ORC; Lehman et al. 2002) to measure perceived organizational factors, and lastly the Therapist Adherence and Competence (TAC) Rating Form (Nishith and Resick 1994; Stirman et al. 2013) to measure early fidelity.
Evidence-Based Practice Attitudes Scale-50 (EBPAS-50; Aarons et al. 2012)
The EBPAS-50 is a 50-item self-report questionnaire that examines clinicians’ attitudes toward EBPs. The EBPAS-50 has a total of 12 subscales that all have moderate to large factor loadings and fair to excellent internal consistency (Aarons et al. 2012). The present study examined seven of the 12 subscales of the EBPAS-50 (openness, divergence, requirements, monitoring, feedback, limitations, and fit) for theoretical reasons. Several subscales (i.e., appeal, balance, burden, job security) were excluded given that there was no evidence suggesting their association with fidelity. Additionally, the organizational support scale was already captured by another measure in the present study and was excluded. The EBPAS-50 demonstrates good internal consistency among the subscales (Cronbach’s alpha = .77 − .92; Aarons et al. 2012). Cronbach’s alphas in the present study ranged from .64 to .92.
Perceptions of Organizational Factors
Organizational variables may be measured at the individual level (e.g., perceptions of the psychological impact of the work environment on an individual’s well-being) or the at the organizational level, in which commonly held individual perceptions are aggregated (Glick 1985; James et al. 2008; Jones and James 1979). In this study, because clinicians were recruited for participation at the individual, rather than at an organization level, we examined their individual perceptions of different organizational attributes. This approach has been used in prior research that has examined predictors of clinician skill in delivering EBPs (Beidas et al. 2014). We selected measures that have been used in previous research with EBPs for mental health that assessed a variety of constructs that may be associated with the use of, and fidelity to, EBPs.
Dimensions of Organizational Readiness-Revised (DOOR-R; Hoagwood et al. 2003)
The DOOR-R is a 21-item self-report measure that examines perspectives on both intra- and extra-organizational implementation readiness that are important to successful implementation of mental health services. The DOOR-R comprises three dimensions: fit with implementation practices, infrastructure support, and organizational mission and support. Cronbach’s alphas in the present study ranged from .70 to .92.
Organizational Readiness to Change Questionnaire (ORC; Lehman et al. 2002)
The ORC is a 129-item self-report questionnaire designed to assess different elements of perceived support related to training implementation for clinicians. The ORC has 18 subscales (Cronbach’s alphas = .56–.92) that fall under four dimensions of readiness: motivation/need for change, adequacy of resources, staff attributes (e.g., job satisfaction, adaptability, perceived influence on others within the organization, agency support for staff’s professional growth), and organizational climate. Constructs measured may appear as individual factors (e.g., staff attributes) but are conceptualized as an organizational variables as these can impact or be influenced by organizational functioning (Lehman et al. 2002). Prior studies have found strong psychometric properties for this measure, and support the concurrent and convergent validity of the ORC (Lehman et al. 2002; Saldana et al. 2007). Cronbach’s alpha in the present study ranged from .85 to .95.
Therapist Adherence and Competence (TAC) Rating Form (Nishith and Resick 1994; Stirman et al. 2013)
A modified version of the TAC, which was employed in previous clinical trials of CPT (e.g., Resick et al. 2008), was used in the present study to measure fidelity of audio-recorded sessions. This measure examined both clinician adherence to, and competence in, delivering prescribed CPT elements in each therapy session. Clinician adherence was rated on a 0–3 scale, with 0 indicating not done at all, 1 indicating partially completed, 2 indicating mostly completed, and 3 indicating completed. Clinician competence was rated on a 7-point Likert scale, from very poor (0) to exemplary (6). The TAC has demonstrated excellent inter-rater agreement (ranging from 92 to 100%) on adherence and competence in previous research (Resick et al. 2002, 2008).
TAC fidelity ratings were conducted by expert-trained graduate students and post-doctoral fellows who were trained in the CPT manual, completed a web-based CPT training, and received 6 months of training on the use of the TAC. The raters were trained until there was 90% inter-rater reliability agreement in ratings of recordings before independent study rating began. Raters remained blind to consultation condition, time spent in consultation, and clients’ outcomes when rating sessions. To assess inter-rater reliability, raters all evaluated 12.5% of the sample of sessions. Intra-class correlations of these ratings indicated a high level of agreement for adherence (.87) and competence (.79).
Data Analytic Strategy
As some clinicians were nested within the same clinic, we examined the variance accounted for by clinic level in adherence and competence and found that rho = .00. This indicated that clinic setting did not account for variance in outcomes, and as a result we did not run multilevel analyses. Missing data were handled using expectation maximization (EM), a form of maximum likelihood estimation that is known for providing unbiased parameter estimates and improving statistical power (Enders 2001; Scheffer 2002). Missing data were found on the EBPAS-50 requirements, EBPAS-50 limitations, all DOOR-R subscales, all ORC subscales, patient age, patient years of education, and session one PTSD Checklist scores. Prior to utilizing the EM method, randomness of missing data was examined using Little’s MCAR test (Little 1988). Little’s MCAR tests supported the assumption that the data were missing completely at random for all missing items and thus, data were imputed using EM. As clinicians were from different organizations and few clinicians were nested within the same clinic, organizational factors were measured as individual clinician perceptions of organizations, rather than aggregating data across clinicians from the same clinic. Means and standard deviations of all variables of interest are listed in Table 2.
Table 2
Means and standard deviations for variables of interest
Variable | Mean (SD) | Minimum | Maximum |
|---|---|---|---|
Adherence | 2.46 (.44) | 1.20 | 3.00 |
Competence | 2.96 (.83) | .60 | 4.80 |
EBPAS-50 | |||
Openness | 2.93 (.59) | 1.50 | 4.00 |
Divergence | .85 (.57) | 0 | 2.75 |
Requirements | 2.48 (.92) | 0 | 4.00 |
Monitoring | 1.07 (.92) | 0 | 4.00 |
Feedback | 3.06 (.78) | 1.00 | 4.00 |
Limitations | .45 (.54) | 0 | 2.29 |
Fit | 3.15 (.57) | 1.86 | 4.00 |
DOOR-R | |||
Fit with implementation practices | 4.59 (1.47) | 1.00 | 7.00 |
Infrastructure support | 4.14 (1.09) | 1.63 | 6.50 |
Organizational mission and support | 5.56 (.91) | 3.00 | 7.00 |
ORC | |||
Motivational needs | 24.67 (7.40) | 10.00 | 41.35 |
Resources | 37.11 (4.97) | 25.00 | 47.94 |
Staff attributes | 39.71 (4.73) | 25.43 | 49.67 |
Organizational climate | 34.41 (5.00) | 18.89 | 44.17 |
First, to evaluate whether consultation condition needed to be included as a covariate in the models, two ANOVAs were conducted to see if consultation condition was predictive of early fidelity. There was no statistically significant difference between the three conditions on TAC adherence (F(1,77) = .12, p = .84) or TAC competence (F(1,77) = 2.36, p = .10), and therefore this variable was not included in the models. To examine associations between the clinician and perceived organizational-level constructs and fidelity, bivariate correlations among TAC adherence and competence with all independent variables were conducted. Assumptions of multiple regression (e.g., multicollinearity) were tested and all were met except TAC adherence, which was not normally distributed. Bootstrapping is a common approach for handling non-normal data (e.g., Sainani 2012), thus, regressions were bootstrapped using 1000 samples with 95% confidence intervals. Variables that were significantly associated with TAC adherence were entered into a regression model as predictors. In a separate model, variables that were significantly associated with TAC competence were entered into a regression model as predictors. Only significant variables were included in the regression models to avoid overfitting (Shmueli 2010). Partial correlation coefficients (pr) were used as measures of an effect size, with 10, .24, and .37 representing small, medium, and large effect sizes, respectively (Kirk 1996).
Results
As shown in Table 3, correlation analyses indicated that two clinician-level factors, EBPAS-50 openness and requirements subscales, were positively associated with TAC adherence and competence. In addition, educational degree was positively associated with competence, indicating that having a doctoral degree was associated with greater competence. No perceptions of organizational factors were associated with TAC adherence or competence.
Table 3
Correlations between independent and dependent variables
Adherence | Competence | |
|---|---|---|
Age | .02 | .01 |
Educational degree type | .12 | .38** |
Years of experience | − .18 | − .16 |
CBT experience | .09 | .12 |
EBPAS-50 openness | .25* | .32** |
EBPAS-50 divergence | − .1 | − .14 |
EBPAS-50 requirements | .27* | .31** |
EBPAS-50 monitoring | .03 | − .11 |
EBPAS-50 feedback | .07 | .17 |
EBPAS-50 limitations | − .06 | − .07 |
EBPAS-50 fit | .15 | .14 |
DOOR-R fit | − .1 | − .02 |
DOOR-R infrastructure support | .01 | − .05 |
DOOR-R organizational mission | − .21 | − .13 |
ORC motivational needs | − .1 | .1 |
ORC resources | − .12 | − .02 |
ORC staff attributes | .03 | .09 |
ORC organizational climate | − .02 | − .01 |
Multiple regression analyses (see Table 4) revealed that the model including EBPAS-50 openness and requirements accounted for 11% of the variability in TAC adherence; EBPAS-50 requirements significantly predicted TAC adherence, though EBPAS-50 openness did not. Effect sizes for these variables were small. Doctoral degree, EBPAS-50 openness, and EBPAS-50 requirements accounted for 25% of the variance in the model predicting TAC competence. Doctoral degree and EBPAS-50 requirements both significantly predicted TAC competence and the effect sizes for these variables were medium. EBPAS-50 openness did not significantly predict competence and the effect size for this variable was small.
Table 4
Multiple regression models
B | SE | p | 95% CI | pr | Model Significance | |
|---|---|---|---|---|---|---|
Outcome: adherence | R2 = .11 F(2,75) = 4.53* | |||||
EBPAS-50 openness | .141 | .081 | .078 | − .009, .309 | .189 | |
EBPAS-50 requirements | .106 | .045 | .024 | .022, .202 | .221 | |
Outcome: Competence | R2 = .25 F(3,74) = 8.09** | |||||
Educational degree type | .509 | .174 | .003 | .160, .837 | .327 | |
EBPAS-50 openness | .281 | .155 | .070 | − .012, .621 | .212 | |
EBPAS-50 requirements | .192 | .087 | .031 | .027, .375 | .228 |
Anzeige
Discussion
The present study examined whether clinician and perceived organizational factors predicted early fidelity in clinicians delivering CPT for PTSD in a randomized controlled implementation trial. Results indicated that the EBPAS-50 requirements subscale emerged as a significant predictor of early adherence and competence. Although the EBPAS-50 openness subscale was significantly correlated with fidelity at the bivariate level, it was not a significant predictor of fidelity in our final analyses. Furthermore, educational degree type with regard to having a doctoral degree was a significant predictor of clinician competence in delivering the treatment, but not clinician adherence. Interestingly, no organizational factors were associated with early fidelity.
With regard to clinician attitudes, the EBPAS-50 requirements subscale significantly predicted both clinician adherence and competence. This subscale measures the extent to which clinicians would adopt an EBP if it were required by an agency, supervisor, or state (Aarons et al. 2012). The positive association indicated that increasing likelihood of adopting an EBP if it was required was associated with delivering CPT with higher fidelity in early sessions. In addition, the EBPAS-50 openness subscale (the extent to which the provider is generally open to trying new interventions or would be willing to try or use new types of therapy) was significantly associated with fidelity in bivariate correlations, but did not emerge as significant predictor of fidelity.
The findings from the present study are in some regards consistent with the results from the study conducted by Beidas et al. (2014), which found that openness to EBPs was not associated with fidelity. Although in the present study the requirements subscale (willingness to adopt EBPs if required to do so) was positively associated with fidelity, Beidas et al. (2014) found that this variable was negatively associated with skill in using CBT for youth anxiety in interviews with a simulated client. The difference in directionality of this finding may be explained by study differences. Beidas et al. (2014) argued that a reason for this finding could be because clinicians were not required to implement the treatment and thus, may have been less motivated to deliver the treatment with high fidelity. Clinicians in the present study had an opportunity to receive a Quality-rated CPT Provider status, if the threshold for satisfactory competence was achieved. This in turn could have motivated clinicians to deliver the treatment with greater fidelity, and may have influenced attitudes regarding requirements of implement EBPs if required to do. Moreover, Beidas et al. (2014) assessed fidelity through a simulated 8-min role play using research assistants who were hired to act as anxious youth, and competence was rated using a single item. Past literature has found that there is low correspondence between fidelity that is measured through role-play and real patient sessions (Decker et al. 2011). As few studies have examined attitudes regarding EBP requirements in relation to fidelity, the current study represents an advance in methodology, but further research in this area is warranted before drawing definitive conclusions.
Only one clinician demographic factor emerged as a significant predictor of competence: having a doctoral degree. However, degree type was not associated with adherence, which suggests clinicians can deliver prescribed elements of CPT regardless of their educational background. Clinicians who hold doctoral degrees may deliver CPT with greater skill in their early attempts to provide the treatment to clients, compared with clinicians with different educational backgrounds. Perhaps the more advanced training and additional years of supervised practica and internship provided to individuals during their doctoral degrees better equips them to deliver EBPs skillfully early on. Prior studies have also found a significant relationship between educational degree type and fidelity (Campbell et al. 2013; Carpenter et al. 2012). However, given that much of the therapy provided in routine care settings is delivered by clinicians who do not have doctoral degrees, these findings may be more indicative that consultation will be necessary to support fidelity in routine care. Previous studies have shown that clinicians in routine practice can achieve high levels of competence with consultation (Creed et al. 2016).
Anzeige
Other clinician demographics that were examined including age, years of experience as a mental health professional, and years of prior CBT experience were not associated with early fidelity. Although some studies have found these to be related to fidelity (e.g., Campbell et al. 2013; Carpenter et al. 2012; Henggeler et al. 2008; James et al. 2001), mixed results in this area may be explained by differences in methodology, client population, interventions delivered, and types of clinicians implementing the interventions. With regard to methodology, fidelity assessment varied across studies (e.g., Carpenter et al. 2012; Henggeler et al. 2008; Schoenwald et al. 2005), with different measures and types of raters conducting fidelity assessments. Some studies only examined one aspect of fidelity, such adherence alone (e.g., Rodriguez 2016), and most did not include observer assessment of an actual therapy session. Differences in the samples of clinicians who participated in these studies should also be considered.
No associations were found between perceived organizational factors and early fidelity. Thus, our hypotheses that the DOOR-R and ORC subscales would predict early fidelity were not supported. It should be noted that there are other measures of organizational constructs such as the Organizational Climate Measure (Patterson et al. 2005) and the Implementation Climate Scale (Ehrhart et al. 2014). It is possible that organizational measures used in previous research (e.g., Glisson and Green 2011; Aarons and Sawitzky 2006) captured aspects of organizational climate, culture, or functioning that we did not capture in the present study, and that these factors may be uniquely related to fidelity. The way organizational factors were measured in the present study could have also contributed to the null findings such that organizational factors were measured as individual clinician perceptions of their organizations, rather than aggregated across clinicians from specific organizations. Becker-Haimes et al. (2019) suggest that there may interaction effects between individual and organizational factors in predicting implementation outcomes, such that the extent to which organizational factors influence clinical practice may be dependent on clinicians attitudes toward EBPs. Thus, examining multilevel interaction effects is a fruitful avenue for future research. Nevertheless, our findings add to a growing body of literature that suggests that organizational characteristics may not be predictive of clinicians’ ability to deliver EBPs soon after participating in training (Creed et al. 2016; Kolko et al. 2012; Stirman et al. 2015). It is possible that organizational characteristics may be more related to EBP reach (Sayer et al. 2017) or long-term sustainment of EBP implementation rather than early treatment fidelity (Glisson et al. 2010; Kolko et al. 2012), particularly if perceptions of EBP fit with the organization leads clinicians to make significant adaptations to the protocol over time (c.f. Cook et al. 2014).
The present study has a number of strengths. Our study employed independent expert observer raters to assess fidelity of randomly selected early sessions, which occurred before the clinicians received substantial consultation on their cases. Additionally, the present study used a variety of clinician factors (both demographics and certain subscales of the EBPAS-50) as well as organizational measures, including the ORC and the DOOR-R. However, a limitation of the present study was a relatively small sample size that was largely characterized by females, and an inability to generalize beyond workshop participants who agreed to participate in a study about consultation, which may be a more motivated population. Additionally, as the focus of the paper was specific to CPT training, our results may not be more generalizable to other EBPs, particularly those with protocols that require more autonomy in selecting appropriate interventions at each session (e.g., Creed et al. 2016). Additional limitations include only a single session rating of fidelity was used, only self-report measures of clinician and organizational constructs were utilized, only early fidelity was examined, and organizational measures were based on individual clinician perceptions of their organizations. Future studies should consider these limitations. Nevertheless, this study contributes to the growing literature on clinician and organizational factors and suggests that certain clinician factors are predictive of fidelity to early therapy sessions in CPT.
Compliance with Ethical Standards
Conflict of interest
Dr. Monson receives book royalties from publication of the Cognitive Processing Therapy Comprehensive Manual (Resick et al. 2016). All other authors declare no conflicts of interest.
Anzeige
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.