Despite robust evidence for efficacy of measurement-based care (MBC) in behavioral health care, studies suggest that adoption of MBC is limited in practice. A survey from Blue Cross–Blue Shield of North Carolina was sent to behavioral health care providers (BHCPs) about their use of MBC, beliefs about MBC, and perceived barriers to its adoption.

Methods:

The authors distributed the survey by using professional networks and snowball sampling. Provider and clinical practice characteristics were collected. Numerical indices of barriers to MBC use were created. Ordered logistic regression models were used to identify associations among practice and provider characteristics, barriers to MBC use, and level of MBC use.

Results:

Of the 922 eligible BHCPs who completed the survey, 426 (46%) reported using MBC with at least half of their patients. Providers were more likely to report MBC use if they were part of a large group practice, had MBC training, had more weekly care hours, or practiced in nonmetropolitan settings. Physicians, self-reported generalists, more experienced providers, and those who did not accept insurance were less likely to report MBC use. Low perceived clinical utility was the barrier most strongly associated with less frequent use of MBC.

Conclusions:

Although evidence exists for efficacy of MBC in behavioral health care, less than half of BHCPs reported using MBC with at least half of their patients, and low perceived clinical utility of MBC was strongly associated with lower MBC use. Implementation strategies that attempt to change negative attitudes toward MBC may effectively target this barrier to use.

HIGHLIGHTS

The authors conducted a survey of behavioral health care providers to understand attitudes toward and current use of measurement-based care (MBC), with a focus on barriers to use.
Only 46% (N=426 of 922) of respondents reported using MBC with at least half of their patients, and 24% (N=224) reported not using MBC at all, although most providers agreed that MBC had some clinical utility.
Providers who were part of a large group practice and had received training in MBC were more likely to use MBC.
Low perceived clinical utility was most strongly associated with less frequent use of MBC.

Measurement-based care (MBC) is defined as the “systematic evaluation of patient symptoms before or during an encounter to inform behavioral health treatment” (1). The adoption of MBC in clinical practice improves patient outcomes (2) and helps agencies achieve accountability requirements for demonstrating care quality and outcomes (3). There are calls in the field to increase the use of MBC (4).

Despite this evidence, adoption of MBC is limited in practice (5). A study from the U.S. Department of Veterans Affairs (VA) found that only 58% of VA providers reported collecting at least one measure for at least half of their patients, despite positive provider attitudes about MBC (6). Community behavioral health care providers (BHCPs) use MBC even less. One national survey of 504 clinicians showed that only 14% reported using standardized progress measures at least monthly, and 62% never used them (7). A survey of 314 psychiatrists revealed that more than 80% indicated that they did not routinely use scales to monitor outcomes when treating depression (8). However, data are limited on MBC utilization among community BHCPs.

In this study, we analyzed data from a survey administered by Blue Cross–Blue Shield of North Carolina (BCBSNC), the largest health plan in the state, as part of an effort to improve the quality of behavioral health care. The survey asked BHCPs in North Carolina about their current use of MBC as well as their demographic and clinical practice characteristics. The survey also assessed multiple domains thought to affect use of MBC, which we call “barriers to use” (9). The aims of the analysis were twofold: to estimate associations between provider and practice characteristics and MBC use and to quantify the relative importance of barriers to use. An assessment of the most common and influential barriers to adoption of MBC among community BHCPs will help payers, health systems, and providers develop implementation strategies to increase adoption of MBC and ultimately improve the overall quality of behavioral health services.

Methods

Survey Instrument

Two of the authors (B.C.K. and I.P.B.) developed a 30-item survey for BHCPs about provider and practice characteristics, self-reported use of MBC, and barriers to its use; they then fielded the survey during February and March 2021. Cognitive pretesting was conducted with five BHCPs to ensure that the survey items were interpreted as intended (10).

Information on clinical and demographic characteristics was collected: sex included male, female, and nonbinary; race-ethnicity included White; Black; Latinx; Asian American, Pacific Islander, Native American, or other; and “prefer not to specify.” Providers could select multiple races and ethnicities. Although we acknowledge that race is a social construct, we expected that provider race-ethnicity may be correlated with unobserved caseload and regional factors. Questions also assessed clinical characteristics, including practice type, professional licensure, years of experience, treatment modalities provided, specialty training, and practice zip code.

MBC was defined in the survey as the systematic administration of symptom rating scales to assist the provider and patient in noting progress and to drive clinical decision making. Respondents were asked to indicate whether they used MBC with 0%, 1%–49%, 50%, 51%–99%, or 100% of their caseload. The survey contained a 16-question instrument designed to measure provider-level barriers to use of MBC (9), and it was adjusted as needed after cognitive pretesting (a table showing theoretical barriers to the use of MBC and the associated survey questions is available in the online supplement to this article) (6, 9). Each of the 16 questions was designed a priori to assess one of five provider-level barriers: attitudes, knowledge and self-efficacy, lack of clarity on the clinical utility of MBC, administrative burden, and concern with data utilization (i.e., whether data will be used to judge providers’ clinical skills or to affect compensation) (9). The barrier constructs also operationalize distinct aspects of a provider’s orientation toward MBC: attitudes represent an evaluative association held toward MBC, self-efficacy involves an assessment of one’s performance while using MBC, and concerns about data use are related to one’s clinical skills potentially being negatively assessed in comparison with injunctive social norms (11–13). Each question was scored on a 5-point Likert scale ranging from 1, strongly disagree, to 5, strongly agree. Secondary analysis of the survey was granted an exemption by the institutional review board of University of North Carolina at Chapel Hill.

Sample

The survey was e-mailed to the BCBSNC behavioral health Listserv of 5,255 providers as well as to leaders of local provider professional societies. Because recipients may have forwarded the survey and the number of recipients in each professional society was unknown, we were unable to calculate the response rate.

Coding and Imputing Responses

We used providers’ primary practice zip code, mapped to the 2013 U.S. Department of Agriculture Economic Research Service Urban Influence Codes (14), to explore the association between their practice Urban Influence Code and their responses. The 12 Urban Influence Codes were collapsed into three categories: large metropolitan areas (≥1 million), small metropolitan areas (<1 million residents), and nonmetropolitan areas (14).

We excluded respondents who reported more than 70 hours per week of outpatient clinical care (N=5). We imputed missing values for the 16 questions regarding beliefs about MBC, including responses of “no experience/not applicable,” through multivariate normal multiple imputations, with practice and provider characteristics as predictors (15). Values were binned into five Likert categories on the basis of absolute distance. We performed a sensitivity analysis by using listwise deletion of records with a missing value for any barrier index (N=30).

Barrier Index Construction

We developed barrier indices with the 16 questions described earlier. We examined correlation matrices between the responses to each of the 16 questions. The correlation pattern was largely consistent with the assignment of the 16 questions to the five barriers described in a previous review (9). However, items regarding “attitudes” and “lack of clarity on the clinical utility” were highly correlated with each other and relatively uncorrelated with other items (9). Therefore, we merged all questions designed to measure these two constructs into a single construct, described as “low perceived clinical utility.” The index for each of the four constructs was the average response to questions included in the given barrier construct, coding “strongly disagree” as 1 and “strongly agree” as 5 and reversing this coding as necessary. The indices had internal validity in item analysis (16) and showed good reliability overall, as measured by Cronbach’s alpha (see online supplement) (17). We calculated the means of the barrier indices to assess how strongly providers agreed, on average, with the component statements in the construct, deemed “endorsement” of the barrier. In the regression models, we standardized each barrier index to have a mean of 0 and an SD of 1 to facilitate comparison.

Data Analysis

In the first analysis, we used an ordered logit model to examine the relationship between practice and provider characteristics and frequency of MBC use. The predictor variables were sex, race-ethnicity, practice Urban Influence Code, years in the field, weekly hours of outpatient behavioral health care, whether the provider bills insurance, insurance distribution of the provider’s caseload, practice type (individual, small group [<10 providers], large group [≥10 providers], or facility based), licensure, specialty training, treatment modalities provided, and whether the provider’s training included instruction in MBC. Average marginal effects (AMEs) were calculated for each variable of interest, and they represent the percentage-point difference in the probability of use of MBC associated with a 10-unit increase (for continuous variables) or change of factor level (for categorical variables). We note values with p<0.10 in tables because of the relatively modest sample size for this model; for brevity, we focus our comments on those characteristics with the larger and statistically significant effects (p<0.05).

In the second analysis, we assessed the relative magnitude of association between the barrier indices and the use of MBC. We estimated an ordered logit model for each standardized barrier index, controlling for provider and practice characteristics. The resulting AMEs represent the percentage-point difference per SD increase of each barrier index. In a secondary analysis, we estimated a single-ordered logit model with all four barrier indices and the full set of provider and practice characteristics to explore whether the barriers mediated associations between these characteristics and use of MBC.

R, version 4.1.3, was used for initial data management (18), and the R package ggplot2 was used for producing figures (19). Stata, version 16.1, was used for all statistical procedures (20). The Stata user-contributed module MIMRGNS was used to estimate AMEs after multiple imputation (21).

Results

A total of 940 respondents completed the survey, with 922 eligible for inclusion. Respondents were excluded if they practiced outside North Carolina, reported weekly clinical care hours exceeding possible limits (e.g., 450 and 1,000 hours), indicated they were not health care providers or not practicing, or indicated they were responding to the survey for their agency in aggregate. Most identified as female (79%), White (72%), or master’s-level (80%) providers. Providers practiced in a variety of settings. Respondents practiced for a mean±SD of 25.1±10.7 hours per week. Table 1 outlines additional provider characteristics.

TABLE 1. Provider and practice characteristics of the 922 eligible behavioral health care providers who completed the measurement-based care survey^a

Characteristic	N	%
Sex
Female	728	79
Male	173	19
Nonbinary	21	2
Race-ethnicity^b
White	664	72
Black	124	13
Latinx	20	2
Asian American, Pacific Islander, Native American, or other race	32	4
Prefer not to specify	99	11
N of years in behavioral health field
<5	59	6
5–10	237	26
11–20	312	34
>20	314	34
Hours per week of outpatient clinical care (M±SD)^c	25.1±10.7
Practice location (Urban Influence Codes)
Large metropolitan	366	40
Small metropolitan	455	49
Nonmetropolitan	101	11
Practice type^b
Solo practice	646	70
Small group (<10 providers)	209	23
Large group (≥10 providers)	79	9
Facility based	14	2
Licensure^b
Master’s	740	80
Advanced practice provider (NP or PA)	20	2
Ph.D. or Psy.D.	154	17
M.D. or D.O.	32	4
Other	18	2
Specialty training^b
Generalist	409	44
Trauma	504	55
Substance use disorders	202	22
Child and adolescent	328	36
Geriatric	84	9
Mood disorders	515	56
Anxiety disorders	647	70
Women’s mental health	356	39
Serious mental illness	140	15
Other	162	18
Treatment modalities provided^b
Medication management	69	8
Medication-assisted treatment (e.g., to treat opioid use disorder)	22	2
Cognitive-behavioral therapy	777	84
Dialectical behavior therapy	276	30
Acceptance and commitment therapy	274	30
Psychodynamic psychotherapy	389	42
Other	323	35
Bill insurance for insured patients?
Yes, all insurance types	510	55
No, only certain insurance types	393	43
No, I do not bill insurance for any of my patients	19	2
Estimated percentage of outpatient caseload with insurance type (M±SD)^c
Medicaid	9±21
Medicare	6±14
Blue Cross–Blue Shield	38±21
Other private	19±19
Pays out of pocket	15±19
State health plan	10±12
TRICARE	3±10

^aNP, nurse practitioner; PA, physician assistant; Psy.D., doctor of psychology; D.O., doctor of osteopathic medicine.

^bFor these questions, respondents could select all that applied to them, so the total percentage may be greater than 100%.

^cVariables are continuous.

TABLE 1. Provider and practice characteristics of the 922 eligible behavioral health care providers who completed the measurement-based care survey^a

Enlarge table

A total of 224 providers (24%) reported never using MBC, 272 providers (30%) reported using MBC with 1%–49% of patients, 138 providers (15%) reported using MBC with 50% of patients, 171 providers (19%) reported using MBC with 51%–99% of patients, and 117 providers (13%) reported using MBC with all patients (see online supplement).

Multivariate analyses showed several factors associated with varying levels of MBC use, although few differences were found by race-ethnicity, sex, licensure type, or insurance billing (Table 2). Providers trained in MBC reported greater use of MBC than those without MBC training (p<0.05). Longer time in practice (≥10 years) was associated with substantially lower frequency of MBC use (p<0.05), even after analyses controlled for MBC training. A practice located in a nonmetropolitan area was associated with a higher probability of providers’ use of MBC than was a practice located in a large metropolitan area (p<0.05). A large practice group was associated with an 11.0–percentage point reduction in no MBC use and with corresponding increases in the probability of using MBC with 50% and 51%–99% of patients (p<0.05), compared with non–large group practices.

TABLE 2. AMEs (in percentage points) of provider and practice characteristics on the probability of the 922 behavioral health care providers using MBC with a given percentage of their caseload^a

Predictor	AME on percentage of caseload with which MBC was used					Overall OR
Predictor	0%	1%–49%	50%	51%–99%	100%	Overall OR
Sex
Female (reference)	—	—	—	—	—	1.00
Male	3.9	1.0	−.7	−2.0	−2.2	.79
Nonbinary	6.7	1.4†	−1.3	−3.2	−3.5	.67
Race-ethnicity^b
White	2.2	.7	−.4	−1.1	−1.4	.87
Black	−10.0†	−4.8	1.2*	5.7	8.0	2.04
Latinx	2.8	.7	−.5	−1.4	−1.6	.84
Asian American, Pacific Islander, Native American, or other race	−8.2	−3.7	.9*	4.5	6.6	1.79
Prefer not to specify	.5	.1	−.1	−.3	−.3	.97
N of years in behavioral health field
<5 (reference)	—	—	—	—	—	1.00
5–10	6.5†	3.5	−.6†	−3.7†	−5.8	.61†
11–20	9.7*	4.6*	−1.1*	−5.4*	−7.8*	.50*
>20	11.3*	4.9*	−1.4*	−6.2*	−8.6*	.46*
Hours per week of outpatient clinical care^c	−3.5*	−1.0*	.6*	1.8*	2.2*	1.25*
Have had training in MBC (reference: no training)	−12.9*	−2.9*	2.6*	6.5*	6.7*	2.15*
Practice location
Large metropolitan (reference)	—	—	—	—	—	1.00
Small metropolitan	−1.2	−.3	.2	.6	.7	1.08
Nonmetropolitan	−10.8*	−5.0*	1.2*	6.0*	8.6*	2.16*
Practice type^b
Solo practice	.9	.3	−.2	−.5	−.6	.94
Small group (<10 providers)	−5.0	−1.7	.8	2.6	3.3	1.39
Large group (≥10 providers)	−11.0*	−5.4	1.1*	6.0*	9.3	2.23†
Facility based	−7.5	−3.3	.9*	4.1	5.9	1.70
Licensure^b
Master’s	1.5	.4	−.2	−.8	−.9	.91
Advanced practice provider (NP or PA)	5.9	1.1	−1.1	−2.8	−3.0	.70
Ph.D. or Psy.D.	7.1	1.5	−1.4	−3.5	−3.7	.65
M.D. or D.O.	17.1	.8	−3.7	−7.3†	−6.9*	.39†
Other	−1.0	−.3	.2	.5	.7	1.07
Specialty training^b
Generalist	4.8*	1.4*	−.8*	−2.5*	−2.8*	.74*
Trauma	4.1†	1.2†	−.7†	−2.1†	−2.5†	.77†
Substance use disorders	−5.5*	−2.0†	.8*	2.9*	3.7*	1.44*
Child and adolescent	−3.4	−1.1	.6	1.8	2.1	1.25
Geriatric	2.4	.6	−.4	−1.2	−1.4	.86
Mood disorders	−4.5†	−1.3†	.8	2.3†	2.7†	1.33†
Anxiety disorders	.0	.0	.0	.0	.0	1.00
Women’s mental health	.8	.2	−.1	−.4	−.5	.95
Serious mental illness	2.4	.6	−.4	−1.2	−1.4	.86
Other	.3	.1	−.1	−.1	−.2	.98
Treatment modalities provided^b
Medication management	−.6	−.2	.1	.3	.3	1.04
Medication-assisted treatment (e.g., to treat opioid use disorder)	−10.0†	−5.0	.9*	5.5†	8.7	2.10
Cognitive-behavioral therapy	−2.4	−.6	.4	1.2	1.4	1.16
Dialectical behavior therapy	−1.7	−.5	.3	.9	1.0	1.11
Acceptance and commitment therapy	−7.0*	−2.4*	1.1*	3.7*	4.7*	1.58*
Psychodynamic psychotherapy	3.9†	1.1†	−.7†	−2.0†	−2.4†	.78†
Other	−1.0	−.3	.2	.5	.6	1.07
Bill insurance for insured patients?
Yes, all insurance types (reference)	—	—	—	—	—	1.00
No, only certain insurance types	3.3	1.0	−.6	−1.7	−2.0	.81
No, I do not bill insurance for any patients	17.1	1.5	−3.6	−7.6†	−7.3*	.38†
Estimated percentage of outpatient caseload with insurance type^c
Medicaid	−.3	−.1	.1	.2	.2	1.02
Medicare	.9	.2	−.1	−.4	−.5	.95
Blue Cross–Blue Shield	.9	.3	−.2	−.5	−.5	.95
Other private	−.5	−.1	.1	.2	.3	1.03
Pays out of pocket	.9	.3	−.2	−.5	−.5	.95
State health plan	.4	.1	−.1	−.2	−.3	.97
TRICARE (omitted because of collinearity)	—	—	—	—	—	—

^aAME, average marginal effect; MBC, measurement-based care; NP, nurse practitioner; PA, physician assistant; Psy.D., doctor of psychology; D.O., doctor of osteopathic medicine.

^bThe comparison group is the absence of the listed category (e.g., non-White).

^cContinuous variable, with AME expressed as percentage-point change in probability per each 10 additional units of the independent variable (e.g., 10 additional weekly clinical care hours or an increase of 10 percentage points in caseload with specific type of insurance).

†p<0.10, *p<0.05.

TABLE 2. AMEs (in percentage points) of provider and practice characteristics on the probability of the 922 behavioral health care providers using MBC with a given percentage of their caseload^a

Enlarge table

Generalists were less likely to use MBC than nongeneralists (p<0.05), and substance use disorder specialists were more likely to use MBC (p<0.05). Providers who offered acceptance and commitment therapy were more likely to use MBC than those who did not offer this service (p<0.05). Providers who do not bill insurance reported lower use of MBC than those who do (p<0.05).

More than half of providers agreed or strongly agreed that research supports use of MBC in practice (54%, N=497), that MBC is useful for tracking symptoms (60%, N=553), that MBC is useful for making treatment decisions (50%, N=463), and that providers are not compensated for MBC (53%, N=489) (Figure 1). Most providers strongly disagreed or disagreed with the following statements: “I don’t think standardized measures are useful” (53%, N=490), “I don’t know what MBC is” (80%, N=738), “I don’t know how to use MBC in my practice” (70%, N=641), “I have trouble interpreting the results” (72%, N=660), and “I do not have an EHR [electronic health record] or platform to collect the results” (55%, N=503). Together, these results suggest that most providers acknowledge the usefulness of MBC and disagree with several proposed barriers to MBC.

FIGURE 1. A heat map of the distributions of answers to the 16 survey questions regarding the 922 survey respondents’ perceptions of measurement-based care (MBC)^a
^aDarker values represent a higher percentage of responses. The response items are grouped by conceptual barrier to MBC use (right y-axis), and the questions are included verbatim (left y-axis). EHR, electronic health record.

More than one-third of providers strongly disagreed or disagreed (42%, N=385) that current measures do not suit patients’ needs and complexity, whereas approximately one-third of providers strongly agreed or agreed (34%, N=316). Providers had a similar lack of consensus about MBC being too time-consuming (33% [N=300] strongly disagreed or disagreed vs. 35% [N=323] strongly agreed or agreed), trouble incorporating measures into their workflow (39% [N=361] strongly disagreed or disagreed vs. 35% [N=321] strongly agreed or agreed), and concern that patients will not complete measures (34% [N=310] strongly disagreed or disagreed vs. 43% [N=393] strongly agreed or agreed).

Data use concern was the most strongly endorsed barrier to MBC (mean index value=3.24, 95% CI=3.16–3.32, α=0.76) (Figure 2; also see online supplement), followed by administrative burden (mean index value=3.09, 95% CI=3.04–3.15, α=0.69), low perceived clinical utility (mean index value=2.62, 95% CI=2.56–2.67, α=0.87), and lack of knowledge and self-efficacy (mean index value=2.02, 95% CI=1.97–2.07, α=0.76). The indices were uncorrelated with each other (correlation coefficient=0.286–0.546), suggesting that they measure distinct barriers to MBC.

Analyses of the association between the barriers and use of MBC, adjusted for the full set of characteristics, showed that low perceived clinical utility was the barrier associated with the largest decrease in use of MBC: a 1-SD increase was associated with an increase of 16.7 percentage points in reported use of MBC with 0% of the caseload, followed by 5.3, −2.5, −8.4, and −11.2 percentage points for the other levels of utilization (p<0.05) (Figure 3; also see online supplement). This barrier was followed by (in order of magnitude) administrative burden, lack of knowledge and self-efficacy, and data use concerns. All four barriers were associated with statistically significant decreases in use of MBC (p<0.001). Listwise deletion instead of multiple imputation returned nearly identical results (see online supplement), likely because of the relatively small number of missing values (N=30).

FIGURE 3. Average marginal effect of each barrier index on the probability of the 922 behavioral health care providers using measurement-based care (MBC) with a given percentage of their caseload^a
^aEstimates are shown as the height of the bars and were derived from ordered logit models (one for each barrier index) adjusted for provider and practice characteristics reported in Tables 1 and 2 (see Methods section for details on predictors). The error bars represent the 95% CIs. All p<0.001.

Finally, we found that including the barriers slightly mediated the association between provider and practice characteristics and use of MBC, with slight attenuation of most AMEs, although largely without changes in statistical significance (see online supplement). However, generalist training and substance use disorder training were no longer associated with a decrease and an increase in MBC use, respectively. In this model containing all four barrier constructs, data use concerns were no longer associated with changes in MBC use.

Discussion

This survey of BHCPs in North Carolina demonstrates that fewer than half of providers use MBC with at least half of their caseload. Our results also suggest that the most prevalent concerns about MBC may not be the most influential in determining its use. Concern about data use (that MBC may be used to judge providers’ clinical skills or affect their compensation) was the most strongly endorsed barrier yet the barrier least associated with real-world reported use of MBC; the barrier associated with the greatest decrease in MBC use was low perceived clinical utility. Together, these results suggest that providers are willing to practice MBC if they perceive it to be in the best interests of their patients, even if they have concerns about how data will be used.

To our knowledge, this study is the first to evaluate utilization of MBC among independent community BHCPs. We hypothesized that community-based providers would practice MBC less regularly than providers in VA settings because of the VA’s long history of collecting behavioral health outcome measures (22). We found that whereas less than half (46%, N=426) of providers reported using MBC for at least half of their patients, providers’ reported use of MBC was greater than prior estimates of MBC use among community BHCPs (20%) (5, 7, 8).

The variation in MBC utilization by provider type may indicate a need for flexibility and heterogeneity in incentives to expand use of MBC. Doctors of medicine and doctors of osteopathic medicine were less likely to use MBC frequently, consistent with prior research (6). Providers trained in substance use disorders were more likely to use MBC, whereas those reporting generalist training were less likely to use MBC. Self-described generalists are a heterogeneous group, and their approach to patient care may be less amenable to systematic and objective assessment because MBC instruments focus on single conditions. A previous study showed that providers with cognitive-behavioral therapy (CBT) orientations held more positive attitudes toward MBC (7). Contrary to our expectations, we did not find an association between CBT and use of MBC despite MBC being a core tenet of CBT (23). Conversely, acceptance and commitment therapy, a more specialized form of therapy that also focuses on measurement, was associated with increased use of MBC.

Providers who had been in the behavioral health field for ≥10 years were less likely to frequently use MBC. It is possible that training on the use of MBC has recently become a standard practice, although providers who have been practicing for a longer time may feel an increased sense of competence and less desire to use manualized interventions and systematic measurement.

Several of the provider and practice characteristics associated with the use of MBC are nonmodifiable, but policy makers and payers should still consider that incentivizing use of MBC may lead to varied results in these distinct groups. Large group practices had higher levels of MBC use possibly because of more robust infrastructure, including EHRs and administrative staff, which may facilitate MBC uptake (24). Surprisingly, providers in nonmetropolitan areas demonstrated more frequent use of MBC, even after adjusting for covariates.

Training in MBC is a modifiable factor that is positively associated with MBC use; however, the cross-sectional nature of our study does not allow us to infer that training has a causal effect on MBC use. Providers who are already aligned with MBC principles may be more likely to seek training in its use. We did not have information on voluntary versus mandatory training.

The most strongly endorsed barrier, concerns about data use, was associated with the smallest decrease in use of MBC. This finding suggests that although providers express concerns about how MBC data will be used, these concerns do not correlate with a decrease in use of MBC.

Administrative burden was associated with a decrease in MBC use. Heavy caseloads and time restrictions limit use of evidence-based practices (25). However, MBC may streamline treatment (1). Advances in technology may decrease administrative burden by allowing patients to complete assessments in the waiting room (26). Measurement feedback systems, or health information technologies that support MBC implementation, have demonstrated that they lead to increases in MBC use (27).

This study showed that the barrier associated with the greatest decrease in MBC use was low perceived clinical utility, which is consistent with previous research (8, 28). Our study produces new estimates of the relative importance of this barrier for a sample of mostly master’s-trained clinicians in the community setting. Of note, providers generally did not feel that they lacked knowledge about MBC. Initiatives that only aim to address knowledge gaps may have less success than those attempting to understand and address providers’ concerns about the clinical utility of MBC.

Limitations of this study include low participation by doctors of medicine and doctors of osteopathic medicine (N=32) as well as by advanced practice providers (N=20). In addition, the data are from a single state, and regional variation in MBC adoption may exist. The survey response rate could not be determined because of our snowball sampling method; thus, we were not able to determine differences between respondents and nonrespondents. It is likely that selection bias favored providers who have stronger opinions about MBC and were willing to respond to a survey from an insurance company. Last, because this study had a cross-sectional design, it is impossible to infer causality.

Conclusions

MBC is essential to addressing the current quality chasm in mental health and substance use treatment (29). Despite its robust evidence base, adoption of MBC in real-world practice continues to be limited. Initiatives to increase use of MBC must address provider attitudes and logistical barriers to use while accounting for the vast heterogeneity in training and practice among behavioral health clinicians. Future research should include longitudinal studies examining how different providers respond to initiatives promoting the use of MBC.

Department of Psychiatry, New York–Presbyterian Hospital, and New York State Psychiatric Institute, Columbia University Irving Medical Center, New York City (Keepers); University of North Carolina at Chapel Hill School of Medicine, Chapel Hill (Easterly); Department of Health Policy and Management, University of North Carolina at Chapel Hill Gillings School of Global Public Health, Chapel Hill (Easterly); Blue Cross–Blue Shield of North Carolina, Durham (Dennis, Bhalla); Duke University School of Medicine, Durham, North Carolina (Dennis); College of Health Solutions and Center for Health Information and Research, Arizona State University, Tempe (Domino).

Send correspondence to Dr. Keepers ([email protected]). Dr. Keepers and Mr. Easterly contributed equally to this article and are coequal first authors.

Mr. Easterly received general support from the University of North Carolina Medical Scientist Training Program (T32-GM-008719).

Dr. Dennis is a clinical adviser for Big Health. Dr. Domino has received institutional salary support for research projects from Arnold Ventures; she receives additional support for editorial efforts from Wiley and the American Hospital Association. The other authors report no financial relationships with commercial interests.

At the time of the study, Dr. Keepers was at the University of North Carolina at Chapel Hill School of Medicine, Chapel Hill, and Blue Cross–Blue Shield of North Carolina, Durham; Dr. Domino was at the Department of Health Policy and Management and Department of Psychiatry and the Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill, Chapel Hill.

The authors acknowledge Lauren Winfrey, B.S., for her help with the manuscript and the North Carolina Nurses Association Psychiatry Council, North Carolina Psychiatric Association, Licensed Professional Counselors Association of North Carolina, North Carolina Neuropsychological Society, National Association of Social Workers–North Carolina Chapter, North Carolina Psychological Association, and all survey participants for making this research possible.

References

1. Scott K, Lewis CC: Using measurement-based care to enhance any treatment. Cogn Behav Pract 2015; 22:49–59Crossref, Medline, Google Scholar

2. Fortney JC, Unützer J, Wrenn G, et al.: A tipping point for measurement-based care. Psychiatr Serv 2017; 68:179–188Link, Google Scholar

3. Connors EH, Douglas S, Jensen-Doss A, et al.: What gets measured gets done: how mental health agencies can leverage measurement-based care for better patient care, clinician supports, and organizational goals. Adm Policy Ment Health 2021; 48:250–265Crossref, Medline, Google Scholar

4. Lambert MJ, Whipple JL, Hawkins EJ, et al.: Is it time for clinicians to routinely track patient outcome? A meta-analysis. Clin Psychol 2003; 10:288–301 Google Scholar

5. Lewis CC, Puspitasari A, Boyd MR, et al.: Implementing measurement-based care in community mental health: a description of tailored and standardized methods. BMC Res Notes 2018; 11:76Crossref, Medline, Google Scholar

6. Oslin DW, Hoff R, Mignogna J, et al.: Provider attitudes and experience with measurement-based mental health care in the VA implementation project. Psychiatr Serv 2019; 70:135–138Link, Google Scholar

7. Jensen-Doss A, Haimes EMB, Smith AM, et al.: Monitoring treatment progress and providing feedback is viewed favorably but rarely used in practice. Adm Policy Ment Health 2018; 45:48–61Crossref, Medline, Google Scholar

8. Zimmerman M, McGlinchey JB: Why don’t psychiatrists use scales to measure outcome when treating depressed patients? J Clin Psychiatry 2008; 69:1916–1919Crossref, Medline, Google Scholar

9. Lewis CC, Boyd M, Puspitasari A, et al.: Implementing measurement-based care in behavioral health: a review. JAMA Psychiatry 2019; 76:324–335Crossref, Medline, Google Scholar

10. Lenzner T, Neuert C, Otto W: GESIS Survey Guidelines: Cognitive Pretesting (Version 2.0). Mannheim, Germany, GESIS–Leibniz Institute for the Social Sciences, 2016. https://doi.org/10.15465/gesis-sg_en_010. Accessed Aug 26, 2022 Google Scholar

11. Bohner G, Dickel N: Attitudes and attitude change. Annu Rev Psychol 2011; 62:391–417Crossref, Medline, Google Scholar

12. Gecas V: The social psychology of self-efficacy. Annu Rev Sociol 1989; 15:291–316 Crossref, Google Scholar

13. Baumeister R, Vohs K: Norms: prescriptive and descriptive; in Encyclopedia of Social Psychology. Edited by Baumeister RF, Vohs KD. New York, SAGE, 2007 Crossref, Google Scholar

14. Urban Influence Codes Documentation. Washington, DC, Department of Agriculture, Economic Research Service, 2013 Google Scholar

15. Allison PD: Missing data; in Quantitative Applications in the Social Sciences. New York, SAGE, 2001 Google Scholar

16. Babbie ER: The Practice of Social Research, 12th ed. Belmont, CA, Wadsworth/Cengage Learning, 2010 Google Scholar

17. Schmitt N: Uses and abuses of coefficient alpha. Psychol Assess 1996; 81:350–353 Crossref, Google Scholar

18. R: A Language and Environment for Statistical Computing. Vienna, R Foundation for Statistical Computing, 2019Google Scholar

19. Wickham H: ggplot2: Elegant Graphics for Data Analysis, 2nd ed. New York, Springer, 2016 Crossref, Google Scholar

20. Stata Statistical Software: Release 16. College Station, TX, StataCorp, 2019Google Scholar

21. Klein D: MIMRGNS: Stata Module to Run Margins After MI Estimate. Boston, Boston College Department of Economics, 2022. https://ideas.repec.org/c/boc/bocode/s457795.html. Accessed Aug 26, 2022 Google Scholar

22. Holliday SB, Hepner KA, Farmer CM, et al.: A qualitative evaluation of Veterans Health Administration’s implementation of measurement-based care in behavioral health. Psychol Serv 2020; 17:271–281Crossref, Medline, Google Scholar

23. Cully JA, Dawson DB, Hamer J, et al.: A Provider’s Guide to Brief Cognitive Behavioral Therapy. North Little Rock, AR, Department of Veterans Affairs, South Central Mental Illness Research, Education and Clinical Center, 2020 Google Scholar

24. Hsiao CJ, Hing E, Ashman J: Trends in electronic health record system use among office-based physicians: United States, 2007–2012. Natl Health Stat Rep 2014; 75:1–18Google Scholar

25. Proctor EK, Knudsen KJ, Fedoravicius N, et al.: Implementation of evidence-based practice in community behavioral health: agency director perspectives. Adm Policy Ment Health 2007; 34:479–488Crossref, Medline, Google Scholar

26. Goldstein LA, Connolly Gibbons MB, Thompson SM, et al.: Outcome assessment via handheld computer in community mental health: consumer satisfaction and reliability. J Behav Health Serv Res 2011; 38:414–423Crossref, Medline, Google Scholar

27. Bickman L: A measurement feedback system (MFS) is necessary to improve mental health outcomes. J Am Acad Child Adolesc Psychiatry 2008; 47:1114–1119Crossref, Medline, Google Scholar

28. Hatfield DR, Ogles BM: Why some clinicians use outcome measures and others do not. Adm Policy Ment Health 2007; 34:283–291Crossref, Medline, Google Scholar

29. Pincus HA, Page AEK, Druss B, et al.: Can psychiatry cross the quality chasm? Improving the quality of health care for mental and substance use conditions. Am J Psychiatry 2007; 164:712–719Link, Google Scholar

Featured in Psychiatric Services From Pages to Practice.

Volume 74
Issue 4

April 01, 2023
Pages 349-357

Metrics

Keywords

PDF download

History

Received 29 December 2021

Revised 6 May 2022

Accepted 22 July 2022

Published online 25 January 2023

Published in print 1 April 2023

Sign In

Change Password

Your password must have 6 characters or more:

Password Changed Successfully

Create your account

Forget yout Password?

Forgot your Username?

A Survey of Behavioral Health Care Providers on Use and Barriers to Use of Measurement-Based Care

Abstract

Objective: