Anxiety Interventions Delivered in Primary Care Behavioral Health Routine Clinical Practice

Abstract

Introduction:

Although anxiety is highly prevalent in primary care and a top reason for referral to Primary Care Behavioral Health (PCBH) services, there are limited data on which anxiety interventions are used in routine PCBH practice. The objective of this study was to identify interventions delivered when treating anxiety in PCBH practice.

Methods:

We conducted an online survey of PCBH providers regarding their clinical practice with patients who present for treatment of anxiety symptoms. The final sample comprised 209 PCBH providers recruited from email listservs of national professional organizations (59.3% psychologists, 23.4% social workers, 12.4% counselors, 4.8% other). Providers reported on use (yes/no) of 17 interventions in their most recent session with their most recent adult patient presenting with a primary concern of non-trauma related anxiety.

Results:

On average, patients were reported to be 42.2 (14.73) years old, White (73.7%), and male (56.5%) with anxiety symptoms of moderate severity (65.6%). Most reportedly had comorbid sleep difficulties (63.6%), depressive symptoms (58.4%), and/or stress/adjustment (56.0%). Providers reported delivering an average of 5.77 (2.05, range: 1–15) interventions, with psychoeducation (94.7%), relaxation training (64.1%), and supportive therapy (60.8%) being most common. Several highly efficacious evidence-based interventions for anxiety, including cognitive therapy (45.0%) and exposure (21.1%), were less common.

Discussion:

While PCBH providers delivered numerous brief interventions for anxiety, cognitive therapy and exposure were underutilized. Furthermore, PCBH patients with anxiety symptoms were complex, with significant severity and comorbidity. These results suggest implications for research, clinical training, intervention design, and future implementation efforts.

Anxiety is prevalent in primary care (Kroenke, Spitzer, Williams, Monahan, & Löwe, 2007) and the second most common reason (after depression) for referral to Primary Care Behavioral Health services (PCBH; Bridges et al., 2015; Funderburk et al., 2011). PCBH comprises focused assessment, intervention, and consultation by mental health providers working within primary care teams (Hunter et al., 2018). PCBH and collaborative care management are the two most popular models of integrated primary care (IPC; Vogel, Kanzler, Aikens, & Goodie, 2016), but PCBH is less established in the literature (see Hunter et al., 2018). Better understanding of the anxiety interventions being delivered in PCBH practice is needed to advance research regarding PCBH’s impact on patient outcomes as well as implementation outcomes such as feasibility, acceptability, and fidelity.

Early work found that interventions drawing from cognitive-behavioral therapy (CBT) were common in PCBH practice (e.g., Funderburk et al., 2011). Subsequent work indicated that the most common interventions specifically for anxiety were exposure, relaxation training, and psychoeducation in Federally Qualified Health Centers (Bridges et al., 2015), relaxation and exposure in a university-based community clinic (Sadock, Auerbach, Rybarczyk, & Aggarwal, 2014), and cognitive therapy and exposure in community clinics (Sawchuk et al., 2018). However, in these studies, clinicians were receiving intensive CBT supervision/training, which limits generalizability. Results from studies with homogenous provider samples in terms of discipline, setting, and/or theoretical orientation may not reflect routine PCBH practice.

The present study utilized a national sample of PCBH providers from a variety of primary care settings to examine which interventions were reportedly delivered when treating anxiety among adult patients in routine PCBH practice.

Method

We used a cross-sectional, web-based survey to collect data from PCBH providers recruited via emails sent in June 2015 to six listservs of national organizations with IPC special interest groups. At that time, the six listservs had a total of 3,429 subscribers.

The target population was behavioral health providers currently working in PCBH. An operational definition of PCBH (based on Strosahl, 1998) was provided: “a population health-based model of care in which the behavioral health provider (BHP) is embedded with the primary care team and serves as a consultant to the primary care provider in the assessment, intervention, and management of the full spectrum of patient concerns; the BHP and medical team share information regarding patients using a shared medical record, treatment plan, and standard of care, and BHP visits are usually 30 minutes or less.” Participants were asked to indicate whether they met all of the following criteria: have at least a master’s degree or equivalent in a mental health field, currently practicing behavioral health provider in IPC (at least once per week over the past month), have worked in IPC for at least six months, currently work in the PCBH model of IPC (as defined above), and not a psychiatric prescriber. Psychiatric prescribers were ineligible due to their reliance on medication management (Funderburk et al., 2011).

After demographic questions, providers were asked to report on their most recent session with their most recent patient presenting with non-trauma-related anxiety, defined as subthreshold or threshold “symptoms of: generalized anxiety disorder, panic disorder, agoraphobia, social anxiety disorder, specific phobia, anxiety not otherwise specified/unspecified or other specified anxiety disorder, adjustment disorder with anxiety or with mixed anxiety and depressed mood.” Providers reported on patient demographics, medication use, anxiety severity and diagnoses, presence and severity of comorbid depressive symptoms, and other symptoms/problems. Providers indicated whether each of 17 interventions were used/discussed in the session (yes/no). Interventions were selected based on the literature, prior research, and PCBH clinical experience (see Table 3 for definitions/examples provided to participants).

Table 3.

Percentage of Sessions in Which Each Intervention was Delivered

Intervention	Definitions/Examples Provided to Participants	n	%
Any psychoeducation1		198	94.74
Anxiety psychoeducation	common symptoms, how anxiety develops	180	86.12
CBT psychoeducation	relationship between thoughts, feelings, and behaviors	168	80.38
Relaxation training	diaphragmatic breathing, progressive muscle relaxation, guided imagery	134	64.11
Supportive therapy	reflective listening, letting patient “vent”	127	60.77
Cognitive therapy	education on cognitive distortions, cognitive restructuring	94	44.98
Behavioral activation	activity log, scheduling pleasurable / meaningful activities	88	42.11
Stress management	general planning related to stress management	83	39.71
Self-monitoring	teaching patient how to use worry or panic log, reviewing prior logs for patterns	68	32.54
Mindfulness meditation	mindfulness and/or meditation training	67	32.06
Motivational interviewing	readiness to change, decisional balance	58	27.75
Medication information	discussion of medication options, trouble-shooting poor adherence	52	24.88
Problem-solving training	structured problem-solving therapy approach to address sources of anxiety	47	22.49
Crisis/risk assessment	suicide risk assessment, safety planning	45	21.53
Exposure	encouraging decreased avoidance, developing anxiety hierarchy, planning gradual exposures	44	21.05
Sleep interventions	stimulus control therapy, sleep restriction	33	15.79
ACT-based interventions	ACT metaphors, values-based living	33	15.79
Other	other interventions	32	15.31

Note. N = 209. ACT = Acceptance and Commitment Therapy, CBT = cognitive-behavioral therapy.

¹Any psychoeducation was not a response option; it was created due to high overlap by collapsing anxiety psychoeducation and CBT psychoeducation.

For consistency with the population of interest and most common PCBH session format (Shepardson, Buchholz, Weisberg, & Funderburk, 2018), we limited the sample to providers treating adults individually and face-to-face. Those reporting couples/family sessions (n = 18), patients younger than 18 years of age (n = 16), or telephone/video sessions (n = 9) and those not responding to the intervention items (n = 25) or reporting they did not provide intervention (i.e., assessment only; n = 14) were excluded. The final sample comprised 209 providers. We collapsed anxiety psychoeducation (86.12%) and CBT (80.38%) psychoeducation given high overlap.

Results

Of the 329 providers who accessed the survey, 92% were eligible, and 96% of eligible providers consented and participated (N = 291). Providers (age M = 42.47 years, SD = 14.7) identified as mostly White, female, psychologists (see Table 1). A conservative estimate is that 9.6% accessed the survey; the true response rate is likely higher given that other, ineligible, professionals were also subscribed.

Table 1.

Provider Demographics and Professional Characteristics (N = 209)

	n	%		n	%
Sex			Theoretical Orientation
Female	153	73.21	Cognitive-Behavioral	114	54.55
Male	55	26.32	Eclectic or Integrative	38	18.18
Prefer Not to Answer	1	0.48	Acceptance & Commitment Therapy	20	9.57
Ethnicity			Behavioral	7	3.35
Not Hispanic or Latino	188	89.95	Psychodynamic / Psychoanalytic	7	3.35
Hispanic or Latino	20	9.57	Family Systems	6	2.87
Prefer Not to Answer	1	0.48	Humanistic / Existential	6	2.87
Race			Insight-oriented	4	1.91
White	184	88.04	Interpersonal	4	1.91
Black or African American	12	5.74	Other	3	1.44
Asian	4	1.91	Work Setting
American Indian or Alaska Native	3	1.44	Department of Veterans Affairs	86	41.15
Native Hawaiian or Pacific Islander	0	0	Outpatient clinic or hospital	55	26.32
Other	6	2.87	Federally qualified health center	39	18.66
Professional Discipline			Community mental health center	6	2.87
Psychologist	124	59.33	Military / Department of Defense	1	0.48
Social Worker	49	23.44	Other	22	10.53
Master’s Level Counselor/Therapist	26	12.44
Nurse/Registered Nurse	1	0.48
Other	9	4.31

Patients (age M = 42.15 years, SD = 14.73, range: 18–79) were reportedly mostly White and male with moderately severe anxiety and sub-threshold/mild depressive symptoms (see Table 2). Patients reportedly had an average of 1.18 anxiety disorder diagnoses (SD = 0.53) and 2.93 comorbid symptoms/problems (SD = 1.56, range: 0–8) aside from anxiety (see Table 2). Providers reported delivering a range of interventions (see Table 3), on average using 5.77 (SD = 2.05, range: 1–15, median = 6).

Table 2.

Patient Demographics and Symptom Profiles as Reported by Providers (N = 209)

	n	%		n	%
Age Range			Anxiety Severity
18–29	47	22.49	Mild	38	18.18
30–39	54	25.84	Moderate	137	65.55
40–54	49	23.44	Severe	32	15.31
55+	59	28.23	Missing	2	0.96
Gender			Depression Severity
Male	118	56.46	None	29	13.88
Female	91	43.54	Subthreshold/Mild	113	54.07
Ethnicity			Moderate	57	27.27
Not Hispanic or Latino	175	83.73	Severe	8	3.83
Hispanic or Latino	26	12.44	Missing	2	0.96
Unsure	7	3.35	Comorbid Symptoms/Problems
Not Reported	1	0.48	Sleep	133	63.64
Race			Depression	122	58.37
White	154	73.68	Stress/adjustment	117	55.98
Black or African American	23	11.00	Pain	52	24.88
American Indian or Alaska Native	3	1.44	Coping with medical condition	43	20.57
Asian	3	1.44	Obesity	36	17.22
Native Hawaiian or Pacific Islander	1	0.48	Alcohol misuse	22	10.53
Other	11	5.26	Tobacco use	21	10.05
Unknown/Not Reported	14	6.70	Other substance misuse	12	5.74
Psychotropic Medication			Suicidal ideation	11	5.26
No	110	52.63	Medical regimen adherence	11	5.26
Yes	94	44.98	Posttraumatic stress disorder	10	4.78
Unsure	4	1.91	None	9	4.31
Not Reported	1	0.48	Mania/hypomania	5	2.39
Any Anxiety Diagnosis	196	93.78	Psychosis	3	1.44
Generalized Anxiety Disorder	70	33.49	Dementia	1	0.48
Unspecified/Other Specified	54	25.84	Other	13	6.22
Panic Disorder	43	20.57
Adjustment Disorder with Anxiety	33	15.79
Social Anxiety Disorder	15	7.18
Agoraphobia	10	4.78
Obsessive-Compulsive Disorder	7	3.35
Specific Phobia	5	2.39
None	4	1.91
Unsure	8	3.83

Discussion

This study explored routine PCBH clinical practice in treating anxiety among a national BHP sample.

Patient Complexity

Our results demonstrate the complexity of PCBH patients with anxiety symptoms. Nearly all (94%) were believed to meet diagnostic criteria for an anxiety disorder. Although seeking treatment in primary care, rather than specialty care, for anxiety, four-fifths were assessed as having at least moderate severity anxiety, and comorbid symptoms/problems were nearly universal (96%). Over half reportedly had at least subthreshold/mild depressive symptoms as well as sleep or stress/adjustment difficulties. Comorbid behavioral health concerns including chronic pain and illness were each reportedly present in approximately 20%.

Interventions Delivered

Psychoeducation was delivered almost universally, likely because it is a quick, easy way to help patients understand symptoms and provide a rationale for further intervention (Strosahl, 1996). Relaxation training was the only other evidence-based anxiety intervention used in more than half of sessions. The third most commonly used intervention was supportive therapy. Thus, while CBT interventions have strong empirical support for anxiety, many did not figure into the top three interventions delivered.

Several highly efficacious CBT interventions were less commonly delivered. The gold standard intervention of exposure (Arch & Craske, 2009) was used in only one-fifth of sessions. This result is not too surprising, however, since exposure is not widely delivered even in specialty care (Wolitzky-Taylor, Zimmermann, Arch, De Guzman, & Lagomasino, 2015). Barriers to using exposure in PCBH are likely similar to those in specialty care, including training deficiencies, time constraints, and provider concerns about patients feeling anxious during exposure (Wolitzky-Taylor et al., 2018). These concerns may be exacerbated in IPC, where even less time is available for developing fear hierarchies and conducting exposures.

Cognitive therapy was delivered in fewer than half of sessions. Mixed findings on its frequency of use are likely related to differences in setting, patient population, and training/supervision (Bridges et al., 2015; Sawchuk et al., 2018). Providers may perceive that cognitive therapy is too complex to deliver briefly. Mindfulness meditation and ACT-based interventions being infrequently used may be related to their newness relative to CBT. The prevalence of behavioral activation and stress management interventions, which are not evidence-based for anxiety specifically, suggests that providers may consider comorbid symptoms in selecting interventions.

Methodological Limitations and Strengths

While a conservative estimate of the response rate in our convenience sample is comparable to other provider surveys (e.g., Cook, Biyanova, Elhai, Schnurr, & Coyne, 2010), the true response rate is unknown. Providers self-identified as working in the PCBH model, and retrospective self-reports of provider behavior may be subject to reporting biases. However, the survey was anonymous to limit social desirability and focused on the most recent session to limit recall difficulties. Although we provided brief descriptions of each intervention, providers may have interpreted them differently, and it may be difficult to parse out when multiple similar/related interventions are used (e.g., CBT psychoeducation related to cognitive restructuring). Nonetheless, key strengths include the national sample of PCBH providers from myriad training backgrounds and work settings who reported on routine practice with a wide range of patients. Our findings shed light on the range of severity, variety of diagnoses, and complexity of comorbidity seen in PCBH patients referred for anxiety.

Clinical and Research Implications

Findings suggest future directions for PCBH research, clinical training, implementation, and intervention design. Qualitative research elucidating barriers and facilitators to implementing evidence-based anxiety interventions in PCBH could inform improvements in clinical training. Implementation support is needed to assist providers in adapting evidence-based anxiety interventions, especially exposure, for feasible delivery in PCBH. The high number of interventions being delivered within sessions indicates a need to examine the dose and fidelity of intervention, which likely impact efficacy. While PCBH is intended to provide lower-intensity treatment for subthreshold, mild, and moderate symptoms (Strosahl, 1996), our results suggest that significant severity and complex comorbidity are common. Thus, interventions targeting specific anxiety diagnoses may have limited utility (see Shepardson et al., 2018). Transdiagnostic interventions that can accommodate a range of anxiety presentations and comorbid symptoms, such as process-based therapy approaches (Hofmann & Hayes, 2019), may be most efficient (Gros, Allan, & Szafranski, 2016). This study provides preliminary data on what interventions PCBH providers deliver in routine practice, which can inform future research on PCBH fidelity, effectiveness, and implementation.