The Effectiveness of Problem-Solving Therapy for Primary Care Patients' Depressive and/or Anxiety Disorders: A Systematic Review and Meta-Analysis

Anao Zhang; Sunyoung Park; John E. Sullivan; Shijie Jing

doi:10.3122/jabfm.2018.01.170270

Article Figures & Data

Figures

Tables

Download figure
Open in new tab
Figure 1.
Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) chart of literature search for Problem-solving therapy (PST) studies for treating primary care patients' depression and/or anxiety.
Download figure
Open in new tab
Figure 2.
Funnel Plot for Publication Bias in Problem-solving therapy (PST) Studies for Treating Primary Care Patients' Depression an/or Anxiety.
Download figure
Open in new tab
Figure 3.
Forest Plot of PST Treatment Effect Size Estimates for Treating Primary Care Patients' Depression and/or Anxiety per Study.

Tables

Figures

View popup

Table 1.

Study Characteristics for Problem-Solving Therapy as Intervention for Treating Depression and/or Anxiety Among Primary Care Patients (n = 11)

Author	Sample^*	Demographics^{^†}	Control^{^‡}	Provider and PCP's role in PST (if applicable)	PST/PST-PC Dosage	Diagnostic or Symptom Severity Criteria	Depression and/or Outcome Measures
Barrett et al. (2001)	T = 80 C1 = 80 C2 = 81^{^§}	44.1 year old (SD NR),	MED Placebo	Ph.D-level psychologists. PCP no involvement	6 PST-PC sessions, lasting about 1 hour for the first visit and 30 minutes for subsequent visits.	DSM-III-R, HDRS, PRIME-MD	HSCL-D-20 HDRS
		36.1% male,					HSCL-D-20 HDRS
		90% white.
Chibanda et al. (2014)	T = 30 C = 28	24.5 years old (SD = 4.9)	MED	Trained Peer Counselor. PCP no involvement	12 sessions (60 mins per session) group PST session which were modeled after a 7-step management plan for depression published earlier (Abbas et al., 1994)	DSM-IV	EPDS
		% male NR
		Race NR
Katon et al. (2004)	T = 164 C = 165	58.3 years old (SD = 12),	TAU	Registered nurses in collaboration with the PCP	Medication OR PST-PC, there is a stepped-care algorithm^{^¶}	PHQ-9 Did not require diagnostic criteria	SCL-90 depression
		35% male,
		75.4% white.
Lam et al. (2009)	T = 149 C = 150	71.8 years old (SD = 7.0)	AC^{^‖}	Primary care physicians	3 sessions of modified PST-PC (Mynors-Wallis et al., 2000), first session 30 to 45 minutes. session 2 & 3 20 to 30 minutes.	HADS score	HADS (AS), HADS (DS)
		43.14% male,					SF-36 mental
		Race NR					SF-36 mental
Lynch et al. (2004)	T = 9 C1 = 9 C2 = 13	38.5 years old (SD = 13.7),	AC^** TAU	Registered nurses. PCP referral, no other involvement	6 sessions of telephone-based PST (adopted Nezu, Nezu, & Perri, 1989)	PRIME-MD HRSD	PRIME-MD, HRSD BDI, DHP-D-A
		17% male
		Race NR
McCusker et al. (2008).	T = 36 C = 32	73.3 years old (SD = 8.6),	TAU	Depression care practitioner supervised by (and in collaboration with) PCP	4 sessions PST intervention (60- minute first session, 30 mins for the rest) developed based on IMPACT	PHQ-2	SCL-20, SF-12 SCID
		33.8% male
		Race NR
Mynors-Wallis et al. (2000)	T1 = 80 T2 = 35 C = 36	35 years old (SD = NR),	MED	General practitioner	6 sessions PST-PC, with first session 1 hr, others 30 minutes	RDC	HDRS
		23% male,		Nurse and General Practitioner (PCP)		HDRS score	BDI-I
		95% white		Nurse and General Practitioner (PCP)
Oxman et al. (2008)	T = 72 C = 69	55.2 years old (SD = 16),	TAU	Masters level counselor. PCP referral, no other involvement	6 sessions PST-PC, with first session 1 hr, others 30 minutes	DSM-IV, HAM-D,	HAM-D, MADRS HSCL-D-20
		41.8% male,				PRIME-MD	HAM-D, MADRS HSCL-D-20
		96.5% white
Reynolds et al. (2014)	T = 125 C = 122	36.5 years old (SD = 10.9)	TAU	Social workers and mental health nurses	6 to 8 sessions PST-PC, with first session 1 hr, the rest 30 minutes	CES-D, DSM-IV;	SCID/DSM-IV;
		28.7% male		PCP referral, no other involvement		MMSS	BDI, SF-12
		62.3% white		PCP referral, no other involvement		MMSS	CIRSG, BSI - Anxiety
Schmaling et al. (2002)	T = 31^{^††} C1 = 31 C2 = 30	42.8 years old (SD = 10.7)	MED Placebo	Trained therapists with no further specification	6 sessions PST-PC, with first session 1 hr, others 30 minutes	DSM-III-TR PRIME-MD, HRSC	HAM-D (17-item) HSCL-D (20-item)
		39.1% male		PCP referral, no other involvement
		88.0% white		PCP referral, no other involvement
Williams et al. (2000)	T = 138 C1 = 137 C2 = 140	71 years old (SD = 7.0),	MED Placebo	PhD Psychologists, Social workers, and Psychology Counselors	6 sessions PST-PC, with first session 1 hr, others 30 minutes	DSM-III-R, HDRS DSM-IV, PRIME-MD	HSCL-D-20 HDRS
		58.5% male,
		78.2% white		PCP no involvement

↵* Sample size: T, treatment; T₂, treatment 2 if applicable; C, control.
↵† Demographic: NR, not reported.
↵‡ Control: TAU, treatment as usual, W/NT, waitlist or no treatment; MED, medication; Placebo, placebo medication.
↵§ C1, medication paroxetine; C2, Placebo.
↵¶ Participants in the treatment group (68.7% ) received PST. Therefore, the authors believed the effect of intervention can be attributed to PST. Because sensitivity analysis that excluded this study did not alter the overall treatment effect, we included and presented this study in final analysis.
↵‖ AC, active control (health education video).
↵** Active control (stress management).
↵†† Specific breakdown of the numbers was not reported in article, thus assigned arbitrarily.
BSI, Brief Symptoms Inventory; CES-D, Center for Epidemiology Scale–Depression; DCS = depression care specialist; DFD; depression-free days; DHP-D-A, Duke Health Profile-Depression–Anxiety; EPDS, 10-item Edinburgh Postnatal Depression Scale; HAM-D, Hamilton Rating Scale for Depression; HADS, Hospital Anxiety and Depression Scale; HDRS, Hamilton Depression Rating Scale; HRSD, Hamilton Rating Scale for Depression; HSCL-D-20, Hopkins Depression self-report scale; MADRS, Montgomery-Åsberg Depression Rating Scale; MED, Medication Management; PCP, Primary Care Physician; PHQ-2, Patient Health Questionnaire, 2-item; PHQ-9, Patient Health Questionnaire, 9-item; PRIME-MD, Primary Care Evaluation of Mental Disorders; RDC, Research Diagnostic Criteria; SCL-20, Hopkins depression symptom checklist; SCL-90 depression, Hopkins Symptom Checklist—90 depression questions; SCID/DSM-IV, Structured Clinical Interview for DSM-IV Axis II Personality Disorders (SCID-II); SF-12, SF-36 Health Survey 12-item version; TAU, Treatment as Usual; CIRSG, Cumulative Illness Rating Scale for Geriatrics.
Abbas M, Broadhead JC, Mbape P, Khumalo-Sakatukwa G. Defeating depression in the developing word: A Zimbabwean model. Br J Psychiatry 164(3):293–296.
Mynors-Wallis LM, Gath DH, Day A, Baker F. Randomised controlled trial of problem solving treatment, antidepressant medication, and combined treatment for major depression in primary care. BMJ 320:26–30.
Nezu A, Nezu C, Perri M. Problem-solving therapy for depression. New York: Wiley; 1989.

View popup

Table 2.

PST for Treating Primary Care Patients' Depression and/or Anxiety; Results of Univariate Meta-regression

Parameter^*	Estimate	95%CI	t (df)	P Value
Overall Effect (β₀)	0.673	0.467 to 0.879	z = 6.41	.000
Depression (β₀)	0.601	0.224 to 0.978	t (11) = 3.12	.007
Anxiety (β₁)	−0.249	−1.015 to 0.516	t (11) = −0.64	.317
Age^{^†} (β₀)	0.696	0.477 to 0.915	t (10) = 6.23	.000
Age (β₁)	0.020	0.007 to 0.033	t (10) = 2.97	.012
% Male (β₀)	−1.408	−2.161 to −0.655	t (10) = −3.66	.004
% Male (β₁)	0.053	0.035 to 0.071	t (10) = 5.63	.000
% White (β₀)	0.741	−0.030 to 1.511	t (7) = 1.88	.075
% White (β₁)	0.001	−0.009 to 0.011	t (7) = 0.14	.381
% Married (β₀)	0.398	0.066 to 0.729	t (5) = 2.35	.041
% Married (β₁)	0.007	−0.013 to 0.026	t (5) = 0.66	.296
Individual (β₀)	0.668	0.455 to 0.880	t (11) = 6.16	.000
Group (β₁)	0.092	−0.769 to 0.954	t (11) = 0.21	.381
Family (β₂)	—		—	—
In-person (β₀)	0.722	0.494 to 0.950	t (10) = 6.22	.000
Tele-health (β₁)	−0.189	−0.846 to 0.469	t (10) = −0.56	.328
Combined (β₂)	—		—	—
Session No. (β₀)	0.465	−0.213 to 1.142	t (11) = 1.34	.157
Session No. (β₁)	0.035	−0.074 to 0.144	t (11) = 0.63	.315
Min per session (β₀)	−0.005	−1.443 to 1.433	t (10) = −0.01	.389
Min per session (β₁)	0.017	−0.018 to 0.052	t (10) = 0.96	.241
Master Level (β₀)	1.569	1.181 to 1.957	t (9) = 7.92	.000
Doctoral Level (β₁)	−1.334	−2.112 to −0.557	t (9) = −3.36	.007
Multi-Discipline (β₂)	−1.280	−1.757 to −0.802	t (9) = −5.26	.000
No Physician (β₀)	1.058	0.755 to 1.362	t (11) = 6.83	.000
Yes Physician (β₁)	−0.711	−1.124 to −0.298	t (11) = −3.38	.005

↵* β₀ should be interpreted as an intercept in a regression, that is the overall average (effect size) of the reference group. β₁ should be interpreted as a regression coefficient in a regression, that is the difference (in effect size) between the reference group and the predicting group, noted as d(diff.) in the text for categorical variable (moderator). For cells with no numeric value, it was either because of missing data or not enough variation for a statistical estimate to be calculated.
↵† Participant age was mean-centered.
PST, problem-solving therapy.

View popup

Table 3.

Results of Subgroup Analysis of Overall Treatment Effect (by Moderator) of PST for Treating Primary Care Patients' Depression and/or Anxiety

Parameter	Estimate	95% CI	t (df)	P Value
Depression	0.601	0.224 to 0.978	t (11) = 3.12	.007
Anxiety	0.352	−0.314 to 1.018	t (16) = 1.04	.226
Individual	0.668	0.455 to 0.880	t (11) = 6.16	.000
Group	0.760	−0.075 to 1.595	t (12) = 1.78	.085
Family	—	—	—	—
In-person	0.722	0.494 to 0.950	t (10) = 6.22	.000
Tele-health	0.533	−0.083 to 1.150	t (11) = 1.70	.097
Combined	—	—	—	—
Master level	1.569	1.181 to 1.957	t (9) = 7.92	.000
Doctoral level	0.235	−0.439 to 0.909	t (11) = 0.68	.304
Multi-discipline	0.290	−0.012 to 0.567	t (11) = 2.04	.056
Physician not involved	1.058	0.755 to 1.362	t (11) = 6.83	.000
Physician involved	0.347	0.068 to 0.627	t (12) = 2.43	.029

For cells with no numeric value, it was either because of missing data or not enough variation for a statistical estimate to be calculated.
CI, confidential interval; PST, problem-solving therapy.

View popup

Table 4.

PST for Treating Primary Care Patients' Depression and/or Anxiety; Results of the Cochrane Collaboration's Tool for Assessing Risk of Bias^*

Study/Year	Random Sequence Generation	Allocation Concealment	Blinding of Participants and Personnel	Blinding of Outcome Assessment	Incomplete Outcome Data	Selective Reporting	Other Source of Bias
Barrett et al. (2001)	+	+	?	?	?	+	+
Chibanda et al. (2014)	+	?	?	?	?	+	?
Katon et al. (2004)	+	+	−	+	?	+	−
Lam et al. (2009)	+	+	+	?	?	+	+
Lynch et al. (2004)	?	?	?	?	?	?	?
McCusker et al. (2008)	+	+	?	?	?	+	+
Mynors-Wallis et al. (2000)	−	+	+	?	?	?	?
Oxman et al. (2008)	+	?	−	?	?	+	?
Reynolds et al. (2014)	+	+	+	+	+	+	+
Schmaling et al. (2002)	+	?	−	−	?	?	+
Williams et al. (2000)	+	+	?	?	?	?	+
Number of “+”s	9	7	3	2	1	7	6

↵* “+” = criteria were met in primary studies, thus no bias present; “?” = unclear whether or not criteria met from reading of primary studies; and “−” = criteria were not met in primary studies, thus bias present.

View popup

Table 5.

Quality Assessment of Controlled PST Intervention Studies for Primary Care Patients' Depression and/or Anxiety (n=11)

Study/Year	Good	Fair
Barrett et al. (2001)⁴¹	✓
Chibanda et al. (2014)⁴²		✓
Katon et al. (2004)⁴³	✓
Lam et al. (2009)⁴⁴	✓
Lynch et al. (2004)⁴⁵		✓
Mccusker et al. (2008)⁴⁶	✓
Mynors-Wallis et al. (2000)⁴⁷		✓
Oxman et al. (2008)⁴⁸		✓
Reynolds et al. (2014)⁴⁹	✓
Schmaling et al. (2002)⁵⁰		✓
Williams et al. (2000)⁵¹	✓