Summary of findings 6. Primary health professional‐led psychological interventions compared to usual or no care for treating adults with post‐traumatic stress or common mental disorders in humanitarian settings in low‐ and middle‐income countries.
| What are the effects of primary health professional‐led psychological interventions vs usual or no care for treating adults with post‐traumatic stress or common mental disorders in humanitarian settings in low‐ and middle‐income countries? | ||||||
|
Patient or population: adults with post‐traumatic stress or common mental disorders Setting: humanitarian settings in low‐ and middle‐income countries (Democratic Republic of Congo (1 study), Iraq (3 studies), Thailand (1 study)) Intervention: primary health professional‐led psychological interventions Comparison: usual (including 1 of the following: psychosocial support, identification and referral to mental health specialist, monthly follow‐up, poorly accessed counselling service) or no care (wait list) | ||||||
| Outcomes | Anticipated absolute effects*(95% CI) | Relative effect (95% CI) | №. of participants (studies) | Certainty of the evidence (GRADE) | Comments | |
| Risk with usual care | Risk with primary health professional‐led psychological intervention | |||||
| Recovery from PTSD | No studies that reported on this outcome were identified | |||||
|
Prevalence of adults with probable PTSD (1 to 6 months post intervention) Diagnosis defined by HTQ ≥ 1.75 (RR > 1 denotes lower prevalence compared to control) |
417 per 1000 participants |
87 per 1000 participants (15 to 152) |
RR 5.50 (2.50 to 12.10) | 313 (1 RCTa) | ⨁⨁⊝⊝ LOWb | PHPs delivering psychological interventions may reduce the number of people with probable PTSD compared to usual or no care |
|
Prevalence of adults with probable depression or anxiety (1 to 6 months post intervention) Diagnosis defined by HSCL‐25 ≥ 1.75 (RR > 1 denotes lower prevalence compared to control) |
417 per 1000 participants |
87 per 1000 participants (15 to 152) |
RR 4.60 (2.10 to 10.08) | 313 (1 RCTa) | ⨁⨁⊝⊝ LOWb | PHPs delivering psychological interventions may reduce the number of people with depression compared to usual or no care |
|
PTS symptoms (1 to 6 months post intervention) Harvard Trauma Questionnaire (higher score = higher severity) |
Mean HTQ score with usual care was 1.5a | Mean HTQ score in the intervention group was 0.5 (1.0 to 0.1) lower | SMD ‐0.78 (‐1.43 to ‐0.13) | 680 (2 RCTsa,c) | ⨁⊝⊝⊝ VERY LOWd | Scores estimated based on an SMD of ‐0.78 (95% CI ‐1.43 to ‐0.13). It is uncertain whether PHPs delivering interventions have any effect on PTS symptoms compared to usual or no care 1 to 6 months post intervention |
|
Depression symptoms (1 to 6 months post intervention) Hopkins Symptom Checklist ‐ depression (higher score = higher severity) |
Mean HSCL score with usual care was 1.5a | Mean HSCL score in the intervention group was 0.5 (1.0 to 0.1) lower | SMD ‐0.91 (‐1.73 to ‐0.1) | 680 (2 RCTsa,c) | ⨁⊝⊝⊝ VERY LOWe | Scores estimated based on an SMD of ‐0.91 (95% CI ‐1.73 to ‐0.1). It is uncertain whether PHPs delivering interventions have any effect on depression symptoms compared with usual or no care 1 to 6 months post intervention |
| Quality of life | No studies that reported on this outcome were identified | |||||
|
Functional impairment (1 to 6 months post intervention) Locally developed functional impairment scale (higher score = higher functional impairment) |
Mean functional impairment score with usual care was 1.8a | Mean functional impairment score in the intervention group was 0.6 (1.2 lower to 0.04 higher) lower | SMD ‐0.64 (‐1.31 to 0.04) | 680 (2 RCTsa,c) | ⨁⊝⊝⊝ VERY LOWf | Scores estimated based on an SMD of ‐0.64 (95% CI ‐1.31 to 0.04). It is uncertain whether PHPs delivering psychological interventions have any effect on functional impairment 1 to 6 months post intervention compared to usual or no care |
| Service utilisation | 1/66 participants referred to a psychiatrist for worsening symptoms (Bolton 2014 (Iraq)) | 1/223 hospitalised for severe depression and 1/223 self‐referred to a psychiatrist (Weiss 2015) | Similar in both arms | 1572 (5 RCTsa,c,g,h,i) |
⨁⊝⊝⊝ VERY LOWj | It is uncertain whether primary health professionals delivering psychological interventions have any effect on service utilisation up to 6 months post intervention compared to usual or no care |
|
Adverse events (RR > 1 denotes greater risk of harm) |
1/50 deaths (Bolton 2014 (Iraq)) 1 participant died of a heart attack (not stated in which arm) (Weiss 2015, deemed unrelated to study) No adverse events detected (Bass 2013; Bass 2016; Bolton 2014 (Thailand)) |
1/215 participants died; 1/215 participants reported being verbally abused by her husband for getting treatment (Bolton 2014 (Iraq)) 1/223 attempted suicide (Weiss 2015), 1 had a heart attack (not mentioned in which arm; deemed unrelated to study) (Weiss 2015) Bolton 2014 (Thailand); Bass 2013; Bass 2016 1/157 participants died (Bass 2013) 1/182 participants died (Bolton 2014 (Thailand) (deemed unrelated to study)) No adverse events detected (Bass 2016) |
Deaths: RR 2.22 (0.23 to 21.34) | 1242 (4 RCTsa,g,h,i) | ⨁⊝⊝⊝ VERY LOWk | It is uncertain whether PHPs delivering psychological interventions have any effect on adverse events up to 6 months post intervention compared to usual or no care |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; HSCL‐25: Hopkins Symptom Checklist ‐ depression; HTQ: Harvard Trauma Questionnaire; PHP: primary health professional; PTS: post‐traumatic stress; PTSD: post‐traumatic stress disorder; RCT: randomised controlled trial; RR: risk ratio; SMD: standardised mean difference. | ||||||
| GRADE Working Group grades of evidence. High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect. | ||||||
aBass 2013. Cognitive processing therapy vs psychosocial support.
bDowngraded by one level for indirectness: results are from a single study done in a low‐income country in which participants were female survivors of sexual violence. Study population may not be generalisable to other adults with PTSD in LMICs. Downgraded by one level for imprecision: low total number.
cWeiss 2015. Cognitive processing therapy vs transdiagnostic Common Elements Treatment Approach (CETA) vs identification and referral.
dDowngraded by one level for limitations in design: high risk of detection bias ‐ Bass 2013 ‐ and contamination ‐ Weiss 2015. Downgraded by one level for inconsistency: large unexplained statistical heterogeneity (I² = 93%). Downgraded by one level for imprecision: confidence interval of SMD ranges from large clinical effect favouring intervention to no effect. Note that a small clinically appreciable benefit was set at SMD 0.2 to 0.5, a moderate benefit at SMD of 0.5 to 0.8, and a large benefit at > 0.8 (Cohen 1988).
eDowngraded by one level for limitations in design: high risk of detection bias ‐ Bass 2013 ‐ and contamination ‐ Weiss 2015. Downgraded by one level for imprecision: confidence interval of SMD ranges from large clinical effect favouring intervention to no effect. Downgraded by one level for inconsistency: large unexplained statistical heterogeneity (I² = 96%).
fDowngraded by one level for limitations in design: high risk of detection bias ‐ Bass 2013 ‐ and contamination ‐ Weiss 2015. Downgraded by one level for imprecision: confidence interval of SMD ranges from large clinical effect favouring intervention to no effect. Downgraded by one level for inconsistency: large unexplained statistical heterogeneity (I² = 94%).
gBass 2016. Locally designed psychological intervention vs wait‐list control.
hBolton 2014 CRCT Iraq. Cognitive processing therapy vs behavioural activation vs monthly follow‐up.
iBolton 2014 RCT Thailand. CETA vs usual care (poorly accessed counselling service).
jDowngraded by one level for risk of bias: high risk of detection bias in Bass 2013 and Bass 2016; high risk of contamination bias in Bolton 2014 (Iraq); Bolton 2014 (Thailand); and Weiss 2015. Downgraded by two levels for imprecision: very few events.
kDowngraded by one level for risk of bias: high risk of detection bias in Bass 2013 and Bass 2016; high risk of contamination bias in Bolton 2014 (Iraq) and Bolton 2014 (Thailand). Downgraded by one level for indirectness: it is unclear if any of the deaths were related to post‐traumatic stress symptoms or study procedures. Downgraded by two levels for imprecision: very few events. The confidence interval of the risk ratio ranged from indicating harm by PHP‐delivered interventions to indicating benefit.