Skip to main content

Table 2 Reported results on validity and reliability (unless otherwise reported, all values are for depression subscales)

From: Selection of depression measures for use among Vietnamese populations in primary care settings: a scoping review

Measure

Citation

Validated

Population

Sample size

Content validity

Construct validity

Criterion validity

Reliability

Beck Depression Inventory (BDI)

 

Lin and Hung [39]

No

VN immigrant women living in Taiwan

n = 143

Translated and back translated in VN

α = 0.80

Centre for Epidemiologic Studies-Depression Scale (CES-D)

 

Tran et al. [47]

Yes

VN immigrants in the United States

Boston community sample (n = 324); Nationwide sample (n = 452)

One item (“I felt that I was just as good as other people”) excluded for poor conceptual equivalence

Community sample α = 0.90; Nationwide sample α = 0.91

 

Leggett et al. [64]

No

VN adults 55 years and older in Da Nang and surrounding areas

n = 600

Translated and back translated

α = 0.85

 

Nguyen et al. [14]

No

Secondary students in Can Tho, Vietnam

n = 1161

 

Gellis et al. [42]

No

VN immigrants in the US receiving psychiatric services

n = 79

α = 0.85 and α = 0.82 at two time points

 

To et al. [35]

No

Older adults receiving cataract surgery in HCMC

n = 413 patients with 40 % loss to follow up

Translated and back translated

Composite International Diagnostic Interview (CIDI)

 

Rees et al. [33]

No

Random sample in one rural and one urban district in the MKD region

n = 3039

CIDI underreported depression prevalence (1.6 %) vs. the PVPS (7.4 %)

 

Steel et al. (2009) [34]

Yes

VN population living in Vietnam and one living in Australia compared with an Australian-born population

n = 3039 in the MKD region, n = 1161 VN people living in Australia

Translated and back translated

MKD sample: CIDI and PVPS combined prevalence of 8.8 %. CIDI identified 42 unique cases, the PVPS 208, and both identified 16 cases. AUS sample: combined prevalence 11.7 %, CIDI indentified 38 cases, the PVPS 58 cases and both 40 cases

MKD Sample-depression subscale: AUC = 0.65 [95 % CI 0.56–0.73]. AUS sample-depression subscale: AUC = 0.73, (95 % CI 0.64–0.81)

 

Liddell et al. [50]

No

A VN sample from the MKD, a VN immigrant population in Australia and an Australian-born sample

n = 3039 in the MKD, n = 1161 VN people living in Australia

CIDI translated and back translated

Depression and Anxiety Stress Scale (DASS)

 

Tran et al. [29]

Yes

Mothers in a rural northern Vietnam

n = 221

One factor (emotional state) significant, eigenvalue = 1.86

For cut-off of >10: Se: 80.8 %, Sp: 77.4 %; AUC 80.4 %

α = 0.72

 

Nguyen et al. [14]

No

Students in Hue, Vietnam

n = 623

Translated and back translated

α = 0.81

 

Fisher [25]

No

Pregnant women in Ha Nam, Vietnam

n = 6

Four Measures of Mental Health (FMoMH)

 

Phan, Steel and Silove [40]

No

Patients in Australia attending public psychiatric services and patients of general primary healthcare services

n = 86 psychiatric patients, n = 99 primary care patients

In cultural sensitivity questionnaire, participants responded: Words easy to understand = 30 %; Idioms that are familiar = 33 %; Individual questions constructed in meaningful way = 15 %; Symptoms that are similar to your or people you know = 15 %; Helpful in assisting a doctor identify mental illness: 22 %

MT-MM assessment showed a reliability of 0.94 for the depression subscale and showed that the 3 depression measures used in the study (PVPS, HSCL, FMoMH had the highest level of convergent validity

General Health Questionnaire (GHQ)

 

McKelvey, Webb, and Strobel [38]

Yes

“Amerasians” in Vietnam before their emigration to the US

n = 42 assessed for DSM-III depression, n = 5 cases

GHQ identified 2 of 5 DSM cases and 2 of 35 subjects without a DSM diagnosis as being in clinical range

.

Indochinese Hopkins Symptom Checklist (iHSCL)

 

Hinton et al. [43]

Yes

Newly-arrived adult VN refugees undergoing health screening in US

n = 206

Excluded item on loss of sexual interest, as deemed culturally inappropriate

Se = 86 %; Sp = 93 %; PPV = 48 %; AUC = 0.91 (SE = 0.06)

 

Smith Fawzi et al. [46]

Yes

VN former POWs in US

n = 62

Using cut-off of 1.75: Se = 0.87; Sp = 0.7; AUC = 0.8916 (SD = 0.0448)

 

McKelvey, Webb and Strobel [38]

Yes

“Amerasians” in Vietnam before emigration to the US

n = 42 assessed for DSM-III depression, n = 5 cases

The HSCL-25 identified 4 out 5 of the DSM-III diagnosed cases

 

 

Phan, Steel and Silove [40]

No

Patients in Australia attending public psychiatric services and patients of general primary healthcare services

n = 86 psychiatric patients, n = 99 primary care patients

In cultural sensitivity questionnaire, participants responded: words easy to understand = 30 %; Idioms that are familiar = 34 %; Individual questions constructed in meaningful way = 18 %; Symptoms that are similar to your or people you know = 19 %; Helpful in assisting a doctor identify mental illness: 27 %

MT-MM assessment showed a reliability of 0.94 for the depression subscale and showed that the 3 depression measures used in the study (PVPS, HSCL, FMoMH had the highest level of convergent validity

  
 

McKelvey and Webb [38]

No

VN migrants to the US, pre and post migration

n = 161

Phan Vietnamese Psychiatric Scale (PVPS)

 

Phan, Steel and Silove [40]

Yes

Sample 1 recruited from mental health service and local primary care services and sample 2 recruited from two psycho-education classes in Australia

Sample 1: n = 185 and Sample 2: n = 64

Extensive review of traditional literature and an ethnographic survey. In cultural sensitivity questionnaire, participants responded: Words easy to understand = 43 %; Idioms that are familiar = 57 %; Individual questions constructed in meaningful way = 32 %; Symptoms that are similar to your or people you know = 32 %; Helpful in assisting a doctor identify mental illness: 42 %

CFA performed on responses from Study 1 and Study 2 showed four-factor model most appropriate (Chi square results decreased from 3858 for 1 factor model to 214 for 4 factor model in 1st administration and from 3862 to 66 in 2nd administration). MT-MM showed reliability of 0.95 for the depression subscale and showed that the 3 depression measures used in the study (PVPS, HSCL, FMoMH) had the highest level of convergent validity compared with other sub-scales

Compared to psychiatrist diagnosis: Se = 83; Sp = 80; Overall agreement = 81; K = 0.62; PVPS cut-off = 1.95. Compared with DIS interview: Se = 84; Sp = 78; Overall agreement = 81; PVPS cut-off = 1.95. Compared with naturalist assessment: Se = 84; Sp = 87; Overall agreement = 85; PVPS cut-off = 1.85

Test–retest correlations for depression scale = 0.89; α for depression subscale in psychiatric sample: 0.93 at baseline and 0.94 at follow-up; in primary care sample: 0.95 at baseline and 0.95 at follow-up

 

Steel et al. [34]

Yes

VN population living in Vietnam and one living in Australia, compared with an Australian-born population.

n = 3039 in the MKD region, n = 1161 VN people living AUS

MKD sample: CIDI and PVPS combined prevalence of 8.8 %. CIDI identified 42 unique cases, the PVPS 208, and both identified 16 cases. AUS sample: combined prevalence 11.7 % with the CIDI indentifying 38 cases, the PVPS 58 cases and both 40 cases

MKD sample- depression subscale: AUC = 0.65 [95 % CI 0.56–0.73]. AUS sample = depression subscale: AUC = 0.73, [95 % CI 0.64–0.81]

Self Reporting Questionnaire-20 (SRQ-20)

 

Tuan, Harpham and Huong [30]

Yes

Portion of random sample of female participants from child poverty study in Hung Yen, Vietnam.

n = 32 cases and n = 34 control

Translated and back translated

 

Compared with psychiatrist diagnosis (based on average of 3 interviews): Se = 73 %; Sp = 82 %; PCC = 79 %; AUC = 0.84 (95 % CI 0.75–0.94)

Inter-rater reliability at cut-off 7/8: K = 0.79 (z = 11.13, p < 0.001).

 

Richardson et al. [31]

No

Adults in Da Nang and Khanh Hoa, Vietnam

n = 4981

 

Son et al. [28]

No

Male MMORPG players in Hanoi, Vietnam

n = 344 players and n = 344 non-players

 

Harpham et al. [27]

No

Mothers in 20 community sites in Vietnam

n = 1570

 

Giang et al. [26]

Yes

Adult patients at district hospital and a sample from general community in rural northern Vietnam.

District hospital: n = 52. General community: n = 485

Translated and back translated. Researchers and health workers modified as needed. Piloted with patients, staff and community members at NIMH, district hospital and Bavi district. Report high face validity, but indicate that certain questions might lead to false “no” responses based on gender of respondent

A psychiatrist’s diagnosis using the CIDI used for comparison. Community sample: Optimal cut-off = 6/7; Se = 85 %; Sp = 61 %; Misclassification rate = 29 %; AUC = 0.86 [95 % CI 0.81–0.93]. Hospital sample: Optimal cut-off = 5/6; Se = 85 %; Sp = 46 %; Misclassification rate = 44 %; AUC = 0.74 [95 % CI 0.59- 0.89]

 

De Silva et al. [24]

No

Primary caregivers of children in 20 semi-purposefully selected clusters in each of four countries (including Vietnam)

100 households in each cluster

 

Stratton et al. [6]

Yes

VN adults in Da Nang and Khanh Hoa

n = 4980

1 factor EFA: CFI = 0.924; TLI = 0.915; RMSEA = 0.065; Bi-Factor Model: CFI = 0.977, TLI = 0.971, RMSEA = 0.030

α = 0.87

Vietnamese Depression Scale (VDS)

 

Hinton et al. [43]

Yes

Newly-arrived adult VN refugees undergoing routine, mandatory health screening in US

n = 206

  

Using cut-off of 12: Sp: 98 %; Se: 64 %; PPV: 75 %; NPV: 97 % (based on prevalence of major depression of 7 %). Cut-off of 11: Se = 79 %. AUC = 0.93 (SE = 0.05)

 
 

Dinh et al. [41]

Yes

Adult VN refugees in Houston, US

n = 180

PCA showed three factors with eigenvalues of >1, which were supported by the analysis of a Scree plot. For the three extracted factors: factor 1 (depressed affect): 40.8 % of variance, α = 0.92; factor 2 (somatic) 14.2 % of variance, α = 0.80; factor 3 (cultural specific) 10.1 % of variance, α = 0.81

 

 

Nguyen et al. [15]

Yes

People who had already been diagnosed by a psychiatrist and patients presenting at a primary care clinic in HCMC, Vietnam

Previously-diagnosed patients: n = 115; Screened primary care patients: n = 177

Added item on sleep disturbance. Six items chosen more frequently by depressed subjects: feeling downhearted/low-spirited, low-spirited/bored, bothered and sad/bothered

   
 

McKelvey, Webb and Strobel [45]

Yes

“Amerasians” in Vietnam before emigration to the US

n = 42 assessed for DSM-III depression, n = 5 cases

VDS did not identify any DSM-III cases or other subjects as being in clinical range

 

McKelvey and Webb [38]

No

VN migrants to the US, before and after migration

n = 161