Social determinants of the changing tuberculosis prevalence in Việt Nam: Analysis of population-level cross-sectional studies.

BACKGROUND: An ecological relationship between economic development and reduction in tuberculosis prevalence has been observed. Between 2007 and 2017, Việt Nam experienced rapid economic development with equitable distribution of resources and a 37% reduction in tuberculosis prevalence. Analysing consecutive prevalence surveys, we examined how the reduction in tuberculosis (and subclinical tuberculosis) prevalence was concentrated between socioeconomic groups. METHODS AND
FINDINGS: We combined data from 2 nationally representative Việt Nam tuberculosis prevalence surveys with provincial-level measures of poverty. Data from 94,156 (2007) and 61,763 (2017) individuals were included. Of people with microbiologically confirmed tuberculosis, 21.6% (47/218) in 2007 and 29.0% (36/124) in 2017 had subclinical disease. We constructed an asset index using principal component analysis of consumption data. An illness concentration index was estimated to measure socioeconomic position inequality in tuberculosis prevalence. The illness concentration index changed from -0.10 (95% CI -0.08, -0.16; p = 0.003) in 2007 to 0.07 (95% CI 0.06, 0.18; p = 0.158) in 2017, indicating that tuberculosis was concentrated among the poorest households in 2007, with a shift towards more equal distribution between rich and poor households in 2017. This finding was similar for subclinical tuberculosis. We fitted multilevel models to investigate relationships between change in tuberculosis prevalence, individual risks, household socioeconomic position, and neighbourhood poverty. Controlling for provincial poverty level reduced the difference in prevalence, suggesting that changes in neighbourhood poverty contribute to the explanation of change in tuberculosis prevalence. A limitation of our study is that while tuberculosis prevalence surveys are valuable for understanding socioeconomic differences in tuberculosis prevalence in countries, given that tuberculosis is a relatively rare disease in the population studied, there is limited power to explore socioeconomic drivers. However, combining repeated cross-sectional surveys with provincial deprivation estimates during a period of remarkable economic growth provides valuable insights into the dynamics of the relationship between tuberculosis and economic development in Việt Nam.
CONCLUSIONS: We found that with equitable economic growth and a reduction in tuberculosis burden, tuberculosis became less concentrated among the poor in Việt Nam.

Introduction

The relationship between tuberculosis disease and economic development is well documented [1-3], while the association between subclinical tuberculosis and economic development has received comparatively little attention to date [4]. Ecological studies analysing historical trends have attributed sustained reductions in the prevalence of tuberculosis to a combination of reductions in crowded living conditions [5], effective anti-tuberculosis chemotherapy [6], and the improvements in housing, access to health services, and nutrition that accompany economic development, poverty reduction, and social policies such as social protection [7]. The World Health Organization’s End TB Strategy and the UN’s Sustainable Development Goals 1 and 3 recognise the importance of healthcare and the control of communicable diseases, including tuberculosis, as outcomes of, and crucial contributors to, economic development [8-10].

Recent empirical work has attempted to quantify the effect of social protection, as an intervention to reduce poverty, on programmatic indicators such as tuberculosis prevalence and case detection rates [1,7,11-16]. Carter et al. considered how social protection, as a component of economic development policy, may affect tuberculosis incidence [11]. Social protection refers to policies designed to reduce poverty through improvements in the labour market, and support for poor and sick individuals. They found that social protection may reduce the incidence of tuberculosis by 76% [11]. In evaluating the relationship between social protection and economic development, Siroka et al. found that tuberculosis prevalence is reduced with increased spending on social protection, though this effect plateaued when countries spent more than 11% of gross domestic product (GDP) on social protection [14]. Although these studies provided evidence that economic growth and social protection are associated with reductions in tuberculosis burden, they did not explore how the distribution of tuberculosis prevalence changes during economic growth.

Việt Nam is an example of a country that has experienced notable sustained economic growth. In 2006, the smear-positive tuberculosis incidence in Việt Nam was estimated to be 260 per 100,000 population, and the treatment success rate was 92% [17]. National tuberculosis prevalence surveys were conducted in Việt Nam in 2007 and 2017 [18,19]. When differences in tuberculosis screening and diagnostic practices were accounted for, a comparative study showed a decline in tuberculosis prevalence over the 10-year period [20]. The study found a 37% reduction in the prevalence of culture-positive tuberculosis, a 53% reduction in the prevalence of smear-positive tuberculosis, and no significant reduction in smear-negative or subclinical tuberculosis. The change in tuberculosis prevalence was more pronounced among men, among people living in rural areas, and in provinces in the north and south of the country [20].

In 1986 a series of economic reforms, the Dổi Mới Policy, were introduced that included investments in health and education [21]. Since then, Việt Nam has experienced rapid and sustained economic growth, with GDP per capita rising from US$230 in 1985 to US$906 in 2007 and US$2,343 in 2017. During this period, income inequality as measured by the Gini coefficient has remained stable for over a decade (35.8 in 2006 and 35.7 in 2018) [22,23]. The increase in GDP per capita with an unchanging Gini coefficient suggests that the benefits of the rapid economic development observed in Việt Nam have been distributed equitably among the population, an example of shared prosperity. Given that the individual risk of tuberculosis disease is increased by poor household socioeconomic position (SEP), the rapid and sustained economic growth in Việt Nam was an opportunity to examine simultaneous changes in tuberculosis prevalence and economic growth.

In the analysis presented here, we used the opportunity of measured longitudinal changes in both poverty and tuberculosis burden to estimate the differential prevalence of tuberculosis by SEP and to examine the individual, household, and neighbourhood social determinants of the reduction in tuberculosis prevalence in Việt Nam.

Methods

We combined individual-level data from 2 cross-sectional nationally representative tuberculosis prevalence surveys to measure the social determinants of changes in tuberculosis prevalence in Việt Nam [18,19]. The SEP of households was estimated by constructing an asset index from consumption data; illness concentration curves and an illness concentration index represented the distribution of illness. Associations between tuberculosis prevalence, individual risk factors, and household SEP within neighbourhoods were estimated by fitting mixed effects multilevel models (MLMs) [24-26].

Possible causal pathways

Fig 1 shows the causal model for the analysis arranged by individual risks and household and neighbourhood effects [27]. Causal models are representations of assumed causal structures and provide a framework for discussing study design, variables included and how this may affect our understanding of the measure of interest [28].

Fig 1

Causal diagram of social determinants of tuberculosis prevalence in Việt Nam.

SEP, socioeconomic position; TB, tuberculosis.

Social determinants of health are the socioeconomic, cultural, and political aspects of the community that affect the health of populations [29]. These determinants include people’s living and working conditions, water and sanitation, housing, unemployment, and political drivers of health. An individual’s biological risk of developing tuberculosis is influenced by age, gender, and previous treatment and how these intersect with household risk and household economic position in the neighbourhood [30]. Transmission of tuberculosis is spatially concentrated in neighbourhoods [16]. Similarly, economic development leads to increased opportunities in neighbourhoods, and depending on how wealth is distributed, there may be a reduction in unemployment, greater assistance to households in need, and therefore more resources per capita. Equitable economic development improves neighbourhood economy, which improves living conditions through reduced crowding, increased availability of windows to improve air circulation, and reduced periods of transmission in neighbourhoods [31]. Furthermore, improvements in the neighbourhood economy increase household resources, reducing malnutrition and improving households’ ability to seek healthcare [1]. Comparatively wealthier households would have greater ability to negotiate access to neighbourhood resources such as housing and health services, therefore lowering their risk of tuberculosis. If symptomatic (clinical tuberculosis), individuals would be more likely to seek and receive tuberculosis care, reducing transmission periods. However, if not symptomatic (subclinical), diagnosis within health services focused on passive tuberculosis case finding may be delayed until onset of clinical disease, leading to increased tuberculosis prevalence in the population [4,32].

Việt Nam national tuberculosis prevalence surveys and case definitions

Nationally representative Việt Nam tuberculosis prevalence surveys were conducted in 60 of 63 provinces in 2007 and 2017 [18,19]. A detailed explanation of the research procedures and differences between the 2 surveys are published elsewhere [20]. In summary, individuals were identified for inclusion in the surveys by multistage sampling whereby first districts and then communes were selected proportional to population size. Clusters (geographical sub-communes) were selected by random sampling, and all households in the selected sub-communes (70 sub-communes in survey 1, and 82 in survey 2) were included. Individuals were eligible for screening if they were 15 years of age or older. Screening procedures included questions on cough and treatment history followed by chest radiography, sputum smear microscopy, and solid Löwenstein–Jensen (LJ) culture. Individuals reporting a cough for at least 2 weeks, haemoptysis, or previous tuberculosis treatment, or who had an abnormal chest X-ray, were considered screen positive. In the first survey, 8.0% (7,529/94,156) of respondents screened positive, compared to 7.4% (4,595/61,763) in the second survey [20]. A flow diagram of participant selection is included in S1 Text.

There were improvements in diagnostic technology between the 2 surveys. For comparability, an individual was considered to have microbiologically confirmed tuberculosis if they were screen positive, had a smear microscopy test, and had at least 1 positive LJ culture. Individuals were considered to have subclinical tuberculosis if they had not reported any symptoms but had at least 1 positive LJ culture.

Data from the prevalence surveys were matched to provincial-level measures of poverty using data from the World Bank, the percentage of people living on less than US$2 per day in 2009, and the 2013 Ministry of Labour, Invalids and Social Affairs (MOLISA) metric [22,23]. The MOLISA metric is used for determining eligibility for the national anti-poverty programme and uses income as an indicator.

Statistical analysis

The asset index was calculated using principal component analysis of 6 variables: the presence of clay floors in the home, wood used as fuel for cooking, and ownership of a stereo system, television, motorbike, and car. In the 2017 survey, the presence of a fridge, computer, air conditioner, washing machine, and water heater were also included in the survey. We restricted the asset index to the same 6 consumption variables in 2007 and 2017 [33]. Using the index, households were divided into groups of relative wealth (SEP groups), and disease prevalence was compared between these groups. We assigned consumption data responses as provided by the self-declared head of the household to all members of the household. To adjust for the relative sampling probability of each participant, we used survey sampling weights based on age, gender, cluster size, geographical area, and post-stratification adjustment. Data were analysed using Stata 16.1 and RStudio 1.3.1093 [34].

The distribution of disease between SEP groups is represented by constructing illness concentration curves [35]. These are used to quantify whether inequality in SEP exists for a health sector variable, such as tuberculosis prevalence [36]. We then quantified the position of the geometric mean on the curve by estimating the concentration index, which is defined as twice the area between the concentration curve and the line of equality (the 45-degree line on the graph) [37].

The relationships between tuberculosis prevalence and subclinical tuberculosis prevalence and SEP are explained not only by individual-level risks, but also by interactions between hierarchical levels including the household and the wider neighbourhood. In our analyses, we investigated the association between the change in tuberculosis prevalence, relative household SEP, and absolute provincial poverty [38]. We used log-binomial models to examine dependencies between variables nested in each group. We used MLMs with group- and individual-level intercepts as random effects. MLMs aim to explain the change in tuberculosis prevalence over time while considering that poverty and the risk of contracting tuberculosis are clustered geographically and in households. MLMs allow us to analyse how neighbourhood effects explain variation in change in tuberculosis prevalence over time.

By partially pooling varying coefficients, we quantified the relationship between variables where we expected the coefficients to vary between neighbourhoods. The Hausman test was used to test the correlation between random error and individual effects (regressors) in the model (see S1 Text).

This study is reported as per the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline (S1 STROBE Checklist).

Ethics

The Việt Nam national tuberculosis prevalence surveys were approved by the National Hospital of Tuberculosis and Respiratory Diseases in Hanoi (2007) and the institutional review board of the Việt Nam National Lung Hospital (2017; 62/17/CTHKH). This analysis was approved by the ethics committee of the London School of Hygiene & Tropical Medicine (16396).

Results

The characteristics of study participants are summarised in Table 1. Data from 155,919 participants were included in the study, 94,156 from survey 1 and 61,763 from survey 2, of which 0.23% (218/94,156) in survey 1 and 0.20% (124/61,763) in survey 2 had microbiologically confirmed tuberculosis. Of the patients with microbiologically confirmed tuberculosis, 21.6% (47/218) in survey 1 and 29.0% (36/124) in survey 2 reported no cough and were therefore considered to have subclinical tuberculosis. The average age of study participants was 40.1 and 46.6 years old, respectively. The gender balance was similar between the 2 surveys, with 54.8% (51,560/94,156) of survey 1 participants and 56.0% (34,613/61,763) of survey 2 participants being male. Similar proportions of patients in the 2 surveys reported at least 1 tuberculosis-associated symptom: 21.7% (20,474/94,156) in survey 1 and 19.3% (11,917/61,763) in survey 2, and the proportions with previous tuberculosis treatment were similar, 1.3% (1,228/94,156) in survey 1 and 1.3% (789/61,763) in survey 2.

Table 1

Comparison of the characteristics of study participants between survey 1 and survey 2.

Characteristic	Survey 1 (2007)		Survey 2 (2017)
	Percent	n/N participants or mean (SD)	Percent	n/N participants or mean (SD)
Microbiologically confirmed tuberculosis	0.23%	218/94,156	0.20%	124/61,763
Microbiologically confirmed tuberculosis—subclinical	21.6%	47/218	29.0%	36/124
Individual
Age (years)
15–24	22.2%	20,934/94,156	10.6%	6,542/61,763
25–34	19.8%	18,681/94,156	16.5%	10,191/61,763
35–44	21.0%	19,790/94,156	18.6%	11,508/61,763
45–54	17.3%	16,285/94,156	21.5%	13,289/61,763
55–64	8.6%	8,138/94,156	18.0%	11,143/61,763
≥65	11.0%	10,328/94,156	14.7%	9,090/61,763
Gender
Male	54.8%	51,560/94,156	56.0%	34,613/61,763
Female	45.2%	42,596/94,156	44.0%	27,150/61,763
Of all participants, those with at least 1 tuberculosis-associated symptom	21.7%	20,474/94,156	19.3%	11,917/61,763
Previous tuberculosis treatment	1.3%	1,228/94,156	1.3%	789/61,763
Household
Absolute wealth estimate		US$2,403.80 (US$27)		US$2,399.60 (US$26)
Household socioeconomic position
Lowest	24.9%	22,677/90,975	35.1%	19,739/56,260
Lower middle	34.5%	31,419/90,975	25.3%	14,207/56,260
Upper middle	16.8%	15,284/90,975	22.7%	12,777/56,260
Highest	23.7%	21,595/90,975	17.0%	9,537/56,260
Region
North	48.5%	45,669/94,156	41.4%	25,575/61,763
Centre	15.6%	14,646/94,156	21.9%	13,525/61,763
South	35.9%	33,841/94,156	36.7%	22,663/61,763
Type of residence
Urban	28.0%	26,353/94,156	30.2%	18,656/61,763
Remote	29.2%	27,532/94,156	25.7%	15,882/61,763
Rural	42.8%	40,271/94,156	44.1%	27,225/61,763
Province
Provincial poverty headcount percent (2009)		22.0 (14.6)		21.6 (15.9)

n, sample size; N, population size; SD, standard deviation.

When comparing household SEP between surveys 1 and 2, a greater proportion of households were in the lowest SEP category (35.1%; 19,739/56,260) in the 2017 compared to the 2007 survey (24.9%; 22,677/90,975). This measure is not consistent with the absolute wealth estimate (AWE), which is similar between the 2 surveys. The AWE per household is based on the household SEP, country measures of production, and the distribution of wealth between rich and poor individuals. Therefore, these measures are related, but the AWE can be compared between time periods. The mean AWE for survey 1 was US$2,403.80 (SD US$27.00) and for survey 2 was US$2,399.60 (SD US$26.00).

The proportion of households sampled from the central region of Việt Nam was slightly larger in survey 2 compared to survey 1 (21.9% versus 15.6%), and survey 2 included more urban (30.2% versus 28.0%) and rural areas (44.1% versus 42.8%) than remote areas (25.7% versus 29.2%). The percentage of households below the poverty line (living on less than US$2 per day) was 22.0% (SD 14.6%) in 2007 compared to 21.6% (SD 15.9%) in 2017.

Fig 2 shows the proportion of study participants with microbiologically confirmed tuberculosis by SEP for each of the surveys (2007 and 2017). A shift in the distribution of tuberculosis disease from a left-leaning slope, where the disease is concentrated among poor households, to a right-leaning slope (concentrated among the wealthy) is observed. The proportion of participants from each of the surveys who are represented by each of the SEP groups is shown in Fig 3. In 2007, there was a similar proportion of households in the lowest and highest SEP groups. In contrast, in 2017, a greater proportion of households were categorised based on their consumption data as poor rather than wealthy.

Fig 2

The distribution of tuberculosis (TB) prevalence by socioeconomic position (SEP) as measured in the 2007 and 2017 tuberculosis prevalence surveys.

The plot shows the average household asset index and the confidence intervals around the mean.

Fig 3

Proportion of participants by socioeconomic position (SEP) in 2007 and 2017.

In Fig 4, we show illness concentration curves, which represent the cumulative tuberculosis prevalence ordered by SEP, relative to the equal distribution line. In the 2007 survey, the concentration curve for tuberculosis prevalence lies above the equal distribution line; therefore, tuberculosis prevalence was concentrated among poorer households in the 2007 survey. In the 2017 survey, the concentration curve lies below the equal distribution line, indicating that tuberculosis prevalence was more equitably distributed among the population, with a higher concentration of tuberculosis in wealthier patients. These results are supported by the estimates of the illness concentration index (see Fig 5). The illness concentration index for tuberculosis disease was −0.10 (95% CI −0.08, −0.16; p = 0.003) in 2007 and 0.066 (95% CI 0.06, 0.18; p = 0.158) in 2017. When we restrict the case definition to subclinical tuberculosis, we see similar results, though with a more pronounced shift towards the wealthier households in 2017.

Fig 4

Illness concentration curves.

The red dashed line represents the equal distribution line, while the blue curve is the cumulative tuberculosis (TB) prevalence in the population ranked by household socioeconomic position (SEP). The blue shaded area is the uncertainty interval. A curve above the equal distribution line means that TB is concentrated among poor households, and a curve below the equal distribution line means that TB is concentrated among wealthy households. Concentration curves for TB-associated symptoms are included in S1 Text.

Fig 5

Illness concentration index for 2007 and 2017 Việt Nam tuberculosis prevalence surveys.

Sampling weights were applied. A negative concentration index means that the health outcome (tuberculosis illness) is concentrated in those who are poor, while a positive index value means that the disease is concentrated in those who are wealthier. The concentration index is an expression of the area between the concentration curve (Fig 4) and the line of perfect equality. In the figure, the bars represent the mean, with the whiskers representing the distribution of data around the mean. The outlying data points are shown as circles above and below the whiskers.

In Table 2, the results of evaluations of the associations between tuberculosis prevalence and individual and household risks for each survey are shown separately. In the 2007 survey, we found that older age (prevalence ratio [PR] = 2.79; 95% CI 2.09, 3.49; p < 0.001) and male gender (PR = 1.61; 95% CI 1.29, 1.92; p < 0.001) were associated with increased tuberculosis prevalence. Living in a remote area was negatively associated with tuberculosis prevalence in 2007 (PR = −0.16; 95% CI −0.53, 0.20; p = 0.387) and in 2017 (PR = −0.09; 95% CI −0.46, 0.29; p = 0.644). The wealthiest households were less likely (PR = −0.41; 95% CI −0.81, −0.00; p = 0.048) to have tuberculosis than the poorest households in 2007. These associations were similar in direction in the 2017 survey, except for the association with household SEP, where the wealthiest participants were more likely to have tuberculosis (PR = 0.76; 95% CI 0.76, 1.16; p < 0.001).

Table 2

Associations between individual- and household-level variables and tuberculosis prevalence at each timepoint (2007 and 2017).

Variable	2007 survey (n = 94,156)		2017 survey (n = 61,763)
	PR (95% CI)	p-Value	PR (95% CI)	p-Value
Age (years)
15–24	Ref		Ref
25–34	1.04 (0.26; 1.81)	0.009	1.19 (−0.06; 2.43)	0.063
35–44	1.83 (1.12; 2.54)	<0.001	1.90 (0.71; 3.08)	0.002
45–54	2.05 (1.35; 2.76)	<0.001	2.10 (0.93; 3.27)	<0.001
55–64	2.36 (1.63; 3.09)	<0.001	2.57 (−0.51; 0.62)	<0.001
≥65	2.79 (2.09; 3.49)	<0.001	2.81 (1.64; 3.97)	<0.001
Gender
Female	Ref		Ref
Male	1.61 (1.29; 1.92)	<0.001	1.59 (1.25; 1.92)	<0.001
Region
North	Ref		Ref
Centre	−0.34 (−0.77; 0.98)	0.129	0.34 (−0.42; 0.72)	0.081
South	0.19 (−0.08; 0.47)	0.170	0.64 (0.30; 0.98)	<0.001
Type of residence
Urban	Ref		Ref
Rural	0.08 (−0.23; 0.39)	0.600	−0.46 (−0.79; −0.14)	0.005
Remote	−0.16 (−0.53; 0.20)	0.387	−0.09 (−0.46; 0.29)	0.644
Household socioeconomic position
Lowest	Ref		Ref
Lower middle	−0.22 (−0.54; 0.10)	0.183	0.08 (−0.30; 0.47)	0.671
Upper middle	0.19 (−0.17; 0.55)	0.309	0.39 (0.01; 0.76)	0.042
Highest	−0.41 (−0.81; −0.00)	0.048	0.76 (0.36; 1.16)	<0.001

PRs and CIs are estimated using log-binomial mixed effects statistical models. Coefficients are weighted for sampling stratification (differential cluster size, participation by age and gender, stratification by areas and post-stratification weight). CI, confidence interval; PR, prevalence ratio; Ref, reference value; TB, tuberculosis.

The results of the MLMs are shown in Table 3. We present the results of 3 mixed effects models with random intercepts. Model A is used to investigate the association between individual characteristics and tuberculosis prevalence while controlling for time and the provincial poverty rate, which is the measured as the percentage of the population below the poverty line. In Model B, we control for the district poverty rate as well as the absolute wealth of the household. In Model C, the district and relative SEP of households are controlled for to understand how individual and household risks explain the change in tuberculosis prevalence between the 2 surveys. We find that the difference in tuberculosis prevalence over time (effect size) reduces when we include indicators of household SEP and provincial poverty, which suggests that some of the observed change in tuberculosis prevalence can be explained by changes in provincial poverty.

Table 3

Multilevel analyses examining associations between individual-, household-, and neighbourhood-level explanatory variables and change in tuberculosis prevalence.

Variable	Model A		Model B		Model C
	PR (95% CI)	p-Value	PR (95% CI)	p-Value	PR (95% CI)	p-Value
Tuberculosis prevalence
Time (comparator: 2007)	−0.35 (−0.58; −0.12)	0.003	−0.35 (−0.69; −0.01)	0.041	−0.37 (−0.70; −0.04)	0.030
Individual
Age (years)
15–24	Ref		Ref		Ref
25–34	1.34 (0.60; 2.08)	<0.001	1.75 (0.97; 2.54)	<0.001	1.75 (0.97; 2.53)	<0.001
35–44	1.94 (1.25; 2.64)	<0.001	2.26 (1.46; 3.07)	<0.001	2.26 (1.46; 3.07)	<0.001
45–54	2.12 (1.42; 2.81)	<0.001	2.48 (1.59; 3.38)	<0.001	2.48 (1.58; 3.37)	<0.001
55–64	2.41 (1.70; 3.12)	<0.001	2.77 (1.93; 3.61)	<0.001	2.77 (1.93; 3.61)	<0.001
≥65	2.73 (2.04; 3.42)	<0.001	3.01 (2.24; 3.77)	<0.001	3.01 (2.24; 3.77)	<0.001
Gender male	1.42 (1.17; 1.68)	<0.001	1.33 (1.06; 1.60)	<0.001	1.33 (1.06; 1.60)	<0.001
Region
North			Ref		Ref
Centre			−0.02 (−0.65; 0.60)	0.944	−0.02 (−0.65; 0.61)	0.950
South			0.24 (−0.20; 0.67)	0.289	0.23 (−0.21; 0.67)	0.304
Type of residence
Urban	Ref		Ref		Ref
Rural	−0.13 (−0.38; 0.12)	0.313	−0.16 (−0.56; 0.24)	0.156	−0.17 (−0.56; 0.23)	0.410
Remote	−0.28 (−0.63; −0.06)	0.107	−0.28 (−0.65; 0.11)	0.429	−0.29 (−0.67; 0.09)	0.138
Household
Household SEP
Lowest					Ref
Lower middle					0.02 (−0.47; 0.51)	0.931
Upper middle					0.25 (−0.05; 0.56)	0.104
Highest					0.23 (−0.19; 0.66)	0.287
Household AWE			0.004 (−0.00; 0.01)	0.140
Province
Provincial poverty rate (2009)	−0.01 (−0.02; 0.01)		0.11 (0.02; 0.70)		0.10 (0.02; 0.68)

The dataset includes a total of 155,919 participants. PRs were estimated using mixed effects multilevel models with random intercepts. Model A shows individual-level regressors only, Model B shows individual-level and household-level variables while using the AWE to understand the impact of household wealth, while Model C uses a relative measure of household SEP. Provincial poverty rate is the percentage of the population living below US$2 per day. AWE, absolute wealth estimate; CI, confidence interval; PR, prevalence ratio; Ref, reference value; SEP, socioeconomic position.

Discussion

We found that in the context of rapid economic growth and equitable distribution of resources in Việt Nam, there was a shift in the distribution of tuberculosis from being concentrated among poor households to a more equal distribution among households of different SEP. In the 2007 survey, older age, being male, and living in an urban centre was associated with increased tuberculosis prevalence. Conversely, in the 2017 survey, the association between older age and tuberculosis prevalence decreased with urban living. MLMs showed the importance of provincial poverty in explaining some of the change in tuberculosis prevalence observed. Similar results were found when restricting the analysis to subclinical tuberculosis.

Studies investigating the association between reductions in tuberculosis incidence and economic development have been conducted in a range of settings [1,6,7,11,12,14]. Relationships between economic development and tuberculosis prevalence are challenging to examine given distal relationships that influence the causal pathway. Economic development may be measured as an increase in country GDP, which represents market productivity, but this is only one aspect of economic development. If economic development increases wealth inequality in a population, patients’ vulnerability to tuberculosis disease may increase [1]. The role of improved healthcare in mediating that relationship is unclear. In a multi-country analysis, Dye et al. showed that rates of decline in tuberculosis incidence were associated with biological, social, and economic determinants [7]. Focusing on poverty alleviation and social protection policies, Carter et al. found that reducing extreme poverty may reduce the global incidence of tuberculosis by 33%; simultaneously expanding social protection coverage may reduce incidence by 84% [11]. While Dye et al. found that health service programmatic indicators did not explain the reduction in tuberculosis incidence [7], Reeves et al. found that reductions in public spending (through economic recession) reduced spending on tuberculosis control and argued that this may lead to increased tuberculosis prevalence [12]. These studies examined associations between different components of economic development and tuberculosis prevalence, but empirical data were limited. In contrast, Siroka et al. used tuberculosis prevalence survey data from 8 countries to examine the association between household-level poverty and tuberculosis prevalence [15]. The study was cross-sectional and did not find a consistent association between household SEP and tuberculosis prevalence. From these studies we therefore understand that it is possible that the relationship between economic development and change in tuberculosis prevalence is not simply dependent on the household or on investment in health services but rather on a combination of risk factors across neighbourhood interactions.

However, economic growth and reduction in poverty may not be the only explanation for the change, as there were also improvements in tuberculosis diagnostics and health service access through an expansion of health insurance in Việt Nam [19,20]. Possible explanations for the results of our study therefore include that the rapid economic development in Việt Nam led to tuberculosis patients being wealthier in the second survey. However, this is mediated by lower participation in the second survey by wealthy households because of the expansion of the Việtnamese National Health Insurance, making the free health check-ups offered for participation in the tuberculosis prevalence survey less attractive [7,12]. Despite lower participation from relatively wealthier households, we found that tuberculosis burden was more concentrated among wealthy households in the 2017 survey than in the 2007 survey, suggesting selective participation.

We found that relative household SEP was weakly associated with tuberculosis prevalence after controlling for known individual-level risk factors such as age and gender [39]. Our finding of a tendency towards tuberculosis being less concentrated among poor households when measured over time corresponds to the findings of Ataguba et al. in South Africa [33]. However, in South Africa, economic development has been accompanied by persistently high levels of income inequality, and the effect was likely mediated by the expansion of the national ART programme, which disproportionally will have benefitted the poor. Our findings suggest that neighbourhood-level (provincial) poverty explains much of the variation in tuberculosis prevalence over time. Neighbourhood-level poverty may be a signal of fewer economic opportunities and therefore a greater vulnerability to tuberculosis [1].

A limitation of our study is the measure of household SEP used, which may have led to the misclassification of household SEP given the limited set of consumption data. Our measure of household-level SEP was primarily based on consumption data collected during the prevalence survey, and some of the important factors predicting poverty in Việt Nam such as education were not included in this measure. Furthermore, consumption data are sensitive to change over time; for example, an item that was a signal of prosperity in 2007 may no longer be a good indicator of wealth in 2017. This limitation was mitigated to some extent by using different household- and neighbourhood-level measures of poverty including rural residence, the region where a district is situated in Việt Nam, and the percentage of people in the district who are considered poor (the district poverty rate). We furthermore did not only rely on results based on the consumption-based asset index, but also estimated the absolute wealth of households, and the primary results of the study held across the different measurements used. A further limitation of the study is that while tuberculosis prevalence surveys are valuable for understanding socioeconomic differences in tuberculosis prevalence in countries, given that tuberculosis is a relatively rare disease in the population studied, there is limited power to explore the socioeconomic drivers of tuberculosis prevalence [15]. However, combining repeated cross-sectional surveys with provincial deprivation estimates during a period of remarkable economic growth provides valuable insights into the dynamics of the relationship between tuberculosis and economic development in Việt Nam. Lastly, it is possible that there may have been selection bias due to non-participation in the sampling by individuals during the second survey, in that, with improved economic welfare, there is less incentive for individuals to attend the free screening service provided by the survey, with therefore a bias towards poorer households enrolling. Our findings may therefore be underestimating the true population-level shift in the tuberculosis burden towards wealthier households.

Our results show the potential for tuberculosis prevalence reductions with general and equitable improvements in socioeconomic circumstances in a population. While promoting economic growth, and ensuring that resource distribution is equitable, falls outside the remit of specific national tuberculosis programmes, our study strengthens the case for a multi-sectoral response to tuberculosis [40], which we hope gives further encouragement to policies that aim to achieve this.

Conclusions

To our knowledge, this is the first study to use repeat direct measurements of tuberculosis burden to empirically examine the relationship between equitable economic development and a reduction in tuberculosis prevalence. We found that with equitable economic growth and a reduction in tuberculosis burden, tuberculosis became less concentrated among poor households in Việt Nam. The study highlights the important contribution of shared resources to not only reduce poverty but also shifting tuberculosis away from differentially impacting the poorest households.

Additional information related to the analysis.

(DOCX)

Click here for additional data file.

STROBE Checklist.

(DOC)

Click here for additional data file.

28 Jul 2021

Dear Dr Foster,

Thank you for submitting your manuscript entitled "Reducing the burden of tuberculosis among the poor: social determinants of the changing tuberculosis prevalence in Viet Nam" for consideration by PLOS Medicine.

Your manuscript has now been evaluated by the PLOS Medicine editorial staff and I am writing to let you know that we would like to send your submission out for external peer review.

However, before we can send your manuscript to reviewers, we need you to complete your submission by providing the metadata that is required for full assessment. To this end, please login to Editorial Manager where you will find the paper in the 'Submissions Needing Revisions' folder on your homepage. Please click 'Revise Submission' from the Action Links and complete all additional questions in the submission questionnaire.

Please re-submit your manuscript within two working days, i.e. by Jul 30 2021 11:59PM.

Login to Editorial Manager here: https://www.editorialmanager.com/pmedicine

Once your full submission is complete, your paper will undergo a series of checks in preparation for peer review. Once your manuscript has passed all checks it will be sent out for review.

Feel free to email us at plosmedicine@plos.org if you have any queries relating to your submission.

Kind regards,

Beryne Odeny

Associate Editor

PLOS Medicine

7 Oct 2021

Dear Dr. Foster,

Thank you very much for submitting your manuscript "Reducing the burden of tuberculosis among the poor: social determinants of the changing tuberculosis prevalence in Viet Nam" (PMEDICINE-D-21-03248R1) for consideration at PLOS Medicine.

Your paper was evaluated by a senior editor and discussed among all the editors here. It was also discussed with an academic editor with relevant expertise, and sent to independent reviewers, including a statistical reviewer. The reviews are appended at the bottom of this email and any accompanying reviewer attachments can be seen via the link below:

[LINK]

In light of these reviews, I am afraid that we will not be able to accept the manuscript for publication in the journal in its current form, but we would like to consider a revised version that addresses the reviewers' and editors' comments. Obviously we cannot make any decision about publication until we have seen the revised manuscript and your response, and we plan to seek re-review by one or more of the reviewers.

In revising the manuscript for further consideration, your revisions should address the specific points made by each reviewer and the editors. Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments, the changes you have made in the manuscript, and include either an excerpt of the revised text or the location (eg: page and line number) where each change can be found. Please submit a clean version of the paper as the main article file; a version with changes marked should be uploaded as a marked up manuscript.

In addition, we request that you upload any figures associated with your paper as individual TIF or EPS files with 300dpi resolution at resubmission; please read our figure guidelines for more information on our requirements: http://journals.plos.org/plosmedicine/s/figures. While revising your submission, please upload your figure files to the PACE digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at PLOSMedicine@plos.org.

We expect to receive your revised manuscript by Oct 28 2021 11:59PM. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

We ask every co-author listed on the manuscript to fill in a contributing author statement, making sure to declare all competing interests. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. If new competing interests are declared later in the revision process, this may also hold up the submission. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT. You can see our competing interests policy here: http://journals.plos.org/plosmedicine/s/competing-interests.

Please use the following link to submit the revised manuscript:

https://www.editorialmanager.com/pmedicine/

Your article can be found in the "Submissions Needing Revision" folder.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

We look forward to receiving your revised manuscript.

Sincerely,

Beryne Odeny,

PLOS Medicine

plosmedicine.org

-----------------------------------------------------------

Requests from the editors:

1) Please revise your title according to PLOS Medicine's style. Your title must be nondeclarative and not a question. It should begin with main concept if possible. Please place the study design in the subtitle (i.e., after a colon). For example, “Changing tuberculosis prevalence by socioeconomic status in Viet Nam: Analysis of population-level cross-sectional studies.”

2) The Data Availability Statement (DAS) requires revision. Please include an appropriate weblink to the LSHTM Data Compass website.

3) Abstract:

a) Please structure your abstract using the PLOS Medicine headings (Background, Methods and Findings, Conclusions).

b) Please combine the Methods and Findings sections into one section, “Methods and findings”.

c) Please ensure that all numbers presented in the abstract are present and identical to numbers presented in the main manuscript text.

d) Please include the important dependent variables that are adjusted for in the analyses.

e) In the last sentence of the Abstract Methods and Findings section, please describe the main limitation(s) of the study's methodology.

4) Please place the Author Summary after the Abstract.

5) Did your study have a prospective protocol or analysis plan? Please state this (either way) early in the Methods section:

a) If a prospective analysis plan (from your funding proposal, IRB or other ethics committee submission, study protocol, or other planning document written before analyzing the data) was used in designing the study, please include the relevant prospectively written document with your revised manuscript as a Supporting Information file to be published alongside your study and cite it in the Methods section. A legend for this file should be included at the end of your manuscript.

b) If no such document exists, please make sure that the Methods section transparently describes when analyses were planned, and when/why any data-driven changes to analyses took place.

c) In either case, changes in the analysis - including those made in response to peer review comments-- should be identified as such in the Methods section of the paper, with rationale.

6) Please ensure that the study is reported according to the STROBE guideline for observational studies, and include the completed STROBE checklist as Supporting Information. Please add the following statement, or similar, to the Methods: "This study is reported as per the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline (S1 Checklist)." The STROBE guideline can be found here: http://www.equator-network.org/reporting-guidelines/strobe/

7) Your study is observational and therefore causality cannot be inferred. Please remove language that implies causality throughout the manuscript, such as “determinants”, “contributed to”, “explanation for”, and so forth. Refer to associations instead.

8) Please avoid assertions of primacy ("We report for the first time....") by stating, “to our knowledge” or similar

9) Please clarify whether population case detection rates, treatment success rates, or earlier diagnosis changed during the study period in the country and may have affected prevalence.

10) Did you account for population changes in your multilevel models?

11) Figures and tables

a) Please include the total N at the top of table 2 & 3

b) Please define abbreviations such in the table footnotes. For example, SEP, AWE, CI

c) Please provide a meaning for the bars and whiskers in figures 2 and 5.

12) The terms gender and sex are not interchangeable (as discussed in http://www.who.int/gender/whatisgender/en/ ); please use the appropriate term.

13) In the Discussion, please include the limitations of the study, implications and next steps for research, clinical practice, and/or public policy.

14) References:

a) Please ensure in-text reference call outs are enclosed in square brackets. For example, “... countries [1,2]."

b) Please include access dates for all weblinks (e.g., Ref #14, 26), and ensure that all weblinks are current and accessible.

c) Please update ref #23 or delete if not published.

15) To help us extend the reach of your research, please provide any Twitter handle(s) that would be appropriate to tag, including your own, your coauthors’, your institution, funder, or lab.

Comments from the reviewers:

Reviewer #1: Dear authors,

Firstly, I congratulate the authors for the dedication and focus to fortify the science in tuberculosis thematic area. The manuscript is important to the public health and provides high importance to better understanding the impact of socio-economic factors on decrease the burden of tuberculosis, and in this way, for health decision making.

I strongly suggest to the authors to follow the PLOS Medicine's submission guidelines. Please use continuous line numbers, do not restart the numbering on each page. In the text, cite the reference number in square brackets according the PLOS Medicine's submission guidelines.

Please, organize the level 1 and 2 heading in the manuscript body according to https://journals.plos.org/plosmedicine/s/file?id=9cba/PLOS%20Manuscript%20Body%20Formatting%20Guidelines.pdf

I suggest reading the manuscript carefully, since there are many little mistakes that could be solved easily by the authors.

The manuscript requires professional editing in the English language to make it easy to read and understand.

I strongly suggest to the authors the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: Guidelines for Reporting Observational Studies.

It is available from von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP, et al. (2007) The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: Guidelines for Reporting Observational Studies. PLoS Med 4(10): e296. https://doi.org/10.1371/journal.pmed.0040296

I also suggest to the authors to verify the United to End TB: Every Word Counts launched by Stop TB Partnership. Available from: http://www.stoptb.org/assets/documents/resources/publications/acsm/LanguageGuide_FtorWeb20131110.pdf

The language guide supports the call for change in the upcoming Global Plan to End TB 2016-2020, which includes changing the mindset, language, and dialogue on TB as one of the key paradigm shifts required to reach the End TB goals.

Title page

There are unnecessary information in the title page. Please, follow the PLOS Medicine's submission guidelines (https://journals.plos.org/plosmedicine/s/file?id=3fac/PLOS%20Affiliations%20Formatting%20Guidelines.pdf).

Line 21 - 23: Delete these information

Line 32 - 36: Delete these information

Line 1 - 3: Delete these information

Line 5: Key words are not necessary according to the PLOS Medicine's submission guideline.

Title

Line 2: Please, replace the word "Viet Nam" by "Vietnam", according to the English spelling in the whole manuscript body.

I suggest to the authors the title "Social determinants and changing prevalence in decrease of the burden of tuberculosis in Vietnam".

Author summary

Page 3: Please delete the author summary heading. This section is not necessary in the manuscript and it does not follow the PLOS Medicine's submission guidelines. Merge this information in the background section, methods section, discussion section and conclusion section.

Abstract

The authors must follow the submission guidelines to provide a better abstract of the study. I suggest to the authors to minimize the use of abbreviations.

Line 10: Please, replace the word ˜Methods" by "Methods and findings".

Line 27: Please, replace the word "interpretation" by "conclusion"

Line 31: Please, delete the "funding" heading.

Background

The Background section should explain the background to the study, its aims, a summary of the existing literature and why this study was necessary or its contribution to the field. There are many blanks about the existing literature about the theme in the manuscript.

I strongly suggest an explanation how the Social Determinants of Health impact the prevalence and how they increase the burden of tuberculosis. A deep literature review about the theme is necessary to understand the contribution to the public health field.

Line: 2: Replace the word "Introduction" by "Background".

Line 4: Replace the abbreviation "TB" after the word "tuberculosis".

Line 5: The authors stated: "Ecological studies…", but the authors just cited one reference. Please, insert references in this statement.

Line 7 - 10: Please, rewrite this statement and insert a reference.

Line 24: Replace the word "Viet Nam" by "Vietnam".

Line 33 - 34: Please insert a reference in this statement.

Line 1 - 2: Please insert a reference in this statement.

Line 2 - 5: Rewrite the paragraph adding the aim of the study following the scientific writing criteria (verb in the infinite tense).

Methods

The authors must follow the STROBE statement to better report the sections in the manuscript body, specially in the methods section.

Add new headings according to the STROBE statement (study design, setting…)

The paragraphs do not have a logical sequential that allow to understand the aim of the study.

Line 23: TB is not associated only with the crowding places. Rewrite this statement and insert a reference.

Please, organize the level 2 heading in the methods section according to https://journals.plos.org/plosmedicine/s/file?id=9cba/PLOS%20Manuscript%20Body%20Formatting%20Guidelines.pdf

Line 20: Move the meaning of the abbreviation for the first time it appears.

In the Statistical Analysis section, why did the authors only use six variables to asset indices?

Results

The results section is well reported. There are major revisions in this section.

The results reported in this section are not clear. I suggest rearranging this section to make it clearer and more readable according to the STROBE Statement.

I suggest to the authors to use the comma to separate the number every third digit from the right.

Discussion

The discussion section is well reported. There are major revisions in this section.

I suggest to the authors a discussion based in the data reported in the results section and, how the findings can subsidize the policy makers in different scenarios to obtain the best outcomes in decrease TB burden.

I suggest to the authors to insert the limitations of the study in the last paragraphs in the discussion section.

Please, discuss the data from survey one and two. Is there a point of integration or similarity?

Line 11 - 13: Please include a reference in this statement.

Line 17: Insert a point after the letter "l" in the "at al."

Line 18: Reference correctly the citation "Carter at al.".

Line 35 - 36: Please include a reference in this statement.

Line 10: Insert a point after the letter "l" in the "at al."

Conclusion

Please include in the conclusion section how this study will support, for example, policy makers to better understanding the impact of socio-economic factors on increase the burden of tuberculosis orientating the efforts to accomplish the aims of The End TB strategy.

References

In the text, cite the reference number in square brackets according the PLOS Medicine's submission guidelines.

I suggest to the authors to insert some other references to be included in the discussion section. These studies were carried out in different scenarios and could improve this study.

1. https://doi.org/10.1371/journal.pone.0249822

2. https://doi.org/10.5588/ijtld.18.0149

3. http://dx.doi.org/10.2471/BLT.06.038331

4. https://doi.org/10.1016/j.jiph.2020.03.010

5. https://doi.org/10.1016/S1473-3099(09)70041-6

6. https://doi.org/10.1111/tmi.13409

Reviewer #2: Well done to the team. My comments are in the attached manuscript.

Reviewer #3: Congratulation for the excellent work and the well written manucript

Reviewer #4: I confine my remarks to statistical aspects of this paper.

First, there is the general issue of causal language. This is an observational study and causality cannot be inferred. Even in the title the word "determinants" is used. Not only should causal words be changed, but the "causal pathways" section should make it clear that these are *possible* causal pathways.

p. 9 line 4 This should really be factor analysis, not PCA. The results are often similar, but the goals are different. The goals here are clearly those of factor analysis - to uncover a latent factor from several measures that are related to it.

line 8-9 Why were these categorized? Categorizing a continuous variable is nearly always a mistake. The index can be used as is. Nonlinearity can be explored with splines. If the authors then want to compare various percentiles, they can do so.

line 15 Please explain what an "illness concentration curve" is.

p. 11 line 17-19 The two measures are also not inconsistent. But these results are inconsistent with the GINI results cited earlier.

Table 2 Do not categorize age or SES. See my blog post https://medium.com/@peterflom/what-happens-when-we-categorize-an-independent-variable-in-regression-77d4c5862b6c

The AIC is of no use here. It is only useful for comparing different models on the same data set.

Peter Flom

Any attachments provided with reviews can be seen via the following link:

[LINK]

Submitted filename: PMEDICINE-D-21-03248_reviewed_SN.pdf

Click here for additional data file.

30 Nov 2021

Submitted filename: responses V2.docx

Click here for additional data file.

24 Jan 2022

Dear Dr. Foster,

Thank you very much for re-submitting your manuscript "Social determinants of the changing tuberculosis prevalence in Việt Nam: analysis of population-level cross-sectional studies" (PMEDICINE-D-21-03248R2) for review by PLOS Medicine.

I have discussed the paper with my colleagues and the academic editor and it was also seen again by two reviewers. I am pleased to say that provided the remaining editorial and production issues are dealt with we are planning to accept the paper for publication in the journal.

The remaining issues that need to be addressed are listed at the end of this email. Any accompanying reviewer attachments can be seen via the link below. Please take these into account before resubmitting your manuscript:

[LINK]

***Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.***

In revising the manuscript for further consideration here, please ensure you address the specific points made by each reviewer and the editors. In your rebuttal letter you should indicate your response to the reviewers' and editors' comments and the changes you have made in the manuscript. Please submit a clean version of the paper as the main article file. A version with changes marked must also be uploaded as a marked up manuscript file.

Please also check the guidelines for revised papers at http://journals.plos.org/plosmedicine/s/revising-your-manuscript for any that apply to your paper. If you haven't already, we ask that you provide a short, non-technical Author Summary of your research to make findings accessible to a wide audience that includes both scientists and non-scientists. The Author Summary should immediately follow the Abstract in your revised manuscript. This text is subject to editorial change and should be distinct from the scientific abstract.

We expect to receive your revised manuscript within 1 week. Please email us (plosmedicine@plos.org) if you have any questions or concerns.

We ask every co-author listed on the manuscript to fill in a contributing author statement. If any of the co-authors have not filled in the statement, we will remind them to do so when the paper is revised. If all statements are not completed in a timely fashion this could hold up the re-review process. Should there be a problem getting one of your co-authors to fill in a statement we will be in contact. YOU MUST NOT ADD OR REMOVE AUTHORS UNLESS YOU HAVE ALERTED THE EDITOR HANDLING THE MANUSCRIPT TO THE CHANGE AND THEY SPECIFICALLY HAVE AGREED TO IT.

Please ensure that the paper adheres to the PLOS Data Availability Policy (see http://journals.plos.org/plosmedicine/s/data-availability), which requires that all data underlying the study's findings be provided in a repository or as Supporting Information. For data residing with a third party, authors are required to provide instructions with contact information for obtaining the data. PLOS journals do not allow statements supported by "data not shown" or "unpublished results." For such statements, authors must provide supporting data or cite public sources that include it.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript.

Please note, when your manuscript is accepted, an uncorrected proof of your manuscript will be published online ahead of the final version, unless you've already opted out via the online submission form. If, for any reason, you do not want an earlier version of your manuscript published online or are unsure if you have already indicated as such, please let the journal staff know immediately at plosmedicine@plos.org.

If you have any questions in the meantime, please contact me or the journal staff on plosmedicine@plos.org.

We look forward to receiving the revised manuscript by Jan 24 2022 11:59PM.

Sincerely,

Beryne Odeny,

PLOS Medicine

plosmedicine.org

------------------------------------------------------------

Requests from Editors:

1) Thank you for providing a link to the description of the dataset; however, this link is inactive and DOI cannot be found.

2) Please situate the Author Summary after the Abstract.

3) In the last sentence of the Abstract’s Methods and Findings section, please describe the main limitation(s) of the study's methodology.

4) In the Abstract conclusion, please include a sentence on implications and next steps for clinical practice and public policy.

5) Thank you for providing your STROBE checklist. Please replace the page numbers with paragraph numbers per section (e.g. "Methods, paragraph 1"), since the page numbers of the final published paper may be different from the page numbers in the current manuscript.

6) References:

a) Please include access dates for all weblinks (e.g., Ref #32, and ensure that all weblinks are current and accessible.

b) Please reformat the citation style into PLOS Medicine's format and ensure journal name abbreviations consistently match those found in the National Center for Biotechnology Information (NCBI) databases

Comments from Reviewers:

Reviewer #1: Dear authors,

Thank you for revising the manuscript considering the suggestions.

The manuscript is now technically sound.

I do not have additional comments on the revision in this manuscript version.

Reviewer #4: The authors have addressed my concerns and I now recoommend publication.

Peter Flom

Any attachments provided with reviews can be seen via the following link:

[LINK]

31 Jan 2022

Submitted filename: response.docx

Click here for additional data file.

3 Feb 2022

Dear Dr Foster,

On behalf of my colleagues and the Academic Editor, Dr. Amitabh Bipin Suthar, I am pleased to inform you that we have agreed to publish your manuscript "Social determinants of the changing tuberculosis prevalence in Việt Nam: analysis of population-level cross-sectional studies" (PMEDICINE-D-21-03248R3) in PLOS Medicine.

Before your manuscript can be formally accepted you will need to complete some formatting changes, which you will receive in a follow up email. Please be aware that it may take several days for you to receive this email; during this time no action is required by you. Once you have received these formatting requests, please note that your manuscript will not be scheduled for publication until you have made the required changes.

In the meantime, please log into Editorial Manager at http://www.editorialmanager.com/pmedicine/, click the "Update My Information" link at the top of the page, and update your user information to ensure an efficient production process.

PRESS

We frequently collaborate with press offices. If your institution or institutions have a press office, please notify them about your upcoming paper at this point, to enable them to help maximise its impact. If the press office is planning to promote your findings, we would be grateful if they could coordinate with medicinepress@plos.org. If you have not yet opted out of the early version process, we ask that you notify us immediately of any press plans so that we may do so on your behalf.

We also ask that you take this opportunity to read our Embargo Policy regarding the discussion, promotion and media coverage of work that is yet to be published by PLOS. As your manuscript is not yet published, it is bound by the conditions of our Embargo Policy. Please be aware that this policy is in place both to ensure that any press coverage of your article is fully substantiated and to provide a direct link between such coverage and the published work. For full details of our Embargo Policy, please visit http://www.plos.org/about/media-inquiries/embargo-policy/.

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Thank you again for submitting to PLOS Medicine. We look forward to publishing your paper.

Sincerely,

Beryne Odeny

PLOS Medicine