Zhang et al.

BMC Public Health (2020) 20:1557

https://doi.org/10.1186/s12889-020-09627-z

RESEARCH ARTICLE Open Access

Association between economic

development level and tuberculosis
registered incidence in Shandong, China
Qian-yun Zhang1,2†, Dong-mei Yang3†, Lin-qing Cao3, Jin-yue Liu4, Ning-ning Tao5, Yi-fan Li1, Yao Liu1,
Wan-mei Song1,2, Ting-ting Xu1, Shi-jin Li1,2, Qi-qi An1,2, Si-qi Liu1,2, Lei Gao6, Wan-yan Song7 and Huai-chen Li1,8*

Abstract
Background: Tuberculosis (TB) is one of the major infectious diseases that seriously endanger people’s health. In
Shandong province, the relationship between the level of economic development and TB incidence has not been
studied. This study aims to provide more research basis for the government to prevent and control TB by exploring
the impact of different economic factors on TB incidence.
Methods: By constructing threshold regression model (TRM), we described the extent to which different economic
factors contribute to TB registered incidence and differences in TB registered incidence among seventeen cities
with different levels of economic development in Shandong province, China, during 2006–2017. Data were
retrieved from the China Information System for Disease Control and Prevention.
Results: Per capita medical expenditure (regression coefficient, -0.0314462; SD, 0.0079305; P > |t|, 0.000) and per
capita savings (regression coefficient, 0.0001924; SD, 0.0000566; P > |t|, 0.001) passed the significance test at the level
of 1%.They are the two economic indicators that have the greatest impact on TB registered incidence. Through the
threshold test, we selected the per capita savings as the threshold variable. In the three stages of per capita savings
(<9772.8086 China Yuan(CNY); 9772.8086–33,835.5391 CNY; >33,835.5391 CNY), rural per capita income always has a
significant negative impact on the TB registered incidence (The regression coefficients are − 0.0015682, − 0.0028132
and − 0.0022253 respectively. P is 0.007,0.000 and 0.000 respectively.).In cities with good economies, TB registered
incidence was 38.30% in 2006 and dropped to 25.10% by 2017. In cities with moderate economies, TB registered
incidence peaked in 2008 at 43.10% and dropped to 27.1% by 2017.In poorer cities, TB registered incidence peaked
in 2008 at 56.30% and dropped to 28.9% in 2017.
(Continued on next page)

* Correspondence: [email protected]
†
Qian-yun Zhang and Dong-mei Yang are authors contributed equally to this
article and share the first authorship.
1
Department of Respiratory Medicine, Shandong Provincial Hospital Affiliated
to Shandong University, Shandong Provincial Hospital Affiliated to Shandong
First Medical University, No. 324, Jingwuweiqi Road, Huaiyin District, Jinan
250021, Shandong, China
8
Shandong University of Traditional Chinese Medicine, Jinan 250355,
Shandong, China
Full list of author information is available at the end of the article

© The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,
which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if
changes were made. The images or other third party material in this article are included in the article's Creative Commons
licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons
licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain
permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the
data made available in this article, unless otherwise stated in a credit line to the data.

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Zhang et al. BMC Public Health (2020) 20:1557 Page 2 of 9

(Continued from previous page)

Conclusion: We found that per capita savings and per capita medical expenditure are most closely related to the
TB incidence. Therefore, relevant departments should formulate a more complete medical system and medical
insurance policy to effectively solve the problem of “difficult and expensive medical treatment”. In order to further
reduce the TB incidence, in addition to timely and accurate diagnosis and treatment, it is more important for
governments to increase investment in medicine and health care.
Keywords: Tuberculosis, Registered incidence, Economic development

Background development and the incidence of TB, especially in

Although the target of the Millennium Development China [9–11]. It can be seen from foreign studies that
Goals to halve tuberculosis prevalence and mortality by increasing social protection and public health expend-
2015, compared to 1990 levels, has been achieved iture is closely related to reducing the incidence, mortal-
around the world, TB remains a major infectious disease ity, and cure success rate of TB, but whether the results
that threatens global health [1].The World Health of foreign studies are applicable to China or Shandong
Organization (WHO) estimates that 10 million people Province is not yet known. Foreign countries mainly ex-
worldwide suffer from TB and 1.6 million people die in plore the impact of factors such as per capita GDP and
2017.The social and economic impacts are devastating population density on TB, as well as the differences in
[2, 3].The End TB Strategy which was adopted by the TB incidence among countries with different income
World Health Assembly in May, 2014, outlines an over- levels [10, 12], and our research has refined economic
all target of reducing global TB incidence and mortality factors based on their researches and increased per
by 90 and 95% respectively by the year of 2035 [4] .This capita Indexes such as per capita savings, number of
strategy emphasizes the importance of social determi- doctors per capita, number of beds per capita, and Eng-
nants for prevention and care of TB, including policies el’s coefficient, etc., to explore whether these factors will
to alleviate poverty, and social protection programme. affect the incidence of TB. This study aims to examine
The non-spatial ecological regression analyses conducted the association between the registered incidence of TB
by Guy Harling and Marcia C. Castro found higher TB and the level of economic development in Shandong
incidence associated with urbanization, population dens- Province with the data from seventeen cities in Shan-
ity, poor economic conditions, household crowding, dong province from 2006 to 2017, and to find out which
worse health and healthcare indicators [5]. The content indicators in economic aspects contribute to TB preven-
of this strategy is supported by many research results [6, tion and control and to what extent, so as to provide a
7]. Hence the 2010 lancet tuberculosis series appeal to a basis for the prevention and control of tuberculosis in
renewed emphasis on tackling poverty and promoting Shandong Province.
social protection to curb epidemics [8].At present, China
is still one of the world’s 30 countries with a high bur-
den of TB, with about 900,000 new cases of TB every Methods
year, ranking the third in the world [1]. In China, people Study population and setting
either seek medical treatment voluntarily because of ill- Short-course treatment and free drug delivery have led
ness or routine medical examinations before employ- to a significant reduction in TB mortality. However,
ment and enrollment, and are diagnosed with there are still many problems and challenges in TB con-
tuberculosis through various laboratory tests and im- trol in China because of its huge population. The num-
aging. Once the patient is diagnosed with tuberculosis, it ber of TB cases is still high. China’s current TB control
will be immediately reported to the Center for Disease service system and control capacity cannot meet the
Control and Prevention (CDC) at all levels. Patients can needs of the new situation. According to Shandong pro-
get anti-tuberculosis drugs for free at designated hospi- vincial bureau of statistics in 2017, Shandong’s GDP
tals, but the cost of their examinations is not free. And (gross domestic product) is 7.26 trillion yuan, accounting
we found in the medical activities that most of the tu- for 8.8% of the country’s total, ranking third in the coun-
berculosis patients in outpatient clinics came from rural try, with a total population of 10.06 million and 7.2% of
areas. Therefore, we want to study whether there is a re- the country’s total, ranking second in the country
lationship between the economy and the incidence of [13].Therefore, reducing the TB incidence in Shandong
TB and what this relationship is like. Despite a lot of re- Province is of great significance to the prevention and
search on TB abroad, only a few studies have investi- control of tuberculosis in China. There are a total of 17
gated the relationship between the level of economic cities in Shandong Province, and we have included all

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Zhang et al. BMC Public Health (2020) 20:1557 Page 3 of 9

rural and urban populations of these 17 cities into the The original tuberculosis registered incidence and eco-
scope of the study. nomic indicator data we obtained are from 2005 to
2017.However, due to the large missing data in 2005, the
Data sources and study design data for this year were not included in the scope of our
The annual incidence of TB is one of the most import- study. In order to ensure the comprehensiveness and
ant indicators in TB burden, which can reflect the public availability of the collected data, we finally select the
health significance of TB. We obtained data at the city panel data of seventeen cities in Shandong province
level for TB registered incidence per 100,000 population from 2006 to 2017.Due to some missing observations,
from the Tuberculosis Information Management System, interpolation method was used to supplement the data
Chinese Center for Disease Control and Prevention in the actual analysis process, and consumer price index
(CDC). Once the hospital confirms the patient as a tu- (CPI) was used for the flattening of all economic indica-
berculosis patient, it will be reported to the disease con- tors to make the data more representative. The
trol center of each region within the prescribed time. interpolation method is that an interpolation line is
The system of the CDC contains information about all established based on the previous data and the next data
tuberculosis patients in Shandong Province. In the panel of the missing value in the column, and then the missing
data, we take the total registered incidence of urban and value is filled in by the function value of the missing
rural areas in each city every year as the TB registered point in the linear interpolation function. CPI refers to
incidence in that city. Misdiagnosis and missed diagnosis the consumer price index. It is an important data used
of tuberculosis, incomplete coverage of surveillance are by the government to measure inflation and an index
the main reasons for uncertainty about published esti- that measures the overall level of prices in the market. It
mates. In clinical work, misdiagnosis and missed diagno- reflects the supply-demand relationship and price trends
sis of diseases are inevitable. Misdiagnosis or missed of consumer goods. Flattening variables is that we calcu-
diagnosis will inevitably lead to high or low registered late the ratio of the CPI in 2007 to the CPI in 2006, and
incidence of tuberculosis. However, the data provided by then divide the corresponding economic data by this ra-
the CDC is the only official data, and we believe they tio to get the final data, and so on for other years. By
have minimized the error. And data on population and using CPI to flatten the data, the impact of price changes
economy were provided by Shandong Statistical Year- on the values of economic indicators can be eliminated
book annually. Economic indicators included per capita so that the data is more representative. In this paper, the
GDP, urban per capita disposable income and rural an- economics software STATA 14.0 is used for regression,
nual net income, population density, urbanization rate, and the regression results are shown in Table 1. In
per capita medical expenditure and the per capita sav- Table 1, the regression coefficient ± indicates that the
ings, number of beds per capita, number of doctors per economic indicator is positively or negatively correlated
capita, rural engel coefficient, urban engel coefficient. In with the registered incidence rate, and the value repre-
the panel data, values of all economic indicators in this sents the degree of increase or decrease.
study are the average of the year.The data used in this Panel data analysis of factors related to TB registered
paper are panel data, covering the years 2006–17 from incidence was carried out in the above-mentioned art-
seventeen cities in Shandong Province. icle, but has the effect of different variables on TB regis-
tered incidence been linear? In fact, the nonlinear
Procedures relationship between variables is also common. Based on
We consider the fixed effect model due to the different the linear analysis, we extended the original research
situations of each city. The fixed effect model is con- from the linear framework to the nonlinear. We used
structed as follows: PTRit = αi + β1GDPit + β2TIit + the threshold regression model (TRM) to explore
β3RIit + β4DOPit + β5URit + β6MEit + β7HSit + β8BNit + whether the influence of relevant factors on the regis-
β9DNit + β10RECit + β11TECit + εit. Wherein, αi is the tered incidence of TB under the effect of threshold vari-
fixed effect of individuals, representing the special effect ables is different in stages. We adopted panel threshold
of the city i and reflecting the individual differences of model proposed by Hansen (1999). Taking per capita
each city. β is the coefficient of various variables.εit rep- savings as a threshold, we studied the influence of other
resents the random disturbance term. In order to further independent variables on the dependent variable under
test the advantages of the fixed effect model, the Haus- the condition of different per capita savings. In order to
mann test was carried out. The Hausmann test results consider the heterogeneity among different sections, an
showed that the P value was 0.0051, rejecting the null indicative function of threshold variable was introduced
hypothesis. Therefore, compared with the mixed model into the model. The specific equation is as follows yit =
and the random effect model, the fixed effect model was μit + β1xitI(qit ≤ γ) + β2xitI(qit > γ) + eit. . t represents time,
the best. i represents different cities, γ is the unknown threshold,

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Zhang et al. BMC Public Health (2020) 20:1557 Page 4 of 9

Table 1 Regression Results of Economic Indicators

Variable RC SD t P > |t|
Per capita GDP 0.0001319 0.0000768 1.72 0.088*
Urban per capita disposable income 0.0003896 0.0001897 2.05 0.041**
Rural annual net income -0.0017531 0.0007222 −2.43 0.016**
Population density 0.0983965 0.0492288 2.00 0.047**
Urbanization rate −0.0544054 0.0990188 −0.55 0.583
Per capita medical expenditure −0.0314462 0.0079305 −3.97 0.000***
Per capita savings 0.0001924 0.0000566 3.40 0.001***
Number of beds per capita 0.0005271 0.0014147 0.37 0.710
Number of doctors per capita −0.0002612 0.0026471 −0.10 0.921
Rural engel coefficient −0.161335 0.2303641 −0.70 0.485
Urban engel coefficient 0.1908406 0.0991439 1.92 0.056*
Note. RC Regression coefficient, SD Standard deviation, GDP Gross domestic product
The number of * indicates the strength of the correlation

qit is the threshold variable, eit~iid(0, δ2) is the random This method divides different cross-section individuals
disturbance term, and I(*) is the index function. In fact, into different regions according to the threshold, and
this model is equivalent to a piecewise function model. different intervals have different regression equation ex-
When the threshold variable qit exceeds the threshold pressions. By separately estimating the corresponding
value γi, structural mutation occurs in the model. The coefficients, the relationship between the independent
model uses the Bootstrap method to retrieve different variable and the dependent variable in each area is more
threshold variables (ie, various economic indicators) in accurate. Firstly, the indicator per capita savings is used
order to determine whether there is a threshold effect. as the threshold, and the single threshold test shows

Fig. 1 LR Statistics (Two thresholds at different levels of per capita savings have a significant impact on TB registered incidence).
LR = likelihood ratio

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Zhang et al. BMC Public Health (2020) 20:1557 Page 5 of 9

γ1 = 9772.8086. Meanwhile, the null hypothesis that results of dynamic changes in TB registered incidence in
there is no threshold is rejected. The test of the two the three regions and the results are shown in Fig. 2.
thresholds continued, showing that the second threshold
value was 33,835.5391 China Yuan(CNY), which also Results
rejected the null hypothesis with only one threshold. At The results of fixed effect in Table 1 show that among
this point, the first threshold value is back-checked, and all the variables, per capita medical expenditure (RC,
fixed γ2 is back-checked for the new threshold value, -0.0314462; SD, 0.0079305; P > |t|, 0.000) and per capita
and the threshold value is also 33,835.5391 CNY, indi- savings(RC, 0.0001924; SD, 0.0000566; P > |t|, 0.001)
cating that there are two thresholds in this model, passed the significance test at the level of 1%. That is to
9772.8086 CNY and 33,835.5391 CNY respectively. After say, these two variables are most closely related to the
determining the threshold value, we divided the per registered incidence of TB. Per capita medical expend-
capita savings into three intervals: low savings < iture is negatively correlated with TB registered inci-
9772.8086 CNY; moderate savings 9772.8086 CNY- dence, and per capita savings is positively correlated
33835.5391 CNY; high savings > 33,835.5391CNY) and with TB registered incidence.The three variables of
discussed on different intervals to obtain more conclu- urban per capita disposable income(RC, 0.0003896; SD,
sions. We can also clearly see that there are two different 0.0001897; P > |t|, 0.041), rural annual net income(RC,
threshold values of the model from the Fig. 1, both of -0.0017531; SD, 0.0007222; P > |t|, 0.016) and population
which are very significant, proving that the calculated density(RC, 0.0983965; SD, 0.0492288; P > |t|, 0.047)
threshold value is representative to some extent. We passed the test at the significance level of 5%.The rela-
imported the above data into statistical software STAT tionship between these variables and the registered inci-
A14.0 and the TRM proposed by Yujun Lian (2015) was dence of TB is significant. Urban per capita disposable
used for modeling. Based on the fixed effect model in income and population density are positively correlated
the above-mentioned paper, some insignificant explana- with TB registered incidence, while rural annual net in-
tory variables were deleted, and the regression results come is negatively correlated with TB registered inci-
with per capita savings as the threshold variable were dence. Per capita GDP(RC, 0.0001319; SD, 0.0000768;
shown in Table 2. P > |t|, 0.088)and urban engel coefficient(RC, 0.1908406;
To study the relationship between regional economic SD, 0.0991439; P > |t|, 0.056) pass the test at the signifi-
level and tuberculosis registered incidence, we artificially cance level of 10%.Relatively speaking, the statistical sig-
divided cities into three groups based on the GDP levels nificance of these two variables is not so obvious.
of 17 cities over the 12 years. The cities with good eco- Table 2 shows the impact of other indicators on TB
nomic development include Jinan, Dongying, Weihai, registered incidence at different stages of per capita sav-
Qingdao, Yantai and Zibo; Cities with moderate eco- ings. It can be seen from Table 1 that the four economic
nomic development include Binzhou, Laiwu, Rizhao, indicators and registered incidence, such as urbanization
Weifang, Zaozhuang and Tai’an; The cities with poor rate, number of beds per capita, number of doctors per
economic development include Jining, Dezhou, Liao- capita, and rural engel coefficient, have no obvious stat-
cheng, Linyi and Heze. By calculating the average of TB istical significance. Therefore, they are eliminated when
registered incidence in three regions, we obtained the we perform threshold regression statistics. We take rural
per capita income as an important variable for threshold
Table 2 The Regression Results with Per Capita Savings as The regression, and the results are shown in Fig. 1. In the
Threshold Variable case of different stages of per capita savings, rural per
Variable RC SD P > |t| capita income always has a significant negative impact
Per capita GDP 0.0003178 0.000067 0.000 on the registered incidence of TB (The regression coeffi-
Urban per capita disposable income 0.0002412 0.0001623 0.139 cients are − 0.0015682, − 0.0028132 and − 0.0022253 re-
The population density 0.125963 0.0375508 0.001 spectively. P > 0.007, 0.000, 0.000 respectively.) That is to
Per capita medical expenditure −0.0253607 0.0059952 0.000
say, with the continuous increase of per capita income
of rural residents, the registered incidence of TB gener-
Urban engel coefficient 0.2440182 0.0862448 0.005
ally decreases. However, at the same time, the impact of
Rural per capita income 1 −0.0015682 0.0005788 0.007 rural per capita income on the registered incidence of
Rural per capita income 2 −0.0028132 0.0005677 0.000 TB has a significant staged difference. From Fig. 1, we
Rural per capita income 3 −0.0022253 0.0005483 0.000 can see that when per capita savings reached the first
Note. TB Tuberculosis, RC Regression coefficient, SD Standard deviation, GDP threshold of 33,835.5391 CNY and the second threshold
Gross domestic product of 9772.8086 CNY, the impact of rural per capita income
Rural per capita income 1, 2 and 3 means that rural per capita income has
different effects on incidence under three different thresholds, so it is divided
on TB registered incidence changed. At the medium sav-
into three levels ings level (9772.8086 CNY- 33835.5391CNY), rural per

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Zhang et al. BMC Public Health (2020) 20:1557 Page 6 of 9

Fig. 2 Dynamic changes in TB registered incidence in regions with different levels of economic development during
2006–2017. Note.TB = tuberculosis

capita income had the greatest impact on the total regis- examinations, and tuberculosis patients are not allowed
tered incidence of TB.Rural per capita income had the to enter school and enter the job. These measures are
least impact on the overall registered incidence of TB at helpful for the early detection and treatment of tubercu-
the low savings level (less than 9772.8086 CNY). losis patients, thereby reducing transmission and making
Dynamic changes in TB registered incidence in regions the registered incidence rate gradually decline. But in
with different levels of economic development during 2007–2008, the registered incidence increased in all
2006–2017 are shown in Fig. 2. In cities with good econ- three regions. By 2017, the differences between the three
omies, the peak was 38.30% in 2006.Then it fell steadily, regions were small.
to 25.10% by 2017. In cities with moderate economic de-
velopment, the registered incidence peaked in 2008 at Discussion
43.10%.After that, there was a big fluctuation, and by To our knowledge, ecological studies of the TB burden
2017, it had dropped to 27.1%.In poorer cities, the regis- have been carried out in America, Europe and Africa [5,
tered incidence peaked in 2008 at 56.30%. Then it fell 9–11].But such research is limited in China. This paper
rapidly, with fluctuations, to 28.9% in 2017. The average provides original, large population and long-term based
TB registered incidence fell fastest in poorer regions, data on the relationship between economy and TB regis-
from 48.30% in 2006 to 28.90% in 2017.First of all, for tered incidence in the second largest province located in
more than 10 years, in order to achieve the goal of fully the eastern coast of China. Our findings show a clear
building a well-off society in 2020, the Chinese govern- ecological association between the economic develop-
ment has continuously increased investment in eco- ment and TB registered incidence. Rural annual net in-
nomic support and medical and health care in poor come and per capita medical expenditure have
areas. Secondly, once a tuberculosis patient is diagnosed significant negative effects on the registered incidence.
in China, he can get anti-TB drugs for free. Finally, With the increase of the above two indicators, the regis-
people are engaged in routine pre-employment medical tered incidence shows a decreasing trend. The influence

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Zhang et al. BMC Public Health (2020) 20:1557 Page 7 of 9

of medical expenditure on the registered incidence rate is gradually decreasing, and the registered incidence of TB
greater than that of rural annual net income. This result in all regions was gradually stable. The following reasons
corroborates the findings of a recent global analysis that might explain this phenomenon. First, areas with high
social protection spending is strongly associated with economic development have better medical and educa-
lower tuberculosis case notification, incidence, and mor- tional conditions. On the one hand, the hospital’s fund-
tality rates [12]. And it also confirms the report that eco- ing for tuberculosis can be adequate. On the other hand,
nomic development and increased access to health care people’s health care awareness is relatively stronger, and
both reduce TB incidence [14]. Therefore, the government the living environment is better. They will use more in-
should formulate more policies to benefit the people, es- come for health care, medicines, nourishment, etc., so
pecially in rural areas, to promote the economic develop- that the incidence of tuberculosis can be effectively pre-
ment of rural areas and alleviate poverty. More vented and controlled. Second, Due to economic con-
importantly, the government should increase spending on straints, people in poor areas will not have a strong
medicine and health care especially in the economically awareness to see a doctor unless there is a major illness.
backward countryside, and actively screen for suspicious Even if they are diagnosed with tuberculosis, compared
TB cases to effectively reduce the TB incidence [15, 16]. with people in economically developed areas, their com-
Population density, per capita savings and urban engel pliance with treatment will be worse. After the country
coefficient have significant positive effects on the regis- introduced policies launched by WHO such as short-
tered incidence. This conclusion confirms previous re- course chemotherapy, free medication, and active
search that the increased population density increases the screening for tuberculosis before entering school, poor
burden of TB [17, 18].With the increase of such indica- areas benefited more than developed areas. Therefore,
tors, the registered incidence shows an increasing trend. while the overall registered incidence has decreased, the
The urban engel coefficient has the greatest impact on the registered incidence in poor areas has become more sig-
TB registered incidence. Engel’s Coefficient refers to the nificant. The studies conducted by Kui Liu and Dong D
proportion of food consumption in households. It means also found that economically underdeveloped areas with
that the lower the household income, the larger the pro- a high TB incidence benefit more [22, 23].Their research
portion of the household income or the total household results also confirm our discovery and explanation.
expenditure spent on food, the larger the engel coefficient, As far as we know, this is the first study in Shandong
which means the poorer the family is. Naturally, residents province to assess the impact of economic development
will spend less on health care. levels on TB registered incidence. Our sample data is large,
Per capita GDP, urban per capita disposable income including seventeen cities in Shandong province. We exam-
also have positive effects on the registered incidence. ined the impact of the following economic factors on TB
This result is not consistent with previous studies abroad registered incidence: per capita GDP, urban per capita dis-
[12]. Their data show that countries with higher GDP posable income and rural annual net income, population
per capita have lower incidence and mortality of TB density, urbanization rate, per capita medical expenditure
than others among countries with different income and the per capita savings, number of beds per capita, num-
levels. For this result, we take into account the increase ber of doctors per capita, rural engel coefficient, urban
of per capita GDP and urban disposable income, which engel coefficient. Both linear and nonlinear patterns were
means that people’s material living conditions become explored to determine the relationship between levels of
better and more unhealthy lifestyles are derived, and the economic development and TB registered incidence. First,
increasing prevalence of diabetes increases the popula- we used panel data regression to examine the impact of dif-
tion’s susceptibility to tuberculosis bacteria. Supporting ferent economic factors on TB registered incidence. And
this relationship is a wealth of evidence, with many stud- then we used the TRM to extend the study from a linear
ies showing that diabetes affects TB registered incidence framework to a nonlinear one. Through the threshold test,
and treatment outcomes [19–21]. Moreover, per capita we selected per capita savings as a threshold variable and
GDP is an average indicator. When the gap between rich examined the impact of other economic factors on TB reg-
and poor increases, per capita GDP will no longer be istered incidence at different levels of savings. Furthermore,
representative. Therefore, we cannot exclude the inter- we divided seventeen cities in Shandong province into three
ference of the above factors. regions with different levels of economic development, and
As we can see from Fig. 2, on the whole, except for conducted a horizontal and vertical comparison to more
small fluctuations, the registered incidence of TB in all directly explore the impact of different levels of economic
cities showed a downward trend and it fell fastest in development on the registered incidence of TB. Overall, in-
poorer regions. The registered incidence of TB in areas creased economic input is critical, and many studies have
with good economic development was lower than that in emphasized the importance of social protection interven-
areas with poor economic development, but the gap was tions to reduce the burden of TB [24, 25].

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Zhang et al. BMC Public Health (2020) 20:1557 Page 8 of 9

There are still some limitations in our research. First, control. For example, the government should focus on
there is a bias in the diagnosis and registration of TB pa- increasing economic investment in poor areas to shorten
tients. Most TB cases are diagnosed on the basis of clin- the gap between rich and poor. Relevant health depart-
ical and radiological evidence, which may lead to some ments should formulate the point-to-point support pol-
misdiagnosis and missed diagnosis. In poor areas, due to icy of the third-class hospitals for rural township
economic constraints, some people with insignificant hospitals, provide special training for tuberculosis diag-
clinical symptoms may not seek medical care, resulting nosis and treatment for doctors in township hospitals,
in fewer TB cases registered than the true number. We and regularly carry out tuberculosis free diagnosis activ-
have to acknowledge that some cities provided high or ities, and increase the publicity of tuberculosis-related
low registered incidence of TB. However, the extent of knowledge, and reduce some tuberculosis special inspec-
these inaccuracies is not known. Underestimates might tion costs. The above measures are helpful for the early
arise from under-reporting or under-diagnosis of cases. detection, early diagnosis and early treatment of tuber-
Nonetheless, we believe CDC estimates are the best data culosis patients. A more refined measurement of eco-
available for TB registered incidence. In future research, nomic investments on a provincial scale would be
perhaps we can reduce the bias caused by misdiagnosis helpful to determine which programmes or policies are
and missed clinics by improving the diagnostic level of best at reducing TB burden. The benefits of increased
imaging doctors and encouraging people to seek medical economic investments go beyond TB and would be
treatment in time. Additionally, the economic develop- likely to affect the incidence and mortality of other com-
ment might not have an immediate effect on TB regis- municable and non-communicable diseases, especially
tered incidence. We considered the problem of those with a well documented association with poverty.
hysteresis, but we did not know the specific length of Our findings suggest that economic investments could
time. Fortunately, the economic development in Shan- contribute to a reduced TB incidence. However, further
dong province has been relatively stable in recent years studies are needed to confirm the relationship between
due to its traditional and conservative culture. In future the level of economic development and TB incidence
studies, we may be able to establish a model of the delay both in developed and developing countries.
effect between the economy and the TB registered inci-
dence, making the study more scientific. In addition, Conclusion
from 2007 to 2008, the TB registered incidence in- In conclusion, we found the effects of different economic fac-
creased in all areas in Shandong province and it peaked tors on TB registered incidence and to what extent. Our re-
in 2008 in both poor and moderate regions. But we sults contribute to the understanding of which economic
didn’t find a possible reason for this, so further research policies can better reduce the TB burden. In addition to de-
is needed. Moreover, if we can completely separate veloping new accurate and timely diagnostic methods, in-
urban and rural areas and explore the relationship be- creased economic input is important for future TB control.
tween economic indicators and TB registered incidence In the economically underdeveloped areas with high preva-
separately, it will be more scientific. But according to lence in Shandong province, on the one hand, medical insti-
our national conditions, it is difficult to distinguish be- tutions should regularly train radiologists in various places to
tween rural and urban populations completely and ac- improve their diagnostic skills in order to reduce missed
curately. As far as most provinces in China (including diagnosis and misdiagnosis. On the other hand, the govern-
Shandong Province) are concerned, many people live in ment should increase the publicity of the basic knowledge of
cities although they are registered in rural areas. How- tuberculosis in the community to encourage the population
ever, we do not know the specific proportion. So the TB to see a doctor in time and actively screen suspected tuber-
incidence in each city is the total incidence of the city’s culosis cases and realize early detection and treatment by in-
urban and rural areas. Finally, TB is preventable and creasing medical and health investment, which will greatly
controllable both in groups and individuals, but we ac- reduce the morbidity and mortality of TB.
knowledge that the study might not necessarily be ap-
plicable on an individual level. It can only provide Supplementary information
certain research basis for the prevention and control of Supplementary information accompanies this paper at https://doi.org/10.
1186/s12889-020-09627-z.
TB by relevant departments. We have not estimated the
registered incidence in China. Whether these results can Additional file 1. Panal data. The panel data contains the registered
be used in other provinces of China and globally is incidence of tuberculosis and various economic indicators in 17 cities in
controversial. Shandong Province from 2005 to 2017.
These results would strengthen the analysis and pro- Additional file 2. The original data of threshold regression model. All
data in the excel are the median of 12-year data (95% confidence
vide information for the government departments, guid- interval).
ing them to make economic investment decisions in TB

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Zhang et al. BMC Public Health (2020) 20:1557 Page 9 of 9

Abbreviations 2. WHO. Global TB report 2018. Geneva: World Health Organization; 2018.
TB: Tuberculosis; TRM: Threshold regression model; WHO: World Health http://www.who.int/tb/publications/global_report/en/.
Organization; GDP: Gross domestic product; RC: Regression coefficient; 3. World Health Organization. Global tuberculosis report 2018. Geneva: The
SD: Standard deviation; CDC: Chinese Center for Disease Control and Organization; 2018. WHO/CDS/TB/2018.20.
Prevention; CNY: China Yuan; CPI: Consumer price index; LR: Likelihood ratio 4. World Health Assembly Secretariat. Global strategy and targets for
tuberculosis prevention, care and control after 2015. Geneva: World Health
Acknowledgments Organization; 2014. http://www.who.int/tb/post2015strategy/en/.
We thank Shandong Provincial Hospital, Shandong Tuberculosis Control Center,17 5. Harling G, Castro MC. A spatial analysis of social and economic
municipal-level health departments for TB registered incidence data and sociode- determinants of tuberculosis in Brazil. Health Place. 2014;25:56–67.
mographic data. This work was supported by Department of Science & Technol- 6. Oxlade O, Murray M. Tuberculosis and poverty: why are the poor at greater
ogy of Shandong Province (CN)(No.2007GG30002033; No.2017GSF218052) and risk in India? PLoS One. 2012;7:e47533.
Jinan Science and Technology Bureau (CN) (No.201704100). 7. Hargreaves JR, Boccia D, Evans CA, et al. The social determinants of
tuberculosis: from evidence to action. Am J Public Health. 2011;101:654–62.
Authors’ contributions 8. Lonnroth K, Castro KG, Chakaya JM, et al. Tuberculosis control and elimination
H.C.L. and D.M.Y. conceived and designed the study. H.C.L., J.Y.L, N.N.T., Y.F.L.(YL1), 2010–50: cure, care, and social development. Lancet. 2010;375:1814–29.
Y.L.(YL2), and L.G. directed its implementation including the data analysis and writing 9. Myers WP, Westenhouse JL, Flood J, et al. An ecological study of
of the paper. Q.Y.Z. and L.Q.C. analyzed the data. W.M.S., T.T.X, S.J.L. and S.Q.L., Q.Q.A., tuberculosis transmission in California. Am J Public Health. 2006;96:685–90.
and W.Y.S. were responsible for completing the interpretation of data and the 10. Reeves A, Basu S, McKee M, et al. Social protection and tuberculosis control
production of figures and Tables. Q.Y.Z., L.Q.C., D.M.Y. and H.C.L. wrote and revised in 21 European countries, 1995–2012: a cross-national statistical modelling
the manuscript. All authors read and approved the final manuscript. analysis. Lancet Infect Dis. 2014;14:1105–12.
11. Pelissari DM, Rocha MS, Bartholomay P, et al. Identifying socioeconomic,
Funding epidemiological and operational scenarios for tuberculosis control in Brazil:
This study is a project supported and developed by the government. It was an ecological study. BMJ Open. 2018;8:e018545. https://doi.org/10.1136/
financed by Department of Science & Technology of Shandong Province bmjopen-2017-018545.
(CN)(No.2007GG30002033; No.2017GSF218052) and Jinan Science and 12. Siroka A, Ponce NA, Lönnroth K. Association between spending on social
Technology Bureau (CN) (No.201704100).Department of Science & protection and tuberculosis burden: a global analysis. Lancet Infect Dis.
Technology of Shandong Province provided financial support during our 2016;16:473–9.
data collection. Jinan Science and Technology Bureau financed the purchase 13. National Bureau of Statistics of China.China Statistical Yearbook (2018) 3-9:
of software for the study. The funders had no role in study design, data Gross Regional Product (2017). http://www.stats.gov.cn/tjsj/ndsj/2018/
collection and analysis, decision to publish, or preparation of the manuscript. indexch.htm.
14. Zhang WH, Xiao C, Ren T. Ecological Trendofthe registered incidence of
Availability of data and materials Tuberculosis in Mianyang City During 2004–2013. Sichuan Da Xue Xue Bao
The datasets used and analysed during the current study are available from the Yi Xue Ban. 2016;47(5):759.
corresponding author on reasonable request. The panel data that contains the 15. Minh HV, Mai VQ, Nhung NV, et al. Costs of providing tuberculosis diagnosis
registered incidence of tuberculosis and various economic indicators in 17 cities and treatment services in Viet Nam. Int J Tuberc Lung Dis. 2017;21(9):1035.
in Shandong Province from 2005 to 2017 is detailed in Additional file 1. The 16. Liu K, Peng Y, Zhou Q, et al. Assessment of Active Tuberculosis Findings in
original data of threshold regression model is in Additional file 2. the Eastern Area of China: A 3-year sequential screening study. Int J Infect
Dis. 2019;S1201-9712(19):30314–5. https://doi.org/10.1016/j.ijid.2019.07.029.
Ethics approval and consent to participate 17. Alba S, Rood E, Bakker MI, et al. Development and validation of a predictive
The ethical approval of our study was obtained from the Ethics Committee of ecological model for TB prevalence. Int J Epidemiol. 2018;47(5):1645.
Shandong Provincial Hospital, affiliated with Shandong University (SPH), China. Before 18. Amsalu E, Liu M, Li Q, et al. Spatial-temporal analysis of tuberculosis in the
data analysis and reporting, any personal identifiers of TB patients were removed. geriatric population of China: An analysis based on the Bayesian conditional
autoregressive model. Arch Gerontol Geriatr. 2019:83. https://doi.org/10.
1016/j.archger.2019.05.011.
Consent for publication
19. Awad SF, Dargham SR, Omori R, et al. Analytical exploration of potential
Not applicable.
pathways by which diabetes mellitus impacts tuberculosis epidemiology.
Sci Rep. 2019;9(1). https://doi.org/10.1038/s41598-019-44916-7.
Competing interests 20. Golub JE, Mok Y, Hong S, et al. Diabetes mellitus and tuberculosis in Korean
The authors declare that they have no competing interests.
adults: impact on tuberculosis registered incidence, recurrence and mortality.
Int J Tuberc Lung Dis. 2019;23(4):507–13. https://doi.org/10.5588/ijtld.18.0103.
Author details 21. Jun C, Hui Z, Yanlin Z, et al. Mutual Impact of Diabetes Mellitus and
1
Department of Respiratory Medicine, Shandong Provincial Hospital Affiliated Tuberculosis in China. Biomed Environ Sci. 2017;30(5):384–9. https://doi.org/
to Shandong University, Shandong Provincial Hospital Affiliated to Shandong
10.3967/bes2017.051.
First Medical University, No. 324, Jingwuweiqi Road, Huaiyin District, Jinan 22. Liu K, Peng Y, Zhou Q, et al. Assessment of active tuberculosis findings in
250021, Shandong, China. 2Cheeloo College of Medicine, Shandong the eastern area of China: A 3-year sequential screening study. Int J Infect
University, Jinan 250012, Shandong, China. 3College of statistics, Shandong Dis. 2019;88:34–40.
University of Finance and Economics, Jinan 250202, Shandong, China.
4 23. Dong D, Jiang WX, et al. Impact of an innovative tuberculosis financing and
Department of Intensive Care Unit, Shandong Provincial Third Hospital, payment model on health service utilization by tuberculosis patients in
Jinan 100191, Shandong, China. 5Peking Union Medical College, Beijing China: do the poor fare better than the rich? Infect Dis Poverty. 2019;8:44.
100005, China. 6NHC Key Laboratory of Systems Biology of Pathogens, https://doi.org/10.1186/s40249-019-0559-z.
Institute of Pathogen Biology, and Center for Tuberculosis Research, Chinese
24. Nery JS, Rodrigues LC, Rasella D, et al. Effect of Brazil’s conditional cash
Academy of Medical Sciences and Peking Union Medical College, 100730 transfer programme on tuberculosis registered incidence. Int J Tuberc Lung
Beijing, P. R. China. 7International College, Jiangxi University of Finance and Dis. 2017;21(7):790–6.
Economics, Nanchang, Jiangxi, China. 8Shandong University of Traditional 25. Shete PB, Reid M, Goosby E, et al. Message to world leaders: we cannot end
Chinese Medicine, Jinan 250355, Shandong, China.
tuberculosis without addressing the social and economic burden of the disease.
Lancet Glob Health. 2018;6(12). https://doi.org/10.1016/S2214-109X(18)30378-4.
Received: 4 September 2019 Accepted: 29 September 2020

Publisher’s Note
References Springer Nature remains neutral with regard to jurisdictional claims in
1. WHO. The global plan to stop TB. Geneva: World Health Organization; 2006. published maps and institutional affiliations.
http://www.stoptb.org/global/plan/.

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:

1. use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
2. use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
3. falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
4. use bots or other automated methods to access the content or redirect messages
5. override any security feature or exclusionary protocol; or
6. share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at

[email protected]