-
Tuberculosis (TB) remains a major public health threat; approximately 10 million new TB cases are recorded worldwide, and more than 1.2 million deaths were caused by TB in 2019[1]. The diagnostic delay presents a major obstacle to the control of the TB epidemic. This is a serious concern because each patient may infect 10–15 people each year, eventually creating a public health problem and even leading to economic difficulties[2]. Ensuring universal access to quality-assured and early diagnosis of TB is critical to diminish transmission, avert deaths, and prevent TB infections, which can help countries move toward the elimination of the disease.
China is one of the countries with TB high burden, accounting for 8.4% of the global TB burden and 50% of the burden in the Western Pacific region[1-2]. Considering the high TB burden and limited human resources, the passive case-finding (PCF) strategy has always been the main approach applied by the National Tuberculosis Programme (NTP) in China, which contributes > 95% of TB patients’ notification. Under the PCF principle, patients with symptomatic TB voluntarily visit healthcare facilities to seek medical care, and all patients with presumptive TB and confirmed TB should be referred to TB-designated medical facilities. Otherwise, cases undetected for a long time or late detection may worsen the disease, increase the mortality rate, and result in the sustained spread of TB in the community.
Identifying reasons for the diagnostic delay is important for the health system to discover methods of diagnosing and treating patients timely, thus reducing disease suffering and transmission. Information on diagnostic delay is also important for the evaluation and improvement of TB control programs. Some small-scale studies have reported that diagnostic delay varies significantly from around 30 days in Shanghai[3] to 71 days in Yunnan province[4]. Factors attributed to diagnostic delay are also likely to vary because of different settings, poverty, non-residence, default health insurance, among others. However, nationally representative data are unavailable. We conducted a nationwide survey in 22 counties to evaluate the extent and risk factors of TB diagnostic delay within NTP in China.
This cross-sectional study used primary data collected from patients with drug-susceptible pulmonary TB. Individuals receiving treatment in facilities other than NTP were excluded.
A multi-stage stratified cluster sampling was adopted. The stratifying factors were region (east/central/west) and residence (rural/urban). The sampling steps are summarized in Supplementary File 1 (available in www.besjourna.com). Assuming a diagnostic delay of 30%[5], relative precision of 0.2, and α error of 0.05, average cluster size of 50, between-cluster variation of 0.4, design effect of 4.36, and anticipating a nonresponse rate of 10%, the final sample size was 960, to be sampled from 22 clusters (Supplementary File 2, available in www.besjourna.com). Table S1. Multi-stage stratified cluster sampling adopted in China’s TB* patient diagnosis survey (2017)
Step 1
Two provinces from east/middle/west regions each (six provinces in total) were chosen by simple random sampling: Liaoning and Fujian (east); Henan and Hunan (middle); and Yunnan and XinJiang (west) respectively.Step 2
The number of clusters in each stratum was determined, letting them to be proportional to number of TB patients registered in every stratum in the previous year. There were 6 strata in total, which are east urban area, east rural area, middle urban area, middle rural area, west urban area, west rural area. But since the total cluster numbers were limited, probability proportional to size sampling (PPS sampling) was adopted.Step 3
A list of NTP’s BMU in the 6 provinces selected in Step 1 was obtained. Each BMU was classified into predominantly rural or urban areas.Step 4 According to the lists of BMU created in Step 3 and the number of BMU decided in Step 2, simple random sampling with replacement was done for 6 strata separately Step 5 Within sampled BMU facilities, consecutive patients on TB treatment were eligible for inclusion. Note. NTP-National tuberculosis programme; BMU- basic management units. Table S2. The parameters used for sample size calculation under the nationally representative TB survey, China (Jan–June 2017)
We calculated sample size using the formula below. $N=N_{S R S} * D E F F \rightarrow N=\left[1.96^{2} \dfrac{\left(1-\pi_{g}\right)}{d^{2} \pi_{g} }\right] \times\left[1+(m-1) \dfrac{k^{2} \pi_{g} }{\left(1-\pi_{g}\right)}\right]$ Parameter Meaning Value estimated in
this studyN Number of people included in the patient survey 960 NSRS Simple Random Sampling size 220 DEFF Design effect 4.36 πg “Prior guess” of the true proportion of patients experiencing diagnosis delay (expressed as a proportion) 30% d Relative precision (expressed as a proportion). Recommended 0.20 or 0.25 0.2 m Cluster size (=number of targeted individuals), assumed to be constant across clusters 50 k Coefficient of between-cluster variation. Recommended to assume is in the range 0.4 – 0.6 0.4 A structured questionnaire was developed, which addresses the health-seeking behaviors of patients with TB. The questionnaire collected the following information: 1) demographic and socioeconomic variables, including sex, age, education level, area of residence, and economic activity; 2) clinical profiles, TB category, sputum status at diagnosis, hospitalization, human immunodeficiency virus (HIV) infection status, and comorbidities, such as chronic nephrosis and chronic liver disease; and 3) insurance status and healthcare-seeking experience, which means that patients with presumptive TB seek healthcare services in healthcare facilities. Written consent was obtained from each eligible patient with TB before enrollment. Informed consent and responses for children aged < 18 years were obtained from the accompanying relative. A face-to-face interview was performed by trained investigators. Diagnostic delay was defined as the time interval from symptom onset to confirmed TB diagnosis longer than 14 days[6].
Data were double entered and validated using EpiData (version 3.1 EpiData Association, Odense, Denmark). Analysis was conducted using STATA (version 12.1, copyright 1985–2011 StataCorp LP USA). Univariate and multivariate analyses were applied to analyze the risk factors associated with diagnostic delay. Variables with P < 0.20 in the univariate analysis were used in the stepwise (entry and removal probability of 0.05 and 0.10, respectively) multivariate logistic regression analysis. The association of predictor variables with dependent variables was assessed by using a 95% confidence interval (CI) and adjusted odds ratio (aOR). A P-value of < 0.05 was regarded as significant.
A total of 974 patients with TB were included in this study (Table 1). Among them, 387 (39.7%) and 380 (39.0%) patients came from the Eastern and Western regions, respectively. Moreover, 396 (40.7%) patients were 45–64 years old, and 682 (70.0%) patients were male. In addition, 573 (58.5%) patients had primary and middle school education, 641 (65.81%) were rural residents, and 810 (87.4%) were symptomatic. In total, 728 patients were under the new rural cooperative medical scheme, accounting for 74.7%.
Table 1. Socio-demographic characteristics of pulmonary tuberculosis patients in diagnostic delay survey in China, 2017 (n = 974)
Characteristics FREQ N (%) Overall 974 (100.00) Region West 380 (39.01) East 387 (39.73) Central 207 (21.25) Age group in years < 15 3 (0.31) 15–44 330 (33.88) 45–64 396 (40.66) ≥ 65 245 (25.15) Gender Male 682 (70.02) Female 292 (29.98) Education Illiterate or not completed primary school 177 (18.17) Completed primary school 259 (26.59) Completed middle school 314 (32.24) Completed high school 154 (15.81) Completed college and above 70 (7.19) Economic activity Regular salary 225 (23.10) Irregular earning 467 (47.95) Economically inactive 282 (28.95) Prime income earner No 408 (41.89) Yes 566 (58.11) Marital status Married 155 (15.91) Unmarried 606 (62.22) Live alone 103 (10.58) Unknown 110 (11.29) Residence Urban 333 (34.19) Rural 641 (65.81) TB Category Previously treated 80 (8.21) New 894 (91.79) Sputum status at diagnosis Negative 655 (67.25) Positive 279 (28.64) Unknown 40 (4.11) Hospitalization No 615 (63.14) Yes 359 (36.86) Symptom No 117 (12.62) Yes 810 (87.38) Hemoptysis 133 (14.35) overall symptoms 143 (15.43) Cough&Expectoration 534 (57.61) HIV Negative 332 (34.09) Positive 11 (1.13) Unknown 631 (64.78) Comorbidity No 589 (60.47) Yes 385 (39.53) Insurance UBMIa 180 (18.48) NCMSb 728 (74.74) Otherc 66 (6.78) Place of registration TB designated hospital 659 (67.66) CDC hospital 127 (13.04) TB dispensary 188 (19.30) Note. FREQ: frenqency; N: Number; TB: tuberculosis; HIV: human immunodeficiency vrus. aUBMI: Urban employee and residence basic medical insurance; bNCMS: New rural cooperative medical scheme; cOther: Commercial insurance, mixed insurance and none. The median diagnostic time for all cases was 23 (IQR 5–53) days. The mean diagnostic time was 37.1 ± 43.3 days, which was similar to results of previous studies in Shanghai and Yunnan provinces in China[4-5]. Compared with other developing counties such as India[7] and South Africa[8], in which the average diagnostic delay was 44–151 days, a shorter delay was recorded in China. The comparatively good results could be attributed to the determination and corresponding measures taken by the Chinese government to control TB, such as training of professionals, scaling up of laboratory equipment, and capacity building. However, 594 (61.0%) patients still experienced a diagnostic delay, and this was still a challenge of controlling TB.
Tables 2 and 3 showed the results of the univariate and multivariate logistic regression analyses. The odds of having a diagnostic delay were higher in patients from central (aOR = 3.03, 95% CI 1.96–4.69) and eastern (aOR = 1.69, 95% CI 1.24–2.31) regions than in those from the western region. Given the relatively good economic situation of patients with presumptive TB in eastern and central regions, they usually choose senior general hospitals as the first treatment facilities; however, recent studies have found that the ability of these hospitals to identify TB is not as sensitive as expected[9] because some general hospitals do not routinely perform TB-related laboratory tests. Therefore, presumptive TB cases are referred to TB-designated hospitals for further TB-specific examination. The tortuous medical-seeking path increases the risk of diagnostic delay.
Table 2. Univariate logistic regression for factors associated with diagnostic delay among patients in China, 2017 (n = 974)
Characteristics Diagnosis time (days)
Median (IQR)OR 95% CI Overall 23.0 (5.0–53.0) Region West 17.0 (3.0–50.5) ref East 29.0 (10.0–56.0) 1.85 1.38–2.48 Central 21.0 (3.0–53.0) 1.41 1.00–2.00 Age group in years ≤ 44 21.0 (4.0–49.0) ref 45–64 23.0 (4.0–51.5) 1.09 0.81–1.46 ≥ 65 24.0 (9.0–59.0) 1.25 0.89–1.75 Gender Male 22.5 (5.0–52.0) ref Female 23.5 (5.0–59.5) 0.92 0.70–1.22 Education Illiterate or not completed primary school 28.0 (6.0–67.0) 1.86 1.20–2.89 Completed primary school 23.0 (7.0–57.0) 1.60 1.07–2.39 Completed middle school 25.5 (7.0–52.0) 1.98 1.34–2.93 Completed high school 13.0 (1.0–34.0) ref Completed college and above 27.0 (3.0–61.0) 1.90 1.06–3.40 Economic activity Regular salary 20.0 (4.0–40.0) ref Irregular earning 23.0 (6.0–53.0) 1.04 0.76–1.45 Economically inactive 25.5 (5.0–61.0) 1.15 0.80–1.64 Prime income earner No 21.0 (5.0–54.0) ref Yes 23.0 (5.0–52.0) 1.05 0.81–1.36 Marital status Married 26.0 (6.0–51.0) 0.90 0.54–1.51 Unmarried 24.0 (7.0–57.0) 0.87 0.56–1.35 Live alone 27.0 (4.0–59.0) ref Unknown 14.5 (1.0–37.0) 0.54 0.31–0.93 Residence Urban 17.0 (3.0–51.0) ref Rural 27.0 (6.0–57.0) 1.44 1.10–1.88 TB Category Previously treated 16.0 (1.0–34.5) ref New 24.0 (6.0–54.0) 1.24 0.78–1.96 Sputum status at diagnosis Negative 20.0 (4.0–48.0) ref Positive 30.0 (6.0–61.0) 1.29 0.97–1.73 Unknown 20.5 (6.5–56.5) 1.03 0.54–1.98 Hospitalization No 21.0 (4.0–52.0) ref Yes 25.0 (7.0–57.0) 1.35 1.03–1.77 Symptom No 3.0 (0.0–18.0) ref Yes 28.5 (10.0–59.0) 4.73 3.11–7.20 Hemoptysis 21.0 (3.0–57.0) ref Overall symptoms 28.0 (9.0–59.0) 1.58 0.97–2.56 Cough&Expectoration 30.0 (12.0–60.0) 2.07 1.40–3.05 HIV Negative 21.0 (5.0–51.0) ref Positive 22.0 (2.0–94.0) 1.83 0.48–7.01 Unknown 24.0 (5.0–53.0) 1.10 0.84–1.45 Comorbidity No 22.0 (5.0–46.0) ref Yes 24.0 (5.0–61.0) 1.08 0.83–1.41 Insurance UBMIa 18.0 (5.0–50.0) 1.43 0.81–2.52 NCMSb 26.0 (6.0–55.5) 1.67 1.01–2.76 Otherc 14.5 (2.0–52.0) ref Place of registration TB designated hospital 20.0 (4.0–53.0) ref CDC hospital 22.0 (2.0–57.0) 1.07 0.72–1.57 TB dispensary 30.0 (14.0–52.0) 2.16 1.50–3.10 Note. *IQR: interquartile range; OR: odds ratio; CI: confidence interval; TB: tuberculosis; HIV: human immunodeficiency vrus. aUBMI: Urban employee and residence basic medical insurance; bNCMS: New rural cooperative medical scheme; cOther: Commercial insurance, mixed insurance and none. Table 3. Multivariate logistic regression for factors associated with diagnostic delay among patients in China, 2017 (n = 974)
Characteristics Diagnosis time (days)
Median (IQR)aOR 95% CI Region West 17.0 (3.0–50.5) ref East 29.0 (10.0–56.0) 1.69 1.24–2.31 Central 21.0 (3.0–53.0) 3.03 1.96–4.69 Residence Urban 17.0 (3.0–51.0) ref Rural 27.0 (6.0–57.0) 1.45 1.07–1.96 Symptom Hemoptysis 21.0 (3.0–57.0) ref Overall symptoms 28.0 (9.0–59.0) 1.51 0.92–2.48 Cough&Expectoration 30.0 (12.0–60.0) 2.20 1.48–3.28 Note. The variables in the univariate logistic analysis whose significance level was less than 0.2 were introduced into the model for the stepwise multivariate logistic regression analysis. *IQR: interquartile range; aOR: adjusted odds ratio; CI: confidence interval. The odds of having a diagnostic delay were higher in rural residents than in urban residents (aOR = 1.45, 95% CI 1.07–1.96). The Fifth National TB Epidemiology Survey showed that the rates of public awareness of TB in rural and urban areas were 51.7% and 63.4%, respectively[10]; the lower awareness of TB in rural areas might lead to a longer diagnostic delay. By contrast, rural residents faced various barriers to access to timely TB diagnoses, such as the lack of diagnostic facilities, lack of quality services and trained staff, and poor economic situation.
The odds of having diagnostic delay were higher for patients with TB experiencing coughs or expectoration (aOR = 2.20, 95% CI 1.48–3.28) than for patients with hemoptysis. This finding was mainly due to the current PCF strategy adopted in China, i.e., once patients with presumptive TB present with TB-related symptoms, he/she will go to the healthcare institutions initiatively for medical consultation. However, this behavior depends on the patient’s awareness of TB and the seriousness of the symptoms. Previous studies have also shown that mild initial symptoms such as cough could prolong the diagnostic time[3], whereas severe initial symptoms such as hemoptysis and bloody sputum could shorten the diagnostic time. As asymptomatic TB cases were mostly found through active case finding (ACF), such as health examination, patients were diagnosed at the early stage, which might shorten the diagnostic delay.
The result of the univariate logistic regression analysis also suggested that patients without education or who have not completed primary school (OR = 1.86, 95% CI 1.20–2.89) or have completed middle school (OR = 1.98, 95% CI 1.34–2.93) have higher odds of having diagnostic delay than patients who have completed high school. Among rural residents with TB, those with hospitalization history have higher odds of experiencing diagnostic delay than patients without hospitalization history (OR = 1.35, 95% CI 1.03–1.77). The odds of having a diagnostic delay in patients registered in TB dispensary were higher than the value in those registered in TB-designated hospitals (OR = 2.16, 95% CI 1.50–3.10). Sociodemographic factors such as age, sex, marital status, and type of medical insurance were not significantly associated with diagnostic delay.
The results of this study may have some important policy implications for TB control in China. To our knowledge, patients with TB are still predominantly passively identified in China, which relies heavily on the patient’s medical treatment behavior, so various approaches should be adopted to improve public awareness of TB and thereby reduce the patient delay. Besides the patient delay, the diagnostic delay includes health system delay, which is affected by the TB-recognizing ability of the doctors and the diagnostic capacity of the facility. Thus, improving the recognizing ability of TB symptoms by all healthcare workers, scaling up of rapid diagnostic tests, and optimizing the diagnostic algorithm are important to reduce the health system delay. In addition, China should gradually scale up ACF among high-risk groups, including close contacts of smear-positive TB cases, presence of HIV/AIDS, and those aged > 65 years.
In brief, this study presents a significant delay in TB diagnosis in China. Major factors associated with diagnostic delay are localization in central and eastern regions, rural residence, and being symptomatic. Multiple measures should be taken to shorten the diagnostic delay in view of the patient and health system and thus improve TB awareness of the public, scale up ACF in high-risk groups, and improve diagnostic capacity in economically less-developed regions in the near future.
The study was approved by the Ethics Committee of the Chinese Center for Disease Control and Prevention [No. 201625, dated 22 November 2016].
No potential conflict of interest was disclosed.
The authors would like to thank all supporters of Liaoning CDC, Henan CDC, Xinjiang CDC, Yunnan CDC, Fujian CDC, and Hunan Chest Hospital.
XU Cai Hong and ZHANG Hui conceived the study, LIU Yan and WANG Li supervised the survey, HU Dong Mei and XIA Yin Yin analyzed the data, XU Cai Hong and ZHANG Xiao Meng wrote the manuscript. ZHANG Hui reviewed the manuscript.
doi: 10.3967/bes2022.010
-
&These authors contributed equally to this work.
注释: -
S1. Multi-stage stratified cluster sampling adopted in China’s TB* patient diagnosis survey (2017)
Step 1
Two provinces from east/middle/west regions each (six provinces in total) were chosen by simple random sampling: Liaoning and Fujian (east); Henan and Hunan (middle); and Yunnan and XinJiang (west) respectively.Step 2
The number of clusters in each stratum was determined, letting them to be proportional to number of TB patients registered in every stratum in the previous year. There were 6 strata in total, which are east urban area, east rural area, middle urban area, middle rural area, west urban area, west rural area. But since the total cluster numbers were limited, probability proportional to size sampling (PPS sampling) was adopted.Step 3
A list of NTP’s BMU in the 6 provinces selected in Step 1 was obtained. Each BMU was classified into predominantly rural or urban areas.Step 4 According to the lists of BMU created in Step 3 and the number of BMU decided in Step 2, simple random sampling with replacement was done for 6 strata separately Step 5 Within sampled BMU facilities, consecutive patients on TB treatment were eligible for inclusion. Note. NTP-National tuberculosis programme; BMU- basic management units. S2. The parameters used for sample size calculation under the nationally representative TB survey, China (Jan–June 2017)
We calculated sample size using the formula below. $N=N_{S R S} * D E F F \rightarrow N=\left[1.96^{2} \dfrac{\left(1-\pi_{g}\right)}{d^{2} \pi_{g} }\right] \times\left[1+(m-1) \dfrac{k^{2} \pi_{g} }{\left(1-\pi_{g}\right)}\right]$ Parameter Meaning Value estimated in
this studyN Number of people included in the patient survey 960 NSRS Simple Random Sampling size 220 DEFF Design effect 4.36 πg “Prior guess” of the true proportion of patients experiencing diagnosis delay (expressed as a proportion) 30% d Relative precision (expressed as a proportion). Recommended 0.20 or 0.25 0.2 m Cluster size (=number of targeted individuals), assumed to be constant across clusters 50 k Coefficient of between-cluster variation. Recommended to assume is in the range 0.4 – 0.6 0.4 Table 1. Socio-demographic characteristics of pulmonary tuberculosis patients in diagnostic delay survey in China, 2017 (n = 974)
Characteristics FREQ N (%) Overall 974 (100.00) Region West 380 (39.01) East 387 (39.73) Central 207 (21.25) Age group in years < 15 3 (0.31) 15–44 330 (33.88) 45–64 396 (40.66) ≥ 65 245 (25.15) Gender Male 682 (70.02) Female 292 (29.98) Education Illiterate or not completed primary school 177 (18.17) Completed primary school 259 (26.59) Completed middle school 314 (32.24) Completed high school 154 (15.81) Completed college and above 70 (7.19) Economic activity Regular salary 225 (23.10) Irregular earning 467 (47.95) Economically inactive 282 (28.95) Prime income earner No 408 (41.89) Yes 566 (58.11) Marital status Married 155 (15.91) Unmarried 606 (62.22) Live alone 103 (10.58) Unknown 110 (11.29) Residence Urban 333 (34.19) Rural 641 (65.81) TB Category Previously treated 80 (8.21) New 894 (91.79) Sputum status at diagnosis Negative 655 (67.25) Positive 279 (28.64) Unknown 40 (4.11) Hospitalization No 615 (63.14) Yes 359 (36.86) Symptom No 117 (12.62) Yes 810 (87.38) Hemoptysis 133 (14.35) overall symptoms 143 (15.43) Cough&Expectoration 534 (57.61) HIV Negative 332 (34.09) Positive 11 (1.13) Unknown 631 (64.78) Comorbidity No 589 (60.47) Yes 385 (39.53) Insurance UBMIa 180 (18.48) NCMSb 728 (74.74) Otherc 66 (6.78) Place of registration TB designated hospital 659 (67.66) CDC hospital 127 (13.04) TB dispensary 188 (19.30) Note. FREQ: frenqency; N: Number; TB: tuberculosis; HIV: human immunodeficiency vrus. aUBMI: Urban employee and residence basic medical insurance; bNCMS: New rural cooperative medical scheme; cOther: Commercial insurance, mixed insurance and none. Table 2. Univariate logistic regression for factors associated with diagnostic delay among patients in China, 2017 (n = 974)
Characteristics Diagnosis time (days)
Median (IQR)OR 95% CI Overall 23.0 (5.0–53.0) Region West 17.0 (3.0–50.5) ref East 29.0 (10.0–56.0) 1.85 1.38–2.48 Central 21.0 (3.0–53.0) 1.41 1.00–2.00 Age group in years ≤ 44 21.0 (4.0–49.0) ref 45–64 23.0 (4.0–51.5) 1.09 0.81–1.46 ≥ 65 24.0 (9.0–59.0) 1.25 0.89–1.75 Gender Male 22.5 (5.0–52.0) ref Female 23.5 (5.0–59.5) 0.92 0.70–1.22 Education Illiterate or not completed primary school 28.0 (6.0–67.0) 1.86 1.20–2.89 Completed primary school 23.0 (7.0–57.0) 1.60 1.07–2.39 Completed middle school 25.5 (7.0–52.0) 1.98 1.34–2.93 Completed high school 13.0 (1.0–34.0) ref Completed college and above 27.0 (3.0–61.0) 1.90 1.06–3.40 Economic activity Regular salary 20.0 (4.0–40.0) ref Irregular earning 23.0 (6.0–53.0) 1.04 0.76–1.45 Economically inactive 25.5 (5.0–61.0) 1.15 0.80–1.64 Prime income earner No 21.0 (5.0–54.0) ref Yes 23.0 (5.0–52.0) 1.05 0.81–1.36 Marital status Married 26.0 (6.0–51.0) 0.90 0.54–1.51 Unmarried 24.0 (7.0–57.0) 0.87 0.56–1.35 Live alone 27.0 (4.0–59.0) ref Unknown 14.5 (1.0–37.0) 0.54 0.31–0.93 Residence Urban 17.0 (3.0–51.0) ref Rural 27.0 (6.0–57.0) 1.44 1.10–1.88 TB Category Previously treated 16.0 (1.0–34.5) ref New 24.0 (6.0–54.0) 1.24 0.78–1.96 Sputum status at diagnosis Negative 20.0 (4.0–48.0) ref Positive 30.0 (6.0–61.0) 1.29 0.97–1.73 Unknown 20.5 (6.5–56.5) 1.03 0.54–1.98 Hospitalization No 21.0 (4.0–52.0) ref Yes 25.0 (7.0–57.0) 1.35 1.03–1.77 Symptom No 3.0 (0.0–18.0) ref Yes 28.5 (10.0–59.0) 4.73 3.11–7.20 Hemoptysis 21.0 (3.0–57.0) ref Overall symptoms 28.0 (9.0–59.0) 1.58 0.97–2.56 Cough&Expectoration 30.0 (12.0–60.0) 2.07 1.40–3.05 HIV Negative 21.0 (5.0–51.0) ref Positive 22.0 (2.0–94.0) 1.83 0.48–7.01 Unknown 24.0 (5.0–53.0) 1.10 0.84–1.45 Comorbidity No 22.0 (5.0–46.0) ref Yes 24.0 (5.0–61.0) 1.08 0.83–1.41 Insurance UBMIa 18.0 (5.0–50.0) 1.43 0.81–2.52 NCMSb 26.0 (6.0–55.5) 1.67 1.01–2.76 Otherc 14.5 (2.0–52.0) ref Place of registration TB designated hospital 20.0 (4.0–53.0) ref CDC hospital 22.0 (2.0–57.0) 1.07 0.72–1.57 TB dispensary 30.0 (14.0–52.0) 2.16 1.50–3.10 Note. *IQR: interquartile range; OR: odds ratio; CI: confidence interval; TB: tuberculosis; HIV: human immunodeficiency vrus. aUBMI: Urban employee and residence basic medical insurance; bNCMS: New rural cooperative medical scheme; cOther: Commercial insurance, mixed insurance and none. Table 3. Multivariate logistic regression for factors associated with diagnostic delay among patients in China, 2017 (n = 974)
Characteristics Diagnosis time (days)
Median (IQR)aOR 95% CI Region West 17.0 (3.0–50.5) ref East 29.0 (10.0–56.0) 1.69 1.24–2.31 Central 21.0 (3.0–53.0) 3.03 1.96–4.69 Residence Urban 17.0 (3.0–51.0) ref Rural 27.0 (6.0–57.0) 1.45 1.07–1.96 Symptom Hemoptysis 21.0 (3.0–57.0) ref Overall symptoms 28.0 (9.0–59.0) 1.51 0.92–2.48 Cough&Expectoration 30.0 (12.0–60.0) 2.20 1.48–3.28 Note. The variables in the univariate logistic analysis whose significance level was less than 0.2 were introduced into the model for the stepwise multivariate logistic regression analysis. *IQR: interquartile range; aOR: adjusted odds ratio; CI: confidence interval. -
[1] WHO. Global tuberculosis report 2020. Geneva: World Health Organization, 2020. [2] Fg C, J M, Frg H, et al. Early detection of Tuberculosis. British Med J, 1942; 1, 535. [3] Meyssonnier V, Li X, Shen X, et al. Factors associated with delayed tuberculosis diagnosis in China. Eur J Public Health, 2013; 23, 253−7. doi: 10.1093/eurpub/cks037 [4] Lin X, Chongsuvivatwong V, Geater A, et al. The effect of geographical distance on TB patient delays in a mountainous province of China. Int J Tuberc Lung Dis, 2008; 3, 99−105. (In Chinese [5] CDC C. Baseline survey report of Gates foundation TB project (Phase II). Beijing, 2018. (In Chinese). [6] Li Y, Ehiri J, Tang SL, et al. Factors associated with patient, and diagnostic delays in Chinese TB patients: a systematic review and meta-analysis. BMC Med, 2013; 11, 156. doi: 10.1186/1741-7015-11-156 [7] Sreeramareddy CT, Qin ZZ, Satyanarayana S, et al. Delays in diagnosis and treatment of pulmonary tuberculosis in India: a systematic review. Int J Tuberc Lung Dis, 2014; 18, 255−66. doi: 10.5588/ijtld.13.0585 [8] Meintjes G, Schoeman H, Morroni C, et al. Patient and provider delay in tuberculosis suspects from communities with a high HIV prevalence in South Africa: a cross-sectional study. BMC Infect Dis, 2008; 8, 72. doi: 10.1186/1471-2334-8-72 [9] Yu YM, Pang XW, Fu YY. Investigation of an outbreak of tuberculosis in an adult education Centre in Tianjin. J Tuberc Lung Health, 2019; 8, 65−8. (In Chinese [10] Wang LX, Cheng SM, Chen MT, et al. The fifth national tuberculosis epidemiological survey in 2010. Chin J Antituberc, 2012; 34, 485−508. (In Chinese -
Supplementary Materials21228.pdf
-