Article Text
Abstract
Objectives This study aims to estimate tuberculosis (TB) incidence, mortality rates and survival HRs in Kazakhstan, using large-scale administrative health data records during 2014–2019.
Design A retrospective cohort study.
Settings Data for patients with TB in Kazakhstan during 2014–2019, reported in the Unified National Electronic Healthcare System.
Participants Patients with TB in Kazakhstan (ICD-10 (The International Classification of Diseases, 10th revision) codes: A15–A19).
Outcome measures Demographic factors, diagnoses and comorbidities were analysed using descriptive, bivariate and multivariable statistical analyses. TB incidence and mortality rates were calculated, and Cox regression and Kaplan-Meier survival analysis were performed to assess risk factors for survival rates.
Results Of the 149 122 patients with TB, 91 437 (61%) were males, and 139 931 (94%) had respiratory TB. From 2014 to 2019, TB incidence declined from 227 to 15.2 per 100 000 individuals, while all-cause mortality increased from 8.4 to 15.2 per 100 000. Age-specific TB incidence was lowest for 0–10 years of age and highest for 20 years of age. Being older, man, urban residence versus rural, retired versus employed, having HIV and having diabetes versus no comorbidities were associated with lower survival rates.
Conclusion To date, this is the largest TB published study for Kazakhstan, characterising TB incidence and mortality trends by demographic factors, and risk factors for survival rates. The findings highlight the need for targeted interventions to address the growing burden of TB, particularly among older adults, men, urban residents and those with HIV and diabetes. The study underscores the importance of using administrative health data to inform policy and health system responses to TB in Kazakhstan.
- tuberculosis
- epidemiology
- epidemiology
- epidemiologic studies
- public health
- public health
Data availability statement
Data may be obtained from a third party and are not publicly available.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
STRENGTHS AND LIMITATIONS OF THIS STUDY
The study provides novel and comprehensive data on tuberculosis (TB) incidence and mortality in Kazakhstan using nationwide hospital data.
Assesses the influence of demographic factors and comorbidities on TB survival probability.
Limitations: lack of data on TB prevalence, drug resistant TB burden and TB-specific mortality.
Discrepancies between databases could lead to missing variables and observations.
Introduction
Tuberculosis (TB) is a significant public health concern globally, and the WHO is working to eliminate it by reducing TB incidence, mortality rates and costs through their End TB Strategy and the United Nations Sustainable Development Goals. Despite this effort, TB still remains a major problem, with 10 million individuals estimated to have TB disease in 2019, resulting in 1.2 million TB-related deaths among HIV-negative individuals. Moreover, the COVID-19 pandemic adversely affected the progress made to reduce the TB global burden. WHO estimated that in 2020, the global number of TB deaths increased by 0.2–0.4 million and the detection rate of TB cases decreased.1
Despite numbers falling slowly in recent years, TB remains a significant health problem in the WHO European region. In 2016, 85% of new cases and 90% of deaths due to TB occurring in the 18 high-priority countries. Kazakhstan is among those high-priority countries with TB incidence>20 cases per 100 000 population (20 per cent mille (pcm)).2 The average annual decline in TB notifications and TB mortality rates during 2005–2016 was 8.7% and 17.8%, respectively.3 According to a report by the Ministry of Healthcare of Kazakhstan, in 2019, the estimated TB incidence was 45.6 pcm (48.2 in 2018) and the mortality rate was 2 pcm (2.4 in 2018)4; however, these numbers might be underestimated due to uncaptured cases.
In 1998, Kazakhstan’s Ministry of Health developed the National Tuberculosis Program for detection and treatment of TB, including directly observed therapy, short-course (DOTS), which significantly reduced TB incidence rates.5 Despite WHO recommendations to treat TB in an outpatient setting, Kazakhstan still has a high hospitalisation rate at TB centres.5 The Ministry of Health attempts to shift TB care more towards outpatient services, including daily provision of medications, day care, video-controlled therapy and mobile teams for those who are not compliant to DOTS.5 Once TB is diagnosed, the patient is registered in the Electronic Registry of Dispensary Patients, National registry of patients with TB, which is a part of the Unified National Electronic Healthcare System (UNEHS).6
Regardless of the public health impact of TB, there are limited comprehensive reports and publications on the epidemiology of TB in Kazakhstan. Existing studies have been conducted covering only limited disperse geographic regions.7–11 The digitalisation of the national healthcare system and the establishment of the UNEHS over the last decade make it feasible to conduct nationwide large-scale epidemiological studies in Kazakhstan.12–14
Therefore, the objective of this in-depth epidemiological study is to estimate the incidence and mortality rates of TB in Kazakhstan, using large-scale administrative health data records between 2014 and 2019, available in the UNEHS database.
Methods
Study design and selection criteria
This retrospective study used data from the UNEHS database for the years 2014–2019. Outpatient and inpatient hospital records were used to retrieve data for patients with TB diagnosis, identified by the following ICD-10 (The International Classification of Diseases, 10th revision) codes: A15 (respiratory TB confirmed bacteriologically and histologically), A16 (respiratory TB, not confirmed bacteriologically or histologically), A17 (TB of the nervous system), A18 (TB of other organs) and A19 (miliary TB). Data entry for UNEHS was performed by hospital personnel and entered on every visit of each patient. Patients were included if they had at least one hospital visit with a TB diagnosis during the study period. Duplicate entries were removed using the population registry ID (RPN ID) used in UNEHS, ensuring that each patient was only counted once. This resulted in a final sample of 1 49 122 patients for analysis. The patient selection process is detailed in online supplemental figure 1.
Supplemental material
Supplemental material
Exposure and covariates
Demographic data including RPN ID, age, sex, region, ethnicity, location, ICD-10 codes for main diagnosis and social status were collected from the UNEHS database. Date of birth, death and comorbidities, including HIV, diabetes mellitus (DM) and hepatitis, were also recorded. Age was categorised into five groups: <18 years, 18–34 years, 35–44 years, 45–50 years and ≥51 years. Ethnicity was categorised into Kazakhs, Russians and others (including 42 ethnicities). Regions were categorised into 14 state regions and 3 cities that have the equivalent status of state region (Astana, Almaty, Shymkent). Patients were divided into either urban or rural residencies based on their location. Social status was classified into unemployed, employed, retired, disabled and others.
Outcome assessment
The incidence and all-cause mortality rates of patients with TB were assessed during 2014–2019. The date of the first registered hospital case (inpatient or outpatient) of the patient is considered the beginning of follow-up, with the death or the censoring date of the available data (31 December 2019) used as the end of follow-up. Survival probability estimates were performed by demographic factors and various diagnoses over the follow-up period. Total time at risk of death was calculated as a sum of all TB patients’ follow-up time.
TB diagnosis is confirmed in hospitals using standard procedures based on the guidelines of the WHO, including microscopy, LJ culture, GeneXpert, line probe assay and drug susceptibility testing (phenotypic or genotypic molecular analysis such as PCR and LAMP (Loop-mediated isothermal amplification)).
We calculated each year’s TB incidence as the number of new cases over total population size for the year. Likewise, the mortality rate was defined as the number of deaths over population size.15
Statistical analysis
Descriptive, bivariate and multivariable analyses were performed on the merged dataset of outpatient and inpatient hospital records. A number of patients were provided for descriptive and bivariate analyses, and median values were provided instead of mean values for age distribution, since age distribution did not follow normal (Gaussian) distribution. Missing data is kept as it is. T-test and χ2 test were used for crude categorical variables’ comparison. For all-cause mortality risk assessment, crude and adjusted Cox regression models and Kaplan-Meier survival estimates were made using the Wald’s test and log-rank test for statistical significance. Cox regression models were adjusted for demographic factors, main diagnoses and comorbidities. Loss of follow-up was recorded and used in Kaplan-Meier survival estimates. Time at risk was calculated using date of hospitalisation, date of death, date of loss of follow-up. P values are two sided and reported as statistically significant at <0.05 for all analyses.
Statistical analyses, data cleaning and data management were performed with the use of Stata V.16 MP2 (Stata Corporation, College Station, Texas, USA).
More information about the study data and methodology has been published by our research team.6
Patient and public involvement
None.
Results
Demographic data
The demographic data is summarised in table 1. During 2014–2019, there were a total of 57 685 female and 91 437 male patients with TB with a median age of 36. The majority of patients were ethnic Kazakhs (69.9%) living in urban area (58%) and had respiratory TB (60.8% for ICD-10 code A16 and 33.0% for A15). The following comorbid conditions such as HIV, DM, stroke and chronic kidney disease (CKD) were present in 0.3%, 4.9%, 0.9 and 3.8% of patients with TB, respectively. The overall number of patients in Kazakhstan’s regions and big cities is presented in online supplemental table 1.
Supplemental material
TB incidence and all-cause mortality rate among patients
From 2014 to 2019, the TB incidence decreased constantly from 227.0 to 69.1 pcm. However, mortality rate remained incrementally higher between the study periods (figure 1). TB incident cases for each TB ICD-10 code are shown in online supplemental figure 2.
Supplemental material
The highest age-specific respiratory TB incidence was in patients’ younger ages, corresponding to the 18–34 age group (online supplemental figure 3). Male patients have had significantly greater incidence of both pulmonary and extrapulmonary TB compared with female patients (online supplemental file 3).
Supplemental material
Among regions of Kazakhstan, the highest incidence was observed in Akmola region and the lowest one in Shymkent city in 2019. North Kazakhstan region had the highest all-cause mortality (online supplemental figure 4 and table 1). Incidence and mortality for regions between 2014 and 2019 are provided in online supplemental figure 5.
Supplemental material
Supplemental material
Survival probability by sex, ethnicity, main diagnosis and comorbidities
In Kaplan-Meier survival analysis, the probability of survival was higher in female patients compared with male patients, and in Kazakh ethnicity compared with others, including Russian ethnicity (figure 2). Patients with comorbid diseases such as HIV, diabetes, stroke and CKD had lower chances to survive than those without comorbidities (figure 3).
Associations between all-cause mortality and demographic data are shown in table 2.
In Kaplan-Meier analysis, women had a higher survival probability than men (Adj. HR=0.7). Russian people showed a 60% higher risk of death compared with Kazakhs (Adj. HR=1.6) (figure 2 and table 2). Patients with confirmed pulmonary TB (A15) tend to have lower survival compared with non-confirmed cases of pulmonary TB (A16) (Adj. HR=1.8) (table 2). Individuals with TB of the nervous system and miliary TB (A17 and A19, respectively) showed about seven and three times higher risk of death, respectively, (Adj. HR=6.7 and 2.8, respectively) than patients with pulmonary TB (A16).
Patients with TB and comorbid HIV had almost eight times lower survival probability in comparison to those patients without HIV (Adj. HR=7.9). Individuals with DM had a 10% greater risk of dying than those without DM (Adj. HR=1.1). Patients with TB and stroke had two times lower survival probability than patients without stroke (Adj. HR=2.0). Finally, patients with TB and CKD showed a 70% lower survival probability when compared with patients with normal kidney function (Adj. HR=1.7) (figure 3 and table 2). In the cohort, there were 15 433 death events recorded, the calculated average follow-up was 3.7 years per patient and the total time at risk was 552 105 years.
Discussion
This is one of the largest studies in Kazakhstan to examine the epidemiology of TB, including its incidence and mortality rates, using the large-scale administrative health data available in Kazakhstan. Our results showed a decrease in TB incidence and increase in all-cause mortality among patients with TB over the last 5 years, while WHO reported decreased TB mortality rate.3 16 In addition, the survival probability was statistically significantly lower for male patients compared with female patients as in line with published data.17 Kazakhs had higher survival than other ethnic groups. Finally, patients with TB and comorbid conditions such as HIV, DM, stroke and CKD showed lower survival than patients without such comorbidities that were supported by other studies.18–20
TB prevention and treatment in Kazakhstan is monitored and financed by the Ministry of Healthcare and controlled by regional primary healthcare organisations and TB control services. The main prophylaxis measures include administration of Bacille Calmette-Guerin vaccines to infants and children, annual TB skin testing for school-age children, radiography of the chest for the adult population and potential TB contacts tracings.21 22 There are also many pilot projects sponsored by non-governmental organisations for studying TB in a special category of people such as injecting drug users, patients with HIV/TB, ex-prisoners and migrants.21 According to the clinical guideline for diagnosis and treatment of TB, patients with smear-positive or culture-positive TB and those requiring thoracoscopy with biopsy for confirmation of TB are subject to inpatient therapy.23 24 Whereas those patients who have smear-negative or culture-negative TB are allowed to obtain their treatment on the outpatient basis.22 25
Our data on TB incidence and survival probability were well supported by official information and previous research findings. The latest WHO report on TB epidemiology in Kazakhstan in 2017 indicated a gradual decline in TB incidence since 2000.4 16 Positive trends in TB epidemiology were also reported in neighbouring countries such as the Russian Federation, Kyrgyzstan and Uzbekistan. For instance, Atadjan et al stated that TB incidence showed a stable decline in the Russian Federation over the past 20 years, while Kozhoyarova et al revealed that there was a slow decline in TB mortality in Kyrgyzstan from 2007 to 2017.26 27 Safaev and coworkers informed that reported TB cases dropped significantly in Uzbekistan over the last 20 years.28
The official WHO report (2017) highlighted that adjusted TB mortality decreased substantially during 2005–2015, on the contrary, we found out that all-cause mortality almost doubled during 2014–2019.3 However, we did not obtain TB-specific mortality in our study, as the UNEHS databases did not provide us with information regarding the cause of death. A potential explanation might be the increase in the number and life-expectancy of the entire population, which led to a proportional increase in TB population mortality.
Increase in the mortality could be due to excessive hospitalisation of patients with TB in Kazakhstan, despite WHO recommendations for outpatient treatment to decrease the risk of in-hospital transmission of drug-resistant strains.5 WHO declares that multidrug-resistant (MDR) TB incidence in Kazakhstan stays at a high stable level. For instance, MDR to rifampicin (MDR/RR-TB) accounts for about 27% of new cases of TB and 44% of treated TB cases.29 Moreover, increased mortality among patients with TB might be related to comorbid conditions. Lin et al claim that risk factors for death among patients with TB are non-infective comorbidities, especially liver cirrhosis, HIV infection and multidrug-resistant TB.30
In line with published data, we observed that female patients with TB had higher survival than male patients. For instance, Feng et al showed that women with TB had a significantly higher survival probability than men.17 Furthermore, Chidambaram et al in their retrospective cohort study concluded that men had higher all-cause mortality compared with women.31 Gender differences in survival could be explained by the fact that women are more likely to visit their healthcare providers and less likely to default from treatment than men.32
In Ukraine, Khalife et al found that patients with central nervous system (CNS) TB had a higher probability of dying compared with a pulmonary form of TB.15 Qian et al observed that extrapulmonary TB was often associated with late diagnosis, female gender, positive HIV status, immunosuppression and end-stage renal disease.33 We also found that people with miliary TB (A19) and TB of the nervous system (A17) showed a lower survival probability compared with pulmonary TB.
Our research revealed that patients with TB and comorbid conditions such as HIV, DM, stroke and CKD tended to have lower survival than patients without those comorbidities. These findings are consistent with previous studies that have investigated the impact of comorbidities on TB outcomes.
Shaweno and Worku analysed patients with TB and HIV on DOTs therapy for 8 months and reported that HIV-positive patients had 60% lower adjusted survival than HIV-negative patients. People with positive HIV status have suppressed immune systems, making them more likely to have HIV/TB coinfection.18 Oursler et al concluded that patients with TB and DM had lower survival than those without DM.19 Moreover, Baker et al in their systematic review revealed that DM was related to a greater risk of TB relapse, treatment failure and death.34 Jeon and Murray remind that DM impairs innate and adaptive immunity, thus putting patients at increased risk of developing TB.23
Patients with TB and stroke showed lower survival in comparison to those without stroke. Sheu et al revealed that patients with CNS-TB were about 50% more likely to develop a stroke than patients with non-TB.35 Zhang et al reported that the mortality of patients with TB and stroke was about three times greater than those without stroke.36 People with TB and CKD had a lower survival probability than patients without CKD. Accumulation of uremic toxins and disturbance of vitamin D metabolism was associated with suppression of the immune system and reactivation of latent TB or development of active TB. Xiao et al reported that the mortality rate for patients with TB and non-dialysis CKD, as well as dialysis CKD, were 37% and 32%, respectively.20 Li et al showed that higher levels of epidermal growth factor receptor were associated with lower risk of TB infection.24
Despite the significant contribution of this study to the epidemiological understanding of TB in Kazakhstan, several limitations were identified that need to be discussed. First, we did not calculate TB prevalence due to the absence of data on recurrent admissions, successful treatment and TB registry exit. However, this missing data has little impact on estimation of incidence, mortality and survival analysis. In addition, the comprehensive nature of the database minimises the risk of under-reporting. Second, the absence of information on the proportion of patients with multidrug-resistant and extensively resistant TB limits our ability to fully understand the burden of disease in Kazakhstan. These patients are more likely to suffer from comorbid conditions, have a worse prognosis and show lower survival. Third, we were unable to analyse TB-specific mortality as the UNEHS databases only provided all-cause mortality data. Fourth, the discrepancies between outpatient and inpatient databases may have resulted in some missing variables and observations, leading to possible inaccuracies in our analysis. Finally, the lack of patient-specific clinical and laboratory data, as well as treatment protocols, could have influenced our multivariable Cox regression analysis.
Despite these limitations, our study provides important insights into the TB incidence and mortality rates among patients with TB in Kazakhstan using large-scale administrative health data. We recommend that the government of Kazakhstan prioritise efforts to improve surveillance systems and health data to accurately estimate the true prevalence and cause-specific mortality of TB and MDR-TB. Additionally, efforts should be made to enhance the management of patients with TB and comorbidities such as HIV, diabetes, stroke and CKD to improve their survival rates.
Conclusion
Our study provides valuable insights into TB epidemiology in Kazakhstan based on large-scale administrative healthcare data for 2014–2019. Results showed declining TB incidence and increasing all-cause mortality among patients with TB. We described higher survival rates for females and Kazakh patients. Patients with TB and comorbidities such as HIV, DM, stroke and CKD had a lower survival probability than patients without comorbidities. These findings urge for evidence-based interventions and policies, such as strengthening TB surveillance, diagnosis and improving treatment adherence, especially for MDR-TB.
Our study highlights the importance of improving the use of health data to inform these interventions and the need for continued investment in TB research. Implementing the recommendations, Kazakhstan’s government can improve public health and contribute to the global effort to end TB.
Data availability statement
Data may be obtained from a third party and are not publicly available.
Ethics statements
Patient consent for publication
Ethics approval
This study was approved by the Nazarbayev University Institutional Review Ethics Committee (NU-IREC 490/18112021), with exemption from informed consent. There was no patient and public involvement in developing this study.
Acknowledgments
Portions of this manuscript were presented as posters at the International scientific–practical conference ‘COVID-19 and other topical problems of Central Asia’ on 23–24 June 2022 (Shymkent, Kazakhstan) and at the 15th European Public Health Conference in 9–12 November 2022 (Berlin, Germany),37 with the conference abstract published elsewhere. We used the RECORD cohort checklist when writing our manuscript.38
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
YSakko and MM are joint first authors.
YSakko and MM contributed equally.
Contributors All authors approved the final version of the manuscript and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. AGaipov, the PI of the project and the guarantor of this article, was involved in conceptualisation and methodology of study, project administration and funding acquisition. YSakko involved in data analysis and visualisation. MM and AK involved in writing—original draft preparation. YSakko, DS, KM, SY, GZ, AGusmanov involved in data curation. YSakko, MM, AK, DS, KM, AGusmanov, GZ, SY, YSemenova, BLC, AS-S and AGaipov involved in article revision.
Funding This work was supported by the Ministry of Education and Science of the Republic of Kazakhstan grant funding (Funder project reference: #AP09259016, title: Epidemiology and forecasting of infectious diseases in Kazakhstan using big healthcare data, mathematical modeling and machine learning).
Map disclaimer The inclusion of any map (including the depiction of any boundaries therein), or of any geographic or locational reference, does not imply the expression of any opinion whatsoever on the part of BMJ concerning the legal status of any country, territory, jurisdiction or area or of its authorities. Any such expression remains solely that of the relevant source and is not endorsed by BMJ. Maps are provided without any warranty of any kind, either express or implied.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.