Forecasting tuberculosis incidence in iran using box-jenkins models

Forecasting tuberculosis incidence: a review of time series and machine learning models for prediction and eradication strategies

Despite efforts by the World Health Organization (WHO), tuberculosis (TB) remains a leading cause of fatalities globally. This study reviews time series and machine learning models for TB incidence prediction, identifies popular algorithms, and highlights the need for further research to improve accuracy and global scope. SCOPUS, PUBMED, IEEE, Web of Science, and PRISMA were used for search and article selection from 2012 to 2023. The results revealed that ARIMA, SARIMA, ETS, GRNN, BPNN, NARNN, NNAR, and RNN are popular time series and ML algorithms adopted for TB incidence rate predictions. The inaccurate TB incidence prediction and limited global scope of prior studies suggest a need for further research. This review serves as a roadmap for the WHO to focus on regions that require more attention for TB prevention and the need for more sophisticated models for TB incidence predictions.

Tuberculosis incidence in Brazil: time series analysis between 2001 and 2021 and projection until 2030

OBJECTIVE

To assess the incidence of tuberculosis in Brazil between 2001 and 2022 and estimate the monthly incidence forecast until 2030.

METHODS

This is a time-series study based on monthly tuberculosis records from the Notifiable Diseases Information System and official projections of the Brazilian population. The monthly incidence of tuberculosis from 2001 to 2022 was evaluated using segmented linear regression to identify trend breaks. Seasonal autoregressive integrated moving average (Sarima) was used to predict the monthly incidence from 2023 to 2030, deadline for achieving the sustainable development goals (SDGs).

RESULTS

There was a decrease in incidence between January/2001 and December/2014 (4.60 to 3.19 cases-month/100,000 inhabitants; β=-0.005; p<0.001), followed by an increase between January/2015 and March /2020 (β=0.013; p<0.001). There was a sharp drop in cases in April/2020, with the onset of the pandemic, and acceleration of the increase in cases since then (β=0.025; p<0.001). A projection of 124,245 cases in 2030 was made, with an estimated incidence of 4.64 cases-month/100,000 inhabitants, levels similar to those in the 2000s. The Sarima model proved to be robust, with error of 4.1% when removing the pandemic period.

CONCLUSION

The decreasing trend in tuberculosis cases was reversed from 2015 onwards, a period of economic crisis, and was also impacted by the pandemic when there was a reduction in records. The Sarima model can be a useful forecasting tool for epidemiological surveillance. Greater investments in prevention and control need to be made to reduce the occurrence of tuberculosis, in line with the SDGs.

Tuberculosis in Poland: Epidemiological and Molecular Analysis during the COVID-19 Pandemic

The COVID-19 pandemic may have a negative impact on the proper implementation of TB control programmes and may increase TB incidence rates in the near future. The aim of this study was to perform an epidemiological and molecular analysis of Mycobacterium tuberculosis strains cultured from tuberculosis patients in Poland in 2020 and to compare the results of monitoring drug-resistant tuberculosis in Poland with previous studies in 2012 and 2016. The analysis was based on questionnaires and strains sent by regional laboratories during the 12 months of 2020. Molecular analysis was performed by spoligotyping 20% of the strains sensitive to the four primary antimycobacterial drugs and all of the drug-resistant strains. The number of strains sent for analysis dropped threefold, from 4136 in 2012 to 1383 in 2020. The incidence of tuberculosis among men was higher than among women. There was an increase in strains’ resistance to antimycobacterial drugs in both newly diagnosed patients, from 4.4% in 2012 to 6.1% in 2020, and previously treated patients, from 11.7% to 12.3%. Four-year resistance increased to 1% and 2.1%, respectively. The spoligotype SIT1 was the most abundant among the resistant strains (17%), and SIT53 (13.9%) was the most common among susceptible strains.

Time series forecasting for tuberculosis incidence employing neural network models

Every effort aimed at stopping the expansion of Tuberculosis is important to national programs' struggle to combat this disease. Different computational tools have been proposed in order to design new strategies that allow managing potential patients and thus providing the correct treatment. In this work, artificial neural networks were used for time series forecasting, which were trained with information on reported cases obtained from the national vigilance institution in Colombia. Three neural models were proposed in order to determine the best one according to their forecasting performance. The first approach employed a nonlinear autoregressive model, the second proposal used a recurrent neural network, and the third proposal was based on radial basis functions. The results are presented in terms of the mean average percentage error, which indicates that the models based on traditional methods show better performance compared to connectionist ones. These models contribute to obtaining dynamic information about incidence, thus providing extra-help for health authorities to propose more strategies to control the disease's spread.

Temporal trends and prediction of bovine tuberculosis: a time series analysis in the North-East of Iran

BACKGROUND

Bovine tuberculosis (BTB) is a disease with high economic relevance.

AIMS

This study aimed to determine a fast alert surveillance system for bTB before the outbreak in the epidemic region of Iran.

METHODS

This cross-sectional study was conducted using the Auto-Regressive Integrated Moving Average (ARIMA) model for monthly bTB detections (reactors). These reactor cases result from the positive Tuberculin Purified Protein Derivative (PPD) test on cattle farms for the period between April 2007 and March 2019 in Razavi Khorasan province. Autocorrelation functions (ACF) and partial autocorrelation functions (PACF) plots were used to determine model parameters. The Akaike Information Criteria (AIC) were employed to select the best-fitted model. The root mean square error (RMSE) was applied for the evaluation of the models. Then, the best-fitted model was hired to predict the cases for 12 oncoming months. The data were analysed by STATA (ver. 14) software with a significant level at P≤0.05.

RESULTS

ARIMA (3, 0, 3) 12 was introduced as a recommended fitted model according to white noise residual test (Q=22.87 and P=0.98), lower AIC (541.85), and more precise model RMSE (1.50). However, the forecast values were more than the observed values.

CONCLUSION

The application and interpretation of ARIMA models are straightforward, and may be used as immediate tools for monitoring systems. However, we proposed an Auto-Regressive Integrated Moving Average with Exogenous Input (ARIMAX) model with some measurable exotic factors such as economic fluctuations, climate changes, and pulmonary tuberculosis to introduce a more precise and accurate model for the fast alert surveillance system.

Forecasting the incidence of mumps in Chongqing based on a SARIMA model

BACKGROUND

Mumps is classified as a class C infection disease in China, and the Chongqing area has one of the highest incidence rates in the country. We aimed to establish a prediction model for mumps in Chongqing and analyze its seasonality, which is important for risk analysis and allocation of resources in the health sector.

METHODS

Data on incidence of mumps from January 2004 to December 2018 were obtained from Chongqing Municipal Bureau of Disease Control and Prevention. The incidence of mumps from 2004 to 2017 was fitted using a seasonal autoregressive comprehensive moving average (SARIMA) model. The root mean square error (RMSE) and mean absolute percentage error (MAPE) were used to compare the goodness of fit of the models. The 2018 incidence data were used for validation.

RESULTS

From 2004 to 2018, a total of 159,181 cases (93,655 males and 65,526 females) of mumps were reported in Chongqing, with significantly more men than women. The age group of 0-19 years old accounted for 92.41% of all reported cases, and students made up the largest proportion (62.83%), followed by scattered children and children in kindergarten. The SARIMA(2, 1, 1) × (0, 1, 1)12 was the best fit model, RMSE and MAPE were 0.9950 and 39.8396%, respectively.

CONCLUSION

Based on the study findings, the incidence of mumps in Chongqing has an obvious seasonal trend, and SARIMA(2, 1, 1) × (0, 1, 1)12 model can also predict the incidence of mumps well. The SARIMA model of time series analysis is a feasible and simple method for predicting mumps in Chongqing.

Forecasting the incidence of acute haemorrhagic conjunctivitis in Chongqing: a time series analysis

Acute haemorrhagic conjunctivitis is a highly contagious eye disease, the prediction of acute haemorrhagic conjunctivitis is very important to prevent and grasp its development trend. We use the exponential smoothing model and the seasonal autoregressive integrated moving average (SARIMA) model to analyse and predict. The monthly incidence data from 2004 to 2017 were used to fit two models, the actual incidence of acute haemorrhagic conjunctivitis in 2018 was used to validate the model. Finally, the prediction effect of exponential smoothing is best, the mean square error and the mean absolute percentage error were 0.0152 and 0.1871, respectively. In addition, the incidence of acute haemorrhagic conjunctivitis in Chongqing had a seasonal trend characteristic, with the peak period from June to September each year.

Prevalence of anemia among patients with tuberculosis: A systematic review and meta-analysis

INTRODUCTION

Anemia is one of the most common hematologic problems occurs among patients with tuberculosis (TB). Many studies have been carried out estimating the prevalence of anemia among TB patients in different countries reported various results. This study aims to estimate the combined estimate of the anemia prevalence among these patients using systematic review and meta-analysis.

METHODS

Required primary studies were provided after a comprehensive and systematic search in PubMed, Scopus, Science direct, Web of Science and also Google scholar search engine. These studies were then quality assessed using Newcastle-Ottawa Scale checklist. Random effects model was applied for combining the point prevalence with 95% confidence intervals.

RESULTS

Of 41 papers entered into the meta-analysis, prevalence (95% confidence interval) of anemia among all TB patients as well as men and women were 61.53% (53.44-69.63), 66.95% (51.75-82.14) and 72.67% (60.79-84.54) respectively. Prevalence (95% confidence intervals) of mild, moderate and severe anemia were 35.67% (27.59-43.46), 31.19% (25.15-37.24) and 11.61% (7.88-15.34) respectively. In addition, prevalence (95% confidence intervals) of chronic disease anemia and iron deficiency anemia were 49.82% (15.58-84.07) and 20.17% (6.68-33.65) respectively.

CONCLUSION

Prevalence of anemia among TB patients was high especially among women. More than 43% of these patients suffered from moderate and severe anemia and about half of them had chronic disease anemia.

Application of a combined model with seasonal autoregressive integrated moving average and support vector regression in forecasting hand-foot-mouth disease incidence in Wuhan, China

Hand-foot-mouth disease (HFMD) is a serious public health problem with increasing cases and substantial financial burden in China, especially in Wuhan city. Hence, there is an urgent need to construct a model to predict the incidence of HFMD that could make the prevention and control of this disease more effective.The incidence data of HFMD of Wuhan city from January 2009 to December 2016 were used to fit a combined model with seasonal autoregressive integrated moving average (SARIMA) model and support vector regression (SVR) model. Then, the SARIMA-SVR hybrid model was constructed. Subsequently, the fitted SARIMA-SVR hybrid model was applied to obtain the fitted HFMD incidence from 2009 to 2016. Finally, the fitted SARIMA-SVR hybrid model was used to forecast the incidence of HFMD of the year 2017. To assess the validity of the model, the mean square error (MSE) and mean absolute percentage error (MAPE) between the actual values and predicted values of HFMD incidence (2017) were calculated.From 2009 to 2017, a total of 107636 HFMD cases were reported in Wuhan City, Hubei Province, and the male-to-female ratio is 1.60:1. The age group of 0 to 5 years old accounts for 95.06% of all reported cases and scattered children made up the large proportion (accounted for 56.65%). There were 2 epidemic peaks, from April to July and September to December, respectively, with an emphasis on the former. High-prevalence areas mainly emerge in Dongxihu District, Jiangxia District, and Hongshan District. SARIMA (1,0,1)(0,0,2)[12] is the optimal model given with a minimum Akaike information criterion (AIC) (700.71), then SVR model was constructed by using the optimum parameter (C = 100000, =0.00001, =0.01). The forecasted incidences of single SARIMA model and SARIMA-SVR hybrid model from January to December 2017 match the actual data well. The single SARIMA model shows poor performance with large MSE and MAPE values in comparison to SARIMA-SVR hybrid model.The SARIMA-SVR hybrid model in this study showed that accurate forecasting of the HFMD incidence is possible. It is a potential decision supportive tool for controlling HFMD in Wuhan, China.

Estimating the incidence of tuberculosis cases reported at a tertiary hospital in Ghana: a time series model approach

Background

The incidence of Tuberculosis (TB) differs among countries and contributes to morbidity and mortality especially in the developing countries. Trends and seasonal changes in the number of patients presenting with TB have been studied worldwide including sub-Saharan Africa. However, these changes are unknown at the Korle-Bu Teaching Hospital (KBTH). The aim of this study was to obtain a time series model to estimate the incidence of TB cases at the chest clinic of the Korle-Bu Teaching hospital.

Methods

A time series analysis using a Box-Jenkins approach propounded as an autoregressive moving average (ARIMA) was conducted on the monthly TB cases reported at the KBTH from 2008 to 2017. Various models were stated and compared and the best was found to be based on the Akaike Information Criterion and Bayesian Information Criterion.

Results

There was no evidence of obvious increasing or decreasing trend in the TB data. The log-transformed of the data achieved stationarity with fairly stable variations around the mean of the series. ARIMA (1, 0, 1) or ARMA (1,1) was obtained as the best model. The monthly forecasted values of the best model ranged from 53 to 55 for the year 2018; however, the best model does not always produce the best results with respect to the mean absolute and mean square errors.

Conclusions

Irregular fluctuations were observed in the 10 -year data studied. The model equation to estimate the expected monthly TB cases at KBTH produced an AR coefficient of 0.971 plus an MA coefficient of − 0.826 with a constant value of 4.127. The result is important for developing a hypothesis to explain the dynamics of TB occurrence so as to outline prevention programmes, optimal use of resources and effective service delivery.

Electronic supplementary material

The online version of this article (10.1186/s12889-018-6221-z) contains supplementary material, which is available to authorized users.

Forecasting the incidence of tuberculosis in China using the seasonal auto-regressive integrated moving average (SARIMA) model

Objectives

The aims of this study were to develop a forecasting model for the incidence of tuberculosis (TB) and analyze the seasonality of infections in China; and to provide a useful tool for formulating intervention programs and allocating medical resources.

Methods

Data for the monthly incidence of TB from January 2004 to December 2015 were obtained from the National Scientific Data Sharing Platform for Population and Health (China). The Box–Jenkins method was applied to fit a seasonal auto-regressive integrated moving average (SARIMA) model to forecast the incidence of TB over the subsequent six months.

Results

During the study period of 144 months, 12,321,559 TB cases were reported in China, with an average monthly incidence of 6.4426 per 100,000 of the population. The monthly incidence of TB showed a clear 12-month cycle, and a seasonality with two peaks occurring in January and March and a trough in December. The best-fit model was SARIMA (1,0,0)(0,1,1)₁₂, which demonstrated adequate information extraction (white noise test, p > 0.05). Based on the analysis, the incidence of TB from January to June 2016 were 6.6335, 4.7208, 5.8193, 5.5474, 5.2202 and 4.9156 per 100,000 of the population, respectively.

Conclusions

According to the seasonal pattern of TB incidence in China, the SARIMA model was proposed as a useful tool for monitoring epidemics.

Estimating the prevalence of Positive Tuberculin Skin Test Reactions in General Population and High-risk Groups: A Meta-analysis

Results of tuberculin skin test (TST) surveys among different populations have been reported in many studies as a method for detecting primary Mycobacterium tuberculosis infection. Combining these results provides reliable estimates of primary latent tuberculosis (TB) infection for health policymakers. This study aims to estimate the prevalence of latent TB infection in general and high-risk populations in Iran. National and international databanks were searched using specific keywords. After restricting the search strategy, duplicates exclusion, reviewing titles, abstracts and full texts, and quality assessment, eligible papers were selected. The heterogeneity between the results was assessed according to Cochrane and I-squared indices. The prevalence of positive TST reactions was estimated using fixed and random effect models. Totally, 33 papers were entered into the meta-analysis reporting the TST results in 12693 people aged over 18. The prevalence (95% confidence intervals) of positive TST reactions in all groups, general population, health staff, medical students, household close contacts, patients with TB, immunocompromised patients, HIV/AIDS patients, and those with risky behaviors was 26.2% (19.6-32.8), 25.4% (4.8-46.1), 38.9% (27.4-50.9), 13.4% (9.9-16.7), 35.9% (16.4-55.5), 13.7% (8.4-18.9), 29.4% (21.2-37.7), and 14.6% (3.9-25.3), respectively. Our study showed great varieties of positive TST results among different Iranian subpopulations. Furthermore, the prevalence of latent TB infection among health professionals and family members of TB patients was considerably different from that of the other subgroup. Since TB control programs such as active case finding are routinely conducted among household close contacts and HIV/AIDS cases, other high-risk groups including health-care workers and immunocompromised patients should be taken into consideration in these preventive programs.

Forecasting the Incidence of Mumps in Zibo City Based on a SARIMA Model

This study aimed to predict the incidence of mumps using a seasonal autoregressive integrated moving average (SARIMA) model, and provide theoretical evidence for early warning prevention and control in Zibo City, Shandong Province, China. Monthly mumps data from Zibo City gathered between 2005 and 2013 were used as a training set to construct a SARIMA model, and the monthly mumps in 2014 were defined as a test set for the model. From 2005 to 2014, a total of 8722 cases of mumps were reported in Zibo City; the male-to-female ratio of cases was 1.85:1, the age group of 1-20 years old accounted for 94.05% of all reported cases, and students made up the largest proportion (65.89%). The main serious endemic areas of mumps were located in Huantai County, Linzi District, and Boshan District of Zibo City. There were two epidemic peaks from April to July and from October to January in next year. The fitted model SARIMA (0, 1, 1) (0, 1, 1)₁₂ was established (AIC = 157.528), which has high validity and reasonability. The SARIMA model fitted dynamic changes of mumps in Zibo City well. It can be used for short-term forecasting and early warning of mumps.

Determinant factors for mortality during treatment among tuberculosis patients: Cox proportional hazards model

BACKGROUND

Investigating the survival of tuberculosis (TB) patients is one of the main parts of the TB control program. It can be related to many factors. This study aimed to estimate the survival experience and its associated factors among these patients.

METHODOLOGY

All TB patients detected during March 2005 to 31 September 2014 were entered into this prospective cohort. Each patient was investigated from the diagnosis date and followed until the last available information during treatment. Data analysis was performed using Kaplan Meier and multivariate Cox regression models.

RESULTS

The survival experience of 2493 TB patients was investigated 73.7% of which were pulmonary type. Mean and median survival time were 6.5 and 6.2 months respectively. The incidence rate of death among patients during the treatment courses was 0.99 (95% confidence interval: 0.84-1.1) per 100 person-months. Controlling the confounders, the incidence (95% confidence interval) of death was significantly higher among men (HR=1.8; 1.2-2.6), diabetic patients (HR=1.7; 1.2-2.6), cancerous patients (HR=4.8; 2.6-8.8) and HIV positive patients (HR=22.1; 7.3-66.4).

CONCLUSION

This study showed that male gender, TB/HIV co-infection and concurrent development of TB and cancer were determinant factors of death during the treatment period of TB.

Estimating Tuberculin Skin Test Reactions among Children and Teenagers Who Received the Bacillus Calmette-Guerin Vaccination at Birth: A Meta-analysis

OBJECTIVES

Tuberculin skin reaction size is one indicator of bacillus Calmette-Guérin (BCG) vaccine efficacy and a way to diagnose latent infection. Several primary studies have examined this issue. Combining the results of these studies using a meta-analysis will provide reliable evidence regarding this indicator for policymakers. This study aimed to estimate the total frequency of different tuberculin skin test reactions among Iranian children and teenagers who received the BCG vaccination at birth.

METHODS

National and international databanks were searched using relevant keywords. After the search strategy was restricted and duplicates were excluded, the titles and abstracts of the remaining papers were screened. All included studies included healthy children who received the BCG vaccine without confirmed tuberculosis exposure. Heterogeneity of the results was assessed using the Cochrane test and I² index showed the random effects model as the best model for estimating the pooled results.

RESULTS

We combined the results of 14 primary studies including purified protein derivative reaction test measures of 26,281 Iranian children. The frequencies (95% confidence intervals) of the reactions were 8.5% (6.2-10.8) for patients with a reaction size ≥ 10 mm, 29.9% (22.3-37.4) for a reaction size of 5-9 mm, and 60% (48.9-71.1) for a reaction size < 5 mm.

CONCLUSION

Our study showed that large numbers of Iranian children and teens have no positive BCG vaccine reaction and a considerable number of children have been exposed to Mycobacterium tuberculosis.

Comparing Seasonal Pattern of Laboratory Confirmed Cases of Pertussis with Clinically Suspected Cases

Objectives

During recent decades, there has been limited attention on the seasonal pattern of pertussis within a high vaccine coverage population. This study aimed to compare the seasonal patterns of clinical suspected pertussis cases with those of laboratory confirmed cases in Iran.

Methods

The current study was conducted using time series methods. Time variables included months and seasons during 2011–2013. The effects of seasons and months on the incidence of pertussis were estimated using analysis of variance or Kruskal–Wallis.

Results

The maximum average incidence of clinically confirmed pertussis was 23.3 in July (p = 0.04), but the maximum incidence of clinical suspected pertussis was 115.7 in May (p = 0.6). The maximum seasonal incidences of confirmed and clinical pertussis cases were reported in summer (average: 12, p = 0.004), and winter (average: 108.1; p = 0.4), respectively.

Conclusion

The present study showed that the seasonal pattern of laboratory confirmed pertussis cases is highly definite and different from the pattern of clinical suspected cases.

The incidence of recurrence of tuberculosis and its related factors in smear-positive pulmonary tuberculosis patients in Iran: A retrospective cohort study

Background and Aim:

Studying the recurrence of smear-positive pulmonary tuberculosis (TB) is a convenient way to evaluate the effectiveness of TB control programs and identify vulnerable patients. In the present study, the rate of recurrence of TB and its predictors were determined in Iran.

Materials and Methods:

This study was a retrospective cohort. Eligible people were patients with smear-positive TB who were diagnosed from 2002 to 2011. The end of the follow-up time was December 2013. The number of people who entered the cohort was 1,271 subjects. In order to determine the predictors of recurrence, multivariate logistic regression was used. Analysis was done using SPSS 20.

Results:

The recurrence incidence was 8.3% and in 85.9% of these patients, it occurred in the time interval of 1-5 years after successful treatment. The recurrence rate was not significantly related to gender, age group, and diabetes. But it was significantly higher in patients whose sputum smear grading before treatment was 2 + or more, patients with positive sputum smear at the end of the second month of the treatment, patients who had completed treatment, and patients who were smokers (P < 0.05).

Conclusions:

Our study showed that a considerable percentage of smear-positive pulmonary TB patients experience recurrence and that some patients are at a higher risk of recurrence.

Predicting the Incidence of Smear Positive Tuberculosis Cases in Iran Using Time Series Analysis

BACKGROUND

Determining the temporal variation and forecasting the incidence of smear positive tuberculosis (TB) can play an important role in promoting the TB control program. Its results may be used as a decision-supportive tool for planning and allocating resources. The present study forecasts the incidence of smear positive TB in Iran.

METHODS

This a longitudinal study using monthly tuberculosis incidence data recorded in the Iranian National Tuberculosis Control Program. The sum of registered cases in each month created 84 time points. Time series methods were used for analysis. Based on the residual chart of ACF, PACF, Ljung-Box tests and the lowest levels of AIC and BIC, the most suitable model was selected.

RESULTS

From April 2005 until March 2012, 34012 smear positive TB cases were recorded. The mean of TB monthly incidence was 404.9 (SD=54.7). The highest number of cases was registered in May and the difference in monthly incidence of smear positive TB was significant (P<0.001). SARIMA (0,1,1)(0,1,1)12 was selected as the most adequate model for prediction. It was predicted that the incidence of smear positive TB for 2015 will be about 9.8 per 100,000 people.

CONCLUSION

Based on the seasonal pattern of smear positive TB recorded cases, seasonal ARIMA model was suitable for predicting its incidence. Meanwhile, prediction results show an increasing trend of smear positive TB cases in Iran.