PATIENTS IN MALAYSIA

THE ASSOCIATION PATTERNS BETWEEN SOCIO-DEMOGRAPHIC, CLINICAL AND TREATMENT- RELATED CHARACTERISTICS IN MULTIDRUG-RESISTANT TUBERCULOSIS

PATIENTS IN MALAYSIA

FAIRUL EZWAN BIN FAHRURAZI

UNIVERSITI SAINS MALAYSIA

2018

THE ASSOCIATION PATTERNS BETWEEN SOCIO-DEMOGRAPHIC, CLINICAL AND TREATMENT- RELATED CHARACTERISTICS IN MULTIDRUG-RESISTANT TUBERCULOSIS

PATIENTS IN MALAYSIA

by

FAIRUL EZWAN BIN FAHRURAZI

Thesis Submitted in Partial Fulfilment of the Requirements for the Degree of Master of Science

(MEDICAL STATISTICS)

UNIVERSITI SAINS MALAYSIA

APRIL 2018

ii

ACKNOWLEDGEMENTS

In the name of Allah, the Most Gracious and the Most Merciful.

I would like to express my deep gratitude to my main supervisor, Dr Wan Nor Arifin bin Wan Mansor for his guidance and assistance during the course of this study. I am grateful for his generosity to share his knowledge and skills especially in R software. Without his advice and explanation, I may not be able to overcome the challenges in completing this research.

I would also like to extend my appreciation to my two co-supervisors, Assoc. Prof Dr. Sarimah Abdullah, the Coordinator of Unit Biostatistics and Research Methodology and Dr. Asmah binti Razali, Senior Assistant Director of Disease Control Division, Ministry of Health, Malaysia. Assoc. Prof Dr. Sarimah Abdullah has given me constant support, constructive comments and idea throughout this study, whereas under Dr Asmah binti Razali, I have received continuous consultation regarding the clinical side of this study and also able to complete my data collection via Tuberculosis Information System (TBIS).

I am thankful to Ministry of Health, especially Dr Chong Chee Kheong, Director of Disease Control Division and also Dr. Mohamed Naim Abdul Kadir who is the Head of Tuberculosis/Leprosy Sector for granting me the permission to access and use the TBIS data for my research.

My sincere thanks to all the lecturers and staffs in the Unit of Biostatistics and Research Methodology for their endless co-operations and guidance. Not to forget all my colleagues for their morale support and knowledge sharing.

Last but not least, I would like to thank my wife (Norazlin binti Zainuddin), as well as my children and family members for their encouragement throughout this study.

iii

To my lovely children, Izzah and Izzuddin, I dedicate this thesis.

iv

TABLE OF CONTENTS

ACKNOWLEDGEMENTS ... ii

TABLE OF CONTENTS ... iv

LIST OF TABLES ... xi

LIST OF FIGURES ... xiii

LIST OF ABBREVIATIONS ... xiv

LIST OF SYMBOLS ... xv

ABSTRAK ... xvi

ABSTRACT ... xviii

CHAPTER 1 INTRODUCTION ... 20

1.1 Tuberculosis ... 20

1.2 Multidrug-resistant Tuberculosis ... 21

1.3 Epidemiology of MDR-TB ... 22

1.3.1 Global Situation ... 22

1.3.2 Situation in Malaysia ... 24

1.4 Tuberculosis Information System (TBIS) Registry ... 24

1.5 Problem Statement ... 25

1.6 Justification of the Study... 26

CHAPTER 2 OBJECTIVES ... 27

2.1 Research Questions ... 27

2.2 Research Objectives ... 27

2.2.1 General Objectives ... 27

2.2.2 Specific Objectives ... 27

v

2.3 Research Hypothesis ... 27

CHAPTER 3 LITERATURE REVIEW ... 28

3.1 Literature Search Strategies ... 28

3.2 Important Characteristics Associated with MDR-TB ... 29

3.2.1 Socio-Demographic Characteristics ... 29

3.2.1.1 Gender ... 29

3.2.1.2 Age ... 29

3.2.1.3 Immigrant status ... 30

3.2.1.4 Ethnicity ... 30

3.2.1.5 Occupation ... 30

3.2.1.6 Education level ... 30

3.2.1.7 History of imprisonment ... 30

3.2.1.8 Residential area/type ... 31

3.2.2 Clinical and Treatment Related Characteristics ... 31

3.2.2.1 Previous TB treatment ... 31

3.2.2.2 HIV infection status ... 31

3.2.2.3 Known TB contact ... 32

3.2.2.4 Diabetic mellitus ... 32

3.2.2.5 Positive sputum smear at the end of two months of treatment ... 32

3.2.2.6 Substance abuse ... 32

3.3 Conceptual Framework ... 33

CHAPTER 4 METHODS ... 34

4.1 Study Design ... 34

4.2 Study Duration ... 34

4.3 Study Location ... 34

4.4 Study Population and Sample ... 34

4.4.1 Inclusion and Exclusion Criteria ... 35

4.4.1.1 Diagnosis of MDR-TB ... 35

4.4.2 Sample Size Determination ... 37

vi

4.4.3 Sampling Method ... 37

4.5 Variables Under Study ... 38

4.5.1.1 Gender ... 39

4.5.1.2 Immigrant status ... 39

4.5.1.3 Age group ... 39

4.5.1.4 HIV status ... 39

4.5.1.5 Sputum status ... 40

4.5.1.6 Previous TB treatment ... 40

4.6 Data Request ... 40

4.7 Definition of Terms Used ... 41

4.7.1 Tuberculosis (TB) ... 41

4.7.2 Multidrug-resistant Tuberculosis (MDR-TB) ... 41

4.7.3 MDR-TB Patient Registration Category ... 41

4.7.3.1 New ... 41

4.7.3.2 Relapse ... 41

4.7.3.3 Treatment after loss to follow-up ... 41

4.7.3.4 After failure of first treatment with first-line drugs ... 42

4.7.3.5 After failure of retreatment regimen with first-line drugs... 42

4.7.3.6 After failure of treatment with second-line drugs ... 42

4.7.3.7 Other previously treated patients ... 42

4.7.3.8 Patients with unknown previous TB treatment history ... 42

4.7.4 Treatment Outcomes ... 42

4.7.4.1 Cured ... 42

4.7.4.2 Treatment completed ... 42

4.7.4.3 Treatment failed ... 43

4.7.4.4 Died ... 43

4.7.4.5 Lost to follow-up ... 43

4.7.4.6 Not evaluated ... 43

4.7.4.7 Treatment success ... 43

4.8 Study Flow Chart ... 44

4.9 Statistical Analysis ... 44

vii

4.9.1 Log-Linear Models for Contingency Tables ... 45

4.9.2 Log-Linear Models for Two-way Contingency Table ... 45

4.9.3 Log-Linear Analysis Using R ... 47

4.9.4 Log-Linear Assumptions (von Eye and Mun, 2012)... 47

4.9.5 Log-Linear Regression Procedure... 48

4.9.5.1 Step 1: Data exploration and cleaning ... 48

4.9.5.2 Step 2: Specify model to be tested ... 48

4.9.5.3 Step 3: Estimate the model ... 50

4.9.5.4 Step 4: Model comparison ... 50

4.9.5.5 Step 5: Regression diagnostic ... 51

4.9.5.6 Step 6: Remedial measures ... 52

4.9.5.7 Step 7: Presentation, interpretation and conclusions ... 52

4.9.6 Statistical Flowchart ... 53

4.10 Ethical Consideration ... 54

4.10.1 Confidentiality ... 54

4.10.2 Conflict of Interest ... 54

CHAPTER 5 RESULTS ... 55

5.1 Descriptive Statistics ... 55

5.1.1 Socio-Demographic Characteristics ... 55

5.1.2 Clinical and Treatment Related Characteristics ... 56

5.2 Association Analysis for MDR-TB ... 57

5.2.1 Model I (Variables based on Expert Opinion) ... 57

5.2.1.1 The data ... 57

5.2.1.2 Model specification ... 58

5.2.1.3 Model Adequacy based on Goodness of Fit G² and X² ... 60

5.2.1.4 Model Estimation (Model 4) ... 62

5.2.1.5 Model comparison with saturated and independence model ... 62

5.2.1.6 Regression diagnostic ... 65

viii

5.2.1.7 Remedial measures ... 65

5.2.1.8 Impact of remedial measures ... 65

5.2.1.9 Model interpretation ... 67

5.2.2 Model II (Based on Expert Opinion and Literature Search) ... 67

5.2.2.1 The data ... 67

5.2.2.2 Model specification ... 70

5.2.2.3 Model Adequacy based on Goodness of Fit G² and X² ... 75

5.2.2.4 Model estimation (Model E3) ... 88

5.2.2.5 Model comparison with saturated and independence model ... 91

5.2.2.6 Regression diagnostic ... 92

5.2.2.7 Remedial Measures. ... 92

5.2.2.8 Model interpretation ... 95

CHAPTER 6 DISCUSSION ... 97

6.1 Socio-Demographic Characteristics ... 97

6.2 Clinical and Treatment Related Characteristics ... 97

6.3 Association patterns between characteristics of MDR-TB patients ... 98

6.3.1 Model I (Based on Expert Opinion) ... 98

6.3.2 Model II (Based on Expert Opinion and Literature Search) ... 98

6.4 Methodological Consideration ... 101

6.4.1 Log-linear Analysis ... 101

6.4.2 Sample Size ... 101

6.4.3 Adequacy of Expected Frequencies ... 102

6.4.4 Empty Cells: Sampling versus Structural Zeros. ... 102

6.4.5 Overdispersion ... 103

6.5 Strengths of the Study ... 103

6.6 Limitations of the Study ... 104

CHAPTER 7 CONCLUSION ... 105

ix

7.1 Conclusion ... 105

7.2 Recommendation ... 105

REFERENCES ... 107

APPENDIX A ... 116

APPENDIX B ... 119

APPENDIX C ... 121

APPENDIX D ... 123

APPENDIX E ... 125

APPENDIX F ... 126

APPENDIX G ... 127

APPENDIX H ... 128

APPENDIX I ... 129

APPENDIX J ... 130

APPENDIX K ... 134

APPENDIX L ... 136

APPENDIX M ... 137

APPENDICES

APPENDIX A : Universiti Sains Malaysia Ethical Approval

APPENDIX B : Medical Research and Ethic Committee (MREC) Ethical Approval

APPENDIX C : Data Collection Form

APPENDIX D : Descriptive Statistics of MDR-TB (R Software)

APPENDIX E : Model to model comparison (Model I) (R Software)

x

APPENDIX F : Expected Frequencies and Standardized Residuals for Model 4 (R Software)

APPENDIX G : Best model comparison with saturated and independence model (Model I) (R Software)

APPENDIX H : Remedial Measures for Model 4 (R Software)

APPENDIX I : Model Interpretation for Model 4 and Model 4a (R Software)

APPENDIX J : Model to model comparison (Model II) (R Software)

APPENDIX K : Expected Frequencies and Standardized Residuals for Model E3 (R Software)

APPENDIX L : Best model comparison with saturated and independence model (Model II) (R Software)

APPENDIX M : Model Interpretation for Model E3 (R Software)

xi

LIST OF TABLES

Table 4.1 Sample size determination for patterns of association in the characteristics of MDR-

TB patients in Malaysia ... 37

Table 4.2 Variables under study based on expert opinion (Model I) and literature and expert opinion (Model II) with respective subgroup ... 40

Table 4.3 Two-way contingency table I x J ... 45

Table 5.1 Socio-demographic characteristics of multidrug-resistant tuberculosis cases in Malaysia from 2012-2016 ... 55

Table 5.2 Clinical and treatment related characteristics of multidrug-resistant tuberculosis cases in Malaysia from 2012-2016 ... 56

Table 5.3 Data set of MDR-TB in Model I (Based on Expert Opinions) ... 57

Table 5.4 Data set in contingency table form ... 57

Table 5.5 Specification of model based on descending model building strategies ... 59

Table 5.6 Checking model adequacy with main effect and interaction term ... 60

Table 5.7 Summary of parameter for Model 2 ... 61

Table 5.8 Summary of parameter for Model 4 ... 61

Table 5.9 Summary of parameter for Model 5 ... 61

Table 5.10 Summary of parameter for Model 6 ... 61

Table 5.11 Expected frequencies and standardized residuals for Model 4 ... 62

Table 5.12 Model comparison between best model and saturated model ... 62

Table 5.13 Model comparison between best model and independence model ... 63

Table 5.14 Association patterns in MDR-TB patient characteristics by log-linear regression for Model 4 before remedial measures ... 64

Table 5.15 Dispersion parameter based on Breslow's suggestion ... 65

Table 5.16 Association patterns in MDR-TB patient characteristics by log-linear regression for Model 4 after remedial measures. ... 66

xii

Table 5.17 Data set of MDR-TB in Model II (Based on Expert Opinions and Literature Search)

... 67

Table 5.18 Data set in contingency table form ... 69

Table 5.19 Specification of model based on descending model building strategies ... 71

Table 5.20 Checking model adequacy with main effect and interaction term ... 75

Table 5.21 Summary of parameter for Model C1 ... 77

Table 5.22 Summary of parameter for Model C5 ... 78

Table 5.23 Summary of parameter for Model C6 ... 79

Table 5.24 Summary of parameter for Model C9 ... 80

Table 5.25 Summary of parameter for Model D1 ... 81

Table 5.26 Summary of parameter for Model D2 ... 82

Table 5.27 Summary of parameter for Model D3 ... 83

Table 5.28 Summary of parameter for Model D6 ... 84

Table 5.29 Summary of parameter for Model E ... 85

Table 5.30 Summary of parameter for Model E2 ... 86

Table 5.31 Summary of parameter for Model E3 ... 87

Table 5.32 Expected Frequencies and standardized residuals for Model E3 ... 88

Table 5.33 Model comparison between best model and saturated model ... 91

Table 5.34 Model comparison between best model and independence model ... 91

Table 5.35 Association patterns in MDR-TB patient characteristics by log-linear regression for Model E3. ... 93

Table 5.36 Contingency table for age category and immigrant status ... 94

Table 5.37 Contingency table for age category and gender ... 94

Table 5.38 Contingency table for HIV status and gender ... 94

Table 5.39 Contingency table for previous TB treatment and gender ... 94

Table 5.40 Contingency table for previous TB treatment and sputum status ... 95

xiii

LIST OF FIGURES

Figure 1.1 The estimated TB incidence rate in 2015 ... 20 Figure 1.2 Estimated incidence of MDR-TB in 2016 by WHO region (World Health Organization, 2016) ... 22 Figure 1.3 The estimated incidence of MDR/RR-TB in 2016 by WHO, for countries with at least 1000 incident cases(World Health Organization, 2016) ... 23 Figure 1.4 MDR-TB Cases in Malaysia from 2004 to 2015(Ministry of Health Malaysia, 2016) ... 24 Figure 1.5 Tuberculosis surveillance system in ... 25 Figure 3.1 List of keywords used in the study with various combination of Boolean operators ... 28 Figure 3.2 Conceptual framework of important characteristics associated with multidrug- resistant tuberculosis (MDR-TB) ... 33 Figure 4.1 Variables under study for association patterns in MDR-TB patient characteristics ... 39 Figure 4.2 Study flow chart of socio-demographic, clinical and treatment related characteristics of multidrug-resistant tuberculosis patients based on TBIS registry ... 44 Figure 4.3 Flow chart of log-linear regression ... 53

xiv

LIST OF ABBREVIATIONS

AIC Akaike Information Criterion

CI Confidence interval

DM Diabetic mellitus

DOTS Directly Observed Therapy Short Course

HIV Human immunodeficiency virus

HREC Human Research Ethics Committee

ID Identification card

LTBI Latent tuberculosis infection MDR-TB Multidrug-resistant tuberculosis MOH Ministry of Health

MREC Medical Research and Ethics Committee MTB Mycobacterium Tuberculosis

OR Odd ratio

RR-TB Rifampicin-Resistant tuberculosis

SD Standard deviation

SPSS Statistical Package for the Social Sciences

TB Tuberculosis

TBIS Tuberculosis Information System USM Universiti Sains Malaysia WHO World Health Organization

xv

LIST OF SYMBOLS

n Sample size

Expected frequency

𝜋 Probability of cell ij

𝜋₊ Marginal probabilities of row effect 𝜋+ Marginal probabilities of column effect

X² Pearson chi-square

G² Likelihood ratio

≤ Less than or equal to

> More than

df Degree of freedom

Δ Difference

Standardized Pearson residuals

z z-statistic distribution

xvi

POLA HUBUNGAN ANTARA CIRI-CIRI DEMOGRAFI SOSIAL, KLINIKAL DAN RAWATAN TUBERKULOSIS RINTANG PELBAGAI UBAT DI ANTARA

PESAKIT DI MALAYSIA

ABSTRAK

Pengenalan: Kemunculan dan peningkatan jumlah tuberkulosis rintang pelbagai ubat (MDR- TB) di seluruh dunia telah menimbulkan ancaman kesihatan awam baru di dalam dan di luar negara. Terdapat beberapa ciri-ciri pesakit yang penting untuk MDR-TB. Walau bagaimanapun, tidak ada kajian yang diterbitkan mengenai hubungan di antara ciri-ciri pesakit MDR-TB di Malaysia.

Objektif: Objektif kajian ini adalah untuk menerangkan ciri-ciri berkaitan demografi, klinikal dan rawatan pesakit MDR-TB di Malaysia dan menentukan hubungan di antara faktor-faktor penyebab MDR-TB.

Metodologi: Analisis data sekunder daripada kajian reka bentuk rentas keratan menggunakan Sistem Maklumat Tuberkulosis (TBIS) Kementerian Kesihatan Malaysia. Ciri-ciri berkaitan demografi, klinikal dan rawatan pesakit MDR-TB diekstrak dari pangkalan data. Analisis log- linear digunakan untuk mengenal pasti pola hubungan di antara ciri-ciri pesakit MDR-TB bersama nisbah odds dan 95% selang keyakinan (SK).

Keputusan: Terdapat 395 kes MDR-TB yang dilaporkan di seluruh Malaysia dari tahun 2012 hingga 2016. Odds pendatang adalah 77% lebih rendah bagi yang berusia tua berbanding muda (nisbah odds = 0.23, 95% SK: 0.11, 0.43). Odds jantina lelaki adalah 2.37 kali bagi yang berusia tua berbanding muda (nisbah odds = 2.37, 95% SK: 1.45, 3.95). Odds jantina lelaki adalah 5.22 kali bagi status HIV positif berbanding status HIV negatif (nisbah odds = 5.22, 95% SK: 1.48, 33.12). Odds sejarah rawatan TB adalah 78% lebih tinggi bagi jantina lelaki

xvii

berbanding wanita (nisbah odds = 1.78, 95% SK: 1.13, 2.82) dan 52% lebih rendah bagi status ujian kahak positif berbanding status ujian kahak negatif (nisbah odds = 0.48, 95% SK: 0.30, 0.78)

Kesimpulan: Analisis log-linear menunjukkan bahawa terdapat hubungan di antara kategori umur dan status pendatang, kategori umur dan jantina, jantina dan HIV, jantina dan sejarah rawatan TB, dan sejarah rawatan TB dan status ujian kahak.

Kata kunci : Penyakit rintang pelbagai ubat, tuberkulosis, analisa log-linear, hubungan di antara faktor

xviii

THE ASSOCIATION PATTERNS BETWEEN SOCIO-DEMOGRAPHIC, CLINICAL AND TREATMENT- RELATED CHARACTERISTICS IN MULTIDRUG-RESISTANT TUBERCULOSIS PATIENTS IN MALAYSIA

ABSTRACT

Introduction: The emergence and increasing number of multidrug-resistant tuberculosis (MDR-TB) worldwide have posed new threats to public health locally and globally. There are a number of important patient’s characteristic for MDR-TB. However, there are no studies published about association between the characteristics of MDR-TB patient in Malaysia.

Objective: The objective of this study was to describe the socio-demographic, clinical and treatment related characteristics of MDR-TB patients in Malaysia and to determine the association patterns between the characteristics of MDR-TB patients in Malaysia

Methods: Analysis of secondary data from a cross-sectional design study on registry of Tuberculosis Information System (TBIS) of Ministry of Health Malaysia. Socio-demographic, clinical and treatment related characteristics of MDR-TB patient were extracted from the databases. Log-linear regression was used to identify association patterns between the characteristics with odd ratios and 95% confidence interval.

Results: There were 395 cases of MDR-TB reported across Malaysia from 2012 to 2016. Odds of immigrant were 77% lower in old age than young age (OR = 0.23, 95% CI: 0.11, 0.43).

Odds of male gender were 2.37 times in old age group than young age group. (OR = 2.37, 95%

CI: 1.45, 3.95). Odds of male gender were 5.22 times in HIV positive status group than negative HIV status (OR = 5.22, 95% CI: 1.48, 33.12). Odds of previous TB treatment were 78% higher in male than female (OR = 1.78, 95% CI: 1.13, 2.82) and 52% lower in positive sputum status than negative sputum status (OR = 0.48, 95% CI: 0.30, 0.78).

xix

Conclusion: Log-linear analysis revealed that there are association between age category and immigrant status, age category and gender, gender and HIV, gender and history of previous TB treatment, and history of previous TB treatment and sputum status.

Keywords: Multidrug-resistant, tuberculosis, log-linear analysis, association between characteristics

20

CHAPTER 1

INTRODUCTION

1.1 Tuberculosis

Tuberculosis (TB) has existed since 1800s and remains a major global health problem and burden (Keshavjee and Farmer, 2012; Frieden et al., 2014; World Health Organization, 2016).

More than 10 million new cases reported in 2015 and almost two million deaths due to TB with about 2 to 3 billion people latently infected despite the fact that TB is curable. (World Health Organization, 2016). Figure 1.1 shows 60% of the new cases are contributed by these countries: India with the highest numbers, followed by Indonesia, China, Nigeria, Pakistan and South Africa. In order to meet their first milestone of the End TB Strategy, WHO has set up a target of reduction of 4% to 5% annually worldwide by 2020. However, the rate of decline persisted at 1.5% from 2014 to 2015 with major steps need to be done in TB prevention and care worldwide especially in these countries (World Health Organization, 2016).

Figure 1.1 The estimated TB incidence rate in 2015

21

Tuberculosis is caused by the rod-shaped, non–spore-forming, aerobic bacterium Mycobacterium tuberculosis (MTB) (Knechel, 2009; Keshavjee and Farmer, 2012). It is transmitted via small airborne droplets and usually formed by the action of talking, coughing, sneezing or even singing by an individual with pulmonary or laryngeal tuberculosis (Knechel, 2009; Turner and Bothamley, 2015). The presence of MTB in the lungs leads to respiratory system infection causes pulmonary tuberculosis and the bacterium may also spread to other organs in the body and thus causing extrapulmonary tuberculosis (Knechel, 2009).

Majority of those who inhaled the droplets containing MTB would develop an effective acquired immune response leading to successful inhibition of MTB growth and thus leads to bacteria becoming latent; this is called latent TB infection (LTBI) (Fogel, 2015). They do not transmit the disease to others nor present any symptoms (Cruz-Knight and Blake-Gumbs, 2013). On the other hand, those with active TB have primarily respiratory symptoms such as cough, chest pain and haemoptysis and systemic symptoms such as fever, night sweat and weight loss (Knechel, 2009; Maher, 2009; Zaman, 2010).

For treatment of a new case, an intensive regimen of a total of six months is recommended (Fox et al., 1999). Daily dose of isoniazid, rifampicin, pyrazinamide and streptomycin (or ethambutol if streptomycin is contraindicated) is given for two months and followed by four- month maintenance regimen of daily isoniazid and rifampicin. By giving medications and monitoring the patients directly via Directly Observed Therapy Short Course (DOTS) programme, it can ensure better success rate of eliminating the disease by more than 85%

(Blumberg et al., 2005; World Health Organization, 2016).

1.2 Multidrug-resistant Tuberculosis

However, the emergence and increasing number of multidrug-resistant tuberculosis (MDR- TB) worldwide have posed new threats to public health locally and globally (Dheda et al., 2017). Based on Global Tuberculosis Report in 2016, almost half a million-people developed MDR-TB and in Malaysia, MDR-TB cases are accounted for 3.1% in previously treated TB cases (World Health Organization, 2016).

22

MDR-TB can be defined as MTB with resistance to at least two most powerful first-line anti- TB drugs which are rifampicin and isoniazid (McBryde et al., 2017). The resistance may arise either by transmission of MDR strain by infected person to another, or by insufficient treatment of individual who has been infected with non-resistant strain or only one single drug resistant (Faustini et al., 2006). The inadequate drug treatment will allow the growth of small number of resistant MTB by spontaneous mutation and subsequently leads to MTB being resistance to many drugs (Faustini et al., 2006).

1.3 Epidemiology of MDR-TB 1.3.1 Global Situation

In 2016, WHO estimated incidence cases of MDR-TB to be approximately 490,000. As depicted in Figure 1.2, the South-East Asia, Europe and Western Pacific regions contribute 37%, 23% and 18% respectively of the total global new cases of MDR-TB. These three regions contribute the largest number of MDR-TB incidence globally.

Figure 1.2 Estimated incidence of MDR-TB in 2016 by WHO region(World Health Organization, 2016)

23

With reference to Figure 1.3, the top three countries with largest number of cases are populous countries led by India with 147,000 cases, followed by China with 73,000 cases and Russian Federation with 63,000 cases which contribute to nearly half of global total.

Moreover, according to a report published in The Lancet Infectious Diseases, the number of MDR-TB are predicted to increase in India, the Philippines, Russia and the South Africa, countries which already have a high burden (Sharma et al., 2017). This is estimated from mathematical model derived from data of WHO reports and drug resistance survey from the respective countries.

Figure 1.3 The estimated incidence of MDR/RR-TB in 2016 by WHO, for countries with at least 1000 incident cases(World Health Organization, 2016)

Also, there is a wide gap between detection and the start of MDR-TB treatment where only 22% of the estimated incidence of MDR/RR-TB cases started treatment in 2016. Most significantly, closing this wide gap requires progress in a subset of countries especially China and India because both countries are responsible for 39% of the total gaps (World Health Organization, 2017). In addition, it has been estimated by the same governing body that 4.1%

of new cases and 19% of previously treated cases have MDR/RR-TB (World Health Organization, 2017).

24

As mentioned previously, the treatment success rate is quite low for MDR-TB as compared to TB. This is evident in the case of death from MDR/RR-TB where almost 240,000 deaths have been reported in 2016 (World Health Organization, 2017).

1.3.2 Situation in Malaysia

Figure 1.4 reveals that there has been a gradual increase in the number of MDR-TB cases in Malaysia from 2004 till 2010. Then, a sharp rise on the number of cases seen and peaked in 2011 with 141 cases. The figure also shows that in 2015, 101 cases of MDR-TB were reported.

The MDR-TB burden for Malaysia in term of its incidence rate is estimated to be 1.8 cases per 100 000 population. Moreover, it has been estimated that 1.5% of new cases and 3.1% of previously treated TB cases have MDR-TB. Furthermore, Malaysia has a lower treatment success rate of 32% as compared to global success rate (54%) based on the cohort study conducted in 2014 (World Health Organization, 2017).

Figure 1.4 MDR-TB Cases in Malaysia from 2004 to 2015(Ministry of Health Malaysia, 2016)

1.4 Tuberculosis Information System (TBIS) Registry

TBIS is an electronic record system that ensures notifications and records of each of TB patient are standardised. This allows continuous monitoring in term of treatment success and overall performance of national TB control programme.

Figure 1.5 shows TB surveillance system in Malaysia. All TB cases including MDR-TB cases must be immediately notified to district, state TB organiser team and TB/Leprosy Sector via