E047023.Draft-Revisions.Pdf

BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from

BMJ Open is committed to open peer review. As part of this commitment we make the peer review history of every article we publish publicly available.

When an article is published we post the peer reviewers’ comments and the authors’ responses online. We also post the versions of the paper that were used during peer review. These are the versions that the peer review comments apply to.

The versions of the paper that follow are the versions that were submitted during the peer review process. They are not the versions of record or the final published versions. They should not be cited or distributed as the published version of this manuscript.

BMJ Open is an open access journal and the full, final, typeset and author-corrected version of record of the manuscript is available on our site with no access controls, subscription charges or pay-per-view fees (http://bmjopen.bmj.com).

If you have any questions on BMJ Open’s open peer review process please email [email protected] http://bmjopen.bmj.com/ on September 29, 2021 by guest. Protected copyright. BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from

Policy changes and the screening, diagnosis and treatment of DRTB patients from 2015-2018 in Zhejiang Province, China: a cascade analysis ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2020-047023

Article Type: Original research

Date Submitted by the 18-Nov-2020 Author:

Complete List of Authors: Jiang, Weixi; Duke Kunshan University, Global Health Research Center Peng, Ying; Zhejiang Provincial Center for Disease Control and Prevention Wang, Xiaomeng; Zhejiang Provincial Center for Disease Control and Prevention Elbers, Chris; Vrije Universiteit Amsterdam Tang, Shenglan; Duke University, Duke Global Health Institute Huang, Fei; Centers for Disease Control China, National Center for Tuberculosis Control and Prevention Chen, Bin; Zhejiang Provincial Center for Disease Control and Prevention Cobelens, Frank ; AIGHD, Amsterdam Institute for Global Health and

Development and Department of Global Health http://bmjopen.bmj.com/

Tuberculosis < INFECTIOUS DISEASES, Health policy < HEALTH Keywords: SERVICES ADMINISTRATION & MANAGEMENT, HEALTH SERVICES ADMINISTRATION & MANAGEMENT

on September 29, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 32 BMJ Open

1 2 3

4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 8 9 I, the Submitting Author has the right to grant and does grant on behalf of all authors of the Work (as defined 10 in the below author licence), an exclusive licence and/or a non-exclusive licence for contributions from authors 11 who are: i) UK Crown employees; ii) where BMJ has agreed a CC-BY licence shall apply, and/or iii) in accordance 12 with the terms applicable for US Federal Government officers or employees acting as part of their official 13 duties; on a worldwide, perpetual, irrevocable, royalty-free basis to BMJ Publishing Group Ltd (“BMJ”) its 14 licensees and where the relevant Journal is co-owned by BMJ to the co-owners of the Journal, to publish the 15 Work in this journal and any other BMJ products and to exploit all rights, as set out in our licence. 16 17 The Submitting Author accepts and understands that any supply made under these terms is made by BMJ to 18 the Submitting Author Forunless you peer are acting as review an employee on behalf only of your employer or a postgraduate 19 student of an affiliated institution which is paying any applicable article publishing charge (“APC”) for Open 20 Access articles. Where the Submitting Author wishes to make the Work available on an Open Access basis (and 21 intends to pay the relevant APC), the terms of reuse of such Open Access shall be governed by a Creative 22 Commons licence – details of these licences and which Creative Commons licence will apply to this Work are set 23 out in our licence referred to above. 24 25 Other than as permitted in any relevant BMJ Author’s Self Archiving Policies, I confirm this Work has not been 26 accepted for publication elsewhere, is not being considered for publication elsewhere and does not duplicate 27 material already published. I confirm all authors consent to publication of this Work and authorise the granting 28 of this licence. 29 30 31 32 33 34 35 36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43 44 on September 29, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 32

1 2 3 Policy changes and the screening, diagnosis and treatment of DRTB patients from 2015-2018 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 in Zhejiang Province, China: a cascade analysis 6 7 Corresponding author: 8 Bin Chen 9 10 Postal address: Zhejiang Provincial Center for Disease Control and Prevention, No. 3399 Binsheng 11 Road, Binjiang District, Hangzhou, China 12 Email: [email protected] 13 14 15 List of authors: 16 Weixi Jiang 17 18 Global Health ResearchFor Center, peer Duke Kunshan review University, Kunshan, only Jiangsu, China 19 Email: [email protected] 20 21 22 Ying Peng 23 24 Zhejiang Provincial Center for Disease Control and Prevention, Hangzhou, China 25 Email: [email protected] 26 27 28 Xiaomeng Wang 29 Zhejiang Provincial Center for Disease Control and Prevention, Hangzhou, China 30 Email: [email protected] 31 32 33 Chris Elbers 34 Faculty of Economics and Business Administration, Vrije Universiteit Amsterdam, Amsterdam, the 35 36 Netherlands http://bmjopen.bmj.com/ 37 Email: [email protected] 38 39 40 Shenglan Tang 41 42 Duke Global Health Institute, Duke University, Durham, NC, US 43 Email: [email protected] 44 on September 29, 2021 by guest. Protected copyright. 45 46 Fei Huang 47 48 National Center for Tuberculosis Control and Prevention, China CDC, Beijing, China 49 Email: [email protected] 50 51 52 Bin Chen 53 Zhejiang Provincial Center for Disease Control and Prevention, Hangzhou, China 54 55 Email: [email protected] 56 57 Frank Cobelens 58 59 The Amsterdam Institute for Global Health and Development, Amsterdam, the Netherlands 60 1

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 32 BMJ Open

1 2 3 Email: [email protected] 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 Word count: 3781 7 8 9 10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43 44 on September 29, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 2

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 32

1 2 3 ABSTRACT 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 Introduction 7 Drug-resistant Tuberculosis (DRTB) has become a serious concern globally and in China. Several 8 policies have been issued to improve DRTB case finding and treatment. We conducted a cascade 9 10 analysis on DRTB case finding and care in Zhejiang during 2015-2018 and explored the impacts of 11 these policies. 12 Methods 13 14 A questionnaire was designed to undertake a survey of all prefecture-level CDCs in Zhejiang on 15 DRTB-replated policies/programs active or introduced during 2015-2018. De-identified registration 16 data of presumptive and diagnosed DRTB patients from 2015-2018 in Zhejiang were used to 17 18 construct a seven-stepFor cascade peer of case finding review and care. Mixed-effect only two-level logistic regressions 19 were conducted to explore factors associated with patient registration, treatment initiation and 20 treatment completion. The DRTB policies active per prefecture per year were categorized according 21 to the combinations of policies on each step of the cascade, and included in the model to explore 22 23 their impacts. 24 Results 25 The number of reported presumptive DRTB patients and the percentage with drug susceptibility 26 27 testing (DST) records largely increased during 2015-2018, and the percentage of registered patients 28 that received anti-DRTB treatment also increased from 59.0% to 86.5%. Patients under the policies 29 of equipping GeneXpert plus expanded criteria for DST had a higher likelihood of being registered 30 31 compared with no GeneXpert (adjusted odds ratio, aOR=2.57), while for treatment initiation the 32 association was only when further expanding the registration criteria (aOR=2.38). Patients with 33 registered residence inside Zhejiang were more likely to be registered (aOR=1.96), treated 34 (aOR=3.83) and complete treatment (aOR=1.92) compared with those outside. 35 36 Conclusion http://bmjopen.bmj.com/ 37 The policy changes on DST and registration have effectively improved DRTB case finding and care. 38 Nevertheless, challenges remain in servicing the vulnerable group like the migrants and improving 39 40 equity in the access to TB care. Future policies should provide comprehensive support for migrants 41 to complete treatment at their current place of residence. 42 43 44 Key words: Tuberculosis, Health policy on September 29, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 3

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 32 BMJ Open

1 2 3 Article summary 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 Strengths and limitations of this study 6 Strengths 7  This study conducted a comprehensive and systematic cascade analysis on the healthcare 8 pathway for drug-resistant tuberculosis (DRTB) patients: screening, diagnosis, treatment and 9 10 management, and examined key factors associated with patient attrition at the different stages 11 of care, using the TB registration data. 12  This study evaluated the effectiveness of different combinations of DRTB related policies in 13 14 all prefectures of one Chinese province, including new technology promotion, management 15 and financial support, and found that expanding the eligibility criteria of presumptive DRTB 16 patients referred for drug-sensitivity testing (DST), together with the increased funding to 17 18 guarantee theFor reagents forpeer rapid assays review like GeneXpert, could only effectively improve the capacity 19 of DRTB case finding compared with equipping the DST facilities alone. 20  This study also examined those patients left-behind during the policy changes from an equity 21 perspective, and found that the inclusion of rifampin mono-resistant (RMR) patients in the 22 23 registration and management of DRTB patients has closed the management gap between RMR 24 and multidrug-resistant (MDR) TB patients. Nevertheless, formidable challenges remain in 25 reducing inequity in accessing TB care, especially for migrant workers and the elderly due to 26 27 the lack of appropriate supporting policies for these patient groups to obtain adequate health 28 services. 29 Limitations 30 31  Almost all prefectures had some form of financing policies since 2015 and there were no 32 individual-level data on whether the patient benefited from these financing policies. Therefore, 33 in this study the impacts of these financing policies could not be explored. 34  Other potentially important socio-economic factors like education and income were not 35 36 available in the dataset used, limiting therefore the scope of the equity analysis to migrant http://bmjopen.bmj.com/ 37 workers. 38 39 40 41 42 43 44 on September 29, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 4

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 32

1 2 3 INTRODUCTION 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 Drug-resistant tuberculosis (DRTB) has become a global concern in recent years. It is estimated that 6 7 8 globally in 2018, 3.4% of new TB patients and 18% of previously treated patients developed 9 10 rifampicin-resistant forms of disease (RRTB) including multi-drug resistant TB (MDR-TB), in 11 12 13 which there is additional resistance to isoniazid, and 7.1% and 21% respectively in China.[1] 14 15 16 Currently China accounts for 14% of the RR/MDR-TB disease burden, and one recent projection 17 18 suggested the incidenceFor of peer RR/MDR-TB review would triple without only interventions to change current 19 20 21 conditions.[2] The RR/MDR-TB epidemic has posed a great challenge to achieve the SDG target of 22 23 ending TB in 2030 in China.[3] 24 25 26 The diagnosis and treatment of DRTB can be very costly in terms of both time and money,[4] and 27 28 29 barriers to accessing DRTB diagnosis and care exist worldwide.[5] The diagnosis of DRTB takes 30 31 1-3 months using traditional technology, and the treatment of RR/MDR-TB lasts for up to two years. 32 33 34 DRTB patients generally have lower socio-economic status, and the cost of treatment is so high that 35 36 the current financing policies in the form of health insurance reimbursement and subsidized http://bmjopen.bmj.com/ 37 38 39 treatment are far from sufficient.[6-8] For China specifically, the public health insurance programs 40 41 42 provide very limited coverage for outpatient services that are required for around 20 months for 43 44 RR/MDR-TB patients, and some second-line anti-TB drugs and auxiliary drugs are often not on September 29, 2021 by guest. Protected copyright. 45 46 47 covered.[9-11] Previous studies have also revealed the long treatment delay, the high pre-diagnosis, 48 49 pre-treatment attrition as well as loss-to-follow-up during the treatment course in several countries 50 51 52 including China.[12-17] Moreover, the MDR-TB treatment success rate in China was less than 53 54 55 50%.[18] 56 57 This formidable barriers for DRTB patients to accessing and adhering to standard treatment call for 58 59 60 5

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 32 BMJ Open

1 2 3

4 strong supporting policies for patients to receive and complete treatment.[19] Previous studies have BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 validated the utility of rapid drug susceptibility testing (DST) technologies such as Genechip and 8 9 GeneXpert in the screening for DRTB, and showed that they could improve DRTB case finding, 10 11 12 shorten treatment delays and decrease pre-diagnosis attrition.[13, 20–24] As for treatment adherence, 13 14 one comprehensive program in China that provided universal health coverage to MDR-TB patients 15 16 17 was shown to improve access to and affordability of diagnosis and quality treatment of MDRTB.[5] 18 For peer review only 19 20 Studies on other intervention strategies showed that directly observed therapy (DOT) or other 21 22 reminding approaches through digital technologies could improve the treatment adherence and 23 24 25 outcomes of TB/MDRTB.[25, 26] Patient counseling alone or combined with financial support 26 27 could also increase the cure rates among MDRTB patients.[27] 28 29 30 In China some provinces, along with or after the roll-out of international donor-funded projects, 31 32 33 have implemented policies to improve DRTB control, including allocating special funds to equip 34 35 DRTB-designated hospitals with DST facilities and reagents, improving health insurance benefit 36 http://bmjopen.bmj.com/ 37 38 packages and providing subsidies to patients.[11,28] However, there is limited evidence on the 39 40 impact of implementing these policies on DRTB case finding and care thereafter. The equity issues 41 42 43 underlying the case finding and care procedures for DRTB patients are also understudied, especially 44 on September 29, 2021 by guest. Protected copyright. 45 46 considering that the eligibility for policies issued in a certain region is often linked with patients’ 47 48 registered residence, work and health insurance status. 49 50 51 This study aims to examine changes in the programmatic performance with regard to screening, 52 53 diagnosis, treatment and management of DRTB patients in Zhejiang from 2015 to 2018 through the 54 55 56 cascade analysis approach.[9] Besides, as the policies and guidelines on this whole procedure of 57 58 59 DRTB care changed during this period, this study systematically summarizes these changes and 60 6

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 32

1 2 3

4 investigates how these policy changes have influenced case finding and treatment of DRTB patients. BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 We also explore whether these policy changes have equally influenced patients of different 8 9 demographic and socio-economic characteristics. 10 11 12 13 14 METHODS 15 16 17 Study settings 18 For peer review only 19 th 20 Zhejiang is a province located in the eastern area with its GDP ranking 4 in China,[30] and has a 21 22 growing migrant population in recent years. Under the current TB control model in Zhejiang, the 23 24 25 prefecture-level designated hospital, usually one in each prefecture, is responsible for the diagnosis 26 27 and treatment of DRTB patients, while the Center for Disease Control (CDC) and primary health 28 29 30 care facilities conduct patient management. Patients who are clinically suspected to have DRTB are 31 32 33 to be referred for DST and should be reported in the Tuberculosis Information Management System 34 35 (TBIMS) as presumptive DRTB patients. Their sputum samples are to be sent to the prefecture- 36 http://bmjopen.bmj.com/ 37 38 level designated hospital for DST. If diagnosed with DRTB, these patients should be registered in 39 40 the TBIMS database for diagnosed patients as such. The criteria regarding what types of DRTB 41 42 43 should be registered have being changing over time. 44 on September 29, 2021 by guest. Protected copyright. 45 46 47 48 Study Design 49 50 51 This study included a questionnaire survey on the DRTB policies/programs among the 11 prefecture 52 53 CDCs and a quantitative analysis of the TBIMS records of presumptive and diagnosed DRTB 54 55 56 patients. 57 58 59 Questionnaire survey on DRTB policies/programs 60 7

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 32 BMJ Open

1 2 3

4 A questionnaire on the DRTB related policies was distributed to the CDC of all prefectures in BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 Zhejiang in collaboration with Zhejiang provincial CDC. After preliminary consultation with the 8 9 provincial CDC, the questionnaire was designed to include policies in four areas: 1) eligibility of 10 11 12 presumptive DRTB patients for DST, 2) eligibility for registering DRTB patients based on the drug- 13 14 resistant type (including rifampin mono-resistance (RMR), multi-drug resistance(MDR), extensive 15 16 17 drug-resistance (XDR) and mono-resistance to other types of drugs), 3) DST technology and 18 For peer review only 19 20 payment for DST and 4) financing policies, including both health insurance benefit packages and 21 22 government subsidies for supporting the DRTB treatment. Detailed questions on the eligibility for 23 24 25 financial support regarding the types of DRTB, the registered permanent residence and the region 26 27 of health insurance enrollment were also included in the questionnaire. If the policies changed at 28 29 30 any time after 2015, we collected information on the policy details before and after change as well 31 32 33 as the year of change. 34 35 TBIMS records 36 http://bmjopen.bmj.com/ 37 38 De-identified TBIMS records of presumptive and diagnosed DRTB patients from 2015-2018 in 39 40 Zhejiang were retrieved from the TB Prevention and Control Center of China CDC. The dataset of 41 42 43 presumptive DRTB patients included demographic information (prefecture of registration, age, sex, 44 on September 29, 2021 by guest. Protected copyright. 45 46 ethnicity, occupation and registered residence), drug-resistance test profile (the date of sending the 47 48 sample, conducting the test and reporting the result, the type of test and the test result), as well as a 49 50 51 unique registration number if registered as a DRTB patient in the diagnosed patient dataset. Only 52 53 those DRTB patients recorded in the diagnosed patient dataset have a trackable treatment history 54 55 56 and are managed under a specialized guideline for DRTB patients. The dataset of diagnosed DRTB 57 58 59 patients contained the same demographic information and diagnostic data as the dataset of 60 8

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 32

1 2 3

4 presumptive patients, plus the treatment information including the starting date of anti-DRTB BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 treatment including both second-line (RR/MDR-TB) and adapted first-line (monoresistance to 8 9 isoniazid) treatment, TB treatment history, current state of treatment (under treatment or ended), 10 11 12 and the ending date of treatment as well as the reason of ending the treatment (cured/treatment 13 14 completed/death/lost-to-follow-up/others) if ended. In TBIMS the treatment status referred to 15 16 17 whether the patient received treatment in Zhejiang. If the patients were registered but returned to 18 For peer review only 19 20 their hometown for treatment, it could be shown as “treatment refused”. 21 22 23 24 25 Patient and Public Involvement Statement 26 27 Patients or the public were not involved in the design, or conduct, or reporting, or dissemination 28 29 30 plans of our research. 31 32 33 34 35 Data analysis 36 http://bmjopen.bmj.com/ 37 38 Using datasets of the presumptive and registered diagnosed patients, a seven-step cascade of 39 40 diagnosis and care was constructed for each year from 2015 to 2018 : 1) reported presumptive DRTB 41 42 43 patients; 2) presumptive patients who had DST; 3) diagnosed DRTB, and RR/MDRTB (including 44 on September 29, 2021 by guest. Protected copyright. 45 46 RMR, MDR and XDR) patients; 4) registered RR/MDRTB patients; 5) RR/MDRTB patients that 47 48 ever received anti-DRTB treatment; 6) RR/MDRTB patients that were under treatment 6 months 49 50 51 after treatment initiated; 7) RR/MDRTB patients that had been cured or completed treatment, as 52 53 shown in Figure 1. Diagnosed DRTB patients were defined as patients with a DST result showing 54 55 56 resistance to at least one type of first-line anti-TB drug. As in all but one prefectures only 57 58 59 RR/MDRTB patients are required to be registered and thus had available data on treatment history 60 9

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 32 BMJ Open

1 2 3

4 before 2019, we only analyzed RR/MDRTB patients for step 4-7. Descriptive analyses were BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 conducted to explore the reasons for attrition at each step based on the relevant records in the 8 9 datasets. Frequencies of patient characteristics were also calculated across patient groups at each 10 11 12 step of the cascade to examine potential factors associated with the attrition at different steps. For 13 14 step 3-5 frequencies of some patient characteristics of interests, including age, sex, drug-resistant 15 16 17 type, occupation and registered residence were calculated separately for each year to examine the 18 For peer review only 19 20 trends from 2015 to 2018. 21 22 Mixed-effect two-level logistic regressions, with fixed slope and random intercept specification 23 24 25 were conducted to explore factors associated with the likelihood of being registered for RRTB 26 27 patients, the likelihood of receiving anti-DRTB treatment if registered, and the likelihood of getting 28 29 30 cured or completing treatment (the latter only for those who initiated treatment before 2017 because 31 32 33 of the 2-year treatment duration). This model showed better fit to the data than the fixed-effect 34 35 model based on the likelihood-ratio test. The DRTB related policies were summarized along the 36 http://bmjopen.bmj.com/ 37 38 four areas in the questionnaires. Besides patient-level variables, the policies implemented in each 39 40 prefecture and each year, categorized according to the different combinations of policies in the four 41 42 43 areas, were included in the model. Per-capita GDP of the prefecture were divided into three groups, 44 on September 29, 2021 by guest. Protected copyright. 45 46 as the numbers assembled in three intervals: RMB 55,000-70,000, 80,000-100,000, and above 47 48 120,000, and also included in the model as a categoric variable. The clustered standard errors were 49 50 51 estimated considering the intra-prefecture correlations. 52 53 54 55 RESULTS 56 57 DRTB policy change from 2015-2018 in Zhejiang 58 59 60 10

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 32

1 2 3

4 Results from the questionnaire survey on DRTB policies showed that the major change in the BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 eligibility of patients for DST was the expansion from high-risk patients (including smear positive 8 9 relapse, treatment failure, positive sputum bacteriology after two months’ treatment), to all 10 11 12 bacteriologically positive TB patients. As for the testing technology, the major change was the 13 14 introduction of the rapid DST assay GeneXpert MTB/RIF that combines testing for Mycobacterium 15 16 17 tuberculosis with screening for rifampicin-resistance with a same-day result, while the costs for the 18 For peer review only 19 20 test were all covered, or mostly covered by the government. The criteria for DRTB patient 21 22 registration (implying eligibility for specialized DRTB patient management) also expanded from 23 24 25 MDR/XDR-TB to RR-TB, and one prefecture even expanded to any DRTB before 2018. In most 26 27 prefectures these changes in DST eligibility and equipment as well as the registration criteria 28 29 30 happened in 2017, while almost all others had adopted the new polices and introduced GeneXpert 31 32 33 earlier (supplemental table 1). 34 35 The financing policies to support DRTB treatment in Zhejiang consisted of both health insurance 36 http://bmjopen.bmj.com/ 37 38 and medical assistance policies, and in most prefectures these policies had been implemented in 39 40 2015. These financing policies aimed to improve the benefit package for DRTB patients through a 41 42 43 three-layer coverage system: 1) increasing the reimbursement rate of outpatient treatment to 70%- 44 on September 29, 2021 by guest. Protected copyright. 45 46 90%, the same level as for inpatient treatment; 2) the out-of-pocket (OOP) expenditure exceeding a 47 48 certain amount, typically RMB 20,000-30,000, could be further reimbursed at a rate of 55%-85% 49 50 51 through a supplemental health insurance program for critical illnesses; 3) the OOP expenditure after 52 53 health insurance reimbursement could be covered by the medical assistance, with a payment limit 54 55 56 varying from 11,000 RMB to over 60,000 RMB across prefectures. As for the eligibility for the 57 58 59 insurance policies, those with public health insurance enrolled outside Zhejiang, usually migrants 60 11

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 32 BMJ Open

1 2 3

4 without formal jobs, were excluded. In 5 out of the 11 prefectures RMR patients were still not BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 eligible for the expanded public health insurance benefit package of up to 2019. For the medical 8 9 assistance, in 6 out of the 11 prefectures only patients with registered residence (“Hukou”) in 10 11 12 Zhejiang were eligible, and for the other 4 prefectures with assistance policies the eligibility had 13 14 been expanded to all residents in this prefecture. Up to 2018, RMR patients in 5 prefectures were 15 16 17 still excluded from the medical assistance (supplemental table 2). 18 For peer review only 19 20 21 22 Cascade analysis of DRTB care from 2015-2018 23 24 25 Table 1 shows the screening and diagnosis of DRTB patients from 2015-2018. In general, the results 26 27 suggested increased capacity to find DRTB patients. The number of reported presumptive DRTB 28 29 30 patients increased from 9285 to 23916, with the largest increase from 2016 to 2017 coinciding with 31 32 33 the change in screening policy in most prefectures. The percentage of patients with test results also 34 35 increased from 69.3% to 78.1%. Along with the increase in the volume of DST there was an increase 36 http://bmjopen.bmj.com/ 37 38 in the number of diagnosed DRTB patients except for 2018, and a decrease in the percentage of 39 40 patients diagnosed among those tested. RR/MDR-TB patients accounted for around 40%-50% of 41 42 43 the total DRTB patients. 44 on September 29, 2021 by guest. Protected copyright. 45 46 Further analysis on the types of diagnostic tests patients received revealed an increasing trend in the 47 48 percentage of patients taking rapid DST as well as the percentage of patients tested with both 49 50 51 conventional culture-based and rapid tests. The percentage of presumptive patients without DST 52 53 profile for unknown reasons as recorded in the TBIMS dataset, which indicates pre-diagnosis 54 55 56 attrition, dropped dramatically from 17.3% to 2.7% (supplemental table 3). 57 58 59 60 12

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 32

1 2 3 Table 1: Diagnostic cascade starting from presumptive DRTB patients. 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 Year No. of Patients with test Diagnosed DRTB Diagnosed 6 presumptive records patients RR/MDR-TB 7 DRTB patients 8 9 patients No. % No. % No. % 10 2015 9285 6434 69.3 1031 16.0 503 48.8 11 2016 10997 8438 76.7 1258 14.9 529 42.1 12 2017 21768 14764 67.8 1729 11.7 716 41.4 13 14 2018 23916 18670 78.1 1580 8.5 663 42.0 15 16 17 18 Table 2 shows theFor registration peer and treatment review for diagnosed RRTB only patients from 2015-2018. While 19 20 21 the actual number of registered RRTB patients increased, the percentage of registered RRTB 22 23 24 patients dropped in 2017, and this percentage increased again to 84.6%. The percentage of registered 25 26 patients that received anti-DRTB treatment increased from 59.0% to 86.5%, and the percentage of 27 28 29 those treated who received at least 6 month’s treatment remained above 90%. Around 70% of 30 31 patients diagnosed in 2015 and 2016 completed treatment. Duration of treatment could be longer 32 33 34 than 2 years, as 38 out of the 283 patients starting treatment in 2016 were still shown as under 35 36 http://bmjopen.bmj.com/ 37 treatment at the time we retrieved the data (see supplemental table 4). 38 39 40 41 Table 2: Registration and treatment cascade for diagnosed RRTB patients. 42 Year Diagnosed Registered Received Under treatment 6 Cured or 43 44 RR/MDR- anti-DRTB months after completed on September 29, 2021 by guest. Protected copyright. 45 TB patients treatment treatment initiated treatment 46 No. % No. % No. % No. % 47 48 2015 503 383 76.1 226 59.0 215 95.1 158 73.5 49 2016 529 410 77.5 283 69.0 263 92.9 182 69.2 50 2017 716 502 70.1 360 71.7 334 92.8 / / 51 52 2018 663 561 84.6 485 86.5 462 95.3 / / 53 54 55 56 Factors associated with the registration and treatment of DRTB patients 57 58 59 Table 3 shows the characteristics of patients across each step of the care cascade. Around 70% of 60 13

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 32 BMJ Open

1 2 3

4 the presumptive and diagnosed patients were male, and around 98% of them were Han People. Only BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 around 40% of the patients had formal jobs other than farming or unemployment. While under the 8 9 assumption of perfect equity we would expect the percentage of patients with different 10 11 12 characteristics to remain the same from the diagnosed to the treated group, it decreased from 30.3% 13 14 to 27.0% for older patients, 28.1% to 22.5% for RMR patients, and from 24.4 to 17.0% for patients 15 16 17 with registered residence outside Zhejiang, i.e., migrants. An increase was observed regarding the 18 For peer review only 19 20 percentage of patients ever taking rapid DST. 21 22 23 Table 3: Patient characteristics across each step of the care cascade 24 Presumptive Diagnosed Diagnosed Registered RR/MDR-TB 25 DRTB DRTB RR/MDR- RR/MDR- patients recevied 26 27 patients patients TB patients TB patients treatment 28 N 65966 5598 2411 1859 1357 29 Age (%) 30 31 >=60 41.3 35.9 30.3 29.8 27.0 32 Gender (%) 33 Male 70.6 73.6 72.5 72.7 72.7 34 35 Ethnicity 36 Han (%) 97.7 98.0 97.9 98.0 98.8 http://bmjopen.bmj.com/ 37 Patient type (%) 38 new patient 70.2 65.2 51.8 51.4 48.8 39 40 Drug-resistanct type (%) 41 RMR / 12.1 28.1 20.8 22.5 42 MDR / 30.4 70.6 78.0 76.1 43 44 XDR / 0.5 1.2 1.2 1.4 on September 29, 2021 by guest. Protected copyright. 45 Non-rifampicin resistant / 56.9 / / / 46 Registered residence (%) 47 48 outside Zhejiang 23.4 25.2 24.4 22.3 17.0 49 Job category (%) 50 farmers 49.0 48.0 43.8 45.0 45.5 51 unemployed 12.4 13.0 14.1 14.2 14.0 52 53 other 38.6 39.0 42.1 40.8 40.5 54 Test type (%) 55 fast test / 38.6 46.5 46.1 49.2 56 57

58 59 60 14

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 32

1 2 3

4 Figure 2 shows the changes in the characteristics of diagnosed, registered and treated RR/MDR-TB BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 patients from 2015-2018. The percentage of diagnosed RMR patients that received adequate 8 9 treatment was low in 2015 and 2016, but increased dramatically from 2017, coinciding with the 10 11 12 policy change. The gaps between the proportion of older patients in the diagnosed and the treated 13 14 group also narrowed. Nevertheless, over the four years there remained a 7-10 percentage decrease 15 16 17 in the proportion of patients with registered residence outside Zhejiang from the diagnosed to the 18 For peer review only 19 20 treated patient group, indicating that migrants were still more likely to drop out after diagnosis 21 22 compared with the local residents. Gender and job category were not obviously associated with the 23 24 25 pre-treatment attrition in this univariate analysis. 26 27 To systematically investigate the impact of policies on DRTB case finding and care, we divided the 28 29 30 combination of policies in different areas of DRTB implemented in each prefecture for each year 31 32 33 during 2015-2018 into four categories, considering both policy implications and the number of cases 34 35 in each category. As shown in Figure 3, since 2017 all prefectures had equipped GeneXpert and 36 http://bmjopen.bmj.com/ 37 38 expanded the criteria for DST as well as registration, except for two prefectures which still did not 39 40 register RMR patients. 41 42 43 Table 4 shows the results of mixed-effect two-level logistic regression of the factors associated with 44 on September 29, 2021 by guest. Protected copyright. 45 46 registration among diagnosed RRTB patients, as well as with receiving anti-DRTB treatment among 47 48 those registered. After adjusting for other factors, patients under the policy of providing GeneXpert 49 50 51 together with expanding eligibility for DST (category 3, adjusted odds ratio (aOR) =2.57, 95% CI: 52 53 1.20, 5.51) had a significantly higher likelihood of being registered compared to patients not being 54 55 56 provided GeneXpert, while this association was not significant for providing GeneXpert testing 57 58 59 without expanding the eligibility (category 2). A positive association with receiving anti-DRTB 60 15

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 32 BMJ Open

1 2 3

4 treatment was significant only for the policy combination of providing GeneXpert and expanding BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 eligibility for both DST and registration (category 4, aOR=2.38, 95%CI: 1.19, 4.79). Not 8 9 surprisingly, MDR/XDR-TB patients were much more likely to be registered compared with RMR 10 11 12 patients due to the registration policy in earlier time periods. In addition, patients with registered 13 14 residence inside Zhejiang were more likely to be registered (aOR=1.96, 95% CI 1.52, 2.52) or 15 16 17 treated (aOR=3.83, 95% CI 2.78, 5.28). Older age was associated with lower likelihood both of 18 For peer review only 19 20 being registered (aOR=0.69,95% CI 0.53, 0.90) and of receiving anti-DRTB treatment (aOR=0.40, 21 22 95% CI 0.30, 0.52). 23 24 25 26 27 Tables 4: Factors associated with registration for diagnosed RRTB patients, and with receiving anti- 28 DRTB treatment for registered RRTB patients. 29 Registration (n=2367) Receiving treatment (n=1824) 30 31 OR P>z 95% CI OR P>z 95%CI 32 Age 33 >=60 0.69 0.006 0.53 0.90 0.40 0.000 0.30 0.52 34 35 Gender 36 male 1.08 0.554 0.85 1.37 0.87 0.425 0.62 1.22 http://bmjopen.bmj.com/ 37 DR type 38 MDR/XDR-TB 5.93 0.000 3.10 11.36 0.91 0.789 0.45 1.84 39 40 Patient type 41 new patient 1.22 0.146 0.93 1.59 0.52 0.000 0.37 0.72 42 Treatment history NA 43 44 no treatment ref. on September 29, 2021 by guest. Protected copyright. 45 1st-line drug only 0.58 0.044 0.34 0.99 46 2nd-line drug used 1.11 0.748 0.58 2.12 47 48 Policy category 49 1 ref. ref. 50 2 1.18 0.626 0.61 2.29 0.87 0.776 0.34 2.25 51 3 2.57 0.015 1.20 5.51 1.44 0.257 0.77 2.68 52 53 4 2.08 0.054 0.99 4.37 2.38 0.015 1.19 4.79 54 Registered residence 55 in Zhejiang 1.96 0.000 1.52 2.52 3.83 0.000 2.78 5.28 56 57 Job category 58 farmers ref. ref. 59 unemployed 1.21 0.017 1.04 1.42 0.90 0.651 0.58 1.40 60 16

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 32

1 2 3 other jobs 0.92 0.281 0.80 1.07 0.94 0.784 0.60 1.47 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 Per capita GDP 6 highest group ref. ref. 7 middle group 1.72 0.122 0.87 3.42 2.20 0.025 1.10 4.38 8 lowest group 1.70 0.099 0.90 3.20 1.15 0.643 0.63 2.12 9 10 _cons 0.32 0.002 0.16 0.66 1.85 0.207 0.71 4.78 11 prefecture var(_cons) 0.18 0.09 0.39 0.18 0.03 0.92 12 13 14 15 16 Tables 5 shows the results of the mixed-effect two-level logistic regression analysis of factors 17 18 associated with theFor likelihood peer of completing review treatment for patients only who initiated treatment before 19 20 21 2017. All prefectures had some kind of financing policies in 2015 and 2016, and most prefectures 22 23 had not yet expanded the eligibility of patients for DST or registration. We therefore used the 24 25 26 individual level data on utilization of rapid DST to assess the impact of rapid testing on treatment 27 28 29 completion. After adjusting for other factors, older patients were less likely to complete treatment 30 31 (aOR=0.24, P<0.001), while registered residence inside Zhejiang was associated with higher 32 33 34 probability of completing treatment (aOR=1.92, P<0.05). Utilization of rapid DST was not 35 36 associated with treatment completion. These two factors had significant impacts of the same http://bmjopen.bmj.com/ 37 38 39 direction on the registration, treatment initiation and treatment completion of RR/MDRTB patients. 40 41 42 Table 5: Factors associated with the likelihood of completing treatment for patients that initiated 43 44 treatment before 2017 on September 29, 2021 by guest. Protected copyright. 45 OR P>z 95% Conf. Interval 46 Age 47 48 >=60 0.24 <0.001 0.14 0.42 49 Gender 50 male 0.79 0.442 0.44 1.44 51 52 Drug-resistance type 53 MDR/XDR-TB 0.50 0.086 0.23 1.10 54 Patient type 55 new patient 1.06 0.820 0.64 1.77 56 57 TB treatment history 58 no treatment history ref. 59 first-line drug only 0.74 0.355 0.39 1.40 60 17

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 32 BMJ Open

1 2 3 second-line drug used 1.07 0.847 0.56 2.02 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 Test type 6 fast test 0.70 0.096 0.46 1.06 7 Registered residence 8 in Zhejiang 1.92 0.040 1.03 3.59 9 10 Job category 11 farmers ref. 12 unemployed 0.84 0.635 0.42 1.71 13 14 other jobs 1.44 0.278 0.74 2.79 15 Per capita GDP 16 highest group ref. 17 18 middle group For peer0.94 review0.939 only0.20 4.39 19 lowest group 1.07 0.921 0.30 3.74 20 Year of sending sample 21 2015 ref. 22 23 2016 1.23 0.164 0.92 1.64 24 _cons 5.58 0.001 2.01 15.50 25 prefecture var(_cons) 0.21 0.04 1.10 26 27 28 29 DISCUSSION 30 31 Results from this study clearly revealed the positive impacts of the combined policy change 32 33 34 regarding the DST of presumptive DRTB patients and the registration of diagnosed patients. 35 36 However, inequity challenges remain in terms of servicing vulnerable groups, e.g. migrant workers http://bmjopen.bmj.com/ 37 38 39 and the older population, in the registration, treatment and management of DRTB. 40 41 42 43 44 Effectiveness and challenges of the policy interventions on September 29, 2021 by guest. Protected copyright. 45 46 47 The expansion of the eligibility criteria of presumptive DRTB patients referred for DST, together 48 49 with the increased funding support to equip the facilities and guarantee the reagents for rapid DST 50 51 52 like GeneXpert, have largely improved the capacity for DRTB case finding. The significant effects 53 54 55 of the combined policies compared to providing GeneXpert alone indicates that in order to 56 57 effectively improve case finding we need to not just introduce new technologies, but also support 58 59 60 18

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 32

1 2 3

4 and expand their use. The inclusion of RMR in the registration and management of DRTB patients BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 closed the management gap between RMR and MDR/XDR-TB patients, and also improved the 8 9 treatment rate for RR/MDR-TB. 10 11 12 It is obvious that these reforms in the DRTB policies and regulations have effectively changed the 13 14 practices in DRTB control. Nevertheless, no policies aiming at migrants were issued between 2015 15 16 17 to 2018, and during these four years the degree of inequity between local and migrant patients did 18 For peer review only 19 20 not seem to be mitigated either. The improved insurance benefit package was only available for 21 22 those with public health insurance enrollment in Zhejiang, and in many prefectures receiving 23 24 25 medical assistance still required local registered residence during 2015-2018. Besides, older people 26 27 were less likely to be registered or receive treatment. This may be because they were more likely to 28 29 30 give up treatment due to the high cost and long course of the standard treatment, and health workers 31 32 33 may not register them once they refused to provide information necessary for registration. These 34 35 findings were consistent with previous studies that age and migration for work, as well as health 36 http://bmjopen.bmj.com/ 37 38 system factors such as lack of clear eligibility criteria for DST and limited capacity to provide DST 39 40 were associated with attrition at different stages of the cascade.[14, 16, 31–33] Nevertheless, no 41 42 43 impacts were observed of other factors, such as association between treatment experience and 44 on September 29, 2021 by guest. Protected copyright. 45 46 treatment outcome, and data on other socio-economic factors such as financial difficulties were not 47 48 available in this study. 49 50 51 52 53 Equity challenges regarding migrant populations 54 55 56 The growing number of migrants in China and other parts of the world are posing a challenge to TB 57 58 59 control.[31, 34–36] Though one systematic review found no significant differences in treatment 60 19

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 32 BMJ Open

1 2 3

4 adherence between migrant and long-term resident MDRTB patients,[37] several studies identified BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 barriers for migrants to accessing TB diagnosis and care mainly in terms of knowledge gaps and 8 9 financial difficulty,[38, 39] and these barriers would likely to be more substantial for migrant 10 11 12 MDRTB patients. In our study, only around 40% of the DRTB patients had a formal job other than 13 14 farmers or self-employment. In China people with a formal job will be compulsively enrolled in the 15 16 17 urban employee basic medical insurance at the place of work as required by the law. However, 18 For peer review only 19 20 migrants without a formal job often choose to participate in the health insurance for residents in 21 22 their hometown which requires lower premium compared to the more developed region they 23 24 25 migrated to, and thus cannot benefit from the reimbursement policy in their working place. Besides, 26 27 they often have to pay the full cost of treatment first before they get reimbursed when they return 28 29 30 home. Therefore, it is expected that many of these DRTB patients would choose to go back to their 31 32 33 hometown for treatment, or even refuse treatment. The need for migrant DRTB patients to travel 34 35 long-distances and the substandard or even absent treatment received would increase the risk of 36 http://bmjopen.bmj.com/ 37 38 disease transmission and treatment failure. A study in Shanghai, a well-developed city in eastern 39 40 China, showed that financial incentives were effective for migrant TB patients to complete 41 42 43 treatment.[40] In terms of these findings, policies aimed at encouraging migrant DRTB patients to 44 on September 29, 2021 by guest. Protected copyright. 45 46 be treated and managed at their current place of residence need to be developed, and this is 47 48 particularly urgent for regions with a large number of migrant populations. 49 50 51 One major limitation of the study is that almost all prefectures had some form of financing policy 52 53 since 2015 and there were no individual-level data on whether the patient benefited from these 54 55 56 financing policies. Therefore, in this study the impacts of these financing policies could not be 57 58 59 explored. Furthermore, other potentially important socio-economic factors like education and 60 20

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 32

1 2 3

4 income were not available in our dataset, limiting the scope of the equity analysis. Nevertheless, BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 these limitations would not influence our conclusions that changes in the screening and registration 8 9 of DRTB patients have largely increased the case finding and management capacity of DRTB 10 11 12 patients, while during the period 2015-2018 differences between the local and migrant patients in 13 14 the registration, treatment initiation and treatment outcome remained. Future research should collect 15 16 17 more individual-level data on the implementation of the health insurance and medical assistance 18 For peer review only 19 20 policies to investigate the impacts of these demand-side policies on DRTB treatment more directly, 21 22 and explore what kind of financing policies provided for migrants could increase their likelihood of 23 24 25 initiating and completing anti-DRTB treatment. 26 27 28 DECALRATIONS 29 Ethical approval 30 31 This study is under the overall evaluation study of China-Gates TB Project Phase III and used data 32 collected from this project. The implementation of this project has received the ethical approval 33 from China CDC (No.201626). The protocol of the overall study design and data collection tools 34 was reviewed and approved by the Institutional Review Board of Duke University (IRB Approval 35 36 code: 2017-0768). This paper does not use data involving human participants. http://bmjopen.bmj.com/ 37 38 Competing interests 39 40 All authors declare no competing interests. 41 42 Funding 43 44 The work was supported by the Bill and Melinda Gates Foundation grant number [OPP1149395]. on September 29, 2021 by guest. Protected copyright. 45 46 Contributorship statement 47 48 The study was designed by Weixi Jiang, Shenglan Tang, Chris Elbers and Frank Cobelens. Fei 49 Huang, Ying Peng, Xiaomeng Wang, Bin Chen and Weixi Jiang coordinated and conducted data 50 collection. Weixi Jiang conducted literature review and wrote the manuscript as the first author. 51 52 Shenglan Tang, Chris Elbers, Bin Chen and Frank Cobelens provided suggestions on data analysis 53 framework and data interpretation, and also revised the manuscript. All authors reviewed the draft 54 manuscript and provided comments on the finalization of the manuscript. All authors have read and 55 approved the manuscript in its current state. 56 57 58 Data sharing statement 59 No additional data available. 60 21

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 32 BMJ Open

1 2 3

4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 Acknowledgements 6 This paper is part of the outputs emanating from the program entitled “China National Health and 7 Family Planning Commission and the Gates Foundation TB Project (Phase III)”-a collaboration 8 between the Government of China and the Melinda and Bill Gates Foundation, and implemented by 9 10 the China Center of Disease Control and Prevention (CDC). The authors of the paper also gratefully 11 acknowledge the officers in Zhejiang Provincial CDC and the 11 prefectural CDCs for helping with 12 collecting DRTB related policies in Zhejiang. 13 14 15 REFERENCES 16 1. World Health Organization. Global Tuberculosis Report 2019: P58 17 18 2. Li BY, Shi WP,For Zhou CM,peer et al. Rising review challenge of multidrug-resistant only tuberculosis in China: 19 A predictive study using Markov modeling. Infect Dis Poverty. 2020;9(1):1-8. 20 doi:10.1186/s40249-020-00682-7 21 3. World Health Organization. accessed August, 2020 https://www.who.int/sdg/targets/en/ 22 23 4. Fitzpatrick C, Floyd K. A Systematic Review of the Cost and Cost Effectiveness of Treatment 24 for Multidrug-Resistant Tuberculosis. Pharmacoeconomics. 2012;30(1):63-80. 25 doi:10.2165/11598640-000000000-00000 26 27 5. Li R, Ruan Y, Sun Q, et al. Effect of a comprehensive programme to provide universal access 28 to care for sputum-smear-positive multidrug-resistant tuberculosis in China: A before-and- 29 after study. Lancet Glob Heal. 2015;3(4):e217-e228. doi:10.1016/S2214-109X(15)70021-5 30 31 6. Long Q, Qu Y, Lucas H. Drug-resistant tuberculosis control in China: progress and challenges. 32 Infect Dis Poverty. 2016;5(1):9. doi:10.1186/s40249-016-0103-3 33 7. Hutchison C, Khan MS, Yoong J, Lin X, Coker RJ. Financial barriers and coping strategies: a 34 qualitative study of accessing multidrug-resistant tuberculosis and tuberculosis care in Yunnan, 35 36 China. BMC Public Health. 2017;17(1):1-11. doi:10.1186/s12889-017-4089-y http://bmjopen.bmj.com/ 37 8. Wingfield T, Boccia D, Tovar M, et al. Defining Catastrophic Costs and Comparing Their 38 Importance for Adverse Tuberculosis Outcome with Multi-Drug Resistance: A Prospective 39 40 Cohort Study, Peru. PLoS Med. 2014;11(7). 41 doi:10.1371/journal.pmed.1001675 42 9. Pan H-Q, Bele S, Feng Y, et al. Analysis of the economic burden of diagnosis and treatment 43 44 of tuberculosis patients in rural China. Int J Tuberc Lung Dis. 2013;17(12):1575-1580. on September 29, 2021 by guest. Protected copyright. 45 doi:10.5588/ijtld.13.0144 46 10. Wang Y, McNeil EB, Huang Z, et al. Household financial burden among multidrug-resistant 47 tuberculosis patients in Guizhou province, China: A cross-sectional study. Med. 48 49 2020;99(28):e21023. doi:10.1097/MD.0000000000021023 50 11. Tang S, Wang L, Wang H, Chin DP. Access to and affordability of healthcare for TB patients 51 in China: Issues and challenges. Infect Dis Poverty. 2016;5(1):1-5. doi:10.1186/s40249-016- 52 53 0096-y 54 12. Hemant Deepak Shewade, Dina Nair, Joel S. Klinton, et al. Pre-diagnosis attrition but high 55 pre-treatment attrition among patients with MDR-TB: An operational research from 56 57 Chennai,India. J Epidemiol Glob Health. 2017;7(4):227-233. 58 13. Boyd R, Ford N, Padgen P, Cox H. Time to treatment for rifampicin-resistant tuberculosis: 59 Systematic review and meta-analysis. Int J Tuberc Lung Dis. 2017;21(11):1173-1180. 60 22

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 32

1 2 3 doi:10.5588/ijtld.17.0230 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 14. Xu C, Li R, Shewade HD, et al. Attrition and delays before treatment initiation among patients 6 with MDR-TB in China (2006-13): Magnitude and risk factors. PLoS One. 2019;14(4):1-16. 7 doi:10.1371/journal.pone.0214943 8 15. Hirasen K, Berhanu R, Evans D, et al. High rates of death and loss to follow-up by 12 months 9 10 of rifampicin resistant TB treatment in South Africa. PLoS One. 2018;13(10):1-14. 11 doi:10.1371/journal.pone.0205463 12 16. Shewade D, Kokane AM, Singh AR, et al. High pre-diagnosis attrition among patients with 13 14 presumptive MDR-TB: an operational research from Bhopal district, India. BMC Health Serv 15 Res. 2017;17(1):1-10. doi:10.1186/s12913-017-2191-6 16 17. Oo T, Kyaw KWY, Soe KT, et al. Magnitude and reasons for pre-diagnosis attrition among 17 18 presumptive Formulti-drug peer resistant tuberculosis review patients in only Bago Region, Myanmar: A mixed 19 methods study. Sci Rep. 2019;9(1):1-10. doi:10.1038/s41598-019-43562-3 20 18. Xu C, Pang Y, Li R, et al. Clinical outcome of multidrug-resistant tuberculosis patients 21 receiving standardized second-line treatment regimen in China. J Infect. 2018;76(4):348-353. 22 23 doi:10.1016/j.jinf.2017.12.017 24 19. Abubakar I, Zignol M, Falzon D, et al. Drug-resistant tuberculosis: Time for visionary political 25 leadership. Lancet Infect Dis. 2013;13(6):529-539. doi:10.1016/S1473-3099(13)70030-6 26 27 20. Radin E, Ariana P, Broekel T et al. Analyzing demand-side efficiency in global health: an 28 application to maternal care in Vietnam. Health Policy Plan. 2016;31(9):1281-1290. 29 doi:10.1093/heapol/czw063 30 31 21. Boehme CC, Nicol MP, Nabeta P, et al. Feasibility, diagnostic accuracy, and effectiveness of 32 decentralised use of the Xpert MTB/RIF test for diagnosis of tuberculosis and multidrug 33 resistance: A multicentre implementation study. Lancet. 2011;377(9776):1495-1505. 34 doi:10.1016/S0140-6736(11)60438-8 35 36 22. Christopher PM, Cucunawangsih, Widysanto A. GeneXpert Mycobacterium tuberculosis/ http://bmjopen.bmj.com/ 37 rifampicin assay for molecular epidemiology of rifampicin-resistant Mycobacterium 38 tuberculosis in an urban setting of Banten province, Indonesia. Int J Mycobacteriol 39 40 2019;8:351-8. 41 23. Atashi S, Izadi B, Jalilian S. et al. Evaluation of GeneXpert MTB / RIF for determination of 42 rifampicin resistance among new tuberculosis cases in west and northwest Iran. New Microbes 43 44 New Infect. 2017;19:117-120. doi:10.1016/j.nmni.2017.07.002 on September 29, 2021 by guest. Protected copyright. 45 24. Pang, Y. , Xia, H. , Zhang, Z., et al. Multicenter evaluation of genechip for detection of 46 multidrug-resistant mycobacterium tuberculosis. Journal of Clinical Microbiology, 51(6), 47 1707-1713. 2013;51(6):1707-1713. doi:10.1128/JCM.03436-12 48 49 25. Yin J, Yuan J, Hu Y, Wei X. Association between directly observed therapy and treatment 50 outcomes in multidrug-resistant tuberculosis: A systematic review and meta-analysis. PLoS 51 One. 2016;11(3):1-14. doi:10.1371/journal.pone.0150511 52 53 26. Narges A , Leah J , Cecily M , et al. Adherence interventions and outcomes of tuberculosis 54 treatment: A systematic review and meta-analysis of trials and observational studies. PLoS 55 Med, 2018, 15(7):e1002595. doi:10.1371/journal.pmed.1002595 56 57 27. Baral SC, Aryal Y, Bhattrai R, et al. The importance of providing counselling and financial 58 support to patients receiving treatment for multi-drug resistant TB: Mixed method qualitative 59 and pilot intervention studies. BMC Public Health. 2014;14(1). doi:10.1186/1471-2458-14-46 60 23

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 32 BMJ Open

1 2 3 28. Wang L, Li R, Xu C, et al. The Global Fund in China: Multidrug-resistant tuberculosis 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 nationwide programmatic scale-up and challenges to transition to full country ownership. PLoS 6 One. 2017;12(6):1-13. doi:10.1371/journal.pone.0177536 7 29. Subbaraman R, Nathavitharana RR, Mayer KH, et al. Constructing care cascades for active 8 tuberculosis: A strategy for program monitoring and identifying gaps in quality of care. PLoS 9 10 Med. 2019;16(2):1-18. doi:10.1371/journal.pmed.1002754 11 30. China National Bureau of Statistics. Accessed August, 2020 12 http://www.stats.gov.cn/tjsj/ndsj/2019/indexch.htm 13 14 31. Truzyan N, Crape B, Grigoryan R, et al. Increased risk for multidrug-resistant tuberculosis in 15 migratory workers, Armenia. Emerg Infect Dis. 2015;21(3):474-476. 16 doi:10.3201/eid2103.140474 17 18 32. Parmar MM,For Sachdeva peerKS, Dewan PK, review et al. Unacceptable only treatment outcomes and associated 19 factors among India’s initial cohorts of multidrug-resistant tuberculosis (MDR-TB) patients 20 under the revised national TB control programme (2007–2011): Evidence leading to policy 21 enhancement. PLoS One. 2018;13(4):1-24. doi:10.1371/journal.pone.0193903 22 23 33. Lalor MK, Greig J, Allamuratova S, et al. Risk factors associated with default from multi- and 24 extensively drug-resistant tuberculosis treatment, Uzbekistan: A retrospective cohort analysis. 25 PLoS One. 2013;8(11). doi:10.1371/journal.pone.0078364 26 27 34. Li X, Yang Q, Feng B, et al. Tuberculosis infection in rural labor migrants in Shenzhen, China: 28 Emerging challenge to tuberculosis control during urbanization. Sci Rep. 2017;7(1):1-8. 29 doi:10.1038/s41598-017-04788-1 30 31 35. Zumla A, Abubakar I. Improving access to multi-drug resistant tuberculosis diagnostic and 32 health services for refugees and migrants. BMC Med. 2018;16(1):7-10. doi:10.1186/s12916- 33 018-1218-0 34 36. Wang W, Wang J, Zhao Q, et al. Contribution of rural-to-urban migration in the prevalence of 35 36 drug resistant tuberculosis in China. Eur J Clin Microbiol Infect Dis. 2011;30(4):581-586. http://bmjopen.bmj.com/ 37 doi:10.1007/s10096-010-1125-6 38 37. Nellums LB, Rustage K, Hargreaves S, et a;. Multidrug-resistant tuberculosis treatment 39 40 adherence in migrants: A systematic review and meta-analysis. BMC Med. 2018;16(1):1-11. 41 doi:10.1186/s12916-017-1001-7 42 38. Long Q, Li Y, Wang Y, et al. Barriers to accessing TB diagnosis for rural-to-urban migrants 43 44 with chronic cough in Chongqing, China: A mixed methods study. BMC Health Serv Res. on September 29, 2021 by guest. Protected copyright. 45 2008;8:1-10. doi:10.1186/1472-6963-8-202 46 39. Wei X, Chen J, Chen P, et al. Barriers to TB care for rural-to-urban migrant TB patients in 47 Shanghai: A qualitative study. Trop Med Int Heal. 2009;14(7):754-760. doi:10.1111/j.1365- 48 49 3156.2009.02286.x 50 40. Wei X, Zou G, Yin J, et al. Providing financial incentives to rural-to-urban tuberculosis 51 migrants in Shanghai: An intervention study. Infect Dis Poverty. 2012;1(1):1-8. 52 53 doi:10.1186/2049-9957-1-9 54 55 56 57 FIGURES AND LEGENDS 58 Figure 1: Seven steps of the DRTB diagnosis and care cascade 59 Figure 2: Characteristics of diagnosed, registered and treated RR/MDR-TB patients, 2015-2018 60 24

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 32

1 2 3 Figure 3: Combination of DRTB policies implemented in each prefecture from 2015-2018 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43 44 on September 29, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 25

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 210x71mm (96 x 96 DPI) 21 22 23 24 25 26 27 28 29 30 31

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

40 on September 29, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 32 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

40 on September 29, 2021 by guest. Protected copyright. 41 42 206x217mm (96 x 96 DPI) 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 29 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 207x102mm (96 x 96 DPI) 25 26 27 28 29 30 31

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

1 2 3 SUPPLEMENTAL MATERIALS 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 Table 1: Year of policy change regarding the screening test and registration in the 11 prefectures 7 Prefecture No. Year of policy change 8 9 screening range test technology registration policy 10 1 2016 2015 2017 11 2 2017 2014 2014 12 3 2017 2017 2017 13 14 4 2017 2017 2017 15 5 2017 2017 2017 16 6 2009 2016 2019 17 18 7 For2017 peer review2017 only2017 19 8 2017 2015 2019 20 9 2017 2017 2017 21 22 10 2017 2017 2015 23 11 2000 2012 2012 24 25 26 Table 2: Starting time and eligibility for health insurance and medical assistance policy coverage, 27 2015-2018 28 Prefecture Health insurance policy Medical assistance 29 No. starting insurance RMR starting registered RMR 30 31 time outside ZJ time residence 32 outside ZJ 33 1 2018 N Y 2013 N N 34 35 2 2013 N Y 2013 Y Y since 2015 36 3 2016 N N 2014 Y N http://bmjopen.bmj.com/ 37 4 2017 N Y 2015 Y N 38 39 5 2016 N N 2015 N N 40 6 2019 N N 2012 N Y 41 7 2014 N N 2014 Y N 42 43 8 2013 N N 2015 N Y 44 9 2014 N Y since 2017 No policy on September 29, 2021 by guest. Protected copyright. 45 10 2013 N Y 2015 N Y 46 11 2011 N Y 2015 Y Y 47 48 49 Table 3: The diagnostic test for presumptive DRTB patients from 2015-2018 50 Year No. of % took % took % took Patients with no DST 51 52 presumptive fast traditional both records (%) 53 patients DST DST tests total not TB unknown 54 reported reason 55 56 2015 9285 18.2 51.3 0.2 30.7 13.4 17.3 57 2016 10997 30.4 46.9 0.5 23.3 11.6 11.7 58 2017 21768 30.9 42.8 5.8 32.2 25.6 6.6 59 60 2018 23916 64.4 33.8 20.1 21.9 19.3 2.7

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 31 of 32 BMJ Open

1 2 3 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 Table 4: Reasons for not completing anti-DRTB treatment according to the registration records 6 Year Total Under Dead lost-to-follow- Other 7 treatment up reasons 8 9 No. % No. % No. % No. % 10 2015 68 11 16.2 17 25.0 13 19.1 27 39.7 11 2016 101 38 37.6 15 14.9 15 14.9 33 32.7 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43 44 on September 29, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 32 of 32

1 2 STROBE Statement—Checklist of items that should be included in reports of cross-sectional studies 3

4 Item Page BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 No Recommendation No 6 Title and abstract 1 (a) Indicate the study’s design with a commonly used term in the title or 1 7 the abstract 8 (b) Provide in the abstract an informative and balanced summary of what 3 9 10 was done and what was found 11 Introduction 12 Background/rationale 2 Explain the scientific background and rationale for the investigation 5 13 14 being reported 15 Objectives 3 State specific objectives, including any prespecified hypotheses 6 16 Methods 17 18 Study design For4 Present peer key elements review of study design earlyonly in the paper 7 19 Setting 5 Describe the setting, locations, and relevant dates, including periods of 7,8 20 recruitment, exposure, follow-up, and data collection 21 22 Participants 6 (a) Give the eligibility criteria, and the sources and methods of selection 8 23 of participants 24 Variables 7 Clearly define all outcomes, exposures, predictors, potential confounders, NA 25 and effect modifiers. Give diagnostic criteria, if applicable 26 27 Data sources/ 8* For each variable of interest, give sources of data and details of methods 8 28 measurement of assessment (measurement). Describe comparability of assessment 29 methods if there is more than one group 30 Bias 9 Describe any efforts to address potential sources of bias NA 31 32 Study size 10 Explain how the study size was arrived at 8 33 Quantitative variables 11 Explain how quantitative variables were handled in the analyses. If 10 34 applicable, describe which groupings were chosen and why 35 Statistical methods 12 (a) Describe all statistical methods, including those used to control for 10

36 http://bmjopen.bmj.com/ 37 confounding 38 (b) Describe any methods used to examine subgroups and interactions 10 39 (c) Explain how missing data were addressed NA 40 41 (d) If applicable, describe analytical methods taking account of sampling NA 42 strategy 43 (e) Describe any sensitivity analyses NA 44 on September 29, 2021 by guest. Protected copyright. 45 Results 46 Participants 13* (a) Report numbers of individuals at each stage of study—eg numbers 12,13 47 potentially eligible, examined for eligibility, confirmed eligible, included 48 in the study, completing follow-up, and analysed 49 50 (b) Give reasons for non-participation at each stage NA 51 (c) Consider use of a flow diagram NA 52 Descriptive data 14* (a) Give characteristics of study participants (eg demographic, clinical, 14 53 54 social) and information on exposures and potential confounders 55 (b) Indicate number of participants with missing data for each variable of NA 56 interest 57 Outcome data 15* Report numbers of outcome events or summary measures 12,13 58 59 Main results 16 (a) Give unadjusted estimates and, if applicable, confounder-adjusted 15,16, 60 estimates and their precision (eg, 95% confidence interval). Make clear 17,18 which confounders were adjusted for and why they were included

1 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 33 of 32 BMJ Open

1 2 (b) Report category boundaries when continuous variables were 10 3 categorized 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 (c) If relevant, consider translating estimates of relative risk into absolute NA 6 risk for a meaningful time period 7 Other analyses 17 Report other analyses done—eg analyses of subgroups and interactions, NA 8 and sensitivity analyses 9 10 Discussion 11 Key results 18 Summarise key results with reference to study objectives 18 12 13 Limitations 19 Discuss limitations of the study, taking into account sources of potential 20 14 bias or imprecision. Discuss both direction and magnitude of any 15 potential bias 16 Interpretation 20 Give a cautious overall interpretation of results considering objectives, 18,19 17 18 Forlimitations, peer multiplicity review of analyses, resultsonly from similar studies, and 20 19 other relevant evidence 20 Generalisability 21 Discuss the generalisability (external validity) of the study results 20 21 22 Other information 23 Funding 22 Give the source of funding and the role of the funders for the present 21 24 study and, if applicable, for the original study on which the present article 25 is based 26 27 28 *Give information separately for exposed and unexposed groups. 29 30 Note: An Explanation and Elaboration article discusses each checklist item and gives methodological background and 31 32 published examples of transparent reporting. The STROBE checklist is best used in conjunction with this article (freely 33 available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of Internal Medicine at 34 http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the STROBE Initiative is 35 available at www.strobe-statement.org.

36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43 44 on September 29, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

2 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from

Policy changes and the screening, diagnosis and treatment of drug-resistant tuberculosis patients from 2015-2018 in Zhejiang Province, China: a retrospective cohort study ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2020-047023.R1

Article Type: Original research

Date Submitted by the 02-Mar-2021 Author:

Complete List of Authors: Jiang, Weixi; Duke Kunshan University, Global Health Research Center Peng, Ying; Zhejiang Provincial Center for Disease Control and Prevention Wang, Xiaomeng; Zhejiang Provincial Center for Disease Control and Prevention Elbers, Chris; Vrije Universiteit Amsterdam, School of Business and Economics, Vrije Universiteit Amsterdam Tang, Shenglan; Duke University, Duke Global Health Institute Huang, Fei; Centers for Disease Control China, National Center for Tuberculosis Control and Prevention Chen, Bin; Zhejiang Provincial Center for Disease Control and Prevention

Cobelens, Frank ; Amsterdam University Medical Centres, Department of http://bmjopen.bmj.com/ Global Health and Amsterdam Institute for Global Health and Development

Primary Subject Health policy Heading:

Secondary Subject Heading: Public health

Tuberculosis < INFECTIOUS DISEASES, Health policy < HEALTH

Keywords: SERVICES ADMINISTRATION & MANAGEMENT, HEALTH SERVICES on September 29, 2021 by guest. Protected copyright. ADMINISTRATION & MANAGEMENT

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 34 BMJ Open

1 2 3

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 34

1 2 3 Policy changes and the screening, diagnosis and treatment of drug-resistant tuberculosis 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 patients from 2015-2018 in Zhejiang Province, China: a retrospective cohort study 6 7 Corresponding author: 8 Bin Chen 9 10 Postal address: Zhejiang Provincial Center for Disease Control and Prevention, No. 3399 Binsheng 11 Road, Binjiang District, Hangzhou, China 12 Email: [email protected] 13 14 15 List of authors: 16 Weixi Jiang 17 18 Global Health ResearchFor Center, peer Duke Kunshan review University, Kunshan, only Jiangsu, China 19 Email: [email protected] 20 21 22 Ying Peng 23 24 Zhejiang Provincial Center for Disease Control and Prevention, Hangzhou, China 25 Email: [email protected] 26 27 28 Xiaomeng Wang 29 Zhejiang Provincial Center for Disease Control and Prevention, Hangzhou, China 30 Email: [email protected] 31 32 33 Chris Elbers 34 School of Business and Economics, Vrije Universiteit Amsterdam, Amsterdam, the Netherlands 35 36 Email: [email protected] http://bmjopen.bmj.com/ 37 38 39 Shenglan Tang 40 Duke Global Health Institute, Duke University, Durham, NC, US 41 42 Email: [email protected] 43 44 on September 29, 2021 by guest. Protected copyright. 45 Fei Huang 46 National Center for Tuberculosis Control and Prevention, China CDC, Beijing, China 47 48 Email: [email protected] 49 50 Bin Chen 51 52 Zhejiang Provincial Center for Disease Control and Prevention, Hangzhou, China 53 Email: [email protected] 54 55 56 Frank Cobelens 57 Department of Global Health and Amsterdam Institute for Global Health and Development, 58 59 Amsterdam University Medical Centers, Amsterdam, the Netherlands 60 1

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 34 BMJ Open

1 2 3 Email: [email protected] 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 Word count: 3823 7 8 9 10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43 44 on September 29, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 2

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 34

1 2 3 ABSTRACT 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 Objectives To examine changes in the screening, diagnosis, treatment and management of drug- 6 resistant tuberculosis (DRTB) patients, and investigate the impacts of DRTB related policies on 7 patients of different demographic and socio-economic characteristics. 8 Design A retrospective cohort study using registry data, plus a survey on DRTB-related policies. 9 10 Setting All prefecture-level Centers for Disease Control in Zhejiang Province, China 11 Main outcome measures Alongside the care cascade we examined: 1) reported number of 12 presumptive DRTB patients; 2) percentage of presumptive patients with drug susceptibility testing 13 14 (DST) records; 3) percentage of DRTB/rifampicin-resistant (RRTB) patients registered; 4) 15 percentage of RR/MDRTB patients that received anti-DRTB treatment; 5) percentage of 16 RR/MDRTB patients cured/completed treatment among those treated. Multivariate logistic 17 18 regressions were conductedFor topeer explore the impactsreview of DRTB policies only after adjusting for other factors. 19 Results The number of reported presumptive DRTB patients and the percentage with DST records 20 largely increased during 2015-2018, and the percentage of registered patients who received anti- 21 DRTB treatment also increased from 59.0% to 86.5%. Patients under the policies of equipping 22 23 GeneXpert plus expanded criteria for DST had a higher likelihood of being registered compared 24 with no GeneXpert (adjusted odds ratio, aOR=2.57, 95% CI: 1.20, 5.51), while for treatment 25 initiation the association was only when further expanding the registration criteria (aOR=2.38, 95% 26 27 CI: 1.19, 4.79). Patients with registered residence inside Zhejiang were more likely to be registered 28 (aOR=1.96, 95% CI 1.52, 2.52), treated (aOR=3.83, 95% CI 2.78, 5.28) and complete treatment 29 (aOR=1.92, 95% CI 1.03, 3.59) compared with those outside. 30 31 Conclusion The policy changes on DST and registration have effectively improved DRTB case 32 finding and care. Nevertheless, challenges remain in servicing vulnerable groups such as migrants 33 and improving equity in the access to TB care. Future policies should provide comprehensive 34 support for migrants to complete treatment at their current place of residence. 35 36 http://bmjopen.bmj.com/ 37 Key words: Tuberculosis, Health policy 38 39 40 41 42 43 44 on September 29, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 3

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 34 BMJ Open

1 2 3 Article summary 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 Strengths and limitations of this study 6 Strengths 7  This study conducted a comprehensive and systematic cascade analysis on the healthcare 8 pathway for drug-resistant tuberculosis (DRTB) patients: screening, diagnosis, treatment and 9 10 management, and examined key factors associated with the case finding and healthcare process. 11  This study evaluated the effectiveness of different combinations of DRTB related policies as 12 they were implemented in all prefectures of one Chinese province, including new diagnostic 13 14 technology promotion, improved patient management and financial support. 15  This study also examined those patients left-behind during the policy changes from an equity 16 perspective, with a focus on rifampin mono-resistant (RMR) patients and patients with 17 18 registered residenceFor (Hukou) peer outside Zhejiang.review only 19 Limitations 20  The impacts of the financing policies could not be explored as almost all prefectures had some 21 form of financing policies since 2015 and there were no individual-level data on whether the 22 23 patient benefited from these financing policies. 24  Other potentially important socio-economic factors like education and income were not 25 available in the dataset used, limiting therefore the scope of the equity analysis to migrant 26 27 workers. 28 29 30 31 32 33 34 35 36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43 44 on September 29, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 4

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 34

1 2 3 INTRODUCTION 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 Drug-resistant tuberculosis (DRTB) has become a global concern in recent years. It is estimated that 6 7 8 globally in 2018, 3.4% of new TB patients and 18% of previously treated patients developed 9 10 rifampicin-resistant forms of disease (RRTB) including multi-drug resistant TB (MDR-TB), in 11 12 13 which there is additional resistance to isoniazid, and 7.1% and 21% respectively in China.[1] 14 15 16 Currently China accounts for 14% of the RR/MDR-TB disease burden, and one recent projection 17 18 suggested the incidenceFor of peer RR/MDR-TB review would triple without only interventions to change current 19 20 21 conditions.[2] The RR/MDR-TB epidemic has posed a great challenge to achieve the SDG target of 22 23 ending TB in 2030 in China.[3] 24 25 26 The diagnosis and treatment of DRTB can be very costly in terms of both time and money,[4] and 27 28 29 barriers to accessing DRTB diagnosis and care exist worldwide.[5] The diagnosis of DRTB takes 30 31 1-3 months using traditional technology, and the treatment of RR/MDR-TB lasts for up to two years. 32 33 34 DRTB patients generally have lower socio-economic status, and the cost of treatment is so high that 35 36 the current financing policies in the form of health insurance reimbursement and subsidized http://bmjopen.bmj.com/ 37 38 39 treatment are far from sufficient.[6-8] For China specifically, the public health insurance programs 40 41 42 provide very limited coverage for outpatient services that are required for around 20 months for 43 44 RR/MDR-TB patients. In addition, some second-line anti-TB drugs and auxiliary drugs are often on September 29, 2021 by guest. Protected copyright. 45 46 47 not covered.[9-11] Previous studies in several countries including China have also revealed long 48 49 treatment delay, high pre-diagnosis and pre-treatment attrition as and high loss-to-follow-up during 50 51 52 the treatment course.[12-17] Moreover, the MDR-TB treatment success rate in China was less than 53 54 55 50%.[18] 56 57 These formidable barriers for DRTB patients to accessing and adhering to standard treatment call 58 59 60 5

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 34 BMJ Open

1 2 3

4 for strong supporting policies for patients to receive and complete treatment.[19] Previous studies BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 have validated the utility of rapid drug susceptibility testing (DST) technologies such as Genechip 8 9 and GeneXpert in the screening for DRTB, and shown that these technologies could improve DRTB 10 11 12 case finding, shorten treatment delays and decrease pre-diagnosis attrition.[13, 20–24] As for 13 14 treatment adherence, one comprehensive program in China that provided universal health coverage 15 16 17 to MDR-TB patients was shown to improve access to and affordability of diagnosis and quality 18 For peer review only 19 20 treatment of MDRTB.[5] Studies on other intervention strategies showed that directly observed 21 22 therapy (DOT) or other reminding approaches through digital technologies could improve the 23 24 25 treatment adherence and outcomes of TB/MDRTB.[25, 26] Patient counseling alone or combined 26 27 with financial support can also increase the cure rates among MDRTB patients.[27] 28 29 30 In China some provinces, along with or after the roll-out of international donor-funded projects, 31 32 33 have implemented policies to improve DRTB control, including allocating special funds to equip 34 35 DRTB-designated hospitals with DST facilities and reagents, improving health insurance benefit 36 http://bmjopen.bmj.com/ 37 38 packages and providing subsidies to patients.[11,28] However, there is limited evidence on the 39 40 impact of implementing these policies on DRTB case finding and care thereafter. The equity issues 41 42 43 underlying the case finding and care procedures for DRTB patients are also understudied, especially 44 on September 29, 2021 by guest. Protected copyright. 45 46 considering that the eligibility for policies issued in a certain region is often linked with patients’ 47 48 registered residence, work and health insurance status. 49 50 51 This study aims to examine changes in the programmatic performance with regard to screening, 52 53 diagnosis, treatment and management of DRTB patients in Zhejiang from 2015 to 2018 through a 54 55 56 cascade analysis approach.[29] In addition, as the policies and guidelines on this whole procedure 57 58 59 of DRTB care changed during this period, this study systematically summarizes these changes and 60 6

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 34

1 2 3

4 investigates how these policy changes have influenced case finding and treatment of DRTB patients. BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 We also explore whether these policy changes have equally influenced patients of different 8 9 demographic and socio-economic characteristics. 10 11 12 13 14 METHODS 15 16 17 Study settings 18 For peer review only 19 th 20 Zhejiang is a province located in the eastern area with its GDP ranking 4 in China,[30] and has a 21 22 growing migrant population in recent years. Under the current TB control model in Zhejiang, the 23 24 25 prefecture-level designated hospital, usually one in each prefecture, is responsible for the diagnosis 26 27 and treatment of DRTB patients, while the Center for Disease Control (CDC) and primary health 28 29 30 care facilities conduct patient management. Patients who are clinically suspected to have DRTB are 31 32 33 to be referred for DST and should be reported in the Tuberculosis Information Management System 34 35 (TBIMS) as presumptive DRTB patients. Their sputum samples are to be sent to the prefecture- 36 http://bmjopen.bmj.com/ 37 38 level designated hospital for DST. If diagnosed with DRTB, these patients should be registered in 39 40 the TBIMS database for diagnosed patients as such. The criteria regarding what types of DRTB 41 42 43 should be registered have being changing over time. 44 on September 29, 2021 by guest. Protected copyright. 45 46 47 48 Study Design 49 50 51 This study included a questionnaire survey on the DRTB policies/programs among the 11 52 53 prefecture-level CDCs and a quantitative analysis of the TBIMS records of presumptive and 54 55 56 diagnosed DRTB patients. 57 58 59 Questionnaire survey on DRTB policies/programs 60 7

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 34 BMJ Open

1 2 3

4 A questionnaire on the DRTB related policies was distributed to the CDC of all prefectures in BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 Zhejiang in collaboration with Zhejiang provincial CDC. After preliminary consultation with the 8 9 provincial CDC, the questionnaire was designed to include policies in four areas: 1) eligibility of 10 11 12 presumptive DRTB patients for DST; 2) eligibility for registering DRTB patients based on the type 13 14 of drug-resistance (including rifampin mono-resistance (RMR), multi-drug resistance(MDR), 15 16 17 extensive drug-resistance (XDR) and mono-resistance to other types of drugs); 3) DST technology 18 For peer review only 19 20 and payment for DST and 4) financing policies, including both health insurance benefit packages 21 22 and government subsidies for supporting the DRTB treatment. Detailed questions on the eligibility 23 24 25 for financial support regarding the types of DRTB, the registered permanent residence and the 26 27 region of health insurance enrollment were also included in the questionnaire (see supplementary 28 29 30 materials). If the policies had changed at any time after 2015, we collected information on the policy 31 32 33 details before and after the change as well as the year of change. 34 35 TBIMS records 36 http://bmjopen.bmj.com/ 37 38 De-identified TBIMS records of presumptive and diagnosed DRTB patients from 2015-2018 in 39 40 Zhejiang were retrieved from the TB Prevention and Control Center of China CDC. The dataset of 41 42 43 presumptive DRTB patients included demographic information (prefecture of registration, age, sex, 44 on September 29, 2021 by guest. Protected copyright. 45 46 ethnicity, occupation and registered residence), drug-resistance test profile (the date of sending the 47 48 sample, conducting the test and reporting the result, the type of test and the test result), as well as a 49 50 51 unique registration number if registered as a DRTB patient in the diagnosed patient dataset. Only 52 53 those DRTB patients recorded in the diagnosed patient dataset have a traceable treatment history 54 55 56 and are managed under a specialized guideline for DRTB patients. The dataset of diagnosed DRTB 57 58 59 patients contained the same demographic information and diagnostic data as the dataset of 60 8

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 34

1 2 3

4 presumptive patients, plus the treatment information including the starting date of anti-DRTB BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 treatment including both second-line (RR/MDR-TB) and adapted first-line (monoresistance to 8 9 isoniazid) treatment, TB treatment history, current state of treatment (under treatment or ended), 10 11 12 and the ending date of treatment as well as the reason of ending the treatment (cured/treatment 13 14 completed/death/lost-to-follow-up/others) if ended. In TBIMS the treatment status referred to 15 16 17 whether the patient received treatment in Zhejiang. If patients were registered but returned to their 18 For peer review only 19 20 hometown for treatment, it could be shown as “treatment refused”. 21 22 23 24 25 Patient and Public Involvement Statement 26 27 Patients or the public were not involved in the design, or conduct, or reporting, or dissemination 28 29 30 plans of our research. 31 32 33 34 35 Data analysis 36 http://bmjopen.bmj.com/ 37 38 Using datasets of the presumptive and registered diagnosed patients, a seven-step cascade of 39 40 diagnosis and care was constructed for each year from 2015 to 2018 : 1) reported presumptive DRTB 41 42 43 patients; 2) presumptive patients who had DST; 3) diagnosed DRTB, and RR/MDRTB (including 44 on September 29, 2021 by guest. Protected copyright. 45 46 RMR, MDR and XDR) patients; 4) registered RR/MDRTB patients; 5) RR/MDRTB patients that 47 48 ever received anti-DRTB treatment; 6) RR/MDRTB patients that were under treatment 6 months 49 50 51 after treatment initiated; 7) RR/MDRTB patients that had been cured or completed treatment, as 52 53 shown in Figure 1. Diagnosed DRTB patients were defined as patients with a DST result showing 54 55 56 resistance to at least one type of first-line anti-TB drug. As in all but one prefectures only 57 58 59 RR/MDRTB patients were required to be registered and thus had available data on treatment history 60 9

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 34 BMJ Open

1 2 3

4 before 2019, we only analyzed RR/MDRTB patients for step 4-7. Descriptive analyses were BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 conducted to explore the reasons for attrition at each step based on the relevant records in the 8 9 datasets. Patients’ ethnicity was categorized as Han and other minority groups. Frequencies of 10 11 12 patient characteristics were also calculated across patient groups at each step of the cascade to 13 14 examine potential factors associated with the attrition. For step 3-5 the frequencies of some patient 15 16 17 characteristics of interests, including age, sex, drug-resistant type, occupation and registered 18 For peer review only 19 20 residence were calculated separately for each year to examine the trends from 2015 to 2018. 21 22 Mixed-effect two-level logistic regressions, with fixed slope and random intercept specification 23 24 25 were conducted to explore factors associated with the likelihood of being registered for RRTB 26 27 patients, the likelihood of receiving anti-DRTB treatment if registered, and the likelihood of getting 28 29 30 cured or completing treatment (the latter only for those who initiated treatment before 2017 because 31 32 33 of the 2-year treatment duration). This model showed better fit to the data than the fixed-effect 34 35 model based on the likelihood-ratio test. The DRTB related policies were summarized along the 36 http://bmjopen.bmj.com/ 37 38 four areas in the questionnaires. Besides patient-level variables, the policies implemented in each 39 40 prefecture and each year, categorized according to the different combinations of policies in the four 41 42 43 areas, were included in the model. Per-capita GDP of the prefecture was divided into three groups, 44 on September 29, 2021 by guest. Protected copyright. 45 46 as the numbers assembled in three intervals: RMB 55,000-70,000, 80,000-100,000, and above 47 48 120,000, and also included in the model as a categoric variable. Standard errors were estimated 49 50 51 accounting for clustering, considering the intra-prefecture correlations. 52 53 54 55 RESULTS 56 57 DRTB policy change from 2015-2018 in Zhejiang 58 59 60 10

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 34

1 2 3

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 34 BMJ Open

1 2 3

4 without formal jobs, were excluded. In 5 out of the 11 prefectures RMR patients were still not BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 eligible for the expanded public health insurance benefit package up to 2019. For the medical 8 9 assistance, in 6 out of the 11 prefectures only patients with registered residence (“Hukou”) in 10 11 12 Zhejiang were eligible, and for the other 4 prefectures with assistance policies the eligibility had 13 14 been expanded to all residents in this prefecture. Up to 2018, RMR patients in 5 prefectures were 15 16 17 still excluded from the medical assistance (supplemental table 2). 18 For peer review only 19 20 21 22 Cascade analysis of DRTB care from 2015-2018 23 24 25 Table 1 shows the screening and diagnosis of DRTB patients from 2015-2018. In general, the results 26 27 suggested increased capacity to find DRTB patients. The number of reported presumptive DRTB 28 29 30 patients increased from 9285 to 23916, with the largest increase from 2016 to 2017 coinciding with 31 32 33 the change in screening policy in most prefectures. The percentage of patients with test results also 34 35 increased from 69.3% to 78.1%. Along with the increase in the volume of DST there was an increase 36 http://bmjopen.bmj.com/ 37 38 in the number of diagnosed DRTB patients except for 2018, and a decrease in the percentage of 39 40 patients diagnosed among those tested. RR/MDR-TB patients accounted for around 40%-50% of 41 42 43 the total DRTB patients. 44 on September 29, 2021 by guest. Protected copyright. 45 46 Further analysis on the types of diagnostic tests patients received revealed an increasing trend in the 47 48 percentage of patients taking rapid DST as well as the percentage of patients tested with both 49 50 51 conventional culture-based and rapid DST. The percentage of presumptive patients without DST 52 53 profile for unknown reasons as recorded in the TBIMS dataset, which indicates pre-diagnosis 54 55 56 attrition, dropped dramatically from 17.3% to 2.7% (supplemental table 3). 57 58 59 60 12

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 34

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 34 BMJ Open

1 2 3

4 the presumptive and diagnosed patients were male, and around 98% of them were Han People. Only BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 around 40% of the patients had formal jobs other than farming or unemployment. While under the 8 9 assumption of perfect equity we would expect that the percentage of patients with different 10 11 12 characteristics remained the same from the diagnosed to the treated group, it decreased from 30.3% 13 14 to 27.0% for older patients, 28.1% to 22.5% for RMR patients, and from 24.4 to 17.0% for patients 15 16 17 with registered residence outside Zhejiang, i.e., migrants. An increase was observed regarding the 18 For peer review only 19 20 percentage of patients ever taking rapid DST. 21 22 23 Table 3: Patient characteristics across each step of the care cascade 24 Presumptive Diagnosed Diagnosed Registered RR/MDR-TB 25 DRTB DRTB RR/MDR- RR/MDR- patients recevied 26 27 patients patients TB patients TB patients treatment 28 N 65966 5598 2411 1859 1357 29 Age (%) 30 31 >=60 41.3 35.9 30.3 29.8 27.0 32 Gender (%) 33 Male 70.6 73.6 72.5 72.7 72.7 34 35 Ethnicity 36 Han (%) 97.7 98.0 97.9 98.0 98.8 http://bmjopen.bmj.com/ 37 Patient type (%) 38 new patient 70.2 65.2 51.8 51.4 48.8 39 40 Drug-resistanct type (%) 41 RMR / 12.1 28.1 20.8 22.5 42 MDR / 30.4 70.6 78.0 76.1 43 44 XDR / 0.5 1.2 1.2 1.4 on September 29, 2021 by guest. Protected copyright. 45 Non-rifampicin resistant / 56.9 / / / 46 Registered residence (%) 47 48 outside Zhejiang 23.4 25.2 24.4 22.3 17.0 49 Job category (%) 50 farmers 49.0 48.0 43.8 45.0 45.5 51 unemployed 12.4 13.0 14.1 14.2 14.0 52 53 other 38.6 39.0 42.1 40.8 40.5 54 Test type (%) 55 fast test / 38.6 46.5 46.1 49.2 56 57

58 59 60 14

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 34

1 2 3

4 Figure 2 shows the changes in the characteristics of diagnosed, registered and treated RR/MDR-TB BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 patients from 2015-2018. The percentage of diagnosed RMR patients that received adequate 8 9 treatment was low in 2015 and 2016, but increased dramatically from 2017, coinciding with the 10 11 12 policy change. The gaps between the proportion of older patients in the diagnosed and the treated 13 14 group also narrowed. Nevertheless, over the four years there remained a 7-10 percentage decrease 15 16 17 in the proportion of patients with registered residence outside Zhejiang from the diagnosed to the 18 For peer review only 19 20 treated patient group, indicating that migrants were still more likely to drop out after diagnosis 21 22 compared with the local residents. Gender and job category were not obviously associated with the 23 24 25 pre-treatment attrition in this univariate analysis. 26 27 To systematically investigate the impact of policies on DRTB case finding and care, we divided the 28 29 30 combination of policies in different areas of DRTB implemented in each prefecture for each year 31 32 33 during 2015-2018 into four categories, considering both policy implications and the number of cases 34 35 in each category. As shown in Figure 3, since 2017 all prefectures had equipped GeneXpert and 36 http://bmjopen.bmj.com/ 37 38 expanded the eligibility criteria for DST as well as registration, except for two prefectures which 39 40 still did not register RMR patients. 41 42 43 Table 4 shows the results of mixed-effect two-level logistic regression of the factors associated with 44 on September 29, 2021 by guest. Protected copyright. 45 46 registration among diagnosed RRTB patients, as well as with receiving anti-DRTB treatment among 47 48 those registered. After adjusting for other factors, patients under the policy of providing GeneXpert 49 50 51 together with expanding eligibility for DST (category 3, adjusted odds ratio (aOR) =2.57, 95% CI: 52 53 1.20, 5.51) had a significantly higher likelihood of being registered compared to patients not 54 55 56 provided with GeneXpert, while this association was not significant for providing GeneXpert testing 57 58 59 without expanding eligibility (category 2). A positive association with receiving anti-DRTB 60 15

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 34 BMJ Open

1 2 3

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 34

1 2 3 other jobs 0.92 0.281 0.80 1.07 0.94 0.784 0.60 1.47 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 Per capita GDP 6 highest group ref. ref. 7 middle group 1.72 0.122 0.87 3.42 2.20 0.025 1.10 4.38 8 lowest group 1.70 0.099 0.90 3.20 1.15 0.643 0.63 2.12 9 10 _cons 0.32 0.002 0.16 0.66 1.85 0.207 0.71 4.78 11 prefecture var(_cons) 0.18 0.09 0.39 0.18 0.03 0.92 12 13 14 15 16 Tables 5 shows the results of the mixed-effect two-level logistic regression analysis of factors 17 18 associated with theFor likelihood peer of completing review treatment for patients only who initiated treatment before 19 20 21 2017. All prefectures had some kind of financing policies in 2015 and 2016, and most prefectures 22 23 had not yet expanded the eligibility of patients for DST or registration. We therefore used the 24 25 26 individual level data on utilization of rapid DST to assess the impact of rapid testing on treatment 27 28 29 completion. After adjusting for other factors, older patients were less likely to complete treatment 30 31 (aOR=0.24, P<0.001), while registered residence inside Zhejiang was associated with higher 32 33 34 probability of completing treatment (aOR=1.92, P=0.04). These two factors had significant impacts 35 36 of the same direction on the registration, treatment initiation and treatment completion of http://bmjopen.bmj.com/ 37 38 39 RR/MDRTB patients. Utilization of rapid DST was not associated with treatment completion in this 40 41 42 study. 43 44 on September 29, 2021 by guest. Protected copyright. 45 Table 5: Factors associated with the likelihood of completing treatment for patients that initiated 46 treatment before 2017 47 48 OR P-value 95% Conf. Interval 49 Age 50 >=60 0.24 <0.001 0.14 0.42 51 52 Gender 53 male 0.79 0.442 0.44 1.44 54 Drug-resistance type 55 MDR/XDR-TB 0.50 0.086 0.23 1.10 56 57 Patient type 58 new patient 1.06 0.820 0.64 1.77 59 TB treatment history 60 17

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 34 BMJ Open

1 2 3 no treatment history ref. 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 first-line drug only 0.74 0.355 0.39 1.40 6 second-line drug used 1.07 0.847 0.56 2.02 7 Test type 8 fast test 0.70 0.096 0.46 1.06 9 10 Registered residence 11 in Zhejiang 1.92 0.040 1.03 3.59 12 Job category 13 14 farmers ref. 15 unemployed 0.84 0.635 0.42 1.71 16 other jobs 1.44 0.278 0.74 2.79 17 18 Per capita GDP For peer review only 19 highest group ref. 20 middle group 0.94 0.939 0.20 4.39 21 lowest group 1.07 0.921 0.30 3.74 22 23 Year of sending sample 24 2015 ref. 25 2016 1.23 0.164 0.92 1.64 26 27 _cons 5.58 0.001 2.01 15.50 28 prefecture var(_cons) 0.21 0.04 1.10 29 30 31 DISCUSSION 32 33 34 Results from this study clearly revealed the positive impacts of the combined policy changes 35 36 regarding DST of presumptive DRTB patients and registration for proper management of diagnosed http://bmjopen.bmj.com/ 37 38 39 patients. However, inequity challenges remain in terms of servicing vulnerable groups, e.g. migrant 40 41 42 workers and the older population, in the registration, treatment and management of DRTB. 43 44 on September 29, 2021 by guest. Protected copyright. 45 46 47 Effectiveness and challenges of the policy interventions 48 49 The expansion of the eligibility criteria of presumptive DRTB patients referred for DST, together 50 51 52 with the increased funding support to equip the facilities and guarantee the supply of reagents for 53 54 55 rapid DST like GeneXpert, have greatly improved the capacity for DRTB case finding. The 56 57 significant effects of the combined policies compared to providing GeneXpert alone indicates that 58 59 60 18

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 34

1 2 3

4 in order to effectively improve case finding we need to not just introduce new technologies, but also BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 support and expand their use. The inclusion of RMR in the registration and management of DRTB 8 9 patients closed the management gap between RMR and MDR/XDR-TB patients, and also improved 10 11 12 the treatment rate for RR/MDR-TB. 13 14 It is obvious that these reforms in the DRTB policies and regulations have effectively changed the 15 16 17 practices in DRTB control. Nevertheless, no policies aiming at migrants were issued between 2015 18 For peer review only 19 20 to 2018, and during these four years the degree of inequity between local and migrant patients did 21 22 not seem to be mitigated either. The improved insurance benefit package was only available for 23 24 25 those with public health insurance enrollment in Zhejiang, and in many prefectures receiving 26 27 medical assistance still required local registered residence during 2015-2018. In addition, older 28 29 30 people were less likely to be registered or receive treatment. This may be because they were more 31 32 33 likely to give up treatment due to the high cost and long course of the standard treatment, and health 34 35 workers may not register them once they refused to provide information necessary for registration. 36 http://bmjopen.bmj.com/ 37 38 All these findings were consistent with previous studies that age and migration for work, as well as 39 40 health system factors such as lack of clear eligibility criteria for DST and limited capacity to provide 41 42 43 DST were associated with attrition at different stages of the cascade.[14, 16, 31–33] Nevertheless, 44 on September 29, 2021 by guest. Protected copyright. 45 46 no impacts were observed of other factors, such as association between treatment experience and 47 48 treatment outcome, and data on other socio-economic factors such as financial difficulties were not 49 50 51 available in this study. 52 53 54 55 56 Equity challenges regarding migrant populations 57 58 59 The growing number of migrants in China and other parts of the world are posing a challenge to TB 60 19

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 34 BMJ Open

1 2 3

4 control.[31, 34–36] Although one systematic review found no significant differences in treatment BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 adherence between migrant and long-term resident MDRTB patients,[37] several studies identified 8 9 barriers for migrants to accessing TB diagnosis and care mainly in terms of knowledge gaps and 10 11 12 financial difficulty,[38, 39] and these barriers would likely to be more substantial for migrant 13 14 MDRTB patients. In our study, only around 40% of the DRTB patients had a formal job other than 15 16 17 farmers or self-employment. In China people with a formal job will be compulsively enrolled in the 18 For peer review only 19 20 urban employee basic medical insurance at the place of work as required by law. However, migrants 21 22 without a formal job often choose to participate in the health insurance for residents in their 23 24 25 hometown which requires lower premium compared to the more developed region they migrated to, 26 27 and thus cannot benefit from the reimbursement policy in their working place. Besides, they often 28 29 30 have to pay the full cost of treatment first before they get reimbursed when they return home. 31 32 33 Therefore, it is expected that many of these DRTB patients would choose to go back to their 34 35 hometown for treatment, or even refuse treatment. The need for migrant DRTB patients to travel 36 http://bmjopen.bmj.com/ 37 38 long-distances and the substandard or even absent treatment received would increase the risk of 39 40 disease transmission and treatment failure. A study in Shanghai, a well-developed city in eastern 41 42 43 China, showed that financial incentives were effective for migrant TB patients to complete 44 on September 29, 2021 by guest. Protected copyright. 45 46 treatment.[40] In terms of these findings, policies aimed at encouraging migrant DRTB patients to 47 48 be treated and managed at their current place of residence need to be developed, and this is 49 50 51 particularly urgent for regions with a large migrant population. 52 53 Limitations 54 55 56 One major limitation of our study is that almost all prefectures had some form of financing policy 57 58 59 since 2015 and there were no individual-level data on whether the patient benefited from these 60 20

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 34

1 2 3

4 financing policies. Therefore, in this study the impacts of these financing policies could not be BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 7 explored. Furthermore, other potentially important socio-economic factors like education and 8 9 income were not available in our dataset, limiting the scope of the equity analysis. Nevertheless, 10 11 12 these limitations would not influence our conclusions that changes in the screening and registration 13 14 policies of DRTB patients have largely increased the case finding and management capacity of 15 16 17 DRTB patients, while during the period 2015-2018 differences between the local and migrant 18 For peer review only 19 20 patients in the registration, treatment initiation and treatment outcome remained. We did not 21 22 consider population changes either as the resident population only increased 5% from 2015 to 2018 23 24 25 in Zhejiang.[41] Future research should collect more individual-level data on the implementation of 26 27 the health insurance and medical assistance policies to investigate the impacts of these demand-side 28 29 30 policies on DRTB treatment more directly, and explore what kind of financing policies provided for 31 32 33 migrants could increase their likelihood of initiating and completing anti-DRTB treatment. 34 35 36 DECALRATIONS http://bmjopen.bmj.com/ 37 Ethical approval 38 This study is under the overall evaluation study of China-Gates TB Project Phase III and used data 39 40 collected from this project. The implementation of this project has received the ethical approval 41 from China CDC (No.201626). The protocol of the overall study design and data collection tools 42 was reviewed and approved by the Institutional Review Board of Duke University (IRB Approval 43 44 code: 2017-0768). This paper does not use data involving human participants. on September 29, 2021 by guest. Protected copyright. 45 46 Competing interests 47 All authors declare no competing interests. 48 49 50 Funding 51 The work was supported by the Bill and Melinda Gates Foundation grant number [OPP1149395]. 52 53 54 Contributorship statement 55 The study was designed by Weixi Jiang, Shenglan Tang, Chris Elbers and Frank Cobelens. Fei 56 57 Huang, Ying Peng, Xiaomeng Wang, Bin Chen and Weixi Jiang coordinated and conducted data 58 collection. Weixi Jiang conducted literature review and wrote the manuscript as the first author. 59 Shenglan Tang, Chris Elbers, Bin Chen and Frank Cobelens provided suggestions on data analysis 60 21

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 34 BMJ Open

1 2 3 framework and data interpretation, and also revised the manuscript. All authors reviewed the draft 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 manuscript and provided comments on the finalization of the manuscript. All authors have read and 6 approved the manuscript in its current state. 7 8 Data sharing statement 9 10 Data may be obtained from a third party and are not publicly available. The data from the TBIMS 11 used in this study are owned by China CDC. The data could only be accessed after obtaining 12 permission from China CDC. 13 14 15 Acknowledgements 16 This paper is part of the outputs emanating from the program entitled “China National Health and 17 18 Family Planning CommissionFor peer and the Gates review Foundation TB onlyProject (Phase III)”-a collaboration 19 between the Government of China and the Melinda and Bill Gates Foundation, and implemented by 20 the China Center of Disease Control and Prevention (CDC). The authors of the paper also gratefully 21 acknowledge the officers in Zhejiang Provincial CDC and the 11 prefectural CDCs for helping with 22 23 collecting DRTB related policies in Zhejiang. 24 25 REFERENCES 26 27 1. World Health Organization. Global Tuberculosis Report 2019: P58 28 2. Li BY, Shi WP, Zhou CM, et al. Rising challenge of multidrug-resistant tuberculosis in China: 29 A predictive study using Markov modeling. Infect Dis Poverty. 2020;9(1):1-8. 30 31 doi:10.1186/s40249-020-00682-7 32 3. World Health Organization. accessed August, 2020 https://www.who.int/sdg/targets/en/ 33 4. Fitzpatrick C, Floyd K. A Systematic Review of the Cost and Cost Effectiveness of Treatment 34 for Multidrug-Resistant Tuberculosis. Pharmacoeconomics. 2012;30(1):63-80. 35 36 doi:10.2165/11598640-000000000-00000 http://bmjopen.bmj.com/ 37 5. Li R, Ruan Y, Sun Q, et al. Effect of a comprehensive programme to provide universal access 38 to care for sputum-smear-positive multidrug-resistant tuberculosis in China: A before-and- 39 40 after study. Lancet Glob Heal. 2015;3(4):e217-e228. doi:10.1016/S2214-109X(15)70021-5 41 6. Long Q, Qu Y, Lucas H. Drug-resistant tuberculosis control in China: progress and challenges. 42 Infect Dis Poverty. 2016;5(1):9. doi:10.1186/s40249-016-0103-3 43 44 7. Hutchison C, Khan MS, Yoong J, Lin X, Coker RJ. Financial barriers and coping strategies: a on September 29, 2021 by guest. Protected copyright. 45 qualitative study of accessing multidrug-resistant tuberculosis and tuberculosis care in Yunnan, 46 China. BMC Public Health. 2017;17(1):1-11. doi:10.1186/s12889-017-4089-y 47 8. Wingfield T, Boccia D, Tovar M, et al. Defining Catastrophic Costs and Comparing Their 48 49 Importance for Adverse Tuberculosis Outcome with Multi-Drug Resistance: A Prospective 50 Cohort Study, Peru. PLoS Med. 2014;11(7). 51 doi:10.1371/journal.pmed.1001675 52 53 9. Pan H-Q, Bele S, Feng Y, et al. Analysis of the economic burden of diagnosis and treatment 54 of tuberculosis patients in rural China. Int J Tuberc Lung Dis. 2013;17(12):1575-1580. 55 doi:10.5588/ijtld.13.0144 56 57 10. Wang Y, McNeil EB, Huang Z, et al. Household financial burden among multidrug-resistant 58 tuberculosis patients in Guizhou province, China: A cross-sectional study. Med. 59 2020;99(28):e21023. doi:10.1097/MD.0000000000021023 60 22

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 34

1 2 3 11. Tang S, Wang L, Wang H, Chin DP. Access to and affordability of healthcare for TB patients 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 in China: Issues and challenges. Infect Dis Poverty. 2016;5(1):1-5. doi:10.1186/s40249-016- 6 0096-y 7 12. Hemant Deepak Shewade, Dina Nair, Joel S. Klinton, et al. Pre-diagnosis attrition but high 8 pre-treatment attrition among patients with MDR-TB: An operational research from 9 10 Chennai,India. J Epidemiol Glob Health. 2017;7(4):227-233. 11 13. Boyd R, Ford N, Padgen P, Cox H. Time to treatment for rifampicin-resistant tuberculosis: 12 Systematic review and meta-analysis. Int J Tuberc Lung Dis. 2017;21(11):1173-1180. 13 14 doi:10.5588/ijtld.17.0230 15 14. Xu C, Li R, Shewade HD, et al. Attrition and delays before treatment initiation among patients 16 with MDR-TB in China (2006-13): Magnitude and risk factors. PLoS One. 2019;14(4):1-16. 17 18 doi:10.1371/journal.pone.0214943For peer review only 19 15. Hirasen K, Berhanu R, Evans D, et al. High rates of death and loss to follow-up by 12 months 20 of rifampicin resistant TB treatment in South Africa. PLoS One. 2018;13(10):1-14. 21 doi:10.1371/journal.pone.0205463 22 23 16. Shewade D, Kokane AM, Singh AR, et al. High pre-diagnosis attrition among patients with 24 presumptive MDR-TB: an operational research from Bhopal district, India. BMC Health Serv 25 Res. 2017;17(1):1-10. doi:10.1186/s12913-017-2191-6 26 27 17. Oo T, Kyaw KWY, Soe KT, et al. Magnitude and reasons for pre-diagnosis attrition among 28 presumptive multi-drug resistant tuberculosis patients in Bago Region, Myanmar: A mixed 29 methods study. Sci Rep. 2019;9(1):1-10. doi:10.1038/s41598-019-43562-3 30 31 18. Xu C, Pang Y, Li R, et al. Clinical outcome of multidrug-resistant tuberculosis patients 32 receiving standardized second-line treatment regimen in China. J Infect. 2018;76(4):348-353. 33 doi:10.1016/j.jinf.2017.12.017 34 19. Abubakar I, Zignol M, Falzon D, et al. Drug-resistant tuberculosis: Time for visionary political 35 36 leadership. Lancet Infect Dis. 2013;13(6):529-539. doi:10.1016/S1473-3099(13)70030-6 http://bmjopen.bmj.com/ 37 20. Radin E, Ariana P, Broekel T et al. Analyzing demand-side efficiency in global health: an 38 application to maternal care in Vietnam. Health Policy Plan. 2016;31(9):1281-1290. 39 40 doi:10.1093/heapol/czw063 41 21. Boehme CC, Nicol MP, Nabeta P, et al. Feasibility, diagnostic accuracy, and effectiveness of 42 decentralised use of the Xpert MTB/RIF test for diagnosis of tuberculosis and multidrug 43 44 resistance: A multicentre implementation study. Lancet. 2011;377(9776):1495-1505. on September 29, 2021 by guest. Protected copyright. 45 doi:10.1016/S0140-6736(11)60438-8 46 22. Christopher PM, Cucunawangsih, Widysanto A. GeneXpert Mycobacterium tuberculosis/ 47 rifampicin assay for molecular epidemiology of rifampicin-resistant Mycobacterium 48 49 tuberculosis in an urban setting of Banten province, Indonesia. Int J Mycobacteriol 50 2019;8:351-8. 51 23. Atashi S, Izadi B, Jalilian S. et al. Evaluation of GeneXpert MTB / RIF for determination of 52 53 rifampicin resistance among new tuberculosis cases in west and northwest Iran. New Microbes 54 New Infect. 2017;19:117-120. doi:10.1016/j.nmni.2017.07.002 55 24. Pang, Y. , Xia, H. , Zhang, Z., et al. Multicenter evaluation of genechip for detection of 56 57 multidrug-resistant mycobacterium tuberculosis. Journal of Clinical Microbiology, 51(6), 58 1707-1713. 2013;51(6):1707-1713. doi:10.1128/JCM.03436-12 59 25. Yin J, Yuan J, Hu Y, Wei X. Association between directly observed therapy and treatment 60 23

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 34 BMJ Open

1 2 3 outcomes in multidrug-resistant tuberculosis: A systematic review and meta-analysis. PLoS 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 One. 2016;11(3):1-14. doi:10.1371/journal.pone.0150511 6 26. Narges A , Leah J , Cecily M , et al. Adherence interventions and outcomes of tuberculosis 7 treatment: A systematic review and meta-analysis of trials and observational studies. PLoS 8 Med, 2018, 15(7):e1002595. doi:10.1371/journal.pmed.1002595 9 10 27. Baral SC, Aryal Y, Bhattrai R, et al. The importance of providing counselling and financial 11 support to patients receiving treatment for multi-drug resistant TB: Mixed method qualitative 12 and pilot intervention studies. BMC Public Health. 2014;14(1). doi:10.1186/1471-2458-14-46 13 14 28. Wang L, Li R, Xu C, et al. The Global Fund in China: Multidrug-resistant tuberculosis 15 nationwide programmatic scale-up and challenges to transition to full country ownership. PLoS 16 One. 2017;12(6):1-13. doi:10.1371/journal.pone.0177536 17 18 29. Subbaraman ForR, Nathavitharana peer RR, Mayerreview KH, et al. Constructing only care cascades for active 19 tuberculosis: A strategy for program monitoring and identifying gaps in quality of care. PLoS 20 Med. 2019;16(2):1-18. doi:10.1371/journal.pmed.1002754 21 30. China National Bureau of Statistics. Accessed August, 2020 22 23 http://www.stats.gov.cn/tjsj/ndsj/2019/indexch.htm 24 31. Truzyan N, Crape B, Grigoryan R, et al. Increased risk for multidrug-resistant tuberculosis in 25 migratory workers, Armenia. Emerg Infect Dis. 2015;21(3):474-476. 26 27 doi:10.3201/eid2103.140474 28 32. Parmar MM, Sachdeva KS, Dewan PK, et al. Unacceptable treatment outcomes and associated 29 factors among India’s initial cohorts of multidrug-resistant tuberculosis (MDR-TB) patients 30 31 under the revised national TB control programme (2007–2011): Evidence leading to policy 32 enhancement. PLoS One. 2018;13(4):1-24. doi:10.1371/journal.pone.0193903 33 33. Lalor MK, Greig J, Allamuratova S, et al. Risk factors associated with default from multi- and 34 extensively drug-resistant tuberculosis treatment, Uzbekistan: A retrospective cohort analysis. 35 36 PLoS One. 2013;8(11). doi:10.1371/journal.pone.0078364 http://bmjopen.bmj.com/ 37 34. Li X, Yang Q, Feng B, et al. Tuberculosis infection in rural labor migrants in Shenzhen, China: 38 Emerging challenge to tuberculosis control during urbanization. Sci Rep. 2017;7(1):1-8. 39 40 doi:10.1038/s41598-017-04788-1 41 35. Zumla A, Abubakar I. Improving access to multi-drug resistant tuberculosis diagnostic and 42 health services for refugees and migrants. BMC Med. 2018;16(1):7-10. doi:10.1186/s12916- 43 44 018-1218-0 on September 29, 2021 by guest. Protected copyright. 45 36. Wang W, Wang J, Zhao Q, et al. Contribution of rural-to-urban migration in the prevalence of 46 drug resistant tuberculosis in China. Eur J Clin Microbiol Infect Dis. 2011;30(4):581-586. 47 doi:10.1007/s10096-010-1125-6 48 49 37. Nellums LB, Rustage K, Hargreaves S, et a;. Multidrug-resistant tuberculosis treatment 50 adherence in migrants: A systematic review and meta-analysis. BMC Med. 2018;16(1):1-11. 51 doi:10.1186/s12916-017-1001-7 52 53 38. Long Q, Li Y, Wang Y, et al. Barriers to accessing TB diagnosis for rural-to-urban migrants 54 with chronic cough in Chongqing, China: A mixed methods study. BMC Health Serv Res. 55 2008;8:1-10. doi:10.1186/1472-6963-8-202 56 57 39. Wei X, Chen J, Chen P, et al. Barriers to TB care for rural-to-urban migrant TB patients in 58 Shanghai: A qualitative study. Trop Med Int Heal. 2009;14(7):754-760. doi:10.1111/j.1365- 59 3156.2009.02286.x 60 24

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 34

1 2 3 40. Wei X, Zou G, Yin J, et al. Providing financial incentives to rural-to-urban tuberculosis 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 migrants in Shanghai: An intervention study. Infect Dis Poverty. 2012;1(1):1-8. 6 doi:10.1186/2049-9957-1-9 7 41. Zhejiang Bureau of Statistics. Statistical Communique on the provincial economy and social 8 development of Zhejiang in 2015 and 2018, Accessed February, 2021. 9 10 http://tjj.zj.gov.cn/col/col1229129205/index.html 11 12 13 14 FIGURES AND LEGENDS 15 Figure 1: Seven steps of the DRTB diagnosis and care cascade 16 Figure 2: Characteristics of diagnosed, registered and treated RR/MDR-TB patients, 2015-2018 17 18 Figure 3: CombinationFor of DRTB peer policies implementedreview in each onlyprefecture from 2015-2018 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43 44 on September 29, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 25

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 34 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 210x71mm (96 x 96 DPI) 21 22 23 24 25 26 27 28 29 30 31

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

40 on September 29, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 34 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

40 on September 29, 2021 by guest. Protected copyright. 41 42 206x217mm (96 x 96 DPI) 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 29 of 34 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 207x102mm (96 x 96 DPI) 25 26 27 28 29 30 31

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

1 2 3 1. Policy survey questionnaire 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 6 Drug-resistant tuberculosis related policy survey in Zhejiang 7 8 Prefecture：______9 10 Institute：______11 Date：______(YY/MM/DD) 12 13 14 1. Eligibility criteria for drug-resistant tuberculosis (DRTB) patient screening 15 1) The current eligibility criteria of patients referred for screening are: 16 ______17 18 2) Are thereFor any changes peer in the eligibility review criteria onlyafter 2014: _____yes/no 19 If yes: year of change______, and the eligibility criteria for patients referred 20 for screening before policy change______21 （If the eligibility criterion is the “high-risk” TB patients, please specify the definition 22 23 of high-risk TB patients in your prefecture, for example, smear positive after two months’ 24 treatment or others）. 25 2. Test tools for drug susceptibility testing (DST) 26 27 1) Has GeneXpert been equipped in the prefecture-level designated hospital? If yes, 28 it was equipped in the year ______29 2) Under what conditions would patients be referred for traditional DRTB tests after 30 31 screening? ______32 3) Please list other types of DST conducted in the prefecture, including both 33 molecular and traditional______. 34 4) Does the government provide funds to secure the supply of the reagents for rapid 35 36 DST now? ______(Yes/no). If yes, at which year did the government begin to http://bmjopen.bmj.com/ 37 provide such funds? ____. 38 3. Financing of DST 39 40 1) Does the patient need to pay out-of-pocket for the DRTB rapid screening test now? 41 _____（yes/no）. If yes, how much does the patient need to pay? RMB_____, 42 accounting for _____% of the total cost 43 44 2) Does the government pay for the DRTB rapid screening test now? （yes/no）. If yes, on September 29, 2021 by guest. Protected copyright. 45 how much does the government pay for each test? RMB_____, accounting for _____% of 46 the total cost 47 3) Does the health insurance pay for the DRTB rapid screening test now? ______48 49 （yes/no）. If yes, _____% of the cost would be reimbursed by the health 50 insurance. 51 4) Does the health insurance pay for the DRTB rapid screening test now? （yes/no）. If 52 53 yes, _____% of the cost would be reimbursed by the health insurance. 54 5) Have the polices on the financing of DST ever been changed after 2014? If yes, the 55 year of change is ______. Before the policy change, 56 57 the patient paid for RMB____ for the screening test, accounting for _____% of the 58 total cost; 59 The government paid for _____% of the total cost; 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 31 of 34 BMJ Open

1 2 3 the health insurance schemes reimbursed ______% of the total cost; 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 other funding resources paid for ______% of the total cost 6 4. DRTB patient registration 7 1) Currently what types of DRTB patients would be registered and managed in the TB 8 information management system (TBIMS)? (Please check all that apply) 9 10 a. XDR and MDRTB b. rifampin mono-resistant (RMR) patients c. Mono-resistant to 11 other type of first-line drug 12 2) Has this policy been changed after 2014? _____(yes/no). If yes, please specify the 13 14 types of DRTB patients that were registered and under management before policy 15 change. (please check all that apply) 16 a. XDR and MDRTB b. rifampin mono-resistant (RMR) patients c. Mono-resistant to 17 18 other typeFor of first peer-line drugs review only 19 5. DRTB treatment 20 1) The standard anti-DRTB treatment regimen includes ______months’ hospitalization, 21 and the total treatment length is _____months 22 23 2) Health insurance policies: 24 a) Has the specialized outpatient reimbursement program for DRTB treatment been 25 implemented in your prefecture now? ______(yes/no) If so, this policy began 26 27 in the year______. The reimbursement rate for outpatient service is ______%, 28 and the annual limit line for compensation is RMB______. 29 b) The reimbursement rate for DRTB inpatient service is ______%, the deductible 30 31 is RMB_____, and the annual limit line for compensation is RMB______. 32 c) Has DRTB treatment been included in the serious disease reimbursement program 33 of the health insurance? ______(yes/no). If yes, this program started in the 34 year______, for the out-of-pocket payment over RMB______, the 35 36 reimbursement rate is ______%, and the annual limit line is RMB______. http://bmjopen.bmj.com/ 37 d) Could RMR patients enjoy the health insurance policies for anti-DRTB treatment 38 now? ______(yes/no). If yes, at which year could the RMR patients start to 39 40 enjoy such policies? ______. 41 e) Are these health insurance policies only eligible for those who enrolled in 42 the local health insurance schemes? ______(yes/no). If no, at which year 43 44 could patients with health insurance outside your prefecture/Zhejiang province on September 29, 2021 by guest. Protected copyright. 45 start to enjoy such policies? ______. 46 f) Could the health insurance policies be enjoyed if patients do not get cured 47 after the standard treatment length? ______(yes/no). If yes, how long could 48 49 patients enjoy these policies? 50 ______. 51 3) Other government financial assistance policies for DRTB patients 52 53 a) Are there any other financial assistance policies for DRTB patients in your 54 prefecture? ______(yes/no). If yes, at which year were they launched? 55 Please briefly describe the policy？（How much money given to DRTB patients, 56 57 how to deliver such assistance, etc.）______58 b) Could RMR patients enjoy the financial assistance policies for now? 59 ______(yes/no). If yes, at which year could the RMR patients start to enjoy 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 32 of 34

1 2 3 such policies? ______. 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 c) Could patients without local registered residence (Hukou) enjoy the policy 6 now? ______(yes/no). If yes, at which year could the RMR patients without 7 local Hukou enjoy such policies？______. 8 d) How long could the patients enjoy these policies after diagnosis? ______9 10 6. Any other DRTB related policies, please describe here______11 12

13 14 2. Supplementary result tables 15 Table 1: Year of policy change regarding the screening test and registration in the 11 prefectures 16 Prefecture No. Year of policy change 17 18 Forscreening peer range test review technology registration only policy 19 1 2016 2015 2017 20 2 2017 2014 2014 21 22 3 2017 2017 2017 23 4 2017 2017 2017 24 5 2017 2017 2017 25 6 2009 2016 2019 26 27 7 2017 2017 2017 28 8 2017 2015 2019 29 9 2017 2017 2017 30 31 10 2017 2017 2015 32 11 2000 2012 2012 33 34 35 Table 2: Starting time and eligibility for health insurance and medical assistance policy coverage, 36 2015-2018 http://bmjopen.bmj.com/ 37 Prefecture Health insurance policy Medical assistance 38 39 No. starting insurance RMR starting registered RMR 40 time outside ZJ time residence 41 outside ZJ 42 43 1 2018 N Y 2013 N N 44 2 2013 N Y 2013 Y Y since 2015 on September 29, 2021 by guest. Protected copyright. 45 3 2016 N N 2014 Y N 46 4 2017 N Y 2015 Y N 47 48 5 2016 N N 2015 N N 49 6 2019 N N 2012 N Y 50 7 2014 N N 2014 Y N 51 52 8 2013 N N 2015 N Y 53 9 2014 N Y since 2017 No policy 54 10 2013 N Y 2015 N Y 55 56 11 2011 N Y 2015 Y Y 57 58 Table 3: The diagnostic test for presumptive DRTB patients from 2015-2018 59 60 Year No. of % took % took % took Patients with no DST

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 33 of 34 BMJ Open

1 2 3 presumptive fast traditional both records (%) 4 BMJ Open: first published as 10.1136/bmjopen-2020-047023 on 12 April 2021. Downloaded from 5 patients DST DST tests total not TB unknown 6 reported reason 7 2015 9285 18.2 51.3 0.2 30.7 13.4 17.3 8 9 2016 10997 30.4 46.9 0.5 23.3 11.6 11.7 10 2017 21768 30.9 42.8 5.8 32.2 25.6 6.6 11 2018 23916 64.4 33.8 20.1 21.9 19.3 2.7 12 13 14 Table 4: Reasons for not completing anti-DRTB treatment according to the registration records 15 Year Total Under Dead lost-to-follow- Other 16 treatment up reasons 17 18 For No.peer % reviewNo. % onlyNo. % No. % 19 2015 68 11 16.2 17 25.0 13 19.1 27 39.7 20 2016 101 38 37.6 15 14.9 15 14.9 33 32.7 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43 44 on September 29, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 34 of 34

1 2 STROBE Statement—Checklist of items that should be included in reports of cross-sectional studies 3

1 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 35 of 34 BMJ Open

36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43 44 on September 29, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

2 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml