The Centers for Disease Control and Prevention System in China: Trends From

The Centers for Disease Control and Prevention System in China: Trends From 2002 2012 Chengyue Li, PhD, Mei Sun, PhD, Ying Wang, PhD, Li Luo, PhD, Mingzhu Yu, Yu Zhang, MBA, Hua Wang, BM, Peiwu Shi, BS, Zheng Chen, Jian Wang, BM, Yueliang Lu, BM, Qi Li, MD, Xinhua Wang, MS, Zhenqiang Bi, MD, Ming Fan, MS, Liping Fu, BM, Jingjin Yu, PhD, and Mo Hao, PhD Objectives. To assess the improvements of the Chinese Centers for Disease Control and Prevention (CDCs) system between 2002 and 2012, and problems the system has encountered. Methods. We obtained data from 2 national cross-sectional surveys in 2006 and 2013, including 32 provincial, 139 municipal, and 489 county-level CDCs throughout China. We performed a pre post comparative analysis to determine trends in resource allocation and service delivery. Results. The overall completeness of public health services significantly increased from 47.4% to 76.6%. Furthermore, the proportion of CDC staff with bachelor s or higher degrees increased from 14.6% to 32.6%, and governmental funding per CDC increased 5.3-fold (1.283 8.098 million yuan). The working area per CDC staff increased from 37.9 square meters to 63.3 square meters, and configuration rate of type A devices increased from 28.1% to 65.0%. Remaining problems included an 11.9% reduction in staff and the fact that financial investments covered only 71.1% of actual expenditures. Conclusions. China s CDC system has progressed remarkably, enabling quicker responses to emergent epidemics. Future challenges include establishing a sustainable financing mechanism and retaining a well-educated, adequately sized public health workforce. (Am J Public Health. 2016;106:2093 2102. doi:10.2105/ AJPH.2016.303508) See also Yu, p. 2086. Between 1953 and 2001, China relied on epidemic prevention stations(epss) for the majority of their public health services (PHSs) from the national to the county levels. EPSs were specifically responsible for disease control and surveillance, health supervision, health education, technical advising, and guideline provision for public health. 1 During health care reform in the 2000s, the lack of capacity to respond to public health emergencies and the changing disease spectrum and the nonseparation nature of health law enforcement led the Chinese central government to restructure the EPS networks to provide better surveillance and guard against disease outbreaks. EPSs have gradually separated their disease control and health inspection functions and changed their names to Centers for Disease Control and Prevention (CDCs). 2 Each administrative division has its own corresponding CDC 2 ; there are 32 provincial (including the Xinjiang Production and Construction Corps), 347 municipal, and nearly 3000 county-level CDCs, which together form a 3-tiered CDC system in mainland China. To carry out the functions of health promotion and disease prevention, China s CDCs must provide 7 core PHSs and 266 activities (Appendix A, available as a supplement to the online version of this article at http://www. ajph.org). 3 ThesecorePHSsaresimilartothe 10 essential PHSs in the United States. 4,5 CDCs are obligated to cooperate with other government departments (e.g., the departments of quality supervision and environmental protection) in the delivery of services, such as occupational health surveillance and environmental health. Before the 1980s, EPSs were fully funded by their local governments. 6 However, around the mid-1980s, the local governments reduced funding for public health and granted local public health agencies the authority to make up for lost revenues by charging for certain PHSs. 7,8 Predictably, EPSs and CDCs concentrated on revenue generation and had little incentive to offer preventive services, 9 leading to a decrease in PHS delivery owing to a lack of financial investment, 10 which was regarded as the primary problem for a wellfunctioning CDC system. 11 This, in turn, generated problems, such as unqualified staff, outdated equipment, and a lack of facilities. ABOUT THE AUTHORS Chengyue Li, Mei Sun, Ying Wang, Li Luo, and Mo Hao are with the Research Institute of Health Development Strategies and Collaborative Innovation Center of Social Risks Governance in Health, Fudan University, Shanghai, China. Mingzhu Yu and Jingjin Yu are with the Department of the Bureau of Disease Control and Prevention of National Health and Family Planning Commission, Beijing, China. Yu Zhang is with the Department of Health and Family Planning Commission of Hubei Province, Wuhan, Hubei, China. Hua Wang is with the Department of Health and Family Planning Commission of Jiangsu Province, Nanjing, Jiangsu, China. Peiwu Shi is with Zhejiang Academy of Medical Sciences, Hangzhou, Zhejiang, China. Zheng Chen is with the Department of National Grassroots Health Prevention Group, Shanghai. Jian Wang is with Chinese Center for Disease Control and Prevention, Beijing. Yueliang Lu is with Jiangsu Provincial Center for Disease Control and Prevention, Nanjing, Jiangsu. Qi Li is with Hebei Provincial Center for Disease Control and Prevention, Shijiazhuang, Hubei. Xinhua Wang is with Gansu Provincial Center for Disease Control and Prevention, Lanzhou, Gansu, China. Zhenqiang Bi is with Shandong Provincial Center for Disease Control and Prevention, Jinan, Shandong, China. Ming Fan is with Jilin Provincial Center for Disease Control and Prevention, Changchun, Jilin, China. Liping Fu is with Xinjiang Provincial Center for Disease Control and Prevention, Urumqi, Xinjiang, China. Correspondence should be sent to Mo Hao, PhD, Research Institute of Health Development Strategies, Fudan University, Box 177, 130 Dong an Road, Shanghai 200032, China (e-mail: haomo03@fudan.edu.cn) or Jingjin Yu, PhD, Director of the Bureau of Disease Control and Prevention, National Health and Family Planning Commission of China, 38 Bei Li Shi Road Jia, Beijing 100044, China (e-mail: yujj@moh.gov.cn). Reprints can be ordered at http://www.ajph.org by clicking the Reprints link. This article was accepted September 30, 2016. doi: 10.2105/AJPH.2016.303508 December 2016, Vol 106, No. 12 AJPH Li et al. Peer Reviewed Research 2093

The 2003 severe acute respiration syndrome (SARS) crisis highlighted the weaknesses of China s CDC system. 12,13 Upon suppressing the epidemic, strengthening the CDC system became the top priority in China s public policy agenda. 14 Since then, multiple policies have been released as guidelines for the CDC system, which include long-term strategies; allocation of personnel, funding, and equipment; and procedures for disease control (Appendix B, available as a supplement to the online version of this article at http://www.ajph.org). 3,15,16 The Chinese governments have become more involved in developing public health infrastructure, including investing 10 billion yuan (1 yuan = US$ 0.156 in 2012) in rebuilding CDC facilities and increasing funding for the delivery of PHSs from 2002 to 2012. 17 During this period, the effects of these reforms have been the focus of considerable public scrutiny and scholarly attention. It is important to scrutinize whether the capacity of the CDC system has increased during the past 10 years. Some researchers have qualitatively introduced or interpreted the policy, evolution, strategy, and progress of the system before or after SARS. 10,12 Others have quantitatively assessed the situation of the CDCs (resource allocation or service delivery) using regional survey data 18,19 or national statistical reports, 20,21 but few have reflected on the changes in the system. Only 1 national survey reported the progress of the CDC system between 2002 and 2005, and it revealed some positive effects in the number of the staff, the funding and infrastructure, and several key performance indicators. 22 Unfortunately, all such quantitative studies used annual or short-term data. Similarly, whereas other countries (e.g., Canada, the United States) have evaluated the staff and financial resources of local public health agencies, 23,24 none did so with a national scope. To our knowledge, little evidence exists to examine the trends in China s CDC system in these 10 years. There is a need to bridge the information gap. We evaluated CDC system changes between 2002 and 2012. Specifically, we quantified the policy effects in terms of the extent to which provision of PHSs and resources used to carry out such services (e.g., personnel, funding, facilities) have increased and the problems they have encountered in attempting to do so. METHODS We administered 2 rounds of national cross-sectional surveys of the CDC system, in 2006 and 2013. We used a multistage sampling strategy to select local CDCs at different levels to obtain a nationally representative, geographically varied sample. The sampling agencies were the same in both survey years to ensure pre post comparison. At the provincial level, we selected all 32 agencies. At the municipal level, we used systematic sampling to select agencies from all municipal CDCs. The amount of governmental funding to CDCs per 1000 population was key for determining sample size (a = 0.05; b = 0.05) on the basis of differences and SDs derived from a pilot study. This process yielded a required sample size of 139 municipal CDCs. Before sampling, we ranked all municipal CDCs in ascending order by their corresponding administrative regional codes. We identified the first sampled institution by using a computergenerated random number; we then selected every third CDC on the list. We treated the sampling progression through the list circularly, with a return to the top once the end of the list was passed. Similarly, we selected every sixth county-level CDC, with a required sample size of 489. This process ensured that all sampled CDCs were well distributed across mainland China (Appendixes C and D, available as supplements to the online version of this article at http://www.ajph.org). Data Collection We performed the 2 data collection rounds in partnership with the Ministry of Health and 32 provincial departments of health. In both surveys we used an electronically structured questionnaire to ensure the comparability of survey results. We ensured the consistency and reliability of the questionnaire with a pilot investigation. We conducted the first round in June September 2006. Provincial quality supervisors, after receiving detailed training in questionnaire administration and the survey process from the Ministry of Health, delivered the information they obtained to investigators at the sampled institutions in their corresponding provinces. The investigators then collected survey data at each agency and submitted it to their provincial supervisors via e-mail or CD-ROM. The questionnaires were then submitted to the research group after being double-checked. We carried out the second round in July October 2013. In this round, the sampled CDCs were required to log in to the National Disease Control and Prevention Management Information Platform to report the data, which was then exported through Microsoft SQL Server 2005 (Microsoft, Redmond, WA). We integrated the quality control process into the platform with logical judgments and audit procedures. The questionnaires collected information on staffing (e.g., number of staff members, educational attainment, professional titles and school majors), funding sources, expenditure (e.g., government funding, profitoriented service revenue, subsidies, general agency expenditures), facilities (e.g., floor areas and housing structure, amount of equipment), and delivery of the 7 core PHSs (e.g., number of immunizations provided and laboratory examinations carried out). The 2006 and 2013 surveys collected data for 2002 2006 and 2012, respectively. We used only data from 2002 and 2012 for our analysis. Trained research team members rechecked the survey data for abnormal values. The overall response rates were 93.8% in 2002 and 88.7% in 2012. Comparing the age distributions (using Myer s index) of staff from 2002 and 2012 survey data against those of all CDC staff in China in the 2003 and 2013, respectively, China s Health Statistics Yearbook 25,26 suggested that the samples were representative of CDCs throughout China. Measures We adopted the conceptual framework of this study from the framework of Handler et al., 27 which can be used as the basis for measuring the performance of public health systems and has been widely applied at multiple levels (national, state, and local). It was derived from the Donabedian model, 28 which links structures, processes, and outcomes. We chose indicators to measure trends 2094 Research Peer Reviewed Li et al. AJPH December 2016, Vol 106, No. 12

in resource allocation (structures) and provision of PHSs (processes) in the Chinese CDC system. A 15-member expert panel, which included national experts who were familiar with public health, performance evaluation, and service measurement, and practice professionals with expertise in the various core services determined and confirmed the indicators. We used 6 indicators to evaluate resource allocation: number of CDC staff members per 100 000 population, proportion of CDC staff with bachelor s or higher degrees, governmental funding per CDC, percentage of funding to general agency expenditures, working area per CDC staff, and configuration rate of type A devices. We adopted 1 continuous indicator completeness of the service (i.e., percentage of the defined tasks that were accomplished of the total required) to comprehensively reflect the provision of PHSs on the basis of the Ministry of Health s National Criteria for Performance Evaluation of Disease Control and Prevention 3 ; this contrasted with previous measurements, which were all categorical. We calculated 8 indexes: 1 index for each of the 7 core PHSs and an eighth index representing overall completeness. The completeness of PHSs ranged from 0% (not provided at all) to 100% (fully provided). The definitions of all indicators are listed in the box on the next page. Statistical Analysis We used Microsoft Excel 2010 (Microsoft, Redmond, WA) to create a database for data processing. We performed statistical analyses using SPSS version 13.0 (SPSS, Inc, Chicago, IL). We used a pre post comparative analysis to measure CDC system changes. 29 Indicators were presented as means and 95% confidence intervals (CIs) by government level (provincial, municipal, and county) and region (eastern, central, and western) for both years. We adjusted the governmental funding per CDC in 2012 using the consumer price index to reflect the actual trend compared with that in 2002. We also calculated the annual change rates for 2002 and 2012. We used the paired sample t test to test differences in means and the Pearson c 2 test to test differences in proportions between 2002 and 2012. All P values were 2 sided, and we considered P <.05 significant. RESULTS As shown in Table 1, the overall completeness of PHS provision in China s CDC system was 76.6% in 2012, compared with 47.4% in 2002 (an increase of 61.7%; P <.001). At the provincial, municipal, and county levels, all CDCs showed increases of more than 50% (56.4%, 82.8%, and 59.2%, respectively; P <.001). From a regional perspective, the average increases in the eastern, central, and western regions were 60.6%, 68.2%, and 74.6%, respectively. Measures of all 7 core PHSs increased to some extent (Table 1). Specifically, the completeness of public health information administration and health education and promotion increased by 135.6% and 122.4%, respectively; the smallest increment (7.3%) was for technical guidance and applied research. Disease control and prevention and management of public health emergencies, the core capacities of dealing with disease epidemics and responding to emergencies, improved by 21.5% and 33.8%, respectively. The largest increases in completeness varied across levels and regions: disease and health risk monitoring at provincial and eastern regional CDCs (increment of 140.0% and 132.7%, respectively); public health information administration at municipal, county-level, and western regional CDCs (increments of 188.0%, 130.2%, and 203.7%, respectively); and laboratory testing in the central regions (increment of 142.8%). Table 2 shows changes in resource allocation between 2002 and 2012. The number of CDC staff per 100 000 showed a downward trend (16.83 in 2002 to 14.82 in 2012, a decrease of 11.9%). The provincial and eastern CDCs showed slight declines (by 1.5% and 2.0%, respectively), whereas central and western CDCs showed greater declines (17.3%; P =.04 and 17.2%; P =.02, respectively). The proportion of CDC staff members with bachelor s or higher degrees increased dramatically, from 14.6% to 32.6% (P <.001). The central CDCs had the largest increases (146.2%), followed by the western (139.4%) and eastern (99.2%) CDCs. The average increment in the proportion differed substantially between provincial (41.7%) and county-level (98.3%) CDCs. Regarding funding, rapid increases in governmental funding per CDC took place between 2002 and 2012. The average funding per CDC in 2002 was 1.283 million yuan; with an annual growth rate of 20.2%, this eventually reached a 5.3-fold increase in 2012 (i.e., 8.098 million yuan; P <.001). We observed similar increasing trends across different levels and regions. Accordingly, the percentage of funding to general agency expenditures grew from 44.4% in 2002 to 71.1% in 2012, meaning that government investments became the main funding source for daily operations. The most rapid increase occurred in eastern CDCs, at 115.0%. Regarding facilities, the working area per CDC staff member increased by 66.9%, to 63.3 square meters (P <.001); these areas at the provincial, municipal, and county levels were 75.6 square meters, 73.4 square meters, and 59.5 square meters (increases of 49.4%, 67.1%, and 68.9%), respectively, which nearly reached the goals of the construction standards for CDCs (70 m 2,65m 2, and 60 m 2, respectively). The configuration rate of type A devices was 65.0% in 2012 and 28.1% in 2002 (P <.001); the largest gains occurred in western and central CDCs (at 154.2% and 143.8%, respectively). DISCUSSION Our study is the first, to our knowledge, to quantify the changes in China s CDC system at the national level between 2002 and 2012, and it serves as a concise overview of the system s dynamics on the basis of the latest evidence. Furthermore, we took advantage of unique data from the CDC system, covering all 32 provincial CDCs as well as 139 municipal and 489 county-level CDCs; thus, our results were highly representative and valid. Our findings indicated that with the implementation of various strategies and policies, China s CDC system has achieved significant improvements in personnel, funding, and facilities, thereby resolving many of the previous problems in these areas. Furthermore, completeness of the core PHSs increased remarkably, particularly the management of public health emergencies among provincial and municipal CDCs (which December 2016, Vol 106, No. 12 AJPH Li et al. Peer Reviewed Research 2095

DEFINITIONS OF THE INDICATORS USED TO ASSESS CENTERS FOR DISEASE CONTROL AND PREVENTION SYSTEM TRENDS: CHINA Group Indicator Measure Definition Resource allocation Number of CDC staff per 100 000 population Average number of CDC staff for every 100 000 residents in the region. Proportion of CDC staff with bachelor s or higher degrees Percentage of staff members who had a bachelor s degree or above among the total staff of the CDC. Governmental funding per CDC The amount of governmental funding allocated to the CDC, including personnel, general operating, regular maintenance, and special disease control funding. Percentage of funding to general agency expenditures Funding to CDC as a percentage of general agency expenditures (including staff salaries and welfare, services and goods spending, and subsidies for families and individuals). Working area per CDC staff Average workspace of every CDC staff member. Configuration rate of type A devices According to the construction standards for CDC, 16 type A devices are defined as equipment required for the completion of routine work. Provincial, municipal, and county-level CDCs should be equipped with 97, 57, and 39, respectively, such devices. We calculated the configuration rate as the proportion of type A devices provided to the total required. Provision of PHSs Completeness of 7 core PHSs (disease control and prevention, management of public health emergencies, public health information administration, disease and health risk monitoring, laboratory testing, health education and promotion, and technical guidance and applied research) and overall completeness of PHSs According to the National Criteria for Performance Evaluation of Disease Control and Prevention, 3 there are corresponding assessment indicators and criteria for evaluating the performance of each of the 266 activities of the core PHSs. The actual value of each assessment indicator can be computed directly with the reported data of the survey. We calculated the completeness of each activity as the percentage of the actual value of the indicator to the performance standard, producing a range from 0% to 100%, with 100% indicating that the activity was fully conducted (i.e., the actual value was equal to or exceeded the standard) and 0% indicating that no activity was performed. Subsequently, we calculated the completeness of each of the 7 core PHSs as the average of the activities included in each core service, and we calculated the overall completeness as the average of all 266 activities. The completeness of the core PHSs ranged from 0% to 100%, with 0% indicating that the services were not provided at all and 100% indicating that the services were fully provided. Note. CDC = Center for Disease Control and Prevention; PHS = public health service. increased to nearly 90%). This suggests that a CDC system with better resource allocation and service delivery has been established in China. This improved system has provided a steady foundation for helping China shift from its disordered response to SARS to its well-organized handling of avian influenza (H7N9) in 2013 in terms of better educational structure, financial support, monitoring, and early warning. 30 The World Health Organization even recognized this response as a good model. 31 A notable improvement was the 5.3-fold increase in governmental funding to CDCs, which afforded sufficient financial support for effective service delivery. This has been corroborated by other studies. 12,32 PHSs should be organized and delivered by governments, 33 because of their role in protecting people s health. Before China s reform and opening up policy, EPSs received sufficient government funding, which was likely responsible for the major gains in controlling infectious diseases through immunization and other traditional public health measures. 34 After the implementation of this policy, however, government public health funding declined, leading CDCs to begin to charge fees. By 2002, the overall completeness was only 47.4%, which indicated that PHS 2096 Research Peer Reviewed Li et al. AJPH December 2016, Vol 106, No. 12

TABLE 1 Trends in Completeness of Public Health Services Provision of the Centers for Disease Control and Prevention by Levels and Regions: China, 2002 and 2012 Characteristic 2002, Mean (95% CI) 2012, Mean (95% CI) Change Between 2002 and 2012, a % Annual Rate of Change Between 2002 and 2012, b % Overall completeness of public health services, % All 47.4 (44.8, 49.9) 76.6 (76.1, 77.0) 61.7 4.9 Provincial 55.3 (46.8, 63.7) 86.4 (83.5, 89.3) 56.4 4.6 Municipal 46.1 (42.3, 50.0) 84.3 (83.3, 85.3) 82.8 6.2 County 47.4 (43.8, 51.0) 75.5 (75.0, 76.0) 59.2 4.8 East 53.4 (48.8, 58.0) 85.8 (84.9, 86.6) 60.6 4.8 Central 45.7 (41.5, 49.9) 76.9 (76.3, 77.5) 68.2 5.3 West 40.7 (37.2, 44.2) 71.1 (70.4, 71.8) 74.6 5.7 Disease control and prevention, % All 66.8 (63.7, 69.9) 81.2 (80.8, 81.6) 21.5 2.0 Provincial 76.7 (63.6, 89.8) 83.1 (79.7, 86.5) 8.4 0.8 Municipal 68.6 (64.0, 73.2) 86.2 (85.3, 87.1) 25.7 2.3 County 66.5 (62.1, 70.9) 80.6 (80.2, 81.0) 21.1 1.9 East 69.6 (64.2, 75.0) 88.1 (87.4, 88.9) 26.6 2.4 Central 65.5 (59.6, 71.4) 82.2 (81.6, 82.7) 25.5 2.3 West 64.7 (59.9, 69.5) 76.3 (75.6, 76.9) 17.9 1.7 Management of public health emergencies, % All 57.4 (54.6, 60.2) 76.8 (76.1, 77.6) 33.8 3.0 Provincial 67.9 (57.5, 78.3) 91.6 (87.1, 96.1) 34.9 3.0 Municipal 55.9 (51.3, 60.4) 88.2 (86.8, 89.6) 57.9 4.7 County 57.5 (53.7, 61.2) 75.3 (74.4, 76.1) 31.0 2.7 East 64.4 (59.6, 69.2) 86.2 (84.9, 87.5) 33.7 2.9 Central 54.8 (49.9, 59.7) 78.3 (77.3, 79.4) 42.9 3.6 West 50.7 (46.8, 54.6) 69.9 (68.5, 71.2) 37.8 3.3 Public health information administration, % All 34.2 (30.9, 37.5) 80.5 (79.8, 81.3) 135.6 8.9 Provincial 45.6 (37.4, 53.8) 95.3 (91.9, 98.8) 109.0 7.6 Municipal 31.1 (26.3, 35.9) 89.5 (87.8, 91.2) 188.0 11.2 County 34.4 (29.8, 39.1) 79.3 (78.4, 80.1) 130.2 8.7 East 39.6 (33.2, 46.0) 90.0 (88.8, 91.1) 127.1 8.5 Central 35.4 (30.1, 40.7) 81.1 (80.0, 82.2) 128.8 8.6 West 24.6 (20.2, 29.0) 74.6 (73.2, 76.0) 203.7 11.7 Disease and health risk monitoring, % All 33.2 (29.8, 36.6) 56.5 (55.6, 57.5) 70.6 5.5 Provincial 34.8 (17.9, 51.8) 83.6 (77.5, 89.7) 140.0 9.2 Municipal 33.5 (28.6, 38.3) 71.8 (69.4, 74.3) 114.7 7.9 County 33.1 (28.2, 38.0) 54.3 (53.3, 55.4) 64.3 5.1 East 33.4 (27.0, 39.8) 77.7 (75.9, 79.5) 132.7 8.8 Central 33.6 (27.6, 39.6) 54.1 (52.7, 55.5) 61.0 4.9 West 32.9 (27.8, 38.0) 47.4 (45.9, 48.8) 43.8 3.7 Laboratory testing, % All 41.2 (37.6, 44.8) 80.9 (80.2, 81.6) 96.3 7.0 Provincial 57.8 (41.2, 74.4) 96.5 (94.8, 98.3) 67.0 5.3 Municipal 44.9 (39.5, 50.4) 89.9 (88.6, 91.3) 100.1 7.2 County 40.6 (35.5, 45.7) 79.6 (78.8, 80.3) 96.1 7.0 East 55.5 (49.4, 61.6) 89.9 (88.7, 91.1) 62.0 4.9 Central 34.2 (28.8, 39.6) 83.1 (82.2, 84.1) 142.8 9.3 Continued December 2016, Vol 106, No. 12 AJPH Li et al. Peer Reviewed Research 2097

TABLE 1 Continued Characteristic 2002, Mean (95% CI) 2012, Mean (95% CI) Change Between 2002 and 2012, a % Annual Rate of Change Between 2002 and 2012, b % West 28.7 (23.0, 34.4) 73.2 (72.0, 74.4) 155.1 9.8 Health education and promotion, % All 39.1 (35.2, 43.0) 86.9 (86.2, 87.6) 122.4 8.3 Provincial 47.0 (26.9, 67.1) 92.3 (87.6, 97.1) 96.3 7.0 Municipal 36.1 (30.4, 41.9) 92.2 (90.7, 93.8) 155.1 9.8 County 39.3 (33.9, 44.8) 86.2 (85.4, 87.0) 119.1 8.2 East 46.4 (39.5, 53.3) 92.8 (91.7, 94.0) 100.1 7.2 Central 38.1 (31.4, 44.8) 88.5 (87.4, 89.5) 132.1 8.8 West 30.5 (25.3, 35.7) 81.8 (80.4, 83.2) 168.3 10.4 Technical guidance and applied research, % All 56.5 (52.6, 60.4) 60.6 (59.9, 61.2) 7.3 0.7 Provincial 59.1 (47.2, 70.9) 77.3 (72.7, 81.8) 30.8 2.7 Municipal 55.5 (49.1, 61.9) 70.4 (68.5, 72.3) 26.8 2.4 County 56.5 (51.2, 61.9) 59.2 (58.5, 59.9) 4.7 0.5 East 61.1 (54.3, 67.9) 71.8 (70.3, 73.2) 17.4 1.6 Central 57.4 (50.6, 64.2) 58.7 (57.8, 59.7) 2.3 0.2 West 49.1 (42.7, 55.5) 56.4 (55.3, 57.4) 14.8 1.4 Note. CDC = Center for Disease Control and Prevention;CI = confidence interval.the completeness of public health services ranged from 0% (not provided at all) to 100% (fully provided). We performed a paired sample t test to examine differences between 2002 and 2012. a Change between 2002 and 2012 (%) = (value in 2012 value in 2002) / value in 2012 100. ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi b Annual rate of change between 2002 and 2012 ð% Þ ¼ p 10 value in 2012=value in 2002 1 100. delivery was insufficient. The Chinese government s slow response to the SARS epidemic reflected this situation. 9 The subsequent financial commitment to CDCs evidently helped increase the completeness (up to 76.6% by 2012), thus indicating the importance of government funding of public health functions. Another trend was the improvement in the qualifications of CDC personnel. These results agree with previous studies in China. 35 Possibly, following the SARS outbreak, the personnel qualifications in CDCs became more specified, and recruitment and requirements of new professionals became more rigorous. Second, nonprofessionals gradually retired or left for other institutions. 36 Additionally, CDCs emphasized training and continuing education to update the current staff s professional knowledge and skills, 37 such as field epidemiology training and techniques for handling emerging infectious diseases 38 to ensure that personnel respond to emergencies appropriately. Public health information administration showed the quickest increase over the past 10 years. Undoubtedly, public health benefited from the progress in modern information technology: the Ministry of Health implemented the world s largest real-time Web-based disease surveillance system in 2004 to monitor notifiable diseases and public health emergencies. This system allows hospitals and clinics to directly report individual case based information immediately through the Internet, which led the mean incidence reporting time to drop from 5 days to 4 hours. This has helped address delays and incomplete reporting of communicable diseases. 39 Furthermore, CDCs can obtain disease informationinrealtime,thusallowing the immediate identification of disease outbreaks and the provision of essential and up-to-date information on epidemics to policymakers. 40 These were demonstrated during the 2013 H7N9 outbreak, when China quickly and successfully detected and rapidly controlled the new cases. 30 Because of the large amount of monitoring data, future analyses should focus on detecting possible outbreaks, alerting relevant officials of them, and identifying the trends in priority diseases. This would help public health officials immediately implement needed containment strategies. 41 Despite the improvements we have outlined, several issues arose regarding the sustainability of the CDC system. First, the amount of government funding was insufficient: there remained a 28.9% gap in being fully funded. Moreover, investment stability was poor: after a strong increase in 2003, the growth rate of government funding slowed. These findings imply a poor financing mechanism, and the CDC system s primary problem has not been resolved. Therefore, the government should establish a sound financing mechanism to fully fund PHSs and avoid arbitrariness in investment. 42 An example of this is setting a steady annual growth rate in financial investment (as with fiscal expenditure), thus eliminating situations such as increased funding after epidemic outbreaks. Second, despite improvements in the educational structure, the number of CDC personnel per 100 000 population has declined, which accords with previous studies reporting an annual decrease in Chinese 2098 Research Peer Reviewed Li et al. AJPH December 2016, Vol 106, No. 12

TABLE 2 Trends in Centers for Disease Control and Prevention Resource Allocation by Levels and Regions: China, 2002 and 2012 Characteristic 2002, Mean (95% CI) 2012, Mean (95% CI) Change Between 2002 and 2012, a % Annual Rate of Change Between 2002 and 2012, b % CDC staff, per 100 000 population, No. All 16.83 (13.71, 19.95) 14.82 (13.47, 16.17) 11.9 1.3 Provincial 0.90 (0.39, 1.41) 0.89 (0.26, 1.52) 1.5 0.2 Municipal 3.56 (1.85, 5.27) 3.13 (2.68, 3.58) 12.1 1.3 County 12.79 (11.61, 13.97) 11.07 (10.59, 11.55) 13.5 1.4 East 18.09 (15.95, 20.23) 17.73 (16.78, 18.68) 2.0 0.2 Central 18.24 (16.91, 19.57) 15.09 (14.59, 15.59) 17.3 1.9 West 14.16 (12.77, 15.55) 11.73 (11.32, 12.14) 17.2 1.9 Proportion of CDC staff with bachelor s or higher degrees, % All 14.6 (14.3, 14.9) 32.6 (32.4, 32.8) 123.0 8.4 Provincial 46.6 (45.7, 47.5) 66.1 (65.2, 67.0) 41.7 3.6 Municipal 28.4 (27.7, 29.1) 48.9 (48.4, 49.4) 72.1 5.6 County 12.6 (12.1, 13.0) 25.0 (24.8, 25.2) 98.3 7.1 East 24.1 (23.3, 24.9) 48.0 (47.5, 48.5) 99.2 7.1 Central 11.9 (11.5, 12.3) 29.3 (28.9, 29.7) 146.2 9.4 West 10.9 (10.4, 11.4) 26.1 (25.8, 26.4) 139.4 9.1 Governmental funding per CDC, 10 000 yuan All 128.3 (92.4, 164.1) 809.8 (741.2, 878.4) 531.2 20.2 Provincial 1 835.4 (1 434.6, 2 236.3) 14 190.9 (10 667.7, 17 714.2) 673.2 22.7 Municipal 311.4 (263.9, 359.0) 1 925.0 (1 652.1, 2 198.0) 518.1 20.0 County 88.4 (76.8, 100.0) 533.7 (504.8, 562.6) 503.9 19.7 East 196.5 (133.3, 259.7) 1521.8 (1247.6, 1796.0) 674.5 22.7 Central 122.5 (89.7, 155.4) 641.3 (567.3, 715.3) 423.5 18.0 West 103.8 (74.2, 133.3) 617.4 (555.5, 679.3) 494.8 19.5 Funding to general agency expenditures, % All 44.4 (41.4, 47.3) 71.1 (70.1, 72.1) 60.2 4.8 Provincial 58.9 (42.4, 75.5) 68.2 (56.9, 79.5) 15.7 1.5 Municipal 42.0 (35.8, 48.2) 78.0 (75.2, 80.8) 85.9 6.4 County 43.0 (39.6, 46.5) 69.1 (68.0, 70.2) 60.7 4.9 East 31.4 (25.5, 37.4) 67.5 (65.0, 70.0) 115.0 8.0 Central 47.2 (42.4, 52.0) 69.6 (67.9, 71.3) 47.5 4.0 West 65.2 (60.7, 69.8) 77.6 (76.2, 79.0) 19.0 1.8 Working area per CDC staff, m 2 All 37.9 (35.4, 40.4) 63.3 (62.1, 64.5) 66.9 5.3 Provincial 50.6 (43.8, 57.4) 75.6 (68.5, 82.7) 49.4 4.1 Municipal 43.9 (38.9, 48.9) 73.4 (69.9, 76.9) 67.1 5.3 County 35.2 (32.2, 38.1) 59.5 (58.2, 60.8) 68.9 5.4 East 53.6 (48.4, 58.8) 75.8 (72.7, 78.9) 41.4 3.5 Central 37.7 (34.9, 40.5) 62.1 (60.2, 64.0) 64.7 5.1 West 30.2 (27.9, 32.5) 57.1 (55.3, 58.9) 89.1 6.6 Continued CDC staff. 20 From 2002 to 2012, CDCs needed to prepare for challenges of newly emerging infectious diseases and simultaneously control existing infectious diseases such as tuberculosis and HIV/AIDS. 40 The pace and spread of lifestyle changes, including nutritional imbalance and decreased physical activity, has accelerated to an unprecedented degree. 43 Furthermore, health risks such as environmental pollution grew rapidly. Taken together, CDCs workloads were greatly increased; however, the number of CDC staff members declined, which indicates a shortage of public health staff members. December 2016, Vol 106, No. 12 AJPH Li et al. Peer Reviewed Research 2099

TABLE 2 Continued Characteristic 2002, Mean (95% CI) 2012, Mean (95% CI) Change Between 2002 and 2012, a % Annual Rate of Change Between 2002 and 2012, b % Configuration rate of type A devices, % All 28.1 (26.8, 29.4) 65.0 (64.2, 65.8) 131.3 8.7 Provincial 53.3 (49.5, 57.1) 83.0 (78.2, 87.7) 55.7 4.5 Municipal 34.7 (32.0, 37.3) 75.2 (73.4, 77.0) 116.7 8.0 County 27.1 (25.7, 28.5) 63.5 (62.7, 64.4) 134.3 8.9 East 39.1 (36.2, 42.0) 74.2 (72.6, 75.8) 89.8 6.6 Central 27.6 (25.9, 29.3) 67.3 (66.2, 68.5) 143.8 9.3 West 22.5 (20.5, 24.5) 57.2 (55.8, 58.5) 154.2 9.8 Note. CDC = Center for Disease Control and Prevention; CI = confidence interval. We performed a paired sample t test or c 2 test to examine differences between 2002 and 2012. a Change between 2002 and 2012 (%) = (value in 2012 value in 2002) / value in 2012 100. ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi b Annual rate of change between 2002 and 2012 ð% Þ ¼ p 10 value in 2012=value in 2002 1 100. There may be several explanations for these results. First, the value of disease control and prevention work was not extensively recognized, which led to the social status of CDC staff being perceived as low. 44 Moreover, as CDC work derives value from prevention by contrast to doctors, whose value is measured through service volume or revenue CDC staff lacked recognition for their achievements. Finally, the average salary of CDC staff was lower than that of local civil servants. 45 Therefore, CDC positions became less attractive, which led to the net loss of CDC staff over time. 20 The workforce is the foundation of public health infrastructure, 46 meaning that any shortage or instability prevents adequate provision of PHSs. Efforts should be made to both retain current CDC staff and recruit new staff through both financial and nonfinancial incentives (e.g., increased salaries, continuing education, career development). 47,48 Furthermore, CDC directors should better use the media to communicate the value of CDCs to the public, because this would help improve recognition of CDCs undertakings. Third, PHS provision has not yet achieved 100% completeness, especially for disease and health risk monitoring, which was the lowest among the 7 core PHSs in 2012. We propose 2 explanations for the inadequate service provision. First, the interagency cooperation between CDCs and other government departments made some CDC functions overlap with others, resulting in mutual evasion of responsibilities and hence low task performance. Second, because of insufficient general operation funding for health risk monitoring, CDCs tended to conduct passive surveillance of health risks upon institutions request, rather than monitoring them actively. 49 Therefore, CDCs should realize their commitment to monitoring and evaluating health-related risks and establish a mechanism for better communication with other departments regarding the division of work and cooperation. Furthermore, general operation funding should be increased to better allow CDCs to actively monitor the distribution of health risks and thereby establish a risk prevention and intervention system. This system would enable regular assessment of the impacts of risk factors on people s health, which would help in updating and developing effective primary prevention and intervention strategies. 50 Our findings also revealed regional disparities in CDC resource allocation and PHS provision. Despite experiencing the highest growth rate between 2002 and 2012, CDCs in western China had less human, fiscal, and facility resources and fewer completed PHSs than did their eastern and central counterparts. This might be because of their relatively lower levels of economic development and staff salary and unrewarding professional environments. 4 Regional disparities might be reduced through more supportive policies, transfer payments from the Chinese central government, greater staff salaries and special subsidies, and better operational management of the agencies. Limitations Our study has several limitations. First, we did not adopt indicators of health outcomes (e.g., incidence of infectious diseases), which could help us obtain a more comprehensive performance evaluation of the CDC system and examine the relationships between structures, processes, and outcomes. Second, we focused on trends in resource allocation and PHS provision using a pre post comparative analysis, but we did not consider the possible influence of other factors, such as the rapid increase in economic development and social progress. Third, data were self-reported and were therefore subject to reporting bias; this could be especially problematic for indicators of service delivery. Nevertheless, we attempted to ensure reliability by adopting quality control measures (e.g., logical error correction and field verification by expert groups). Public Health Implications Our analysis provides up-to-date information on how China s CDC system is developing and will help to prepare for tracking changes in resource allocation and service delivery in the future. It is important for Chinese governments to develop measures that can accurately describe the capacity of the CDC system, which could serve as an evidence-based assessment of the effects 2100 Research Peer Reviewed Li et al. AJPH December 2016, Vol 106, No. 12

of the reform programs and help researchers analyze the causes of the success or failure of the reforms. 42 China s CDC system has shown remarkable improvements in resource allocation and service delivery. This progress highlights the importance of proper financial investment and policies for the development of the CDC system. The governments should continue to progress, especially in establishing a sound financing mechanism and retaining an adequately sized, highly skilled workforce to improve the delivery of PHSs. CONTRIBUTORS C. Li provided administrative and material assistance. C. Li, M. Sun, J. Yu, and M. Hao acquired study funding. C. Li, M. Sun, Y. Wang, and L. Luo drafted the article and analyzed the data. C. Li, M. Sun, Y. Wang, L. Luo, M. Yu, Y. Zhang, H. Wang, P. Shi, Z. Chen, J. Yu, and M. Hao designed and conceptualized the study. C. Li, M. Sun, Y. Wang, L. Luo, M. Yu, Y. Zhang, H. Wang, P. Shi, Z. Chen, J. Wang, Y. Lu, Q. Li, X. Wang, Z. Bi, M. Fan, and L. Fu interpreted the data. C. Li, M. Sun, Y. Wang, L. Luo, M. Yu, Y. Zhang, H. Wang, P. Shi, Z. Chen, J. Wang, Y. Lu, Q. Li, X. Wang, Z. Bi, M. Fan, L. Fu, J. Yu, and M. Hao acquired the data. J. Yu and M. Hao revised the article and supervised the study. All authors approved the final version of the article. ACKNOWLEDGMENTS This study was funded by the Program for Changjiang Scholars and the Innovative Research TeamProgram from theministry ofeducation ofchina (grants IRT13R11and IRT0912); the National Natural Science Foundation of China (grants 70733002, 71373004, 71303058); the Program of National Social Science Fund of China (grant 13AZD081); and the Fourth Round of the Three-Year Public Health Action Plan in Shanghai 2015 2017 (grant GWIV-32). We would like to acknowledge the Bureau of Disease Control and Prevention of the National Health and Family Planning Commission of China, the provincial departments of health, the provincial Centers for Disease Control and Prevention (CDCs), and the sampling CDCs for their data collection support. HUMAN PARTICIPANT PROTECTION The study was approved by the former Ministry of Health in China and reviewed by the Medical Research Ethics Committee at the School of Public Health of Fudan University. REFERENCES 1. Research Group on Disease Control and Prevention System of China. Research Report on the Problems and Strategies of Development of Disease Control and Prevention System of China. Beijing, China: People s Medical Publishing House; 2006. 2. Ministry of Health of the People s Republicof China. Notice on printing and distributing two documents: opinions on implementing Health Inspection System reform and directory opinions on Disease Control and Prevention System reform. Available at: http://law.lawtime.cn/d613524618618_1_p3.html. Accessed March 8, 2012. 3. Ministry of Health of the People s Republic of China. Notice on printing and distributing document of essential public health services for Centers for Disease Control and Prevention, the national criteria for performance evaluation of disease control and prevention. Available at: http://www.nhfpc.gov.cn/jkj/s7914g/ 200812/a9949418dac742239caac4332cbc3275.shtml. Accessed December 1, 2008. 4. Institute of Medicine. The Future of Public Health. Washington, DC: National Academy Press; 1988. 5. Public Health Functions Steering Committee. Public Health in America. Washington, DC: US Public Health Services; 1994. 6. China: Issues and Options in Health Financing. Washington, DC: World Bank; 1996. 7. State Council of the People s Republic of China. Report on policy issues of health care reform. Available at: http://www.china.com.cn/law/flfg/txt/2006-08/08/ content_7060220.htm. Accessed August 8, 2006. 8. China National Health Economics Institute. China National Health Accounts Report. Beijing: China National Health Economics Institute; 2005. 9. Blumenthal D, Hsiao W. Privatization and its discontents the evolving Chinese health care system. NEnglJMed. 2005;353(11):1165 1170. 10. Schwartz J, Evans G, Greenberg S. Evolution of health provision in pre-sars China: the changing nature of disease prevention. China Rev. 2007;7(1):81 104. 11. Yu JJ, Yu MZ, Duan Y, Gong XG. The problem of public health service delivery insufficiency of disease prevention and control system of China. Wei Sheng Yan Jiu. 2005;34(1):8 9. 12. Wang L, Liu J, Chin DP. Progress in tuberculosis control and the evolving public-health system in China. Lancet. 2007;369(9562):691 696. 13. Overby J, Rayburn M, Hammond K, Wyld DC. The China syndrome: the impact of the SARS epidemic in Southeast Asia. Asia Pac J Mark Log. 2004;16(1):69 94. 14. The conference on prevention and control of SARS was held in Beijing, China. Available at: http://www. gov.cn/test/2005-06/28/content_10715.htm. Accessed June 28, 2005. 15. Ministry of Health of the People s Republic of China. Provision on the Development of Disease Control and Prevention System (No. 40 Ministerial Decree of MOH). Available at: http://www.gov.cn/gongbao/ content/2005/content_75219.htm. Accessed January 5, 2005. 16. Ministry of Housing and Urban Rural Development; National Development and Reform Commission of the People s Republic of China. Construction standards for Centers for Disease Control and Prevention. Available at: http://wenku.baidu.com/link? url=mhyxcecwxn2spyp_jcmu36mtms1kut- Az8Bm5bsDqG8JGsZW0GjRa-iPLa-VqHHLf3u SH86vQ6W4dHU5LkcUNxOt6BUXVsUFMzoD5_ lxluw. Accessed August 4, 2015. 17. Bureau of Disease Control and Prevention of National Health and Family Planning Commission of the People s Republic of China. Report on progress of the Disease Control and Prevention system of China. Available at: http://www.moh.gov.cn/jkj/s7915v/201504/ d5f3f871e02e4d6e912def7ced719353.shtml. Accessed April 1, 2015. 18. Jiang Y, Yang XL. Human resources situation and adaptability study on its functions in the disease control and prevention system of Jiangxi province, China. Chin Health Resour. 2013;16(5):362 366. 19. Wang ZF. Study on the Status and Acceptability of Performance Evaluation of Disease Prevention and Control in Linyi City, China [MPH dissertation]. Shandong, China: Shandong University; 2013. 20. Jing YM. Study on the Allocation Fairness and Prediction of the Health Human Resource in CDC [MPH dissertation]. Shandong, China: Shandong University; 2013. 21. Chen H, Wang L. Investigation on current situation of management and operation mechanism of Centers for Disease Control and Prevention in China. Chin J Public Health Manage. 2013;29(1):27 29. 22. Yu JJ, Yu MZ, Su HJ, et al. A brief introduction to the study concerning constructing disease prevention and control system of China. Chin Health Resour. 2006;9(3): 115 117. 23. Regan S, MacDonald M, Allan DE, Martin C, Peroff- Johnston N. Public health human resources: a comparative analysis of policy documents in two Canadian provinces. Hum Resour Health. 2014;12:13. 24. Freedman AM, Kuester SA, Jernigan J. Evaluating public health resources: what happens when funding disappears. Prev Chronic Dis. 2013;10:E190. 25. Ministry of Health. China s Health Statistical Yearbook 2003. Beijing, China: Peking Union Medical College Press; 2003. 26. Ministry of Health. China s Health Statistical Yearbook 2013. Beijing, China: Peking Union Medical College Press; 2013. 27. Handler A, Issel M, Turnock B. A conceptual framework to measure performance of the public health system. Am J Public Health. 2001;91(8): 1235 1239. 28. Donabedian A. Explorations in Quality Assessment and Monitoring: The Definition of Quality and Approaches to its Assessment. Ann Arbor, MI: Health Administration Press; 1980. 29. Roberts MJ, Hsiao W, Berman P, Reich MR. Getting Health Reform Right: A Guide to Improving Performance and Equity. Oxford, UK: Oxford University Press; 2003. 30. Yao L, Chen E, Chen Z, Gong Z. From SARS to H7N9: the mechanism of responding to emerging communicable diseases has made great progress in China. Biosci Trends. 2013;7(6):290 293. 31. China s organization regarding avian influence was praisedbytheworldhealthorganizationasagoodmodel. 2013. Available at: http://www.moh.gov.cn/wsb/pwsyw/ 201304/12a5a6c8bbca4a9785a4418bc6ce14e9.shtml. Accessed April 25, 2013. 32. Liu Y. China s public health-care system: facing the challenges. Bull World Health Organ. 2004;82(7): 532 538. 33. Strengthening Essential Public Health Functions. Washington, DC: World Bank; 2004. 34. Liu Y, Rao K, Hu S. People s Republic of China: Toward Establishing a Rural Health Protection System. Beijing, China: Asian Development Bank; 2002. 35. Zhang XL, Ding F, Li Q. The analysis on the human resource of the Centers for Disease Control and Prevention in China. Chin J Prev Med. 2012;13(5):399 402. 36. Wang WC, Yu JJ, Yu MZ, et al. Reform steps of reestablishing disease prevention and control system of China. Wei Sheng Yan Jiu. 2005;34(2):130 132. December 2016, Vol 106, No. 12 AJPH Li et al. Peer Reviewed Research 2101

37. Griffiths SM, Tang JL. Healthcare reform in China and the challenges for public health education. Public Health. 2011;125(1):3 5. 38. Beaglehole R, Dal Poz MR. Public health workforce: challenges and policy issues. Hum Resour Health. 2003; 1(1):4. 39. Ma JQ, Yang GH, Shi XM. Information technology platform in China s disease surveillance system. Dis Surveill. 2006;21(1):1 3. 40. Wang L, Wang Y, Jin S, et al. Emergence and control of infectious diseases in China. Lancet. 2008;372(9649): 1598 1605. 41. Jin SG. Development of public health information in China. China Med Device Inform. 2010;16(3):9 18, 36. 42. Yip WC, Hsiao WC, Chen W, Hu S, Ma J, Maynard A. Early appraisal of China s huge and complex healthcare reforms. Lancet. 2012;379(9818):833 842. 43. Yang G, Kong L, Zhao W, et al. Emergence of chronic non-communicable diseases in China. Lancet. 2008; 372(9650):1697 1705. 44. Ding J, Li M, Zhang S, et al. Self-orientation and career expectation of public health professionals in Shanghai. Chin Health Resour. 2007;10(1):47 48. 45. Chen H, Wang L. Research on the implementation of performance income of Centers for Disease Control and Prevention in provincial and sub-provincial level cities. Chin J Public Health Manage. 2014;30(2):189 190. 46. Lichtveld MY, Cioffi JP. Public health workforce development: progress, challenges, and opportunities. J Public Health Manag Pract. 2003;9(6):443 450. 47. Huselid M. The impact of human resource management practices on turnover, productivity, and corporate financial performance. Acad Manage J. 1995;38(3): 635 672. 48. Dean HD, Myles RL, Spears-Jones C, Bishop-Cline A, Fenton KA. A strategic approach to public health workforce development and capacity building. Am J Prev Med. 2014;47(5 suppl 3):S288 S296. 49. Tang ZZ, Zhong GM, Lu W. Investigation of the work on health risk monitoring in county-level Centers for Disease Control and Prevention in Guangxi, China. Chin J Public Health Manage. 2008;24(4):457 461. 50. The World Health Report 2013: Research for Universal Health Coverage. Geneva, Switzerland: World Health Organization; 2013. 2102 Research Peer Reviewed Li et al. AJPH December 2016, Vol 106, No. 12