For Peer Review Only

BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

Should iodised and non-iodised salt both be made available in Chinese cities? A cross-sectional survey

ForJournal: peerBMJ Open review only Manuscript ID: bmjopen-2014-005397

Article Type: Research

Date Submitted by the Author: 04-Apr-2014

Complete List of Authors: yu, JUN; Institute for Iodine Deficiency Disorders, Center for endemic disease control, Chines CDC, Harbin medical university LIU, Peng; Institute for Iodine Deficiency Disorders, Center for endemic disease control, Chines CDC, Harbin medical university LIU, YING; Institute for Iodine Deficiency Disorders, Center for endemic disease control, Chines CDC, Harbin medical university LIU, SHOUJUN; Institute for Iodine Deficiency Disorders, Center for endemic disease control, Chines CDC, Harbin medical university SUN, DIANJUN; Center for endemic disease control, Chines CDC, Harbin medical university

Primary Subject Health policy Heading: http://bmjopen.bmj.com/ Secondary Subject Heading: Epidemiology, Health policy, Public health

telephone interview, iodized salt, non-iodized salt, knowledge-attitude- Keywords: practice, iodine deficiency disorders

Note: The following files were submitted by the author for peer review, but cannot be converted to PDF. You must view these files (e.g. movies) online. on September 30, 2021 by guest. Protected copyright. STROBE_checklist_BMJ-Open_cross-sectional-studies.doc

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 26 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 Title page 4 5 Should iodised and non-iodised salt both be made available in 6 7 Chinese cities? A cross-sectional survey 8 9 10 † † 11 Jun Y , Peng L , Ying L, Shou-jun L, Dian-jun S* 12 13 Corresponding author 14 15 Name: Dian-junFor SUNpeer review only 16 17 Degree: M.D. 18 19 Position: Chief of Center for Endemic Disease Control 20 Affiliation: Center for Endemic Disease Control, Chinese Center for 21 22 Disease Control and Prevention, Harbin Medical University, Harbin 23 24 150081, People’s Republic of China 25 26 Full address: No.157 Baojian Road, Nan’gang District, Center for 27 28 Endemic Disease Control, Harbin Medical University, Harbin, China 29 30 Postal code: 150081 31 32 Telephone: 86-451-86675814

33 http://bmjopen.bmj.com/ 34 Fax: 86-451-86657674 35 36 Email: [email protected] 37 38 39 40 First Author

41 † † on September 30, 2021 by guest. Protected copyright. 42 Name: Jun YU , Peng LIU 43 † 44 have made the same contribution 45 46 Other Author Name:Ying LIU, Shou-jun LIU 47 48 Affiliation: Center for Endemic Disease Control, Chinese Center for 49 50 Disease Control and Prevention, Harbin Medical University, Harbin 51 52 150081, People’s Republic of China 53 54 Keywords: telephone interview, iodized salt, non-iodized salt, 55 knowledge-attitude-practice, iodine deficiency disorders 56 57 Word count: 3054 words 58 59 60 1 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 26 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 Shall we supply iodized or non-iodized salt simultaneously? 4 5 Chinese National telephone interview survey 6 7 ABSTRACT 8 9 Objective: To contribute evidence relevant to the policy of supplying 10 11 iodised salt (IS), non Iodised salt NIS) or both, in Chinese cities. 12 13 14 15 Design:For National telephonepeer interview review survey. only 16 17 18 19 Setting: 17 capital cities and 6 coastal cities of 17 iodine deficiency 20 21 disorder (IDD)-eliminated provinces (municipalities) in China. 22 23 24 25 Participants: 24 557 numbers were dialled. Telephone numbers were 26 27 randomly selected by random digit dialing and Mitofsky-Waksberg two 28 29 stages sampling method used. 30 31 32 Results: Among 4833 citizens who accepted the telephone interview, 33 http://bmjopen.bmj.com/ 34 35 3738 (77.3%) citizens chose IS, 481 (10.0%) citizens chose NIS, others 36 37 chose both IS and NIS. Citizens’ awareness rate of IDD and IDD 38 preventive measures were 68.7% and 62.5%, respectively. 39 40

41 on September 30, 2021 by guest. Protected copyright. 42 Conclusions: It was not a suitable time to supply IS and NIS 43 44 simultaneously in the developed cities of China, whereas a pilot project 45 46 may be carried out in the cities where IDD has been sustainably 47 48 eliminated, there is strong awareness of IDD and the population can 49 50 choose IS accurately. IDD health education should be further 51 52 strengthened, especially on its IQ damage. 53 54 55 Keywords: telephone interview, iodized salt, non-iodized salt, 56 57 knowledge-attitude-practice, iodine deficiency disorders 58 59 60 2 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 26 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 Strengths and limitations of this study 12 13 14 The 77.3% population of China developed cities chose iodized rather 15 For peer review only 16 than non-iodized salt. 17 18 19 The 68.7% population of China developed cities knew IDD and 20 21 65.2% knew the preventive measure. 22 23 24 It was not a suitable time to supply IS and NIS simultaneously in the 25 26 developed cities of China, whereas a pilot project may be carried out 27 28 29 in some cities. 30 31 32 33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 30, 2021 by guest. Protected copyright. 42 43 44 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 BMJ Open Page 4 of 26 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 1 Shall we supply iodized or non-iodized salt simultaneously? 4 5 2 Chinese National telephone interview survey 6 7 3 INTRODUCTION 8 9 4 China has about 20% population of the world, and is one of the most 10 11 5 severe iodine deficiency (external environment) countries. The Iodine 12 13 6 Deficiency Disorders (IDD) was once widespread and caused great 14 15 7 damages.For As an effectivepeer measure, review the Universal Saltonly Iodization(USI) has 16 17 8 been adopted for IDD prevention and control and has made great 18 19 9 achievement since 1994. By 2000, 17 provinces (municipalities) of China 20 21 10 had reached the goal of IDD elimination, and by far, have sustained for 22 [1] 23 11 10 years according to IDD eliminating criteria by WHO. 24 25 12 However, along with the discovering of iodine excess and its 26 [2-4] 27 13 side-effects, the increasing rate of thyroid disease, the raising of 28 29 14 residents’ health awareness, the debate of informed choices on choosing 30 31 15 IS or NIS became acute in China, ad hoc in coastal areas. As we know, 32 16 some developed European countries such as Belgium, Finland, France, 33 http://bmjopen.bmj.com/ 34 35 17 Germany, Greece, Ireland, Italy and Netherlands, etc. supply IS and NIS 36 [5] 18 37 simultaneously in the market. Their residents can purchase IS or NIS 38 19 voluntarily. As regards to China, for developed cities in IDD sustainable 39 40 20 eliminated provinces, is it the right time for supplying IS and NIS

41 on September 30, 2021 by guest. Protected copyright. 42 21 simultaneously? If so, can their residents still be free of IDD? And is 43 44 22 there a proper method to demarcate cities of supplying IS and NIS 45 46 23 simultaneously or only supplying IS? To answer these questions, the 47 48 24 status of residents’ subjective preference on choosing IS or NIS and 49 50 25 relevant influencing factors should be made clear first. 51 52 26 Door-to-door interview is usually used to conduct such investigations. 53 54 27 Considering the expenditure of time and money, a cross-sectional 55 56 28 telephone interview is a feasible alternative，it is a widespread method 57 58 59 60 4 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 26 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 [6] 3 29 adopted rather earlier and used by different countries all over the world. 4 5 30 Successful cases in the past in China were KAP surveys on influenza, 6 [7-10] 7 31 smoke and pollinosis, etc. Its application bases on conditions of high 8 9 32 social economic level and telephone popularization rate. 10 11 33 In China, by the end of 2009, the national average Urban Per Capita 12 13 34 Disposable Income was 17,175 RMBY, and some developed cities 14 15 35 (municipalities)For inpeer the IDD eliminatedreview provinces only were all near even 16 [11] 17 36 beyond this(table 1). Besides, the national fixed line penetration had 18 19 37 increased to 314 million (23.5 telephones/100 persons), and in cities, 212 20 21 38 million (34.1 telephones/100 persons). Facts above made the telephone 22 23 39 interview a good representative and feasible way to research on residents’ 24 25 40 KAP on IDD. 26 41 27 As we know, there were still no such records on residents’ subjective 28 42 preference between IS and NIS, and no research on residents’ awareness 29 30 43 of IDD in developed cities (municipalities) in recent years. Hence, the 31 32 44 national Center for Endemic Disease Control (CEDC) launched the

33 http://bmjopen.bmj.com/ 34 45 project in April, 2010, aimed to obtain data on above aspects to decide 35 36 46 whether IS and NIS supply simultaneously was feasible for developed 37 38 47 cities in IDD eliminated provinces. The method and results may provide 39 40 48 evidence or references for the similar regions in the world.

41 on September 30, 2021 by guest. Protected copyright. 42 49 METHODS 43 44 50 45 46 51 City sites to be investigated 47 48 52 The 17 capital cities of the 17 provinces (3 municipalities, 6 coastal 49 50 53 provinces and 8 inland provinces) which had achieved the goal of IDD 51 52 54 elimination in 2000 were selected (table 1), 1 coastal city was also 53 54 55 selected in each above coastal province. Total 23 cities were chosen. 55 56 56 57 58 57 59 60 5 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 26 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 58 4 5 59 Table 1 Provinces and cities surveyed 6 7 Provincial capital Coastal cities 8 Category Provinces Name Tel* Income** Name Tel * Income ** 9 10 Hebei Shijiazhuang - 16,607 Tangshan 24.58 18,053 11 12 Shandong Ji’nan 39.28 22,272 Yantai 28.27 21,125 13 Jiangsu Nanjing 38.91 25,504 Nantong 35.80 21,001 14 Coastal province Zhoushan 53.45 24,082 15 ForZhejiang peer Hangzhou review 48.23 26,864 only 16 Guangdong Guangzhou 61.28 27,610 Shenzhen 52.64 29,244 17 Beihai - 15,134 18 Guangxi Nanning 16.61 16,254 19 Heilongjiang Harbin 21.66 15,887 20 21 Jilin Changchun - 16,277 22 Shanxi Taiyuan 44.31 15,607 23 24 Inland He’nan Zhengzhou 29.28 17,117 25 province Anhui Hefei 32.47 17,158 26 27 Hubei Wuhan 40.57 18,385 28 Jiangxi Nanchang 32.37 16,472 29 30 Hu’nan Changsha 33.28 20,238 31 Beijing Beijing 50.89 26,738 32

33 Municipalities Shanghai Shanghai 48.70 28,838 http://bmjopen.bmj.com/ 34 35 Tianjin Tianjin 31.36 21,430 36 60 All “*” and “**” data were from Statistics Bulletin of the Cities’ Economic and Social 37 38 61 Development of 2009, “*” fix telephones/100 persons, “**” Urban per capita disposable income 39 62 (RMBY), National fixed line penetration was 23.50 telephones/100 persons and Urban per capita 40 63 disposable income was 17,175 RMBY of 2009 for reference.

41 on September 30, 2021 by guest. Protected copyright. 42 43 64 44 65 45 46 66 Target population 47 48 67 Citizens had fixed telephone in the 23 selected cities. 49 50 68 The main contents of questionnaire 51 52 69 If IS and NIS are supplied simultaneously, which one will you choose? 53 54 70 And why? 55 56 71 Have you heard of IDD? If you have heard, what is its damage? And 57 58 72 how to prevent IDD? 59 60 6 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 26 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 73 Can you tell us your occupation and age group? 4 5 74 Sampling method and Sampling size 6 7 75 Random Digit Dialing and Mitofsky-Waksberg two stages sampling 8 [12] 9 76 method were used in this study. In each city, only the main urban zones 10 11 77 were investigated, and two districts were selected randomly from each 12 13 78 main urban zone. At least 200 effective samples were sampled in each 14 15 79 city in theFor survey. peerQuota of samples review was divided intoonly each district evenly. 16 17 80 Sampling size was calculated according to the formula 1, the positive rate 18 19 81 (π ) was about 70% according to the pilot survey, and the allowable 20 21 82 error(δ ) was 0.5, the sampling size calculated should be 170 per city, so 22 23 83 200 effective samples should be enough. 24  2 (1−× ππ ) 25 84 n =  α  π Positive rate δ Allowable error 26  δ  27 28 85 Formula 1 29 30 86 Telephone number acquired 31 87 32 If the fixed telephone of the neighborhood offices in the selected district

33 http://bmjopen.bmj.com/ 88 was 7 or 8 digit numbers, the first 3 or 4 digits were deemed as direction 34 35 89 code, respectively. The last 4 digits were obtained by Random function 36 ® 37 90 (=Rand ()×10000) in the Microsoft Office Excel . If the function gave 38 39 91 randomized numbers less than 4 digits, “0” was supplemented in the front. 40

41 92 The size of telephone number created randomly should be at least 5 times on September 30, 2021 by guest. Protected copyright. 42 43 93 of the effective sample size, in another word, the created number should 44 45 94 be at least 1000 in a city. 46 47 95 Quality control 48 49 96 The survey program and survey result database were designed and 50 51 97 delivered to the project cities by the national CEDC. This survey program 52 53 98 was improved based on a pilot survey, and the questionnaire had clear 54 55 99 questions and simple answers, avoiding hints on the residents. Sensitive 56 57 100 questions, such as age, were designed as “Which age group do you 58 59 60 7 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 26 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 101 belong to? <20, 20-30, 30-40... ”. 4 5 102 In the program, the way of carrying out the investigation was unified, 6 7 103 including the order and the pattern of asking question, the dialing time of 8 9 104 investigation(18:30 - 21:00) , the instruction of making answer note and 10 11 105 filling database, and so on. 12 13 106 To avoid the bias of investigation and report, and to improve the research 14 15 107 objects’For cooperation, peer the interviewers review were trained only on the attitude, the 16 17 108 mood, and the countermeasures for emergency by the project provinces 18 19 109 themselves according to the rule of survey program. For example, some 20 21 110 related questions put forward by residents could be answered only after 22 23 111 the interview finished for avoiding the bias. 24 25 112 When a project city finished telephone interview, to guarantee the 26 113 27 reliability of the research information, a return visit was conducted by the 28 114 quality control officer of its provincial CDC or CEDC among the citizens 29 30 115 who had accepted the interview. At least 10 percent of them were 31 32 116 randomly sampled to answer the return visit questionnaire, when the

33 http://bmjopen.bmj.com/ 34 117 coincidence rate of survey information was greater than 80%, the 35 36 118 investigation results of the city can be regarded as qualified, otherwise, 37 38 119 the city should be reinvestigated. 39 40 120 Definition of telephone number

41 on September 30, 2021 by guest. Protected copyright. 42 121 Success: respondent finished the entire questionnaire; Rejection: 43 44 122 respondent refused to answer or unfinished; No answer: the phone is 45 46 123 busy, re-dials after 5 minute, still busy; Invalid number: the phone 47 48 124 number is vacant, fax, switchboard or phone quit. 49 50 125 Data processing and statistical analysis 51 ® ® 52 126 The Epi info 3.5.1 was used for database design and SPSS 17.0 for 53 54 127 consistency check and chique analysis. The level of significance test is 55 56 128 0.05. 57 58 129 59 60 8 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 26 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 130 RESULTS 4 5 131 Response situation 6 7 132 The flow chart see figure 1. The survey investigated 235 districts of 120 8 9 133 main urban zones in 23 cities totally, 446 direction numbers were chosen 10 11 134 and 24,557 telephone numbers were dialed, including Success 4,833; 12 13 135 Rejection 1,949; No answer 6,433; Invalid number 11,342. Among them, 14 15 136 valid numberFor was peer 13,215, accounted review for 53.8% only of the total dialed. 16 17 137 Answered number (success and rejection) was 6,782, accounted for 18 19 138 51.3% of the valid, and the total response rate was 71.3% (figure 2). 20 21 139 Basic information of the interviewee 22 23 140 Residents who answered the age and occupation questions were 4,861 24 25 141 and 4,857, respectively. The age and occupation proportion were similar 26 142 27 between capital city (including Beijing) and coastal city (including 28 143 Tianjin and Shanghai) residents. Residents’ age mainly distributed 29 30 144 between 20 and 60, accounted for 80%. However, when compared with 31 32 145 the population composition from the ‘2009 China Health Statistical

33 http://bmjopen.bmj.com/ [13] [14] 34 146 Yearbook’ and Chinese population pyramid, the 10-20 age group had 35 36 147 a less constituent ratio. This may lead to bad representativeness for the 37 38 148 whole population, whereas, may be good representativeness for this 39 40 149 survey, because the children rarely buy salt. Regarding the occupation,

41 on September 30, 2021 by guest. Protected copyright. 42 150 the residents distributed evenly in each profession, accounted for about 43 44 151 10%, except serviceman, students and irregular profession for their small 45 46 152 sample size. 47 48 153 49 50 154 Proportion of choosing IS or NIS 51 52 155 Of the residents investigated, 4,865 answered the question ‘If IS and NIS 53 54 156 are supplied simultaneously, which one will you choose?’ The total ratio 55 56 157 of ‘choosing IS’ was 76.8%, ratio of ‘choosing NIS’ was 9.9%, ‘both’, 57 58 158 3.9% and‘does not matter’, 9.3%. The capital cities’ IS choosing rate 59 60 9 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 26 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 159 (79.5%) was higher than that of coastal cities (72.0%), for the NIS 4 5 160 choosing rate was opposite (6.7%, 15.7%), and the ratio of choosing 6 7 161 “both”(<5%) and ‘does not matter’(<10%) were nearly the same in the 8 9 162 two areas. In the 23 cities, there were 12 cities whose IS choosing rate 10 11 163 was higher than 80%, 8 cities between 60-80%, 3 cities (Shanghai: 32.8%, 12 13 164 Zhoushan: 42.8% and Hangzhou:49.3%) lower than 50% (table2, table 3). 14 15 165 After groupingFor thepeer data by agereview and occupation, only it was found that in 16 17 166 different age and occupation residents, their choosing rate had merely 18 19 167 tiny difference, except the service group for their small sample size. 20 21 168 Reasons for “choosing IS” 22 23 169 For 3,738 residents who choosing IS, their reasons were as follows: 24 25 170 61.9% of them accurately understood the benefits of IS, i.e. preventing 26 171 27 IDD and increasing children’s Intelligence Quotient (IQ), 14.7% of them 28 172 merely chose IS for good quality and they know little about the 29 30 173 relationship between IS and IDD prevention, 27.7% had no clear reason, 31 32 174 most of them explained as a custom. (table 2)

33 http://bmjopen.bmj.com/ 34 175 35 36 176 37 38 177 39 40 178

41 on September 30, 2021 by guest. Protected copyright. 42 179 43 44 180 45 46 181 47 48 182 49 50 183 51 52 184 53 54 185 55 56 186 57 58 187 59 60 10 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 26 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 188 4 5 189 Table 2 Reasons for buying IS of survey residents（%） 6 7 Reasons Number 8 IS of Benefit 9 Province City choosing choosing Prevent for Good 10 rate Others 11 IS IDD children’s quality 12 IQ 13 Heilongjiang Harbin 183 91.5 36.6 3.8 1.6 60.1 14 Jilin Changchun 175 86.6 48.6 9.7 6.3 0 15 BeijingFor - peer 202 review 86.7 61.4 only 12.9 9.4 28.2 16 Hebei Shijiazhuang 174 87.0 67.8 7.5 12.1 14.4 Shanxi Taiyuan 189 95.0 13.8 0 14.3 70.9 17 Shandong Ji’nan 187 93.5 73.3 27.3 20.3 11.2 18 He’nan Zhengzhou 177 88.5 54.8 15.3 26.6 27.7 19 Jiangsu Nanjing 180 72.6 64.4 25.0 23.9 7.8 20 Anhui Hefei 178 84.8 32.0 0.6 5.6 61.8 21 Hubei Wuhan 163 75.1 41.7 8.0 24.5 29.4 22 Zhejiang Hangzhou 66 32.8 57.6 16.7 21.2 0 23 Jiangxi Nanchang 184 86.8 52.2 7.6 13.6 8.7 24 Hu’nan Changsha 146 72.3 56.2 6.8 5.5 39.7 25 Guangdong Guangzhou 124 61.7 39.5 1.6 46.0 3.2 26 Guangxi Nanning 170 78.0 38.8 4.1 6.5 46.5 27 Capital city total 2498 79.5 9.8 15.0 29.0 28 Hebei Tangshan 208 94.1 81.7 17.3 9.6 5.8 29 Tianjin - 200 83.7 75.5 28.5 27.0 6.5 30 Shandong Yantai 186 93.0 90.3 3.2 8.1 4.8 31 Jiangsu Nantong 153 73.9 46.4 5.9 2.6 47.7 32 Shanghai - 108 49.3 25.9 6.5 29.6 48.1

33 Zhejiang Zhoushan 86 42.8 68.6 12.8 18.6 0 http://bmjopen.bmj.com/ 34 Guangdong Shenzhen 144 71.6 22.9 0.7 3.5 42.4 35 Guangxi Beihai 155 66.2 21.9 2.6 18.7 58.1 36 Coastal city total 1240 72.0 10.6 14.1 25.0 37 Total 3738 51.9 76.8 14.7 27.7 38 190 39 40 191

41 on September 30, 2021 by guest. Protected copyright. 42 192 Reasons for “choosing NIS” 43 44 193 Most of the 481 residents who chose NIS lived in Shanghai, Hangzhou, 45 46 194 Zhoushan and Beihai. Majority of them considered themselves had taken 47 48 195 enough iodine from food or not lived in the IDD areas. Unexpectedly, 49 50 196 10% of them thought IS was harmful (table 3). 51 52 197 53 54 198 55 56 199 57 58 200 59 60 11 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 26 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 201 4 5 202 Table 3 Reasons for choosing NIS of survey residents（%） 6 7 Reasons 8 Not IS 9 City n % Iodine IS is 10 IDD Patient* bad Cheap Others sufficient harmful 11 endemia taste 12 Capital city Total 211 6.7 6.7 13.3 9.0 10.5 2.9 1.4 23.3 13 14 Coastal city Total 270 15.7 15.7 17.7 5.2 8.1 4.0 1.5 29.9 15 TotalFor 481 peer 9.9 9.9 review16.6 7.3 9.1only 3.5 1.5 27.0 16 203 The “*” meant patients of thyroid disease unfeasible taken IS. Partial provinces had small sample sizes; the 17 204 proportions were referred for large ones; n : Number of choosing NIS; %: NIS choosing rate. 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 30, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 12 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 26 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 205 Reasons for “choosing both” 4 5 206 Most of the 192 residents who chose “both IS and NIS” was from 6 7 207 Hangzhou, Guangzhou and Shanghai. The two major reasons were to 8 9 208 keep appropriate iodine intake and to avoid the harmful effect of excess 10 11 209 iodine, accounted for 27.6% and 30.7%, respectively. 12 13 210 Reasons for “does not matter” 14 15 211 The 454For residents peer chose “does review not matter”, mainly only lived in Nanjing, 16 17 212 Wuhan, Hangzhou, Changsha, Guangzhou, Nanning, Tianjin, Shanghai 18 19 213 and Zhoushan. They tended to choose the IS or NIS basing on which one 20 21 214 is convenient to buy, accounted for 52.0%. 22 23 215 Awareness rate of IDD 24 25 216 Of the residents responded to our phone call, 4,865 answered the question 26 217 27 “Have you heard of Iodine Deficiency Disorders?” Among them 68.7% 28 218 (3,344) have heard, 31.3% (1,521) never heard. There was little 29 30 219 difference between capital cities (68.2%) and coastal cities (69.8%). The 31 32 220 awareness rate in Yantai (94.5%) was the highest, in four cities’

33 http://bmjopen.bmj.com/ 34 221 (Shijiazhuang, Jinan, Hangzhou and Tangshan) awareness rate was 35 36 222 greater than 80%. Beihai and Taiyuan had the lowest rate of 46.2% and 37 38 223 44.7%, respectively. There were no significant difference between age 39 40 224 groups and occupation groups.

41 on September 30, 2021 by guest. Protected copyright. 42 225 Among 3,344 residents of heard of IDD, up to 77.1% of them knew that 43 44 226 iodine deficiency could lead to goiter, but only 11.8% knew that it could 45 46 227 also cause intelligence damage to children, and even 16.9% of them knew 47 48 228 nothing about its hazard. 49 50 229 Awareness rate of IDD preventive measures 51 52 230 Of the 3,344 residents of heard of IDD, 62.5% of them knew IS could 53 54 231 prevent IDD, 43.2% of them knew that kelp and purple seaweed could 55 56 232 prevent IDD, nearly 18.7% residents knew nothing about how to prevent 57 58 233 IDD. 59 60 13 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 26 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 234 The IS choosing rate and IDD awareness rate in each city 4 5 235 All cities could be divided into four groups, Thigh IS choosing rate + 6 7 236 high IDD awareness rate, like Yantai, Shijiazhuang, Beijing, Tianjin, 8 9 237 Nanjing, Changsha and Guangzhou; Thigh IS choosing rate + low IDD 10 11 238 awareness rate, like Taiyuan, Harbin, Nanchang, Hefei, Wuhan and 12 13 239 Beihai; Tapproximate between choosing rate and awareness rate, like 14 15 240 Tangshan,For Ji’nan, peer Zhengzhou, review Changchun, Nanning, only Nantong and 16 17 241 Shenzhen; ④low IS choosing rate + high IDD awareness rate, like 18 19 242 Shanghai, Zhoushan and Hangzhou. Only in Yantai and Nanjing, the IDD 20 21 243 IDD awareness rates were higher than IS choosing rate. The IS choosing 22 23 244 rate of Harbin and Taiyuan was greater than 90%, but their IDD 24 25 245 awareness rate were only 57.0% and 44.7%, respectively. There were 26 27 246 only 4 cities whose IS rate and awareness rate were all greater than 80% 28 29 247 (figure 3). 30 31 248 The chi square was used to analyze the influence of IDD awareness 32 249 situation on IS choosing rate, the result showed no difference between 33 http://bmjopen.bmj.com/ 34 250 35 capital cities’ IS and NIS group, but significant difference between 36 251 coastal cities’ IS and NIS group. This could be explained as for capital 37 38 252 cities, there were weak relations between IS choosing rate and IDD 39 40 253 awareness rate, whereas, for coastal cities, there were some relations

41 on September 30, 2021 by guest. Protected copyright. 42 254 between them, most coastal residents who choosing NIS had lower IDD 43 44 255 awareness rate. 45 46 256 Result of return visit 47 48 257 The total coherence rate of all questions was 85.1%. The Kappa value of 49 50 258 coherence analysis for choosing IS, Choosing NIS and IDD awareness 51 52 259 were 0.752, 0.693 and 0.76, above required 0.4, described good 53 [15] 54 260 coherence. 55 56 261 57 58 262 DISCUSSIONS 59 60 14 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 26 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 263 Telephone interview is a wide-used method all over the world, which has 4 5 264 the advantages of wide coverage, low costs and efficient. The survey 6 7 265 subjects could be units or personal with fixed telephone, the choice, 8 9 266 supervision and training of investigators are convenient. These features 10 11 267 made it feasible for large scale, clear content and easily acceptable survey. 12 [16] 13 268 However, it does have biases and limitations. Only after detailed design, 14 15 269 decreasedFor the error peer as much as possible,review the results only would be convincible. 16 17 270 For some interview like street intercept or telephone interview, if most of 18 19 271 the interviewees reject the survey and the rejection rate is greater than 20 [17] 21 272 70%(i.e. the response rate <30%), the representation of the random 22 23 273 samples will be questioned, hence, the interviewees will be resample. In 24 25 274 our study, the telephone survey was conducted under a unified quality 26 275 27 control, the response rate was 71.3% and the coherence rate of re-survey 28 276 was above 80% (85.1%), in terms of age and occupation, there were no 29 30 277 significant differences. So the results of this telephone survey could 31 32 278 basically reflected the subjective preference on IS or NIS of the

33 http://bmjopen.bmj.com/ 34 279 developed urban residents in China. As the first telephone survey on KAP 35 36 280 of IDD in China, the traditional PAPI method was taken, and the result 37 38 281 would offer references for the survey followed. 39 40 282 The residents’ total IS choosing rate was 76.8% of 6 coastal cities and 14

41 on September 30, 2021 by guest. Protected copyright. 42 283 capital cities in 14 provinces and 3 municipalities (i.e. Beijing, Tianjin 43 44 284 and Shanghai) where IDD elimination had been sustained for nearly ten 45 46 285 years. The reasons, for choosing IS, were mainly ‘IDD prevention 47 48 286 (51.9%)’ and ‘improving children’s IQ (10%)’, other 38% of the residents 49 50 287 had no idea about the relationship between IS and IDD prevention, their 51 52 288 choice of IS was only because the IS had a good quality or just a kind of 53 54 289 custom, another phenomenon emerged in this study was that most 55 56 290 residents had a lower IDD awareness rate than IS choosing rate, for 57 58 291 example, Taiyuan, Hefei, Harbin and Changchun, although the residents’ 59 60 15 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 26 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 292 IS choosing rates were greater than 80%, their awareness rate of IDD 4 5 293 were low, which reflected that although IDD health education activities, 6 7 294 such as 5.15 IDD Day, was carried out widely and continuously by the 8 9 295 relevant departments vigorously, it was still not enough and exhaustive, 10 11 296 the current propaganda strength and method need further improving. 12 13 297 Considering the supply system of iodized salt is divided by province in 14 15 298 China, Forthe KAP ofpeer IDD of the reviewcity residents is more only important. Though 16 17 299 these provinces were all attained the goal of IDD elimination and their 18 19 300 capital cities and coastal cities were all developed cities, their IS choosing 20 21 301 and IDD awareness situation were different, they should be treated 22 23 302 differently according to the reality. In this study, 12 cities had an IS 24 25 303 choosing rate of greater than 80%, they were all northern cities except 26 304 27 Nanchang. Especially, 5 cities had a rate of greater than 90%. The other 28 305 11 cities with a rate of lower than 80%, were all southern cities. The rate 29 30 306 of Hangzhou, Zhoushan and Shanghai, 32.8%, 42.8% and 49.3%, 31 32 307 respectively, were all lower than 50%. The facts above showed that the IS

33 http://bmjopen.bmj.com/ 34 308 choosing rate of south city residents was lower than that of north city 35 36 309 residents. The main reason was that the people in south coastal areas 37 38 310 think that they have received adequate iodine from food other than 39 40 311 iodized salt, and another reason was that they thought they did not lived

41 on September 30, 2021 by guest. Protected copyright. 42 312 in iodine deficiency area, especially in Hangzhou, Zhoushan and 43 44 313 Shanghai. Their IDD awareness rate was greater than the IS choosing rate, 45 46 314 and some residents chose NIS under the situation of knowing the IDD 47 48 315 damage. In fact, these districts had no apparent IDD prevalence should 49 50 316 attribute to widespread IS supply. In the other survey “iodine nutritional 51 [18] 52 317 status of coastal districts residents in 2009”conducted by MOH, 53 54 318 showed that under the condition of consumption rate of qualified IS 55 56 319 greater than 90%, the Median Urinary Iodine (MUI) of Shanghai 57 58 320 residents was merely between 100µg/L and 200µg/L, just at iodine intake 59 60 16 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 26 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 321 appropriate level, meantime, the MUI of the pregnant women were lower 4 5 322 than 150µg/L, at iodine intake insufficient level. The residents of 6 7 323 Xiacheng district of Hangzhou city, Zhejiang province, the consumption 8 9 324 rate of qualified IS was 96.7%, and the MUI of 8-10-year old children 10 [19] 11 325 was between 100 and 200µg/L, was at iodine intake appropriate level. 12 13 326 If allowing these cities’ residents choose IS or NIS freely, the residents 14 15 327 may chooseFor IS blindlypeer and the review achievement of controlonly IDD had been 16 17 328 obtained will be destroyed. 18 19 329 From the aspect of society and economic development, it is an inevitable 20 21 330 trend of supplying IS and NIS simultaneously for some cities or some 22 23 331 persons needed. According to this telephone interview, it was not a 24 25 332 suitable time to supply IS and NIS simultaneously, even in the developed 26 333 27 cities of China. Intake of non-iodized salt should be cautious in China. 28 334 The only thing we can do now is conducting a trial of supplying IS and 29 30 335 NIS simultaneously in appropriate cities. The selected experiment city 31 32 336 should be a developed city with no IDD prevalence in recent years, and

33 http://bmjopen.bmj.com/ 34 337 the residents’ awareness rate of IDD and choosing rate of IS should be 35 36 338 higher. According to our study, Yantai, Tangshan and Ji’nan may be 37 38 339 appropriate. Before the experiment, an emergency planning should be 39 40 340 formulated, extensive IDD education should be followed. Once some

41 on September 30, 2021 by guest. Protected copyright. 42 341 abnormal situation emerges, the emergency program will be started, or 43 44 342 even stop the trail. 45 46 343 47 48 344 Author Affiliation 49 50 345 Jun Y, Peng L, Ying L and Shou-jun L were affiliated in Institute for 51 52 346 iodine deficiency disorders, Center for Endemic Disease Control, Chinese 53 54 347 Center for Disease Control and Prevention, Harbin Medical University, 55 56 348 Harbin, People’s Republic of China. 57 58 349 Dian-jun S was affiliated in Center for Endemic Disease Control, Chinese 59 60 17 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 26 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 350 Center for Disease Control and Prevention, Harbin Medical University, 4 5 351 Harbin, People’s Republic of China. 6 7 352 8 9 353 Acknowledgment We thank the organizations participated in this 10 11 354 survey: national Center for Endemic Disease Control (CEDC), 16 12 13 355 provincial (municipality) Center for Disease Control (CDC) or CEDC, 14 15 356 includingFor Beijing, peer Tianjin, Shanghai, review Jilin province, only Hebei province, 16 17 357 Shandong province, He’nan province, Jiangsu province, Anhui province, 18 19 358 Hubei province, Zhejiang province, Jiangxi province, Hu’nan province, 20 21 359 Guangdong province and Guangxi province, and 6 coastal city CDC, 22 23 360 including Tangshan, Yantai, Nantong, Zhoushan, Shenzhen and Beihai. 24 25 361 The authors of this article thank those colleagues who had taken part in 26 362 27 the survey from above organizations for their hardworking. 28 363 29 30 364 Contributors Jun Y wrote the project report, Peng L analyzed the data 31 32 365 and wrote the manuscript of the article, Ying L and Shou-jun L designed

33 http://bmjopen.bmj.com/ 34 366 the plan and managed the project, Dian-jun S was responsible for the 35 36 367 project and the publication. 37 38 39 368 Funding The Ministry of Health of China funding the project. 40

41 on September 30, 2021 by guest. Protected copyright. 42 369 Competing interests None. 43 44 45 370 Funding Project was funded by Ministry of Health of China. 46 47 371 Ethics approval The Harbin Medical University. 48 49 372 50 51 373 Reference 52 53 374 [1] WHO/UNICEF/ICCIDD. Assessment of iodine deficiency disorders and monitoring their 54 375 elimination. A guide for programme managers. Third edition, France, 2007. 55 56 376 [2] Todd CH, Allain T, Gomo ZA, et al. Increase in thyrotoxicosis associated with iodine 57 377 supplements in Zimbabwe. Lancet. 1995,346:1563-4. 58 59 60 18 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 26 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 378 [3] Stanbury JB, Ermans AE, Bourdoux P, et al. Iodine-induced hyperthyroidism: occurrence 4 379 and epidemiology. Thyroid 1998, 8:83-100. 5 6 380 [4] Parveen S, Latif SA, Kamal MM, et al. Iodized salt induced thyrotoxicosis: Bangladesh 7 381 perspective, Mymensingh Med J. 2009,18:165-8. 8 9 382 [5] ICCIDD. International Council for the Control of Iodine Deficiency Disorder. CIDDS 10 11 383 Database. Current IDD Status Database.2004. 12 384 [6] Groves RM, Biemer PP, Lyberg LE, et al. Telephone survey methodology [J] John Wiley 13 14 385 &Sons Inc.1989, 2-10. 15 386 [7] JianhuaFor L, Hanwu Mpeer , Yongsheng W,review et al. Analysis of know only ledge, attitude and practice 16 17 387 of the influenza A /H1N1of general and immigration population in Shenzhen, Chinese Health 18 388 Education. 2009,25:915-8. 19 20 389 [8] Jin M, Shunxiang Z, Hanwu M, et al. Telephone survey on behavior risk factors of 21 390 Shenzhen residents among six districts in 200. Journal of Disease Control. 2005,9:561-5 22 23 391 [9] Tao W, Su-jun L, Liang Z, et al. Investigation of Pollinosis in Beijing Residents over the 24 392 Age of 15. J Environ Health, 2008,25:403-4. 25 26 393 [10] Blair G. Obstetricians' Receptiveness to Teen Prenatal Patients Who Are Medicaid 27 394 Recipients, Health Services Research.1997,32:265-82. 28 29 395 [11] National Bureau of Statistics of China.Statistics Bulletin of the Cities’ Economic and 30 396 Social Development of 2009. 31 32 397 [12] Potthoff RF. Some Generazations of the Mitofsky-Waksberg techniques for random

33 http://bmjopen.bmj.com/ 398 digital dialing. Journal of the American Statistical Association, 1987, 82:409-41. 34 35 399 [13] Yi MING. http://geo.cersp.com/sJxzy/sc/200706/2658.html. 36 37 400 [14] http://www.moh.gov.cn/publicfiles/business/htmlfiles/zwgkzt/ptjnj/200908/42635.htm. 38 401 [15] Jiezhen W, Jing H, Yanxun L, et al. Kappa statistic value and its application. [J], Chinese 39 40 402 Health Statistical, 1995,12:46-50.

41 403 [16] Ping W, Weijing D. Telephone Survey and Its Application in Public Health. Health on September 30, 2021 by guest. Protected copyright. 42 43 404 education of China. 2004,20(6):549-50. 44 405 [17] Guohua D. A powerful tool of the statistic survey—the computer aid telephone survey 45 46 406 system. Newspaper of Theory Guidance, 2008,12:51-2. 47 407 [18] Chinese CDC, CEDC. Report of coastal residents’ iodine nutrient level,2009. 48 49 408 [19] ICCIDD Current IDD Status Database,2003. 50 409 [http://www.who.int/vmnis/database/iodine/countries/en/index.html]. 51 52 53 54 55 56 57 58 59 60 19 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 26 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 4 Figure legend 5 6 Figure 1 Flow chart 7 8 Figure 2 The answer status of the sampled telephone number 9 10 Figure 3 The relationship between IDD awareness rates and IS choosing 11 12 13 rate 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 30, 2021 by guest. Protected copyright. 42 43 44 Flow chart 45 92x148mm (96 x 96 DPI) 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 22 of 26 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 FIGURE 2 25 211x104mm (96 x 96 DPI) 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 30, 2021 by guest. Protected copyright. 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 Page 23 of 26 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 Figure 3 25 246x121mm (96 x 96 DPI) 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

2 2 8 8 8 8 5-7 5-7

5 5 4-5 4-5

8 8 Reported on page # Reported on page 2 2 7 7

cross-sectional studies cross-sectional

BMJ Open http://bmjopen.bmj.com/ on September 30, 2021 by guest. Protected copyright. For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

) Indicate ) study’sthe design commonlywith useda in term ortitlethe abstract the ) ) theGive eligibility criteria, and sourcesthe methodsand of selection of participants For ) in theProvide abstract informativean and balanced summary of what was donewhat wasand found peer review only a a b For eachof interest,For variable sourcesdatagive of and details of methods of assessment (measurement). Describe comparability of of comparability assessment methods isthanthere more if one group (

applicable collection collection ( Recommendation STROBE 2007 (v4) Statement—Checklist of items that should be included in reports of reports in be included should that items of Statement—Checklist (v4) 2007 STROBE

6 1 ( 3 specific objectives, State including prespecified any hypotheses 5 the Describe setting, locations, relevantand dates, including periods exposure,of recruitment, follow-up, data and 9 Describe any efforts to address sourcespotential of bias 7 define Clearly outcomes,all exposures,predictors, potential confounders,effect and modifiers. diagnostic Give criteria, if 4 key Present of elements study inthedesign early paper 8* # Item Item Bias Data sources/Data measurement Variables Participants Setting Study design Methods Objectives Background/rationale 2 thebackground Explain scientific for rationale investigation the and being reported Introduction Title andTitle abstract Section/Topic

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

9 9 8 8 8 8

9 9 9 9 9 9

8 8 8 8 7 7

9-14 9-14 8 8

BMJ Open http://bmjopen.bmj.com/ on September 30, 2021 by guest. Protected copyright. For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml ) Report) boundaries category when continuous variables were categorized ) ) unadjustedGive estimates if applicable, and, confounder-adjusted estimates their and precision95% (eg, confidence ) ) statisticalDescribe all methods, thoseincluding confoundingusedforcontrol to ) ) Describe any methods used subgroups examine to interactionsand ) If ) applicable, describe methodsanalytical taking account of sampling strategy ) Describe) sensitivity any analyses

For) howExplain missing data were addressed peer review only b a a b e d c confounders ( confirmed eligible,confirmed included in study, the completing follow-up, analysed and why why interval). Make interval). clear confounderswhich were adjusted they why and for were included ( ( (

( number Indicate (b) of participants missingwith eachfor data of interest variable Consider (c) diagram use flow of a reasons Give (b) non-participationfor each at stage 10 10 how Explain studythe size was arrived at 16 16 ( 13* 13* Report (a) numbers of individuals each at stage of study—eg numbers eligible, potentially examinedfor eligibility,

Main resultsMain Outcome Outcome data 15* numbers Report of outcome events or summary measures

Descriptive data 14* characteristicsGive (a) participantsof study (eg demographic, clinical, social) and information on exposures potential and

Participants Results

Statistical Statistical methods 12 ( Quantitative variables 11 how Explain quantitative were variables handled in analyses.the If applicable, describe which groupingschosen were and Study size Page 25 of 26 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from Page 26 of 15-17 15-17

15 15 15 15 17-19 17-19

18 18 14 14

BMJ Open http://bmjopen.bmj.com/ on September 30, 2021 by guest. Protected copyright. For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

) If )relevant, If consider translating estimatesrelativerisk of into absolute forrisk meaningfulperioda time For peer review only c similar studies,othersimilar and relevant evidence magnitude of magnitude potential any bias which the presentthe which is based article

18 18 Summarise results key to reference withstudy objectives 22 22 sourcethe Give of funding and role the of fundersthe thefor studypresent if and, applicable, originalthe for study on 19 19 Discuss of study, the limitations taking into sourcesaccount biasof potential imprecision.or Discussboth direction and An Explanation and Elaboration article discusses itemeach checklist and gives methodological backgroundand published examples of transparent reporting. The STROBE checklist is best used in thiswith conjunction (freely available article onsites the Medicine of PLoS Web http://www.plosmedicine.org/, at Annals of Internal Medicine at http://www.annals.org/, Epidemiologyhttp://www.epidem.com/). at and Information on InitiativeSTROBE the is www.strobe-statement.org.available at information*Give separately cases for and controls studiesin case-controlfor applicable, if and, exposedunexposed and groups in cohort and cross-sectional studies. Note: Funding Other information Generalisability 21 Discuss (externalgeneralisabilitythe of validity) studythe results Interpretation 20 cautious Give interpretationa of overall results consideringlimitations, objectives, multiplicity of resultsanalyses, from Limitations Key Key results Discussion Other analyses 17 other Report analyses done—eg analyses of subgroups sensitivity interactions, and and analyses ( 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

Should both iodised and non-iodised salt be made available in Chinese cities? A cross-sectional survey

ForJournal: peerBMJ Open review only Manuscript ID: bmjopen-2014-005397.R1

Article Type: Research

Date Submitted by the Author: 01-Jun-2014

Primary Subject Health policy Heading: http://bmjopen.bmj.com/ Secondary Subject Heading: Epidemiology, Health policy, Public health

telephone interview, iodized salt, non-iodized salt, knowledge-attitude- Keywords: practice, iodine deficiency disorders

on September 30, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 47 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 Title page 4 5 6 Should both iodised and non-iodised salt be made available in 7 8 Chinese cities? A cross-sectional survey 9 † † 10 Jun Y , Peng L , Ying L, Shou-jun L, Dian-jun S* 11 12 13 Author Affiliation 14 15 Jun Y, PengFor L, Ying peer L and Shou-jun review L are affiliated only with the Institute for 16 17 iodine deficiency disorders, the Centre for Endemic Disease Control, the 18 19 Chinese Centre for Disease Control and Prevention, and Harbin Medical 20 21 University, Harbin, People’s Republic of China. 22 23 Dian-jun S is affiliated with the Centre for Endemic Disease Control, the 24 25 Chinese Centre for Disease Control and Prevention, and Harbin Medical 26 27 University, Harbin, People’s Republic of China. 28 29 30 31 Corresponding author 32

33 Name: Dian-jun SUN http://bmjopen.bmj.com/ 34 35 Degree: M.D. 36 37 Position: Chief of Center for Endemic Disease Control 38 39 Affiliation: Center for Endemic Disease Control, Chinese Center for 40 Disease Control and Prevention, Harbin Medical University, Harbin

41 on September 30, 2021 by guest. Protected copyright. 42 43 150081, People’s Republic of China 44 45 Full address: No.157 Baojian Road, Nan’gang District, Center for 46 47 Endemic Disease Control, Harbin Medical University, Harbin, China 48 Postal code: 150081 49 50 Telephone: 86-451-86612695 51 52 Fax: 86-451-86657674 53 54 Email: [email protected] 55 56 57 58 59 60 1 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 47 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 First Author 4 5 Name: Jun YU†, Peng LIU† 6 7 † have made the same contribution 8 9 Other Author Name:Ying LIU, Shou-jun LIU 10 11 Affiliation: Center for Endemic Disease Control, Chinese Center for 12 13 Disease Control and Prevention, Harbin Medical University, Harbin 14 15 150081,For People’s Republicpeer of Chinareview only 16 17 Keywords: telephone interview, iodized salt, non-iodized salt, 18 19 knowledge-attitude-practice, iodine deficiency disorders 20 21 Word count: 4366 words 22 23 Should both iodised and non-iodised salt be made available in 24 25 Chinese cities? A cross-sectional survey 26 27 ABSTRACT 28 29 Objective: To contribute evidence relevant to the policy of supplying 30 31 iodised salt (IS), non-iodised salt (NIS) or both in Chinese cities. 32 Design: Sub-national telephone interview survey. 33 http://bmjopen.bmj.com/ 34 35 Setting: China. 36 37 Participants: A total of 4833 citizens accepted the telephone interview. 38 The telephone numbers were randomly selected by random digit dialling 39 40 and a Mitofsky-Waksberg two-stage sampling method in 17 capital cities

41 on September 30, 2021 by guest. Protected copyright. 42 and six coastal cities from 17 iodine deficiency disorder (IDD)-eliminated 43 44 provinces (municipalities). 45 46 Results: Among the 4833 citizens who finished the telephone interview, 47 48 3738 (77.3%) citizens chose IS, 481 (10.0%) citizens chose NIS, and the 49 50 others chose both IS and NIS. The citizens’ awareness rates of IDD and 51 52 IDD preventive measures were 68.7% and 62.5%, respectively. 53 54 Conclusions: It is not a suitable time to supply IS and NIS simultaneously 55 56 in the developed cities of China, but a pilot project may be conducted in 57 58 the cities where IDD has been sustainably eliminated, there is strong 59 60 2 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 47 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

33 http://bmjopen.bmj.com/ 34 35 cities knew about IDD, and 65.2% knew measures to prevent IDD. 36 37 It is not an appropriate time to supply IS and NIS simultaneously in 38 39 40 the developed cities of China, but a pilot project may be carried out in

41 on September 30, 2021 by guest. Protected copyright. 42 some cities. 43 44 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 BMJ Open Page 4 of 47 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 1 Should both iodised and non-iodised salt be made available 4 5 2 in Chinese cities? A cross-sectional survey 6 7 3 INTRODUCTION 8 9 4 Approximately 20% of the total global population resides in China, which 10 11 5 has one of the most severe iodine deficiency epidemics, in part due to the 12 13 6 external environment. Iodine Deficiency Disorders (IDD) were once 14 15 7 widespreadFor and causedpeer great damages.review Universal only Salt Iodization (USI) 16 17 8 has been adopted for IDD prevention and control and has proven to be an 18 19 9 effective measure since 1994. By 2000, 17 provinces (municipalities) in 20 21 10 China achieved the goal of eliminating IDD and have sustained this 22 [1] 23 11 outcome for 10 years based on the WHO criteria for IDD elimination. 24 [2-4] 25 12 However, an excess of iodine has been discovered, and new issues , 26 27 13 including the increasing rate of thyroid disease and the elevated health 28 29 14 awareness of the Chinese population, have started a debate about the 30 31 15 ability of the population to make informed choices to use iodised salt(IS) 32 16 or non-iodised salt(NIS). As we know, some developed European 33 http://bmjopen.bmj.com/ 34 35 17 countries, such as Belgium, Finland, France, Germany, Greece, Ireland, 36 18 37 Italy and the Netherlands, supply IS and NIS simultaneously in the 38 [5] 19 market. Residents of these countries can purchase IS or NIS voluntarily. 39 40 20 In developed cities in provinces that have demonstrated the sustainable

41 on September 30, 2021 by guest. Protected copyright. 42 21 elimination of IDD, is it the right time to supply IS and NIS 43 44 22 simultaneously? If so, can these populations remain free of IDD? Finally, 45 46 23 is there a proper method to demarcate cities to which IS and NIS is 47 48 24 supplied simultaneously or to which only IS is supplied? To answer these 49 50 25 questions, the subjective preferences of the residents about choosing IS or 51 52 26 NIS and the relevant influencing factors should be examined. 53 54 27 Door-to-door interviewing is usually used to conduct such investigations. 55 56 28 Considering the associated time and money expenditures, a 57 58 29 cross-sectional telephone interview is a feasible alternative and is a 59 60 4 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 47 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 30 widespread method that was adopted earlier and used by different 4 [6] 5 31 countries all over the world. Successful cases of the use of 6 7 32 cross-sectional telephone interviewing in China include KAP surveys on 8 [7-10] 9 33 influenza, smoking and pollinosis. 10 11 34 The application of cross-sectional telephone interviewing is based on 12 13 35 educational and economic levels, in addition to the telephone 14 15 36 popularizationFor rate. peer In China, byreview the end of 2009, only the national average 16 17 37 Urban Per Capita Disposable Income was 17,175 RMBY, and some 18 19 38 developed cities (municipalities) in the IDD-eliminated provinces were 20 [11] 21 39 all at or beyond the national average (table 1). Additionally, the 22 23 40 national fixed line penetration had increased to 314 million people (23.5 24 25 41 telephones/100 persons) and 212 million people in the cities (34.1 26 42 27 telephones/100 persons). These facts made telephone interview a 28 43 representative and feasible way to research the residents’ KAP toward 29 30 44 IDD. 31 32 45 As far as we know, there were no such records about the residents’

33 http://bmjopen.bmj.com/ 34 46 subjective preferences between IS and NIS, and no research on the 35 36 47 residents’ awareness of IDD in developed cities (municipalities) has been 37 38 48 conducted in recent years. Hence, the National Centre for Endemic 39 40 49 Disease Control (CEDC) launched this project in April, 2010, with the

41 on September 30, 2021 by guest. Protected copyright. 42 50 aim of obtaining data on the above aspects to decide whether supplying 43 44 51 IS and NIS simultaneously is feasible in developed cities in 45 46 52 IDD-eliminated provinces. The methods and results may provide 47 48 53 evidence or references for similar regions in the world. 49 50 54 51 52 55 METHODS 53 54 56 55 56 57 City sites to be investigated 57 58 58 The 17 capital cities of the 17 provinces (3 municipalities, 6 coastal 59 60 5 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 47 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 59 provinces and 8 inland provinces) that had achieved the goal of IDD 4 5 60 elimination in 2000 were selected (table 1). One coastal city was also 6 7 61 selected in each of the above coastal provinces. A total of 23 cities were 8 9 62 included in the study. 10 11 63 Table 1 Provinces and cities surveyed 12 13 Provincial capital Coastal cities Category Provinces 14 Name Tel* Income** Name Tel * Income ** 15 For peer review only 16 Hebei Shijiazhuang - 16,607 Tangshan 24.58 18,053 17 Yantai 28.27 21,125 18 Shandong Ji’nan 39.28 22,272 19 Jiangsu Nanjing 38.91 25,504 Nantong 35.80 21,001 Coastal province 20 Zhoushan 53.45 24,082 21 Zhejiang Hangzhou 48.23 26,864 22 Guangdong Guangzhou 61.28 27,610 Shenzhen 52.64 29,244 23 Guangxi Nanning 16.61 16,254 Beihai - 15,134 24 25 Heilongjiang Harbin 21.66 15,887 26 27 Jilin Changchun - 16,277 28 Shanxi Taiyuan 44.31 15,607 29 30 Inland He’nan Zhengzhou 29.28 17,117 province 31 Anhui Hefei 32.47 17,158 32

33 Hubei Wuhan 40.57 18,385 http://bmjopen.bmj.com/ 34 Jiangxi Nanchang 32.37 16,472 35 36 Hu’nan Changsha 33.28 20,238 37 Beijing Beijing 50.89 26,738 38 39 Municipalities Shanghai Shanghai 48.70 28,838 40 Tianjin Tianjin 31.36 21,430

41 on September 30, 2021 by guest. Protected copyright. 42 64 All data marked with “*” and “**”were obtained from the Statistics Bulletin of the Cities’ 43 65 Economic and Social Development of 2009: “*” fix telephones/100 persons, “**” Urban per 44 45 66 capita disposable income (RMBY). The national fixed line penetration was 23.50 telephones/100 46 67 persons, and the urban per capita disposable income was 17,175 RMBY in 2009. 47 48 68 49 50 69 Target population 51 52 70 Citizens who had fixed telephones in the 23 selected cities. 53 54 71 The main contents of the questionnaire: 55 56 72 If IS and NIS are supplied simultaneously, which will you choose? 57 58 73 And why? 59 60 6 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 47 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 74 Have you heard of IDD? If so, what are the effects? How is IDD 4 5 75 prevented? 6 7 76 Can you tell us your occupation and age group? 8 9 77 Sampling method and sampling size 10 11 78 Random Digit Dialling and the Mitofsky-Waksberg two-stage sampling 12 [12] 13 79 method were used in this study. In each city, only the main urban zones 14 15 80 were investigated,For peer and two districts review were selected only randomly from each 16 17 81 main urban zone. At least 200 effective samples were sampled in each 18 19 82 city included in the survey. The sample quota was divided evenly 20 21 83 between each district. The sample size was calculated according to 22 23 84 formula 1; the positive response rate ( π ) was approximately 70% 24 25 85 according to the pilot survey, and the allowable error (δ ) was 0.5. Thus, 26 86 27 the calculated sample size was 170 responses per city; therefore, 200 28 87 responses per city were considered adequate. 29 30  2 (1−× ππ ) 31 88 n =  α  π Positive rate δ Allowable error 32  δ 

33 http://bmjopen.bmj.com/ 34 89 Formula 1 35 90 Telephone number acquisition 36 37 91 If the fixed telephone of the neighbourhood offices in the selected district 38 39 92 was 7- or 8-digit numbers, the first 3 or 4 digits were deemed to be the 40

41 93 direction code. The last 4 digits were obtained by the Random function on September 30, 2021 by guest. Protected copyright. 42 ® 43 94 (=Rand ()×10000) in Microsoft Office Excel . If the function produced 44 45 95 randomised numbers less than 4 digits, and “0” was supplemented in 46 47 96 front of the generated numbers. The number of telephone numbers 48 49 97 created randomly should be at least 5 times the effective sample size, i.e., 50 51 98 at least 1000 telephone numbers should be created per city. 52 53 99 Quality control 54 55 100 The research was approved by the review board of the Harbin Medical 56 57 101 University. Participant consent was acquired when they answered the 58 59 60 7 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 47 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 102 telephone. The survey program and survey result database were designed 4 5 103 and delivered to the project cities by the national CEDC. The survey 6 7 104 program was improved based on a pilot survey, and the questionnaire had 8 9 105 clear questions that prompted simple answers and avoided providing hints 10 11 106 to the residents. Sensitive questions, such as age, were designed as 12 13 107 “Which age group do you belong to? <20, 20-30, 30-40... ”. 14 15 108 In the program,For thepeer investigation review was carried out only in a unified manner, 16 17 109 including the order and pattern of the questions, the dialling time (18:30 - 18 19 110 21:00), and instructions for making notes of the answers and filling the 20 21 111 database. 22 23 112 To avoid investigation and report biases and to improve the research 24 25 113 subjects’ cooperation, the interviewers were trained on their attitude, 26 114 27 mood, and countermeasures in case of emergency by the project 28 115 provinces themselves according to the rules of the survey program. For 29 30 116 example, some related questions put forward by residents could be 31 32 117 answered only after the interview was finished to avoid bias.

33 http://bmjopen.bmj.com/ 34 118 To guarantee the reliability of the research information, a return visit was 35 36 119 conducted after a project city finished the telephone interviews. The 37 38 120 quality control officer of the provincial CDC or CEDC visited the citizens 39 40 121 who had accepted the interview. At least 10% of the respondents were

41 on September 30, 2021 by guest. Protected copyright. 42 122 randomly sampled to complete the return visit questionnaire. When the 43 44 123 coincidence rate of survey information was greater than 80%, the 45 46 124 investigation results of the city can be regarded as qualified, and 47 48 125 otherwise, the city should be reinvestigated. 49 50 126 Definition of telephone number 51 52 127 Success: respondent finished the entire questionnaire; Rejection: 53 54 128 respondent refused to answer or did not finish the questionnaire; No 55 56 129 answer: the phone was busy both the first time and when re-dialling after 57 58 130 5 minutes; Invalid number: the phone number was vacant, fax, 59 60 8 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 47 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 131 switchboard or phone quit. 4 5 132 Data processing and statistical analysis 6 ® 7 133 Epi info 3.5.1 (CDC, Atlanta, GA USA) was used for the database 8 ® 9 134 design, and SPSS 17.0 was utilized for analysis. The level of 10 11 135 significance was set at 0.05. 12 13 136 14 15 137 RESULTSFor peer review only 16 17 138 Response situation 18 19 139 The flow chart is presented in figure 1. The survey included 235 districts 20 21 140 of 120 main urban zones in 23 cities. In total, 446 direction numbers were 22 23 141 chosen, and 24,557 telephone numbers were dialled, of which 13,215 24 25 142 (53.8%) were valid and 11,342 were invalid numbers. An invalid number 26 143 27 was a number that was not used as a telephone number. Among the 28 144 13,215 valid telephone numbers, 4,833 successfully finished the 29 30 145 questionnaire, 1,949 answered the phone but refused to answer the 31 32 146 questions, and 6,433 telephones were not answered. The answered

33 http://bmjopen.bmj.com/ 34 147 numbers (successes and rejections) were 6,782 and accounted for 51.3% 35 36 148 of the valid numbers. The total response rate (4833/(4833+1949)*100%) 37 38 149 was 71.3% (figure 2).. 39 40 150 Basic information about the interviewee

41 on September 30, 2021 by guest. Protected copyright. 42 151 The age and occupation questions were answered by 4,861 and 4,857 43 44 152 residents, respectively. The age and occupation proportion were similar 45 46 153 between the capital city (including Beijing) and the coastal city (including 47 48 154 Tianjin and Shanghai) residents. The residents’ age were mainly 49 50 155 distributed between 20 and 60 years, and this range accounted for 80% of 51 52 156 the respondents. However, compared with the population composition 53 [13] 54 157 from the ‘2009 China Health Statistical Yearbook’ and the Chinese 55 [14] 56 158 population pyramid, the 10-20 age group had a smaller constituent 57 58 159 ratio. This may lead to a non-representative sample of the Chinese 59 60 9 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 47 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 160 population, although this may be acceptable in this study because 4 5 161 children rarely buy salt. The residents were evenly distributed in each 6 7 162 profession (accounting for approximately 10%), except for the 8 9 163 professions of servicemen, students and irregular professions, which had 10 11 164 small sample sizes. 12 13 165 14 15 166 ProportionFor of respondents peer choosing review IS or NIS only 16 17 167 Of the residents investigated, 4,865 answered the question ‘If IS and NIS 18 19 168 are supplied simultaneously, which will you choose?’ The total 20 21 169 percentage of respondents ‘choosing IS’ was 76.8%, ‘choosing NIS’ was 22 23 170 9.9%, ‘both’ was 3.9% and ‘does not matter’ was 9.3%. The respondents 24 25 171 from the capital cities chose IS at a higher rate (79.5%) than those from 26 172 27 the coastal cities (72.0%). The percentages of respondents choosing NIS 28 173 was opposite between the city and coastal residents (6.7% and 15.7%, 29 30 174 respectively), and the ratio of respondents who chose “both” (<5%) and 31 32 175 ‘does not matter’ (<10%) were nearly the same in the two areas. Of the 23

33 http://bmjopen.bmj.com/ 34 176 cities, there were 12 cities whose IS selection rate was higher than 80%, 8 35 36 177 cities in which the IS selection rate was between 60-80%, and 3 cities 37 38 178 (Shanghai: 32.8%, Zhoushan: 42.8% and Hangzhou: 49.3%) in which the 39 40 179 IS selection rate was lower than 50% (table 2, table 3).

41 on September 30, 2021 by guest. Protected copyright. 42 180 Reasons for “choosing IS” 43 44 181 Of the 3,738 residents who chose IS, their reasons were as follows: 45 46 182 61.9% of them accurately understood the benefits of IS, i.e., preventing 47 48 183 IDD and increasing children’s Intelligence Quotient (IQ), 14.7% of them 49 50 184 merely chose IS for superior-quality reasons and knew little about the 51 52 185 relationship between IS and IDD prevention, and 27.7% had no clear 53 54 186 reason (most explained that it was the custom) (table 2). 55 56 187 57 58 188 59 60 10 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 47 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 189 Table 2 Respondents’ reasons for buying IS (%) 4 5 Reasons Number 6 IS 7 of Benefit Province City choosing Prevent for Good 8 choosing Others rate IDD children’s quality 9 IS IQ 10 Heilongjiang Harbin 183 91.5 36.6 3.8 1.6 60.1 11 Jilin Changchun 175 86.6 48.6 9.7 6.3 0 12 Beijing - 202 86.7 61.4 12.9 9.4 28.2 13 Hebei Shijiazhuang 174 87.0 67.8 7.5 12.1 14.4 14 Shanxi Taiyuan 189 95.0 13.8 0 14.3 70.9 15 ShandongFor Ji’nan peer 187 review 93.5 73.3 only 27.3 20.3 11.2 16 He’nan Zhengzhou 177 88.5 54.8 15.3 26.6 27.7 17 Jiangsu Nanjing 180 72.6 64.4 25.0 23.9 7.8 18 Anhui Hefei 178 84.8 32.0 0.6 5.6 61.8 19 Hubei Wuhan 163 75.1 41.7 8.0 24.5 29.4 20 Zhejiang Hangzhou 66 32.8 57.6 16.7 21.2 0 21 Jiangxi Nanchang 184 86.8 52.2 7.6 13.6 8.7 22 Hu’nan Changsha 146 72.3 56.2 6.8 5.5 39.7 23 Guangdong Guangzhou 124 61.7 39.5 1.6 46.0 3.2 24 Guangxi Nanning 170 78.0 38.8 4.1 6.5 46.5 25 Capital city total 2498 79.5 79.5 9.8 15.0 29.0 26 Hebei Tangshan 208 94.1 81.7 17.3 9.6 5.8 27 Tianjin - 200 83.7 75.5 28.5 27.0 6.5 28 Shandong Yantai 186 93.0 90.3 3.2 8.1 4.8 29 Jiangsu Nantong 153 73.9 46.4 5.9 2.6 47.7 30 Shanghai - 108 49.3 25.9 6.5 29.6 48.1 31 Zhejiang Zhoushan 86 42.8 68.6 12.8 18.6 0 32 Guangdong Shenzhen 144 71.6 22.9 0.7 3.5 42.4

33 Guangxi Beihai 155 66.2 21.9 2.6 18.7 58.1 http://bmjopen.bmj.com/ 34 Coastal city total 1240 72.0 72.0 10.6 14.1 25.0 35 Total 3738 76.8 51.9 10.0 14.7 27.7 36 190 Note: The reason for choosing IS was a multiple choice question. 37 38 191 39 40 192 Reasons for “choosing NIS”

41 on September 30, 2021 by guest. Protected copyright. 42 193 Most of the 481 residents who chose NIS lived in Shanghai, Hangzhou, 43 44 194 Zhoushan and Beihai. The majority thought that they had received an 45 46 195 adequate amount of iodine from food or that they did not live in an IDD 47 196 48 area. Unexpectedly, 10% of the respondents who chose NIS thought IS 49 197 was harmful. Nearly 27.0% chose NIS because of reasons that cannot be 50 51 198 classified, such as habit, parent’s decision, or changing taste (table 3). 52 53 199 54 55 200 56 57 201 58 59 60 11 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 47 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 202 Table 3 Respondents’ reasons for choosing NIS（%） 4 5 Reasons 6 Not IS 7 City n % Iodine IS is 8 IDD Patient* bad Cheap Others sufficient harmful 9 endemia taste 10 Capital city Total 211 40.0 6.7 13.3 9.0 10.5 2.9 1.4 23.3 11 12 Coastal city Total 270 27.7 15.7 17.7 5.2 8.1 4.0 1.5 29.9 13 Total 481 33.1 9.9 16.6 7.3 9.1 3.5 1.5 27.0 14 203 Note: The “*” meant patients with thyroid disease for whom it is not feasible to take IS. Some provinces had small 15 204 sample sizes;For the proportions peer refer to large ones; review n: number that chose NIS;only %: NIS choosing rate, the 16 17 205 reason for choosing NIS was a multiple choice question. 18 19 206 Reasons for “choosing both” 20 21 207 Most of the 192 residents who chose “both IS and NIS” were from 22 208 23 Hangzhou, Guangzhou and Shanghai. The two major reasons were to 24 209 maintain an appropriate iodine intake and to avoid the harmful effect of 25 26 210 excess iodine, accounting for 27.6% and 30.7% of the responses, 27 28 211 respectively. 29 30 212 Reasons for “does not matter” 31 32 213 The 454 residents who chose “does not matter” mainly resided in Nanjing,

33 http://bmjopen.bmj.com/ 34 214 Wuhan, Hangzhou, Changsha, Guangzhou, Nanning, Tianjin, Shanghai 35 36 215 and Zhoushan. They tended to choose IS or NIS based on which was 37 38 216 most convenient to buy (52.0% of the respondents who had no 39 40 217 preference).

41 on September 30, 2021 by guest. Protected copyright. 42 218 Awareness rates of IDD 43 44 219 Of the residents who responded to our phone calls, 4,865 answered the 45 46 220 question “Have you heard of Iodine Deficiency Disorders?”. Among 47 48 221 those, 68.7% (3,344) had heard and 31.3% (1,521) had never heard of 49 50 222 IDD. There was little difference in having heard of IDD between the 51 52 223 capital cities (68.2%) and the coastal cities (69.8%). The awareness rate 53 54 224 in Yantai (94.5%) was the highest, and in four cities (Shijiazhuang, Jinan, 55 56 225 Hangzhou and Tangshan), the awareness rates were greater than 80%. 57 58 226 Beihai and Taiyuan had the lowest awareness rates (46.2% and 44.7%, 59 60 12 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 47 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 227 respectively). There were no significant differences in the levels of 4 5 228 awareness between age groups or occupation groups. 6 7 229 Among the 3,344 residents who had heard of IDD, up to 77.1% of them 8 9 230 knew that iodine deficiency could lead to goitre, but only 11.8% knew 10 11 231 that it could also cause intelligence damage to children, and 16.9% of 12 13 232 them knew nothing about its hazard. 14 15 233 AwarenessFor rates ofpeer IDD preventive review measures only 16 17 234 Of the 3,344 residents who had heard of IDD, 62.5% of them knew IS 18 19 235 could prevent IDD, 43.2% of them knew that kelp and purple seaweed 20 21 236 could prevent IDD, and nearly 18.7% of the residents knew nothing about 22 23 237 how to prevent IDD. 24 25 238 The IS preference rates and IDD awareness rates in each city 26 239 27 All cities could be divided into four groups according to their relationship 28 240 between the IS preference rate and IDD awareness rate: Uhigh IS 29 30 241 preference rate and high IDD awareness rate, such as Yantai, 31 32 242 Shijiazhuang, Beijing, Tianjin, Nanjing, Changsha and Guangzhou;

33 http://bmjopen.bmj.com/ 34 243 Uhigh IS preference rate and low IDD awareness rate, such as Taiyuan, 35 36 244 Harbin, Nanchang, Hefei, Wuhan and Beihai; Uapproximate equal the 37 38 245 choosing rate and the awareness rate, such as Tangshan, Ji’nan, 39 40 246 Zhengzhou, Changchun, Nanning, Nantong and Shenzhen; and ④low IS

41 on September 30, 2021 by guest. Protected copyright. 42 247 choosing rate and high IDD awareness rate, such as Shanghai, Zhoushan 43 44 248 and Hangzhou. Only in Yantai and Nanjing were the IDD awareness rates 45 46 249 higher than the IS preference rates. The IS choosing rates in Harbin and 47 48 250 Taiyuan were greater than 90%, but their IDD awareness rates were only 49 50 251 57.0% and 44.7%, respectively. There were only 4 cities whose IS rates 51 52 252 and awareness rates were all greater than 80% (figure 3). 53 54 253 The chi square test was used to analyse the influence of IDD awareness 55 56 254 on the IS preference rate. The results showed no differences between the 57 58 255 capital cities’ IS and NIS groups but significant differences between the 59 60 13 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 47 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 256 coastal cities’ IS and NIS groups. This could explain the weak relations 4 5 257 between the IS preference rates and the IDD awareness rates in the cities, 6 7 258 whereas for coastal cities, there were associations between the IS 8 9 259 preferences and IDD awareness; specifically, most coastal residents who 10 11 260 chose NIS had lower IDD awareness rates. 12 13 261 Results of the return visits 14 15 262 The totalFor coherence peer rate of all questionsreview was 85.1%. only The Kappa values of 16 17 263 the coherence analysis for choosing IS, choosing NIS and IDD awareness 18 19 264 were 0.752, 0.693 and 0.76, respectively, and were all above the required 20 [15] 21 265 0.4, describing good coherence. 22 23 266 24 25 267 DISCUSSIONS 26 268 27 Telephone interviews are a widely used method all over the world that 28 269 has the advantages of wide coverage, low costs and high efficiency. The 29 30 270 survey subjects could be units or persons with fixed telephones, and the 31 32 271 selection, supervision and training of investigators are convenient. These

33 http://bmjopen.bmj.com/ 34 272 features made it feasible to conduct a large-scale, easily acceptable survey 35 36 273 with clear content. However, telephone surveys do have biases and 37 [16] 38 274 limitations. Only after a detailed design to decrease error as much as 39 40 275 possible are the results convincible. For some interviews, such as street

41 on September 30, 2021 by guest. Protected copyright. 42 276 intercept or telephone interviews, if most of the interviewees reject the 43 44 277 survey and the rejection rate is greater than 70% (i.e., the response rate 45 [17] 46 278 <30%), the representation of the random samples will be questioned; 47 48 279 hence, the interviewees will be resampled. In our study, the telephone 49 50 280 survey was conducted with unified quality control, the response rate was 51 52 281 71.3%, and the coherence rate of re-survey was above 80% (85.1%). 53 54 282 There were no significant differences in terms of age and occupation. 55 56 283 Thus, the results of this telephone survey could basically reflect the 57 58 284 subjective preferences about IS or NIS of the urban residents in China. As 59 60 14 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 47 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 285 the first telephone survey on the KAP toward IDD in China, the 4 5 286 traditional PAPI method was taken, and the results offer references for 6 7 287 future surveys. 8 9 288 The residents’ total IS preference rate was 76.8% in the 6 coastal cities 10 11 289 and 14 capital cities in 14 provinces and 3 municipalities (i.e., Beijing, 12 13 290 Tianjin and Shanghai) where IDD elimination has been sustained for 14 15 291 nearly For ten years. peer The reasons review for choosing ISonly were mainly ‘IDD 16 17 292 prevention (51.9%)’ and ‘improving children’s IQ (10%)’, the remaining 18 19 293 residents (38%) had no idea about the relationship between IS and IDD 20 21 294 prevention and their choice of IS was only because IS is of good quality 22 23 295 or simply a custom. Another phenomenon that emerged in this study was 24 25 296 that most residents had lower IDD awareness rates than IS selection rates, 26 297 27 as was observed in Taiyuan, Hefei, Harbin and Changchun. Although the 28 298 residents’ IS preference rates were greater than 80%, their awareness of 29 30 299 IDD was low. An additional 33.1% of residents chose NIS, 27% of whom 31 32 300 had no clear reason, which reflected that although IDD health education

33 http://bmjopen.bmj.com/ 34 301 activities, such as the 5.15 IDD Day, have been carried out widely and 35 36 302 continuously by the relevant departments, it is still not enough or 37 38 303 exhaustive, and the current propaganda strength and methods need further 39 40 304 improving.

41 on September 30, 2021 by guest. Protected copyright. 42 305 Considering that the supply system of iodised salt is divided by province 43 44 306 in China, the KAP toward IDD of the city residents is more important. 45 46 307 Although these provinces all attained the goal of IDD elimination and 47 48 308 their capital cities and coastal cities are all developed cities, their IS 49 50 309 choosing and IDD awareness rates were different, and they should be 51 52 310 treated differently according to these results. In this study, 12 cities had 53 54 311 an IS preference rate of greater than 80%, and these were all northern 55 56 312 cities with the exception of Nanchang. Five cities had a rate greater than 57 58 313 90%. The other 11 cities with rates lower than 80% were all southern 59 60 15 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 47 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 314 cities. The IS preference rates of Hangzhou, Zhoushan and Shanghai were 4 5 315 32.8%, 42.8% and 49.3%, respectively, and were all lower than 50%. 6 7 316 This result shows that the IS preference rates of southern city residents 8 9 317 were lower than those of northern city residents. The main reason was 10 11 318 that the people in the southern coastal areas think that they have received 12 13 319 adequate iodine from food other than iodised salt, and another reason was 14 15 320 that theyFor thought theypeer did not livereview in an iodine-deficient only area, as was the 16 17 321 case with the residents of Hangzhou, Zhoushan and Shanghai. Their IDD 18 19 322 awareness rates were greater than their IS preference rates, and some 20 21 323 residents chose NIS despite knowing about the potential damage caused 22 23 324 by IDD. In fact, these districts had no apparent IDD prevalence, which is 24 25 325 likely attributable to the widespread IS supply. The survey “Iodine 26 326 27 nutritional status of coastal districts residents in 2009” conducted by the 28 [18] 327 MOH found that under the condition of consumption rates of qualified 29 30 328 IS greater than 90%, the Median Urinary Iodine (MUI) of Shanghai 31 32 329 residents was merely between 100 µg/L and 200 µg/L, i.e., just at the

33 http://bmjopen.bmj.com/ 34 330 iodine intake appropriate level, and the MUI of pregnant women was 35 36 331 lower than 150 µg/L, which corresponds to an insufficient iodine intake 37 38 332 level. Among the residents of the Xiacheng district of Hangzhou city, 39 40 333 Zhejiang province, the consumption rate of qualified IS was 96.7%, and

41 on September 30, 2021 by guest. Protected copyright. 42 334 the MUI of 8-10-year-old children was between 100 and 200 µg/L, which 43 [19] 44 335 is at the iodine intake appropriate level. If allowing these cities’ 45 46 336 residents to freely choose IS or NIS, the residents may choose NIS 47 48 337 blindly, and the achievement of controlled IDD that has been obtained 49 50 338 will be destroyed. 51 52 339 From the society and economic development aspects, supplying IS and 53 54 340 NIS simultaneously to some cities or persons is an inevitable trend. In the 55 56 341 public health domain, if a disease has a prevalence rate below 5%, it will 57 58 342 not become a public health problem. Hence, we deem that if the IDD 59 60 16 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 47 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 343 awareness rates and the correct selection rates are above 95%, iodised and 4 5 344 non-iodised salt can both be made available. According to this telephone 6 7 345 interview study, it is not a suitable time to supply IS and NIS 8 9 346 simultaneously, even in the developed cities of China. The intake of 10 11 347 non-iodised salt should be performed with caution in China. The next step 12 13 348 is to conduct a trial in which IS and NIS is simultaneously supplied in 14 15 349 appropriateFor cities. peer The selected review experimental city onlyshould be a developed 16 17 350 city with no IDD prevalence in recent years, and the residents’ awareness 18 19 351 rate of IDD and selection rate of IS should be high. According to our 20 21 352 study, Yantai, Tangshan and Ji’nan may be appropriate cities for this 22 23 353 experiment. Before the experiment, emergency planning should be 24 25 354 performed, and extensive IDD education should be conducted. If 26 355 27 abnormal situations emerge, the emergency program will be started, or 28 356 the trial will even be stopped. Internationally, universal salt iodisation is 29 30 357 still considered the most cost-effective way to prevent and control iodine 31 32 358 deficiency disorders. Since the universal salt iodisation implemented in

33 http://bmjopen.bmj.com/ 34 359 China in 1996, there has not been documentation of the citizens’ 35 36 360 preferences of salt type in a large-scale survey of the awareness of IDD 37 38 361 and its prevention. This study revealed some important findings as to why 39 40 362 people choose iodised or non-iodised salt. For instance, quality and

41 on September 30, 2021 by guest. Protected copyright. 42 363 convenience, i.e., easy access, were among the reasons people would 43 44 364 purchase iodised salt, whereas concerns about the potential harm and 45 46 365 having thyroid disorders were the reasons for choosing non-iodised salt. 47 48 366 These issues should be taken into consideration when formulating future 49 50 367 health promotion messages. 51 52 368 53 54 369 55 56 370 57 58 371 59 60 17 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 47 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 372 4 5 373 Acknowledgments We thank the organisations that participated in this 6 7 374 survey: the National Centre for Endemic Disease Control (CEDC), 16 8 9 375 provincial (municipality) Centres for Disease Control (CDCs) or CEDC, 10 11 376 including those of the Beijing, Tianjin, Shanghai, and Jilin Provinces, 12 13 377 Hebei Province, Shandong Province, He’nan Province, Jiangsu Province, 14 15 378 Anhui Province,For Hubeipeer Province, review Zhejiang Province, only Jiangxi Province, 16 17 379 Hu’nan Province, Guangdong Province and Guangxi province, and six 18 19 380 coastal city CDCs, including those of Tangshan, Yantai, Nantong, 20 21 381 Zhoushan, Shenzhen and Beihai. The authors of this article thank those 22 23 382 colleagues from the above organisations who took part in the survey for 24 25 383 their hard work. 26 384 27 28 385 Contributorship statement Jun Y wrote the project report, Peng L 29 30 386 analysed the data and wrote the manuscript, Ying L and Shou-jun L 31 32 387 designed the study and managed the project, and Dian-jun S was

33 http://bmjopen.bmj.com/ 34 388 responsible for the project and the publication. 35 36 37 389 Competing interests None. 38 39 40 390 Funding The project was funded by the Ministry of Health of China.

41 on September 30, 2021 by guest. Protected copyright. 42 391 43 44 392 Data sharing No additional data are available. 45 46 393 47 48 394 Ethics approval The review board of the Harbin Medical University. 49 50 395 51 396 52 53 397 54 55 398 56 399 57 58 400 59 60 18 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 47 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 401 References 4 5 402 [1] WHO/UNICEF/ICCIDD. Assessment of iodine deficiency disorders and monitoring their 6 403 elimination. A guide for programme managers. Third edition, France, 2007. 7 404 [2] Todd CH, Allain T, Gomo ZA, et al. Increase in thyrotoxicosis associated with iodine 8 9 405 supplements in Zimbabwe. Lancet. 1995,346:1563-4. 10 406 [3] Stanbury JB, Ermans AE, Bourdoux P, et al. Iodine-induced hyperthyroidism: occurrence 11 12 407 and epidemiology. Thyroid 1998, 8:83-100. 13 408 [4] Parveen S, Latif SA, Kamal MM, et al. Iodized salt induced thyrotoxicosis: Bangladesh 14 409 perspective, Mymensingh Med J. 2009,18:165-8. 15 For peer review only 16 410 [5] ICCIDD. International Council for the Control of Iodine Deficiency Disorder. CIDDS 17 411 Database. Current IDD Status Database.2004. 18 412 [6] Groves RM, Biemer PP, Lyberg LE, et al. Telephone survey methodology [J] John Wiley 19 20 413 &Sons Inc.1989, 2-10. 21 414 [7] Jianhua L, Hanwu M , Yongsheng W, et al. Analysis of know ledge, attitude and practice 22 23 415 of the influenza A /H1N1of general and immigration population in Shenzhen, Chinese Health 24 416 Education. 2009,25:915-8. 25 417 [8] Jin M, Shunxiang Z, Hanwu M, et al. Telephone survey on behavior risk factors of 26 27 418 Shenzhen residents among six districts in 200. Journal of Disease Control. 2005,9:561-5 28 419 [9] Tao W, Su-jun L, Liang Z, et al. Investigation of Pollinosis in Beijing Residents over the 29 420 Age of 15. J Environ Health, 2008,25:403-4. 30 31 421 [10] Blair G. Obstetricians' Receptiveness to Teen Prenatal Patients Who Are Medicaid 32 422 Recipients, Health Services Research.1997,32:265-82.

33 http://bmjopen.bmj.com/ 423 [11] National Bureau of Statistics of China.Statistics Bulletin of the Cities’ Economic and 34 35 424 Social Development of 2009. 36 425 [12] Potthoff RF. Some Generazations of the Mitofsky-Waksberg techniques for random 37 38 426 digital dialing. Journal of the American Statistical Association, 1987, 82:409-41. 39 427 [13] Yi MING. http://geo.cersp.com/sJxzy/sc/200706/2658.html. 40 428 [14] http://www.moh.gov.cn/publicfiles/business/htmlfiles/zwgkzt/ptjnj/200908/42635.htm.

41 on September 30, 2021 by guest. Protected copyright. 42 429 [15] Jiezhen W, Jing H, Yanxun L, et al. Kappa statistic value and its application. [J], Chinese 43 430 Health Statistical, 1995,12:46-50. 44 431 [16] Ping W, Weijing D. Telephone Survey and Its Application in Public Health. Health 45 46 432 education of China. 2004,20(6):549-50. 47 433 [17] Guohua D. A powerful tool of the statistic survey—the computer aid telephone survey 48 49 434 system. Newspaper of Theory Guidance, 2008,12:51-2. 50 435 [18] Chinese CDC, CEDC. Report of coastal residents’ iodine nutrient level,2009. 51 436 [19] ICCIDD Current IDD Status Database,2003. 52 53 437 [http://www.who.int/vmnis/database/iodine/countries/en/index.html]. 54 55 56 57 58 59 60 19 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 47 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 4 Figure legend 5 6 Figure 1 Flow chart 7 8 Figure 2 The answer status of the sampled telephone numbers 9 10 Figure 3 The relationship between IDD awareness rates and IS preference 11 12 13 rates 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

1 2 3 Title page 4 5 6 Should both iodised and non-iodised salt be made available in 7 8 Chinese cities? A cross-sectional survey 9 † † 10 Jun Y , Peng L , Ying L, Shou-jun L, Dian-jun S* 11 12 Corresponding author 13 Name: Dian-jun SUN 14 15 Degree:For M.D. peer review only 16 17 Position: Chief of Center for Endemic Disease Control 18 19 Affiliation: Center for Endemic Disease Control, Chinese Center for 20 21 Disease Control and Prevention, Harbin Medical University, Harbin 22 23 150081, People’s Republic of China 24 25 Full address: No.157 Baojian Road, Nan’gang District, Center for 26 27 Endemic Disease Control, Harbin Medical University, Harbin, China 28 29 Postal code: 150081 30 31 Telephone: 86-451-86612695 32

33 Fax: 86-451-86657674 http://bmjopen.bmj.com/ 34 35 Email: [email protected] 36 37 38 39 First Author 40 Name: Jun YU†, Peng LIU†

41 on September 30, 2021 by guest. Protected copyright. 42 † 43 have made the same contribution 44 45 Other Author Name:Ying LIU, Shou-jun LIU 46 47 Affiliation: Center for Endemic Disease Control, Chinese Center for 48 Disease Control and Prevention, Harbin Medical University, Harbin 49 50 150081, People’s Republic of China 51 52 Keywords: telephone interview, iodized salt, non-iodized salt, 53 54 knowledge-attitude-practice, iodine deficiency disorders 55 56 Word count: 4366 words 57 58 59 60 1 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 47 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 Should both iodised and non-iodised salt be made available in 4 5 Chinese cities? A cross-sectional survey 6 7 ABSTRACT 8 9 Objective: To contribute evidence relevant to the policy of supplying 10 11 iodised salt (IS), non-iodised salt (NIS) or both in Chinese cities. 12 13 Design: NSub-national telephone interview survey. 14 15 Setting:For China. peer review only 16 17 Participants: A total of 4833 citizens accepted the telephone interview24 18 19 557 numbers were dialed. The telephone numbers were randomly 20 21 selected by random digit dialling and a Mitofsky-Waksberg two-stage 22 23 sampling method in 17 capital cities and six coastal cities from 17 iodine 24 25 deficiency disorder (IDD)-eliminated provinces (municipalities). 26 27 Results: Among the 4833 citizens who acceptfinished the telephone 28 29 interview, 3738 (77.3%) citizens chose IS, 481 (10.0%) citizens chose 30 31 NIS, and the others chose both IS and NIS. The citizens’ awareness rates 32 of IDD and IDD preventive measures were 68.7% and 62.5%, 33 http://bmjopen.bmj.com/ 34 35 respectively. 36 37 Conclusions: It is not a suitable time to supply IS and NIS simultaneously 38 in the developed cities of China, but a pilot project may be conducted in 39 40 the cities where IDD has been sustainably eliminated, there is strong

41 on September 30, 2021 by guest. Protected copyright. 42 awareness of IDD and the population can make informed decisions 43 44 regarding IS. IDD health education should be further strengthened, 45 46 especially regarding the potential for IQ damage. 47 48 49 50 Keywords: telephone interview, iodized salt, non-iodized salt, 51 52 knowledge-attitude-practice, iodine deficiency disorders 53 54 55 56 57 58 59 60 2 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 47 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 4 5 6 7 Strengths and limitations of this study 8 9 10 The vast majority (77.3%) of the population of China living in 11 12 developed cities chose iodized rather than non-iodized salt. 13 14 15 OverFor half (68.7%) peer of the populationreview of China only living in developed 16 17 cities knew about IDD, and 65.2% knew measures to prevent IDD. 18 19 20 It is not an appropriate time to supply IS and NIS simultaneously in 21 22 the developed cities of China, but a pilot project may be carried out in 23 24 25 some cities. 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 30, 2021 by guest. Protected copyright. 42 43 44 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 BMJ Open Page 24 of 47 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

41 on September 30, 2021 by guest. Protected copyright. 42 21 elimination of IDD, is it the right time to supply IS and NIS 43 44 22 simultaneously? If so, can these populations remain free of IDD? Finally, 45 46 23 is there a proper method to demarcate cities to which IS and NIS is 47 48 24 supplied simultaneously or to which only IS is supplied? To answer these 49 50 25 questions, the subjective preferences of the residents about choosing IS or 51 52 26 NIS and the relevant influencing factors should be examined. 53 54 27 Door-to-door interviewing is usually used to conduct such investigations. 55 56 28 Considering the associated time and money expenditures, a 57 58 29 cross-sectional telephone interview is a feasible alternative and is a 59 60 4 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 47 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

41 on September 30, 2021 by guest. Protected copyright. 42 50 aim of obtaining data on the above aspects to decide whether supplying 43 44 51 IS and NIS simultaneously is feasible in developed cities in 45 46 52 IDD-eliminated provinces. The methods and results may provide 47 48 53 evidence or references for similar regions in the world. 49 50 54 51 52 55 METHODS 53 54 56 55 56 57 City sites to be investigated 57 58 58 The 17 capital cities of the 17 provinces (3 municipalities, 6 coastal 59 60 5 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 47 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

41 on September 30, 2021 by guest. Protected copyright. 42 64 All data marked with “*” and “**”were obtained from the Statistics Bulletin of the Cities’ 43 65 Economic and Social Development of 2009: “*” fix telephones/100 persons, “**” Urban per 44 45 66 capita disposable income (RMBY). The national fixed line penetration was 23.50 telephones/100 46 67 persons, and the urban per capita disposable income was 17,175 RMBY in 2009. 47 48 68 49 50 69 Target population 51 52 70 Citizens who had fixed telephones in the 23 selected cities. 53 54 71 The main contents of the questionnaire: 55 56 72 If IS and NIS are supplied simultaneously, which will you choose? 57 58 73 And why? 59 60 6 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 47 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

41 93 direction code. The last 4 digits were obtained by the Random function on September 30, 2021 by guest. Protected copyright. 42 ® 43 94 (=Rand ()×10000) in Microsoft Office Excel . If the function produced 44 45 95 randomised numbers less than 4 digits, and “0” was supplemented in 46 47 96 front of the generated numbers. The number of telephone numbers 48 49 97 created randomly should be at least 5 times the effective sample size, i.e., 50 51 98 at least 1000 telephone numbers should be created per city. 52 53 99 Quality control 54 55 100 The research was approved by the review board of the Harbin Medical 56 57 101 University. Participant consent was acquired when they answered the 58 59 60 7 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 47 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

41 on September 30, 2021 by guest. Protected copyright. 42 122 randomly sampled to complete the return visit questionnaire. When the 43 44 123 coincidence rate of survey information was greater than 80%, the 45 46 124 investigation results of the city can be regarded as qualified, and 47 48 125 otherwise, the city should be reinvestigated. 49 50 126 Definition of telephone number 51 52 127 Success: respondent finished the entire questionnaire; Rejection: 53 54 128 respondent refused to answer or did not finish the questionnaire; No 55 56 129 answer: the phone was busy both the first time and when re-dialling after 57 58 130 5 minutes; Invalid number: the phone number was vacant, fax, 59 60 8 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 29 of 47 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 131 switchboard or phone quit. 4 5 132 Data processing and statistical analysis 6 ® 7 133 Epi info 3.5.1 (CDC, Atlanta, GA USA).was used for the database design, 8 ® 9 134 and SPSS 17.0 was utilized for analysis. The level of significance was 10 11 135 set at 0.05. 12 13 136 14 15 137 RESULTSFor peer review only 16 17 138 Response situation 18 19 139 The flow chart is presented in figure 1. The survey included 235 districts 20 21 140 of 120 main urban zones in 23 cities. In total, 446 direction numbers were 22 23 141 chosen, and 24,557 telephone numbers were dialled, of which 13,215 24 25 142 (53.8%) were valid and 11,342 were invalid numbers. An invalid number 26 143 27 was a number that was not used as a telephone number. Among the 28 144 29 13,215 valid telephone numbers, 4,833 successfully finished the 30 145 questionnaire, 1,949 answered the phone but refused to answer the 31 32 146 questions, and 6,433 telephones were not answered. The answered

33 http://bmjopen.bmj.com/ 34 147 numbers (successes and rejections) were 6,782 and accounted for 51.3% 35 36 148 of the valid numbers. The total response rate (4833/(4833+1949)*100%) 37 38 149 was 71.3% (figure 2).The survey investigated 235 districts of 120 main 39 40 150 urban zones in 23 cities totally, 446 direction numbers were chosen and

41 on September 30, 2021 by guest. Protected copyright. 42 151 24,557 telephone numbers were dialed, including Success 4,833; 43 44 152 Rejection 1,949; No answer 6,433; Invalid number 11,342. Among them, 45 46 153 valid number was 13,215, accounted for 53.8% of the total dialed. 47 48 154 Answered number (success and rejection) was 6,782, accounted for 49 50 155 51.3% of the valid, and the total response rate was 71.3% (figure 2). 51 52 156 Basic information about the interviewee 53 54 157 The age and occupation questions were answered by 4,861 and 4,857 55 56 158 residents, respectively. The age and occupation proportion were similar 57 58 159 between the capital city (including Beijing) and the coastal city (including 59 60 9 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 30 of 47 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 160 Tianjin and Shanghai) residents. The residents’ age were mainly 4 5 161 distributed between 20 and 60 years, and this range accounted for 80% of 6 7 162 the respondents. However, compared with the population composition 8 [13] 9 163 from the ‘2009 China Health Statistical Yearbook’ and the Chinese 10 [14] 11 164 population pyramid, the 10-20 age group had a smaller constituent 12 13 165 ratio. This may lead to a non-representative sample of the Chinese 14 15 166 population,For although peer this may review be acceptable inonly this study because 16 17 167 children rarely buy salt. The residents were evenly distributed in each 18 19 168 profession (accounting for approximately 10%), except for the 20 21 169 professions of servicemen, students and irregular professions, which had 22 23 170 small sample sizes. 24 25 171 26 172 27 Proportion of respondents choosing IS or NIS 28 173 Of the residents investigated, 4,865 answered the question ‘If IS and NIS 29 30 174 are supplied simultaneously, which will you choose?’ The total 31 32 175 percentage of respondents ‘choosing IS’ was 76.8%, ‘choosing NIS’ was

33 http://bmjopen.bmj.com/ 34 176 9.9%, ‘both’ was 3.9% and ‘does not matter’ was 9.3%. The respondents 35 36 177 from the capital cities chose IS at a higher rate (79.5%) than those from 37 38 178 the coastal cities (72.0%). The percentages of respondents choosing NIS 39 40 179 was opposite between the city and coastal residents (6.7% and 15.7%,

41 on September 30, 2021 by guest. Protected copyright. 42 180 respectively), and the ratio of respondents who chose “both” (<5%) and 43 44 181 ‘does not matter’ (<10%) were nearly the same in the two areas. Of the 23 45 46 182 cities, there were 12 cities whose IS selection rate was higher than 80%, 8 47 48 183 cities in which the IS selection rate was between 60-80%, and 3 cities 49 50 184 (Shanghai: 32.8%, Zhoushan: 42.8% and Hangzhou: 49.3%) in which the 51 52 185 IS selection rate was lower than 50% (table 2, table 3). Reasons for 53 54 186 “choosing IS” 55 56 187 Of the 3,738 residents who chose IS, their reasons were as follows: 57 58 188 61.9% of them accurately understood the benefits of IS, i.e., preventing 59 60 10 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 31 of 47 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 189 IDD and increasing children’s Intelligence Quotient (IQ), 14.7% of them 4 5 190 merely chose IS for superior-quality reasons and knew little about the 6 7 191 relationship between IS and IDD prevention, and 27.7% had no clear 8 9 192 reason (most explained that it was the custom) (table 2). 10 11 193 12 13 194 Table 2 Respondents’ reasons for buying IS (%) 14 15 195 For peer review only 16 17 Reasons Number 18 IS of Benefit Province City choosing 19 choosing Prevent for Good rate Others 20 IS IDD children’s quality 21 IQ 22 Heilongjiang Harbin 183 91.5 36.6 3.8 1.6 60.1 23 Jilin Changchun 175 86.6 48.6 9.7 6.3 0 24 Beijing - 202 86.7 61.4 12.9 9.4 28.2 25 Hebei Shijiazhuang 174 87.0 67.8 7.5 12.1 14.4 26 Shanxi Taiyuan 189 95.0 13.8 0 14.3 70.9 27 Shandong Ji’nan 187 93.5 73.3 27.3 20.3 11.2 28 He’nan Zhengzhou 177 88.5 54.8 15.3 26.6 27.7 29 Jiangsu Nanjing 180 72.6 64.4 25.0 23.9 7.8 30 Anhui Hefei 178 84.8 32.0 0.6 5.6 61.8 31 Hubei Wuhan 163 75.1 41.7 8.0 24.5 29.4 32 Zhejiang Hangzhou 66 32.8 57.6 16.7 21.2 0

33 Jiangxi Nanchang 184 86.8 52.2 7.6 13.6 8.7 http://bmjopen.bmj.com/ 34 Hu’nan Changsha 146 72.3 56.2 6.8 5.5 39.7 35 Guangdong Guangzhou 124 61.7 39.5 1.6 46.0 3.2 36 Guangxi Nanning 170 78.0 38.8 4.1 6.5 46.5 37 Capital city total 2498 79.5 79.5 9.8 15.0 29.0 38 Hebei Tangshan 208 94.1 81.7 17.3 9.6 5.8 39 Tianjin - 200 83.7 75.5 28.5 27.0 6.5 40 Shandong Yantai 186 93.0 90.3 3.2 8.1 4.8 Jiangsu Nantong 153 73.9 46.4 5.9 2.6 47.7

41 on September 30, 2021 by guest. Protected copyright. 42 Shanghai - 108 49.3 25.9 6.5 29.6 48.1 43 Zhejiang Zhoushan 86 42.8 68.6 12.8 18.6 0 44 Guangdong Shenzhen 144 71.6 22.9 0.7 3.5 42.4 45 Guangxi Beihai 155 66.2 21.9 2.6 18.7 58.1 46 Coastal city total 1240 72.0 72.0 10.6 14.1 25.0 47 Total 3738 76.8 51.9 10.0 14.7 27.7 48 196 Note: The reason for choosing IS was a multiple choice question. 49 197 50 51 198 Reasons for “choosing NIS” 52 53 199 Most of the 481 residents who chose NIS lived in Shanghai, Hangzhou, 54 55 200 Zhoushan and Beihai. The majority thought that they had received an 56 57 201 adequate amount of iodine from food or that they did not live in an IDD 58 59 60 11 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 32 of 47 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 202 area. Unexpectedly, 10% of the respondents who chose NIS thought IS 4 5 203 was harmful. Nearly 27.0% chose NIS because of reasons that cannot be 6 7 204 classified, such as habit, parent’s decision, or changing taste (table 3). 8 9 205 10 11 206 12 13 207 Table 3 Respondents’ reasons for choosing NIS（%） 14 15 For peer reviewReasons only 16 Not IS 17 City n % Iodine IS is IDD Patient* bad Cheap Others 18 sufficient harmful 19 endemia taste 20 Capital city Total 211 40.0 6.7 13.3 9.0 10.5 2.9 1.4 23.3 21 Coastal city Total 270 27.7 15.7 17.7 5.2 8.1 4.0 1.5 29.9 22 23 Total 481 33.1 9.9 16.6 7.3 9.1 3.5 1.5 27.0 24 208 Note: The “*” meant patients with thyroid disease for whom it is not feasible to take IS. Some provinces had small 25 209 sample sizes; the proportions refer to large ones; n: number that chose NIS; %: NIS choosing rate, the 26 27 210 reason for choosing NIS was a multiple choice question. 28 211 Reasons for “choosing both” 29 30 212 Most of the 192 residents who chose “both IS and NIS” were from 31 32 213 Hangzhou, Guangzhou and Shanghai. The two major reasons were to

33 http://bmjopen.bmj.com/ 34 214 maintain an appropriate iodine intake and to avoid the harmful effect of 35 36 215 excess iodine, accounting for 27.6% and 30.7% of the responses, 37 38 216 respectively. 39 40 217 Reasons for “does not matter”

41 on September 30, 2021 by guest. Protected copyright. 42 218 The 454 residents who chose “does not matter” mainly resided in Nanjing, 43 44 219 Wuhan, Hangzhou, Changsha, Guangzhou, Nanning, Tianjin, Shanghai 45 46 220 and Zhoushan. They tended to choose IS or NIS based on which was 47 48 221 most convenient to buy (52.0% of the respondents who had no 49 50 222 preference). 51 52 223 Awareness rates of IDD 53 54 224 Of the residents who responded to our phone calls, 4,865 answered the 55 56 225 question “Have you heard of Iodine Deficiency Disorders?”. Among 57 58 226 those, 68.7% (3,344) had heard and 31.3% (1,521) had never heard of 59 60 12 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 33 of 47 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 227 IDD. There was little difference in having heard of IDD between the 4 5 228 capital cities (68.2%) and the coastal cities (69.8%). The awareness rate 6 7 229 in Yantai (94.5%) was the highest, and in four cities (Shijiazhuang, Jinan, 8 9 230 Hangzhou and Tangshan), the awareness rates were greater than 80%. 10 11 231 Beihai and Taiyuan had the lowest awareness rates (46.2% and 44.7%, 12 13 232 respectively). There were no significant differences in the levels of 14 15 233 awarenessFor between peer age groups orreview occupation groups. only 16 17 234 Among the 3,344 residents who had heard of IDD, up to 77.1% of them 18 19 235 knew that iodine deficiency could lead to goitre, but only 11.8% knew 20 21 236 that it could also cause intelligence damage to children, and 16.9% of 22 23 237 them knew nothing about its hazard. 24 25 238 Awareness rates of IDD preventive measures 26 239 27 Of the 3,344 residents who had heard of IDD, 62.5% of them knew IS 28 240 could prevent IDD, 43.2% of them knew that kelp and purple seaweed 29 30 241 could prevent IDD, and nearly 18.7% of the residents knew nothing about 31 32 242 how to prevent IDD.

33 http://bmjopen.bmj.com/ 34 243 The IS preference rates and IDD awareness rates in each city 35 36 244 All cities could be divided into four groups according to their relationship 37 38 245 between the IS preference rate and IDD awareness rate: Uhigh IS 39 40 246 preference rate and high IDD awareness rate, such as Yantai,

41 on September 30, 2021 by guest. Protected copyright. 42 247 Shijiazhuang, Beijing, Tianjin, Nanjing, Changsha and Guangzhou; 43 44 248 Uhigh IS preference rate and low IDD awareness rate, such as Taiyuan, 45 46 249 Harbin, Nanchang, Hefei, Wuhan and Beihai; Uapproximate equal the 47 48 250 choosing rate and the awareness rate, such as Tangshan, Ji’nan, 49 50 251 Zhengzhou, Changchun, Nanning, Nantong and Shenzhen; and ④low IS 51 52 252 choosing rate and high IDD awareness rate, such as Shanghai, Zhoushan 53 54 253 and Hangzhou. Only in Yantai and Nanjing were the IDD awareness rates 55 56 254 higher than the IS preference rates. The IS choosing rates in Harbin and 57 58 255 Taiyuan were greater than 90%, but their IDD awareness rates were only 59 60 13 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 34 of 47 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 256 57.0% and 44.7%, respectively. There were only 4 cities whose IS rates 4 5 257 and awareness rates were all greater than 80% (figure 3). 6 7 258 The chi square test was used to analyse the influence of IDD awareness 8 9 259 on the IS preference rate. The results showed no differences between the 10 11 260 capital cities’ IS and NIS groups but significant differences between the 12 13 261 coastal cities’ IS and NIS groups. This could explain the weak relations 14 15 262 betweenFor the IS preference peer rates reviewand the IDD awareness only rates in the cities, 16 17 263 whereas for coastal cities, there were associations between the IS 18 19 264 preferences and IDD awareness; specifically, most coastal residents who 20 21 265 chose NIS had lower IDD awareness rates. 22 23 266 Results of the return visits 24 25 267 The total coherence rate of all questions was 85.1%. The Kappa values of 26 268 27 the coherence analysis for choosing IS, choosing NIS and IDD awareness 28 269 were 0.752, 0.693 and 0.76, respectively, and were all above the required 29 [15] 30 270 0.4, describing good coherence. 31 32 271

33 http://bmjopen.bmj.com/ 34 272 DISCUSSIONS 35 36 273 Telephone interviews are a widely used method all over the world that 37 38 274 has the advantages of wide coverage, low costs and high efficiency. The 39 40 275 survey subjects could be units or persons with fixed telephones, and the

41 on September 30, 2021 by guest. Protected copyright. 42 276 selection, supervision and training of investigators are convenient. These 43 44 277 features made it feasible to conduct a large-scale, easily acceptable survey 45 46 278 with clear content. However, telephone surveys do have biases and 47 [16] 48 279 limitations. Only after a detailed design to decrease error as much as 49 50 280 possible are the results convincible. For some interviews, such as street 51 52 281 intercept or telephone interviews, if most of the interviewees reject the 53 54 282 survey and the rejection rate is greater than 70% (i.e., the response rate 55 [17] 56 283 <30%), the representation of the random samples will be questioned; 57 58 284 hence, the interviewees will be resampled. In our study, the telephone 59 60 14 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 35 of 47 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 285 survey was conducted with unified quality control, the response rate was 4 5 286 71.3%, and the coherence rate of re-survey was above 80% (85.1%). 6 7 287 There were no significant differences in terms of age and occupation. 8 9 288 Thus, the results of this telephone survey could basically reflect the 10 11 289 subjective preferences about IS or NIS of the urban residents in China. As 12 13 290 the first telephone survey on the KAP toward IDD in China, the 14 15 291 traditionalFor PAPI methodpeer was taken,review and the results only offer references for 16 17 292 future surveys. 18 19 293 The residents’ total IS preference rate was 76.8% in the 6 coastal cities 20 21 294 and 14 capital cities in 14 provinces and 3 municipalities (i.e., Beijing, 22 23 295 Tianjin and Shanghai) where IDD elimination has been sustained for 24 25 296 nearly ten years. The reasons for choosing IS were mainly ‘IDD 26 297 27 prevention (51.9%)’ and ‘improving children’s IQ (10%)’, the remaining 28 298 residents (38%) had no idea about the relationship between IS and IDD 29 30 299 prevention and their choice of IS was only because IS is of good quality 31 32 300 or simply a custom. Another phenomenon that emerged in this study was

33 http://bmjopen.bmj.com/ 34 301 that most residents had lower IDD awareness rates than IS selection rates, 35 36 302 as was observed in Taiyuan, Hefei, Harbin and Changchun. Although the 37 38 303 residents’ IS preference rates were greater than 80%, their awareness of 39 40 304 IDD was low. An additional 33.1% of residents chose NIS, 27% of whom

41 on September 30, 2021 by guest. Protected copyright. 42 305 had no clear reason, which reflected that although IDD health education 43 44 306 activities, such as the 5.15 IDD Day, have been carried out widely and 45 46 307 continuously by the relevant departments, it is still not enough or 47 48 308 exhaustive, and the current propaganda strength and methods need further 49 50 309 improving. 51 52 310 Considering that the supply system of iodised salt is divided by province 53 54 311 in China, the KAP toward IDD of the city residents is more important. 55 56 312 Although these provinces all attained the goal of IDD elimination and 57 58 313 their capital cities and coastal cities are all developed cities, their IS 59 60 15 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 36 of 47 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 314 choosing and IDD awareness rates were different, and they should be 4 5 315 treated differently according to these results. In this study, 12 cities had 6 7 316 an IS preference rate of greater than 80%, and these were all northern 8 9 317 cities with the exception of Nanchang. Five cities had a rate greater than 10 11 318 90%. The other 11 cities with rates lower than 80% were all southern 12 13 319 cities. The IS preference rates of Hangzhou, Zhoushan and Shanghai were 14 15 320 32.8%, For 42.8% and peer 49.3%, respectively, review and were only all lower than 50%. 16 17 321 This result shows that the IS preference rates of southern city residents 18 19 322 were lower than those of northern city residents. The main reason was 20 21 323 that the people in the southern coastal areas think that they have received 22 23 324 adequate iodine from food other than iodised salt, and another reason was 24 25 325 that they thought they did not live in an iodine-deficient area, as was the 26 326 27 case with the residents of Hangzhou, Zhoushan and Shanghai. Their IDD 28 327 awareness rates were greater than their IS preference rates, and some 29 30 328 residents chose NIS despite knowing about the potential damage caused 31 32 329 by IDD. In fact, these districts had no apparent IDD prevalence, which is

33 http://bmjopen.bmj.com/ 34 330 likely attributable to the widespread IS supply. The survey “Iodine 35 36 331 nutritional status of coastal districts residents in 2009” conducted by the 37 [18] 38 332 MOH found that under the condition of consumption rates of qualified 39 40 333 IS greater than 90%, the Median Urinary Iodine (MUI) of Shanghai

41 on September 30, 2021 by guest. Protected copyright. 42 334 residents was merely between 100 µg/L and 200 µg/L, i.e., just at the 43 44 335 iodine intake appropriate level, and the MUI of pregnant women was 45 46 336 lower than 150 µg/L, which corresponds to an insufficient iodine intake 47 48 337 level. Among the residents of the Xiacheng district of Hangzhou city, 49 50 338 Zhejiang province, the consumption rate of qualified IS was 96.7%, and 51 52 339 the MUI of 8-10-year-old children was between 100 and 200 µg/L, which 53 [19] 54 340 is at the iodine intake appropriate level. If allowing these cities’ 55 56 341 residents to freely choose IS or NIS, the residents may choose NIS 57 58 342 blindly, and the achievement of controlled IDD that has been obtained 59 60 16 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 37 of 47 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 343 will be destroyed. 4 5 344 From the society and economic development aspects, supplying IS and 6 7 345 NIS simultaneously to some cities or persons is an inevitable trend. In the 8 9 346 public health domain, if a disease has a prevalence rate below 5%, it will 10 11 347 not become a public health problem. Hence, we deem that if the IDD 12 13 348 awareness rates and the correct selection rates are above 95%, iodised and 14 15 349 non-iodisedFor salt canpeer both be made review available. According only to this telephone 16 17 350 interview study, it is not a suitable time to supply IS and NIS 18 19 351 simultaneously, even in the developed cities of China. The intake of 20 21 352 non-iodised salt should be performed with caution in China. The next step 22 23 353 is to conduct a trial in which IS and NIS is simultaneously supplied in 24 25 354 appropriate cities. The selected experimental city should be a developed 26 355 27 city with no IDD prevalence in recent years, and the residents’ awareness 28 356 rate of IDD and selection rate of IS should be high. According to our 29 30 357 study, Yantai, Tangshan and Ji’nan may be appropriate cities for this 31 32 358 experiment. Before the experiment, emergency planning should be

33 http://bmjopen.bmj.com/ 34 359 performed, and extensive IDD education should be conducted. If 35 36 360 abnormal situations emerge, the emergency program will be started, or 37 38 361 the trial will even be stopped. Internationally, universal salt iodisation is 39 40 362 still considered the most cost-effective way to prevent and control iodine

41 on September 30, 2021 by guest. Protected copyright. 42 363 deficiency disorders. Since the universal salt iodisation implemented in 43 44 364 China in 1996, there has not been documentation of the citizens’ 45 46 365 preferences of salt type in a large-scale survey of the awareness of IDD 47 48 366 and its prevention. This study revealed some important findings as to why 49 50 367 people choose iodised or non-iodised salt. For instance, quality and 51 52 368 convenience, i.e., easy access, were among the reasons people would 53 54 369 purchase iodised salt, whereas concerns about the potential harm and 55 56 370 having thyroid disorders were the reasons for choosing non-iodised salt. 57 58 371 These issues should be taken into consideration when formulating future 59 60 17 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 38 of 47 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 372 health promotion messages. 4 5 373 6 7 374 Author Affiliation 8 9 375 Jun Y, Peng L, Ying L and Shou-jun L are affiliated with the Institute for 10 11 376 iodine deficiency disorders, the Centre for Endemic Disease Control, the 12 13 377 Chinese Centre for Disease Control and Prevention, and Harbin Medical 14 15 378 University,For Harbin, peer People’s Republic review of China. only 16 17 379 Dian-jun S is affiliated with the Centre for Endemic Disease Control, the 18 19 380 Chinese Centre for Disease Control and Prevention, and Harbin Medical 20 21 381 University, Harbin, People’s Republic of China. 22 23 382 24 25 383 Acknowledgments We thank the organisations that participated in this 26 384 27 survey: the National Centre for Endemic Disease Control (CEDC), 16 28 385 provincial (municipality) Centres for Disease Control (CDCs) or CEDC, 29 30 386 including those of the Beijing, Tianjin, Shanghai, and Jilin Provinces, 31 32 387 Hebei Province, Shandong Province, He’nan Province, Jiangsu Province,

33 http://bmjopen.bmj.com/ 34 388 Anhui Province, Hubei Province, Zhejiang Province, Jiangxi Province, 35 36 389 Hu’nan Province, Guangdong Province and Guangxi province, and six 37 38 390 coastal city CDCs, including those of Tangshan, Yantai, Nantong, 39 40 391 Zhoushan, Shenzhen and Beihai. The authors of this article thank those

41 on September 30, 2021 by guest. Protected copyright. 42 392 colleagues from the above organisations who took part in the survey for 43 44 393 their hard work. 45 46 394 47 48 395 Contributorship statement Jun Y wrote the project report, Peng L 49 50 396 analysed the data and wrote the manuscript, Ying L and Shou-jun L 51 52 397 designed the study and managed the project, and Dian-jun S was 53 54 398 responsible for the project and the publication. 55 56 57 399 Competing interests None. 58 59 60 18 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 39 of 47 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 400 Funding The project was funded by the Ministry of Health of China. 4 5 401 Data sharing No additional data are available. 6 7 402 Ethics approval The review board of the Harbin Medical University. 8 9 403 10 11 404 References 12 405 [1] WHO/UNICEF/ICCIDD. Assessment of iodine deficiency disorders and monitoring their 13 14 406 elimination. A guide for programme managers. Third edition, France, 2007. 15 407 [2] Todd For CH, Allain T,peer Gomo ZA, et review al. Increase in thyrotoxicosis only associated with iodine 16 17 408 supplements in Zimbabwe. Lancet. 1995,346:1563-4. 18 409 [3] Stanbury JB, Ermans AE, Bourdoux P, et al. Iodine-induced hyperthyroidism: occurrence 19 410 and epidemiology. Thyroid 1998, 8:83-100. 20 21 411 [4] Parveen S, Latif SA, Kamal MM, et al. Iodized salt induced thyrotoxicosis: Bangladesh 22 412 perspective, Mymensingh Med J. 2009,18:165-8. 23 413 [5] ICCIDD. International Council for the Control of Iodine Deficiency Disorder. CIDDS 24 25 414 Database. Current IDD Status Database.2004. 26 415 [6] Groves RM, Biemer PP, Lyberg LE, et al. Telephone survey methodology [J] John Wiley 27 28 416 &Sons Inc.1989, 2-10. 29 417 [7] Jianhua L, Hanwu M , Yongsheng W, et al. Analysis of know ledge, attitude and practice 30 418 of the influenza A /H1N1of general and immigration population in Shenzhen, Chinese Health 31 32 419 Education. 2009,25:915-8.

33 420 [8] Jin M, Shunxiang Z, Hanwu M, et al. Telephone survey on behavior risk factors of http://bmjopen.bmj.com/ 34 421 Shenzhen residents among six districts in 200. Journal of Disease Control. 2005,9:561-5 35 36 422 [9] Tao W, Su-jun L, Liang Z, et al. Investigation of Pollinosis in Beijing Residents over the 37 423 Age of 15. J Environ Health, 2008,25:403-4. 38 39 424 [10] Blair G. Obstetricians' Receptiveness to Teen Prenatal Patients Who Are Medicaid 40 425 Recipients, Health Services Research.1997,32:265-82.

41 426 [11] National Bureau of Statistics of China.Statistics Bulletin of the Cities’ Economic and on September 30, 2021 by guest. Protected copyright. 42 43 427 Social Development of 2009. 44 428 [12] Potthoff RF. Some Generazations of the Mitofsky-Waksberg techniques for random 45 429 digital dialing. Journal of the American Statistical Association, 1987, 82:409-41. 46 47 430 [13] Yi MING. http://geo.cersp.com/sJxzy/sc/200706/2658.html. 48 431 [14] http://www.moh.gov.cn/publicfiles/business/htmlfiles/zwgkzt/ptjnj/200908/42635.htm. 49 432 [15] Jiezhen W, Jing H, Yanxun L, et al. Kappa statistic value and its application. [J], Chinese 50 51 433 Health Statistical, 1995,12:46-50. 52 434 [16] Ping W, Weijing D. Telephone Survey and Its Application in Public Health. Health 53 54 435 education of China. 2004,20(6):549-50. 55 436 [17] Guohua D. A powerful tool of the statistic survey—the computer aid telephone survey 56 437 system. Newspaper of Theory Guidance, 2008,12:51-2. 57 58 438 [18] Chinese CDC, CEDC. Report of coastal residents’ iodine nutrient level,2009. 59 60 19 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 40 of 47 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 439 [19] ICCIDD Current IDD Status Database,2003. 4 5 440 [http://www.who.int/vmnis/database/iodine/countries/en/index.html]. 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 30, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 21 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 42 of 47 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 30, 2021 by guest. Protected copyright. 42 43 44 45 46 47 Figure 1 134x218mm (300 x 300 DPI) 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 43 of 47 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 Figure 2 25 77x38mm (300 x 300 DPI) 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 30, 2021 by guest. Protected copyright. 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 44 of 47 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 Figure 3 78x37mm (300 x 300 DPI) 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

2 2 8 8 8 8 5-7 5-7

5 5 4-5 4-5

8 8 Reported on page # Reported on page 2 2 7 7

cross-sectional studies cross-sectional

BMJ Open http://bmjopen.bmj.com/ on September 30, 2021 by guest. Protected copyright. For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

Page 45 of 47 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from Page 46 of 47

9 9 8 8 8 8

9 9 9 9 9 9

8 8 8 8 7 7

9-14 9-14 8 8

Main resultsMain Outcome Outcome data 15* numbers Report of outcome events or summary measures

Descriptive data 14* characteristicsGive (a) participantsof study (eg demographic, clinical, social) and information on exposures potential and

Participants Results

Statistical Statistical methods 12 ( Quantitative variables 11 how Explain quantitative were variables handled in analyses.the If applicable, describe which groupingschosen were and Study size 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 15-17 15-17

15 15 15 15 17-19 17-19

18 18 14 14

BMJ Open http://bmjopen.bmj.com/ on September 30, 2021 by guest. Protected copyright. For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

18 18 Summarise results key to reference withstudy objectives 22 22 sourcethe Give of funding and role the of fundersthe thefor studypresent if and, applicable, originalthe for study on 19 19 Discuss of study, the limitations taking into sourcesaccount biasof potential imprecision.or Discussboth direction and An Explanation and Elaboration article discusses itemeach checklist and gives methodological backgroundand published examples of transparent reporting. The STROBE checklist is best used in thiswith conjunction (freely available article onsites the Medicine of PLoS Web http://www.plosmedicine.org/, at Annals of Internal Medicine at http://www.annals.org/, Epidemiologyhttp://www.epidem.com/). at and Information on InitiativeSTROBE the is www.strobe-statement.org.available at information*Give separately cases for and controls studiesin case-controlfor applicable, if and, exposedunexposed and groups in cohort and cross-sectional studies. Note: Funding Other information Generalisability 21 Discuss (externalgeneralisabilitythe of validity) studythe results Interpretation 20 cautious Give interpretationa of overall results consideringlimitations, objectives, multiplicity of resultsanalyses, from Limitations Key Key results Discussion Other analyses 17 other Report analyses done—eg analyses of subgroups sensitivity interactions, and and analyses ( Page 47 of 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

Should both iodised and non-iodised salt be made available in Chinese cities? A cross-sectional survey

ForJournal: peerBMJ Open review only Manuscript ID: bmjopen-2014-005397.R2

Article Type: Research

Date Submitted by the Author: 18-Jun-2014

Primary Subject Health policy Heading: http://bmjopen.bmj.com/ Secondary Subject Heading: Epidemiology, Health policy, Public health

telephone interview, iodized salt, non-iodized salt, knowledge-attitude- Keywords: practice, iodine deficiency disorders

on September 30, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 46 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

41 on September 30, 2021 by guest. Protected copyright. 42 43 150081, People’s Republic of China 44 45 Full address: No.157 Baojian Road, Nan’gang District, Center for 46 47 Endemic Disease Control, Harbin Medical University, Harbin, China 48 Postal code: 150081 49 50 Telephone: 86-451-86612695 51 52 Fax: 86-451-86657674 53 54 Email: [email protected] 55 56 57 58 59 60 1 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 46 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 First Author 4 5 Name: Jun YU†, Peng LIU† 6 7 † have made the same contribution 8 9 Other Author Name:Ying LIU, Shou-jun LIU 10 11 Affiliation: Center for Endemic Disease Control, Chinese Center for 12 13 Disease Control and Prevention, Harbin Medical University, Harbin 14 15 150081,For People’s Republicpeer of Chinareview only 16 17 Keywords: telephone interview, iodized salt, non-iodized salt, 18 19 knowledge-attitude-practice, iodine deficiency disorders 20 21 Word count: 4366 words 22 23 24 25 ABSTRACT 26 27 Objective: To contribute evidence relevant to the policy of supplying 28 29 iodised salt (IS), non-iodised salt (NIS) or both in Chinese cities. 30 31 Design: Sub-national telephone interview survey. 32 Setting: China. 33 http://bmjopen.bmj.com/ 34 Participants: Totally, 24557 telephone numbers were dialled and 4833 35 36 citizens accepted the telephone interview. The telephone numbers were 37 38 randomly selected by random digit dialling and a Mitofsky-Waksberg 39 40 two-stage sampling method in 17 capital cities and six coastal cities from

41 on September 30, 2021 by guest. Protected copyright. 42 17 iodine deficiency disorder (IDD)-eliminated provinces 43 44 (municipalities). 45 46 Results: The 4833 citizens finished the telephone interview. Among them, 47 48 3738 (77.3%) citizens chose IS, 481 (10.0%) citizens chose NIS, and the 49 50 others chose both IS and NIS. The citizens’ awareness rates of IDD and 51 52 IDD preventive measures were 68.7% and 62.5%, respectively. 53 54 Conclusions: It is not a suitable time to supply IS and NIS simultaneously 55 56 in the developed cities of China, but a pilot project may be conducted in 57 58 the cities where IDD has been sustainably eliminated, there is strong 59 60 2 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 46 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 awareness of IDD and the population can make informed decisions 4 5 regarding IS. IDD health education should be further strengthened, 6 7 especially regarding the potential for IQ damage. 8 9 10 11 Keywords: telephone interview, iodized salt, non-iodized salt, 12 13 knowledge-attitude-practice, iodine deficiency disorders 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 Strengths and limitations of this study 26 (1) The article initially describes the residents’ preferences on salt type (iodised or non-iodised) and their 27 awareness of iodine deficiency disorders and its prevention measures. 28 29 (2) The telephone interview method is biased towards those residents with a fixed-line telephone. The 30 pencil-and-paper interview method, rather than computer-aided telephone interviews, were used. 31 (3) Options for explaining the iodized/non-iodized salt choice could have been more detailed. 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 30, 2021 by guest. Protected copyright. 42 43 44 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 BMJ Open Page 4 of 46 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 1 INTRODUCTION 4 5 2 Approximately 20% of the total global population resides in China, which 6 7 3 has one of the most severe iodine deficiency epidemics, in part due to the 8 9 4 external environment. Iodine Deficiency Disorders (IDD) were once 10 11 5 widespread and caused great damages. Universal Salt Iodization (USI) 12 13 6 has been adopted for IDD prevention and control and has proven to be an 14 15 7 effectiveFor measure peersince 1994. Byreview 2000, 17 provinces only (municipalities) in 16 17 8 China achieved the goal of eliminating IDD and have sustained this 18 [1] 19 9 outcome for 10 years based on the WHO criteria for IDD elimination. 20 [2-4] 21 10 However, an excess of iodine has been discovered, and new issues , 22 23 11 including the increasing rate of thyroid disease and the elevated health 24 25 12 awareness of the Chinese population, have started a debate about the 26 13 27 ability of the population to make informed choices to use iodised salt(IS) 28 14 or non-iodised salt(NIS). As we know, some developed European 29 30 15 countries, such as Belgium, Finland, France, Germany, Greece, Ireland, 31 32 16 Italy and the Netherlands, supply IS and NIS simultaneously in the

33 http://bmjopen.bmj.com/ [5] 34 17 market. Residents of these countries can purchase IS or NIS voluntarily. 35 36 18 In developed cities in provinces that have demonstrated the sustainable 37 38 19 elimination of IDD, is it the right time to supply IS and NIS 39 40 20 simultaneously? If so, can these populations remain free of IDD? Finally,

41 on September 30, 2021 by guest. Protected copyright. 42 21 is there a proper method to demarcate cities to which IS and NIS is 43 44 22 supplied simultaneously or to which only IS is supplied? To answer these 45 46 23 questions, the subjective preference of the residents about choosing IS or 47 48 24 NIS and the relevant influencing factors should be examined. 49 50 25 Door-to-door interviewing is usually used to conduct such investigations. 51 52 26 Considering the associated time and money expenditures, a 53 54 27 cross-sectional telephone interview is a feasible alternative and is a 55 56 28 widespread method that was adopted earlier and used by different 57 [6] 58 29 countries all over the world. Successful cases of the use of 59 60 4 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 46 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 30 cross-sectional telephone interviewing in China include KAP surveys on 4 [7-10] 5 31 influenza, smoking and pollinosis. 6 7 32 The application of cross-sectional telephone interviewing is based on 8 9 33 educational and economic levels, in addition to the telephone 10 11 34 popularization rate. In China, by the end of 2009, the national average 12 13 35 Urban Per Capita Disposable Income was 17,175 RMBY, and some 14 15 36 developedFor cities (municipalities)peer review in the IDD-eliminated only provinces were 16 [11] 17 37 all at or beyond the national average (table 1). Additionally, the 18 19 38 national fixed line penetration had increased to 314 million people (23.5 20 21 39 telephones/100 persons) and 212 million people in the cities (34.1 22 23 40 telephones/100 persons). These facts made telephone interview a 24 25 41 representative and feasible way to research the residents’ KAP toward 26 42 27 IDD. 28 43 As far as we know, there were no such records about the residents’ 29 30 44 subjective preferences between IS and NIS, and no research on the 31 32 45 residents’ awareness of IDD in developed cities (municipalities) has been

33 http://bmjopen.bmj.com/ 34 46 conducted in recent years. Hence, the National Centre for Endemic 35 36 47 Disease Control (CEDC) launched this project in April, 2010, with the 37 38 48 aim of obtaining data on the above aspects to decide whether supplying 39 40 49 IS and NIS simultaneously is feasible in developed cities in

41 on September 30, 2021 by guest. Protected copyright. 42 50 IDD-eliminated provinces. The methods and results may provide 43 44 51 evidence or references for similar regions in the world. 45 46 52 47 48 53 METHODS 49 50 54 51 52 55 City sites to be investigated 53 54 56 The 17 capital cities of the 17 provinces (3 municipalities, 6 coastal 55 56 57 provinces and 8 inland provinces) that had achieved the goal of IDD 57 58 58 elimination in 2000 were selected (table 1). One coastal city was also 59 60 5 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 46 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 59 selected in each of the above coastal provinces. A total of 23 cities were 4 5 60 included in the study. 6 7 61 Table 1 Provinces and cities surveyed 8 9 Provincial capital Coastal cities Category Provinces 10 Name Tel* Income** Name Tel * Income ** 11 12 Hebei Shijiazhuang - 16,607 Tangshan 24.58 18,053 13 Yantai 28.27 21,125 14 Shandong Ji’nan 39.28 22,272 15 ForJiangsu peer Nanjing review 38.91 25,504 onlyNantong 35.80 21,001 16 Coastal province Zhoushan 53.45 24,082 17 Zhejiang Hangzhou 48.23 26,864 18 Guangdong Guangzhou 61.28 27,610 Shenzhen 52.64 29,244 19 Beihai - 15,134 20 Guangxi Nanning 16.61 16,254 21 Heilongjiang Harbin 21.66 15,887 22 23 Jilin Changchun - 16,277 24 Shanxi Taiyuan 44.31 15,607 25 26 Inland He’nan Zhengzhou 29.28 17,117 province 27 Anhui Hefei 32.47 17,158 28 29 Hubei Wuhan 40.57 18,385 30 Jiangxi Nanchang 32.37 16,472 31

32 Hu’nan Changsha 33.28 20,238

33 Beijing Beijing 50.89 26,738 http://bmjopen.bmj.com/ 34 35 Municipalities Shanghai Shanghai 48.70 28,838 36 37 Tianjin Tianjin 31.36 21,430 38 62 All data marked with “*” and “**”were obtained from the Statistics Bulletin of the Cities’ 39 40 63 Economic and Social Development of 2009: “*” fix telephones/100 persons, “**” Urban per

41 64 capita disposable income (RMBY). The national fixed line penetration was 23.50 telephones/100 on September 30, 2021 by guest. Protected copyright. 42 65 persons, and the urban per capita disposable income was 17,175 RMBY in 2009. 43 44 66 45 46 67 Target population 47 48 68 Citizens who had fixed telephones in the 23 selected cities. 49 50 69 The main contents of the questionnaire: 51 52 70 If IS and NIS are supplied simultaneously, which will you choose? 53 54 71 And why? 55 56 72 Have you heard of IDD? If so, what are the effects? How is IDD 57 58 73 prevented? 59 60 6 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 46 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 74 Can you tell us your occupation and age group? 4 5 75 Sampling method and sampling size 6 7 76 Random Digit Dialling and the Mitofsky-Waksberg two-stage sampling 8 [12] 9 77 method were used in this study. In each city, only the main urban zones 10 11 78 were investigated, and two districts were selected randomly from each 12 13 79 main urban zone. At least 200 effective samples were sampled in each 14 15 80 city includedFor in peer the survey. review The sample quota only was divided evenly 16 17 81 between each district. The sample size was calculated according to 18 19 82 formula 1; the positive response rate ( π ) was approximately 70% 20 21 83 according to the pilot survey, and the allowable error (δ ) was 0.5. Thus, 22 23 84 the calculated sample size was 170 responses per city; therefore, 200 24 25 85 responses per city were considered adequate. 26  2 (1−× ππ ) 27 86 n =  α  π Positive rate δ Allowable error 28  δ  29 30 87 Formula 1 31 32 88 Telephone number acquisition

33 http://bmjopen.bmj.com/ 89 If the fixed telephone of the neighbourhood offices in the selected district 34 35 90 was 7- or 8-digit numbers, the first 3 or 4 digits were deemed to be the 36 37 91 direction code. The last 4 digits were obtained by the Random function 38 ® 39 92 (=Rand ()×10000) in Microsoft Office Excel . If the function produced 40

41 93 randomised numbers less than 4 digits, and “0” was supplemented in on September 30, 2021 by guest. Protected copyright. 42 43 94 front of the generated numbers. The number of telephone numbers 44 45 95 created randomly should be at least 5 times the effective sample size, i.e., 46 47 96 at least 1000 telephone numbers should be created per city. 48 49 97 Quality control 50 51 98 The research was approved by the review board of the Harbin Medical 52 53 99 University. Participant consent was acquired when they answered the 54 55 100 telephone. The survey program and survey result database were designed 56 57 101 and delivered to the project cities by the national CEDC. The survey 58 59 60 7 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 46 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 102 program was improved based on a pilot survey, and the questionnaire had 4 5 103 clear questions that prompted simple answers and avoided providing hints 6 7 104 to the residents. Sensitive questions, such as age, were designed as 8 9 105 “Which age group do you belong to? <20, 20-30, 30-40... ”. 10 11 106 In the program, the investigation was carried out in a unified manner, 12 13 107 including the order and pattern of the questions, the dialling time (18:30 - 14 15 108 21:00), Forand instructions peer for making review notes of the answersonly and filling the 16 17 109 database. 18 19 110 To avoid investigation and report biases and to improve the research 20 21 111 subjects’ cooperation, the interviewers were trained on their attitude, 22 23 112 mood, and countermeasures in case of emergency by the project 24 25 113 provinces themselves according to the rules of the survey program. For 26 114 27 example, some related questions put forward by residents could be 28 115 answered only after the interview was finished to avoid bias. 29 30 116 To guarantee the reliability of the research information, a return visit was 31 32 117 conducted after a project city finished the telephone interviews. The

33 http://bmjopen.bmj.com/ 34 118 quality control officer of the provincial CDC or CEDC visited the citizens 35 36 119 who had accepted the interview. At least 10% of the respondents were 37 38 120 randomly sampled to complete the return visit questionnaire. When the 39 40 121 coincidence rate of survey information was greater than 80%, the

41 on September 30, 2021 by guest. Protected copyright. 42 122 investigation results of the city can be regarded as qualified, and 43 44 123 otherwise, the city should be reinvestigated. 45 46 124 Definition of telephone number 47 48 125 Success: respondent finished the entire questionnaire; Rejection: 49 50 126 respondent refused to answer or did not finish the questionnaire; No 51 52 127 answer: the phone was busy both the first time and when re-dialling after 53 54 128 5 minutes; Invalid number: the phone number was vacant, fax, 55 56 129 switchboard or phone quit. 57 58 130 Data processing and statistical analysis 59 60 8 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 46 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 ® 3 131 Epi info 3.5.1 (CDC, Atlanta, GA USA) was used for the database 4 ® 5 132 design, and SPSS 17.0 was utilized for analysis. The level of 6 7 133 significance was set at 0.05. 8 9 134 10 11 135 RESULTS 12 13 136 Response situation 14 15 137 The flowFor chart is presentedpeer in figurereview 1. The survey only included 235 districts 16 17 138 of 120 main urban zones in 23 cities. In total, 446 direction numbers were 18 19 139 chosen, and 24,557 telephone numbers were dialled, of which 13,215 20 21 140 (53.8%) were valid and 11,342 were invalid numbers. An invalid number 22 23 141 was a number that was not used as a telephone number. Among the 24 25 142 13,215 valid telephone numbers, 4,833 successfully finished the 26 143 27 questionnaire, 1,949 answered the phone but refused to answer the 28 144 questions, and 6,433 telephones were not answered. The answered 29 30 145 numbers (successes and rejections) were 6,782 and accounted for 51.3% 31 32 146 of the valid numbers. The total response rate (4833/(4833+1949)*100%)

33 http://bmjopen.bmj.com/ 34 147 was 71.3% (figure 2).. 35 36 148 Basic information about the interviewee 37 38 149 The age and occupation questions were answered by 4,861 and 4,857 39 40 150 residents, respectively. The age and occupation proportion were similar

41 on September 30, 2021 by guest. Protected copyright. 42 151 between the capital city (including Beijing) and the coastal city (including 43 44 152 Tianjin and Shanghai) residents. The residents’ age were mainly 45 46 153 distributed between 20 and 60 years, and this range accounted for 80% of 47 48 154 the respondents. However, compared with the population composition 49 [13] 50 155 from the ‘2009 China Health Statistical Yearbook’ and the Chinese 51 [14] 52 156 population pyramid, the 10-20 age group had a smaller constituent 53 54 157 ratio. This may lead to a non-representative sample of the Chinese 55 56 158 population, although this may be acceptable in this study because 57 58 159 children rarely buy salt. The residents were evenly distributed in each 59 60 9 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 46 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 160 profession (accounting for approximately 10%), except for the 4 5 161 professions of servicemen, students and irregular professions, which had 6 7 162 small sample sizes. 8 9 163 10 11 164 Proportion of respondents choosing IS or NIS 12 13 165 Of the residents investigated, 4,865 answered the question ‘If IS and NIS 14 15 166 are suppliedFor simultaneously, peer review which will you only choose?’ The total 16 17 167 percentage of respondents ‘choosing IS’ was 76.8%, ‘choosing NIS’ was 18 19 168 9.9%, ‘both’ was 3.9% and ‘does not matter’ was 9.3%. The respondents 20 21 169 from the capital cities chose IS at a higher rate (79.5%) than those from 22 23 170 the coastal cities (72.0%). The percentages of respondents choosing NIS 24 25 171 was opposite between the city and coastal residents (6.7% and 15.7%, 26 172 27 respectively), and the ratio of respondents who chose “both” (<5%) and 28 173 ‘does not matter’ (<10%) were nearly the same in the two areas. Of the 23 29 30 174 cities, there were 12 cities whose IS selection rate was higher than 80%, 8 31 32 175 cities in which the IS selection rate was between 60-80%, and 3 cities

33 http://bmjopen.bmj.com/ 34 176 (Shanghai: 32.8%, Zhoushan: 42.8% and Hangzhou: 49.3%) in which the 35 36 177 IS selection rate was lower than 50% (table 2, table 3). 37 38 178 Reasons for “choosing IS” 39 40 179 Of the 3,738 residents who chose IS, their reasons were as follows: 61.9%

41 on September 30, 2021 by guest. Protected copyright. 42 180 of them accurately understood the benefits of IS, i.e., preventing IDD and 43 44 181 increasing children’s Intelligence Quotient (IQ), 14.7% of them merely 45 46 182 chose IS for superior-quality reasons and knew little about the 47 48 183 relationship between IS and IDD prevention, and 27.7% had no clear 49 50 184 reason (most explained that it was the custom) (table 2). 51 52 185 53 54 186 55 56 187 Table 2 Respondents’ reasons for buying IS (%) 57 58 Province City Number IS Reasons 59 60 10 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 46 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 of choosing Benefit choosing rate Prevent for Good 4 Others 5 IS IDD children’s quality 6 IQ 7 Heilongjiang Harbin 183 91.5 36.6 3.8 1.6 60.1 8 Jilin Changchun 175 86.6 48.6 9.7 6.3 0 9 Beijing - 202 86.7 61.4 12.9 9.4 28.2 10 Hebei Shijiazhuang 174 87.0 67.8 7.5 12.1 14.4 11 Shanxi Taiyuan 189 95.0 13.8 0 14.3 70.9 12 Shandong Ji’nan 187 93.5 73.3 27.3 20.3 11.2 13 He’nan Zhengzhou 177 88.5 54.8 15.3 26.6 27.7 14 Jiangsu Nanjing 180 72.6 64.4 25.0 23.9 7.8 15 AnhuiFor Hefei peer 178 review 84.8 32.0 only 0.6 5.6 61.8 16 Hubei Wuhan 163 75.1 41.7 8.0 24.5 29.4 17 Zhejiang Hangzhou 66 32.8 57.6 16.7 21.2 0 18 Jiangxi Nanchang 184 86.8 52.2 7.6 13.6 8.7 Hu’nan Changsha 146 72.3 56.2 6.8 5.5 39.7 19 Guangdong Guangzhou 124 61.7 39.5 1.6 46.0 3.2 20 Guangxi Nanning 170 78.0 38.8 4.1 6.5 46.5 21 Capital city total 2498 79.5 79.5 9.8 15.0 29.0 22 Hebei Tangshan 208 94.1 81.7 17.3 9.6 5.8 23 Tianjin - 200 83.7 75.5 28.5 27.0 6.5 24 Shandong Yantai 186 93.0 90.3 3.2 8.1 4.8 25 Jiangsu Nantong 153 73.9 46.4 5.9 2.6 47.7 26 Shanghai - 108 49.3 25.9 6.5 29.6 48.1 27 Zhejiang Zhoushan 86 42.8 68.6 12.8 18.6 0 28 Guangdong Shenzhen 144 71.6 22.9 0.7 3.5 42.4 29 Guangxi Beihai 155 66.2 21.9 2.6 18.7 58.1 30 Coastal city total 1240 72.0 72.0 10.6 14.1 25.0 31 Total 3738 76.8 51.9 10.0 14.7 27.7 32 188 Note: The reason for choosing IS was a multiple choice question. 33 http://bmjopen.bmj.com/ 34 189 35 36 190 Reasons for “choosing NIS” 37 38 191 Most of the 481 residents who chose NIS lived in Shanghai, Hangzhou, 39 40 192 Zhoushan and Beihai. The majority thought that they had received an

41 on September 30, 2021 by guest. Protected copyright. 42 193 adequate amount of iodine from food or that they did not live in an IDD 43 44 194 area. Unexpectedly, 10% of the respondents who chose NIS thought IS 45 46 195 was harmful. Nearly 27.0% chose NIS because of reasons that cannot be 47 48 196 classified, such as habit, parent’s decision, or changing taste (table 3). 49 50 197 51 52 198 53 54 199 55 56 200 Table 3 Respondents’ reasons for choosing NIS（%） 57 58 City n % Reasons 59 60 11 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 46 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 Not IS Iodine IS is 4 IDD Patient* bad Cheap Others 5 sufficient harmful 6 endemia taste 7 Capital city Total 211 40.0 6.7 13.3 9.0 10.5 2.9 1.4 23.3 8 Coastal city Total 270 27.7 15.7 17.7 5.2 8.1 4.0 1.5 29.9 9 Total 481 33.1 9.9 16.6 7.3 9.1 3.5 1.5 27.0 10 201 Note: The “*” meant patients with thyroid disease for whom it is not feasible to take IS. Some provinces had small 11 12 202 sample sizes; the proportions refer to large ones; n: number that chose NIS; %: NIS choosing rate, the 13 203 reason for choosing NIS was a multiple choice question. 14 15 204 ReasonsFor for “choosing peer both” review only 16 17 205 Most of the 192 residents who chose “both IS and NIS” were from 18 19 206 Hangzhou, Guangzhou and Shanghai. The two major reasons were to 20 21 207 maintain an appropriate iodine intake and to avoid the harmful effect of 22 23 208 excess iodine, accounting for 27.6% and 30.7% of the responses, 24 25 209 respectively. 26 27 210 Reasons for “does not matter” 28 29 211 The 454 residents who chose “does not matter” mainly resided in Nanjing, 30 31 212 Wuhan, Hangzhou, Changsha, Guangzhou, Nanning, Tianjin, Shanghai 32 213 and Zhoushan. They tended to choose IS or NIS based on which was 33 http://bmjopen.bmj.com/ 34 214 35 most convenient to buy (52.0% of the respondents who had no 36 215 preference). 37 38 216 Awareness rates of IDD 39 40 217 Of the residents who responded to our phone calls, 4,865 answered the

41 on September 30, 2021 by guest. Protected copyright. 42 218 question “Have you heard of Iodine Deficiency Disorders?”. Among 43 44 219 those, 68.7% (3,344) had heard and 31.3% (1,521) had never heard of 45 46 220 IDD. There was little difference in having heard of IDD between the 47 48 221 capital cities (68.2%) and the coastal cities (69.8%). The awareness rate 49 50 222 in Yantai (94.5%) was the highest, and in four cities (Shijiazhuang, Jinan, 51 52 223 Hangzhou and Tangshan), the awareness rates were greater than 80%. 53 54 224 Beihai and Taiyuan had the lowest awareness rates (46.2% and 44.7%, 55 56 225 respectively). There were no significant differences in the levels of 57 58 226 awareness between age groups or occupation groups. 59 60 12 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 46 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 227 Among the 3,344 residents who had heard of IDD, up to 77.1% of them 4 5 228 knew that iodine deficiency could lead to goitre, but only 11.8% knew 6 7 229 that it could also cause intelligence damage to children, and 16.9% of 8 9 230 them knew nothing about its hazard. 10 11 231 Awareness rates of IDD preventive measures 12 13 232 Of the 3,344 residents who had heard of IDD, 62.5% of them knew IS 14 15 233 could preventFor IDD, peer 43.2% of themreview knew that kelp only and purple seaweed 16 17 234 could prevent IDD, and nearly 18.7% of the residents knew nothing about 18 19 235 how to prevent IDD. 20 21 236 The IS preference rates and IDD awareness rates in each city 22 23 237 All cities could be divided into four groups according to their relationship 24 25 238 between the IS preference rate and IDD awareness rate: Uhigh IS 26 239 27 preference rate and high IDD awareness rate, such as Yantai, 28 240 Shijiazhuang, Beijing, Tianjin, Nanjing, Changsha and Guangzhou; 29 30 241 Uhigh IS preference rate and low IDD awareness rate, such as Taiyuan, 31 32 242 Harbin, Nanchang, Hefei, Wuhan and Beihai; Uapproximate equal the

33 http://bmjopen.bmj.com/ 34 243 choosing rate and the awareness rate, such as Tangshan, Ji’nan, 35 36 244 Zhengzhou, Changchun, Nanning, Nantong and Shenzhen; and ④low IS 37 38 245 choosing rate and high IDD awareness rate, such as Shanghai, Zhoushan 39 40 246 and Hangzhou. Only in Yantai and Nanjing were the IDD awareness rates

41 on September 30, 2021 by guest. Protected copyright. 42 247 higher than the IS preference rates. The IS choosing rates in Harbin and 43 44 248 Taiyuan were greater than 90%, but their IDD awareness rates were only 45 46 249 57.0% and 44.7%, respectively. There were only 4 cities whose IS rates 47 48 250 and awareness rates were all greater than 80% (table 4). 49 50 251 51 52 252 53 54 253 55 56 254 Table 4 The IDD awareness rates and IS preference rates（%） 57 58 59 60 13 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 46 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 4 City IS rate(%) Acknowledge rate(%) 5 Taiyuan 95.0 44.7 6 7 Tangshan 94.1 85.1 8 Jinan 93.5 81.0 9 10 Yantai 93.0 94.5 11 Harbin 91.5 57.0 12 Zhengzhou 88.5 74.0 13 14 Shijiazhuang 87.0 85.0 15 For peer review only 16 Nanchang 86.8 67.5 17 Beijing 86.7 79.8 18 19 Changchun 86.6 65.8 20 Hefei 84.8 52.9 21 Tianjin 83.7 76.6 22 23 Nanning 78.0 63.8 24 Wuhan 75.1 59.0 25 26 Nantong 73.9 62.8 27 Nanjing 72.6 74.2 28 Changsha 72.3 69.8 29 30 Shenzhen 71.6 60.7 31 Beihai 66.2 46.2 32 Guangzhou 61.7 65.2 33 http://bmjopen.bmj.com/ 34 Shanghai 49.3 57.5 35 36 Zhoushan 42.8 77.6 37 Hangzhou 32.8 81.1 38 39 255 40 256 The chi square test was used to analyse the influence of IDD awareness

41 on September 30, 2021 by guest. Protected copyright. 42 257 on the IS preference rate. The results showed no difference between the 43 44 258 capital cities’ IS and NIS groups but significant difference between the 45 46 259 coastal cities’ IS and NIS groups. This could explain the weak relations 47 48 260 between the IS preference rates and the IDD awareness rates in the cities, 49 50 261 whereas for coastal cities, there were associations between the IS 51 52 262 preferences and IDD awareness; specifically, most coastal residents who 53 54 263 chose NIS had lower IDD awareness rates. 55 56 264 Results of the return visits 57 58 265 The total coherence rate of all questions was 85.1%. The Kappa values of 59 60 14 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 46 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 266 the coherence analysis for choosing IS, choosing NIS and IDD awareness 4 5 267 were 0.752, 0.693 and 0.76, respectively, and were all above the required 6 [15] 7 268 0.4, describing good coherence. 8 9 269 10 11 270 DISCUSSIONS 12 [16] 13 271 Telephone interview has the advantages . However, if the response 14 [17] 15 272 rate<30%,For the representationpeer reviewof the random samples only will be questioned. 16 17 273 In our study, the response rate was 71.3%, and the coherence rate of 18 19 274 re-survey was above 80%. One limitation was that it was the first 20 21 275 telephone survey on the KAP toward IDD in China; the traditional PAPI 22 23 276 method was taken rather than CATI. The other, some “other reason” 24 25 277 should be recorded more carefully. 26 278 27 The 6 coastal cities and 14 capital cities in 14 provinces and 3 28 279 municipalities had IDD elimination for nearly ten years. A phenomenon 29 30 280 that emerged in this study was that most residents had lower IDD 31 32 281 awareness rates than IS selection rates, as was observed in Taiyuan, Hefei,

33 http://bmjopen.bmj.com/ 34 282 Harbin and Changchun. Although IDD health education activities, such as 35 36 283 the 5.15 IDD Day, have been carried out widely and continuously by the 37 38 284 relevant departments, it is still not enough or exhaustive, and the current 39 40 285 propaganda strength and methods need further improving.

41 on September 30, 2021 by guest. Protected copyright. 42 286 Considering that the supply system of iodised salt is divided by province 43 44 287 in China, the KAP toward IDD of the city residents is more important. 45 46 288 Although these provinces all attained the goal of IDD elimination and 47 48 289 their capital cities and coastal cities are all developed cities, their IS 49 50 290 choosing and IDD awareness rates were different, and they should be 51 52 291 treated differently according to these results. The result shows that the IS 53 54 292 preference rates of southern city residents were lower than those of 55 56 293 northern city residents. The main reason was that the people in the 57 58 294 southern coastal areas think that they have received adequate iodine from 59 60 15 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 46 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 295 food other than iodised salt, and another reason was that they thought 4 5 296 they did not live in an iodine-deficient area, as was the case with the 6 7 297 residents of Hangzhou, Zhoushan and Shanghai. Their IDD awareness 8 9 298 rates were greater than their IS preference rates, and some residents chose 10 11 299 NIS despite knowing about the potential damage caused by IDD. In fact, 12 13 300 these districts had no apparent IDD prevalence, which is likely 14 15 301 attributableFor to the peer widespread ISreview supply. The survey only “Iodine nutritional 16 [18] 17 302 status of coastal districts residents in 2009” conducted by the MOH 18 19 303 found that under the condition of consumption rates of qualified IS 20 21 304 greater than 90%, the Median Urinary Iodine (MUI) of Shanghai 22 23 305 residents was merely between 100 µg/L and 200 µg/L, i.e., just at the 24 25 306 iodine intake appropriate level, and the MUI of pregnant women was 26 307 27 lower than 150 µg/L, which corresponds to an insufficient iodine intake 28 308 level. Among the residents of the Xiacheng district of Hangzhou city, 29 30 309 Zhejiang province, the consumption rate of qualified IS was 96.7%, and 31 32 310 the MUI of 8-10-year-old children was between 100 and 200 µg/L, which

33 http://bmjopen.bmj.com/ [19] 34 311 is at the iodine intake appropriate level. If allowing these cities’ 35 36 312 residents to freely choose IS or NIS, the residents may choose NIS 37 38 313 blindly, and the achievement of controlled IDD that has been obtained 39 40 314 will be destroyed.

41 on September 30, 2021 by guest. Protected copyright. 42 315 From the society and economic development aspects, supplying IS and 43 44 316 NIS simultaneously to some cities or persons is an inevitable trend. In the 45 46 317 public health domain, if a disease has a prevalence rate below 5%, it will 47 48 318 not become a public health problem. Hence, we deem that if the IDD 49 50 319 awareness rates and the correct selection rates are above 95%, iodised and 51 52 320 non-iodised salt can both be made available. According to this telephone 53 54 321 interview study, it is not a suitable time to supply IS and NIS 55 56 322 simultaneously, even in the developed cities of China. The intake of 57 58 323 non-iodised salt should be performed with caution in China. The next step 59 60 16 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 46 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 324 is to conduct a trial in which IS and NIS is simultaneously supplied in 4 5 325 appropriate cities. The selected experimental city should be a developed 6 7 326 city with no IDD prevalence in recent years, and the residents’ awareness 8 9 327 rate of IDD and selection rate of IS should be high. According to our 10 11 328 study, Yantai, Tangshan and Ji’nan may be appropriate cities for this 12 13 329 experiment. Before the experiment, emergency planning should be 14 15 330 performed,For and peer extensive IDD review education should only be conducted. If 16 17 331 abnormal situations emerge, the emergency program will be started, or 18 19 332 the trial will even be stopped. Currently, Universal salt iodization was 20 21 333 carried out as a mandatory effective strategy to prevent IDD in China 22 23 334 since 1996. However, no documentation was found about the residents’ 24 25 335 preferences on salt type (IS or NIS) based on their awareness of IDD and 26 336 27 its prevention measure knowledge. Hence, the study tried to find out the 28 337 reason why people choose iodised or non-iodised salt. Some possible 29 30 338 reasons were enlisted including preventing IDD, improve children’s IQ, 31 32 339 good quality and convenience, etc, were main reasons people would

33 http://bmjopen.bmj.com/ 34 340 purchase iodised salt, while, regarding to the potential harm and having 35 36 341 thyroid disorders were the reasons for choosing non-iodised salt. These 37 38 342 results were important for further research and prevention measures and 39 40 343 policy decision.

41 on September 30, 2021 by guest. Protected copyright. 42 344 43 44 345 45 46 346 47 48 347 49 50 348 51 52 349 53 54 350 55 56 351 57 58 352 Acknowledgments We thank the organisations that participated in this 59 60 17 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 46 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 353 survey: the National Centre for Endemic Disease Control (CEDC), 16 4 5 354 provincial (municipality) Centres for Disease Control (CDCs) or CEDC, 6 7 355 including those of the Beijing, Tianjin, Shanghai, and Jilin Provinces, 8 9 356 Hebei Province, Shandong Province, He’nan Province, Jiangsu Province, 10 11 357 Anhui Province, Hubei Province, Zhejiang Province, Jiangxi Province, 12 13 358 Hu’nan Province, Guangdong Province and Guangxi province, and six 14 15 359 coastal For city CDCs, peer including review those of Tangshan, only Yantai, Nantong, 16 17 360 Zhoushan, Shenzhen and Beihai. The authors of this article thank those 18 19 361 colleagues from the above organisations who took part in the survey for 20 21 362 their hard work. We also thank Elsevier English Language Editing for 22 23 363 their English editing. 24 25 364 26 365 27 Contributorship statement Jun Y wrote the project report, Peng L 28 366 analysed the data and wrote the manuscript, Ying L and Shou-jun L 29 30 367 designed the study and managed the project, and Dian-jun S was 31 32 368 responsible for the project and the publication.

33 http://bmjopen.bmj.com/ 34 35 369 Competing interests None. 36 37 38 370 Funding The project was funded by the Ministry of Health of China. 39 40 371 Data sharing No additional data are available.

41 on September 30, 2021 by guest. Protected copyright. 42 372 Ethics approval The review board of the Harbin Medical University. 43 44 373 45 46 374 47 375 48 49 376 50 377 51 52 378 53 379 54 55 380 56 381 References 57 58 59 60 18 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 46 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 382 [1] WHO/UNICEF/ICCIDD. Assessment of iodine deficiency disorders and monitoring their 4 5 383 elimination. A guide for programme managers. Third edition, France, 2007. 6 384 [2] Todd CH, Allain T, Gomo ZA, et al. Increase in thyrotoxicosis associated with iodine 7 385 supplements in Zimbabwe. Lancet. 1995,346:1563-4. 8 9 386 [3] Stanbury JB, Ermans AE, Bourdoux P, et al. Iodine-induced hyperthyroidism: occurrence 10 387 and epidemiology. Thyroid 1998, 8:83-100. 11 388 [4] Parveen S, Latif SA, Kamal MM, et al. Iodized salt induced thyrotoxicosis: Bangladesh 12 13 389 perspective, Mymensingh Med J. 2009,18:165-8. 14 390 [5] ICCIDD. International Council for the Control of Iodine Deficiency Disorder. CIDDS 15 For peer review only 16 391 Database. Current IDD Status Database.2004. 17 392 [6] Groves RM, Biemer PP, Lyberg LE, et al. Telephone survey methodology [J] John Wiley 18 393 &Sons Inc.1989, 2-10. 19 20 394 [7] Jianhua L, Hanwu M , Yongsheng W, et al. Analysis of know ledge, attitude and practice 21 395 of the influenza A /H1N1of general and immigration population in Shenzhen, Chinese Health 22 396 Education. 2009,25:915-8. 23 24 397 [8] Jin M, Shunxiang Z, Hanwu M, et al. Telephone survey on behavior risk factors of 25 398 Shenzhen residents among six districts in 200. Journal of Disease Control. 2005,9:561-5 26 399 [9] Tao W, Su-jun L, Liang Z, et al. Investigation of Pollinosis in Beijing Residents over the 27 28 400 Age of 15. J Environ Health, 2008,25:403-4. 29 401 [10] Blair G. Obstetricians' Receptiveness to Teen Prenatal Patients Who Are Medicaid 30 31 402 Recipients, Health Services Research.1997,32:265-82. 32 403 [11] National Bureau of Statistics of China.Statistics Bulletin of the Cities’ Economic and

33 404 Social Development of 2009. http://bmjopen.bmj.com/ 34 35 405 [12] Potthoff RF. Some Generazations of the Mitofsky-Waksberg techniques for random 36 406 digital dialing. Journal of the American Statistical Association, 1987, 82:409-41. 37 407 [13] Yi MING. http://geo.cersp.com/sJxzy/sc/200706/2658.html. 38 39 408 [14] http://www.moh.gov.cn/publicfiles/business/htmlfiles/zwgkzt/ptjnj/200908/42635.htm. 40 409 [15] Jiezhen W, Jing H, Yanxun L, et al. Kappa statistic value and its application. [J], Chinese

41 on September 30, 2021 by guest. Protected copyright. 42 410 Health Statistical, 1995,12:46-50. 43 411 [16] Ping W, Weijing D. Telephone Survey and Its Application in Public Health. Health 44 412 education of China. 2004,20(6):549-50. 45 46 413 [17] Guohua D. A powerful tool of the statistic survey—the computer aid telephone survey 47 414 system. Newspaper of Theory Guidance, 2008,12:51-2. 48 415 [18] Chinese CDC, CEDC. Report of coastal residents’ iodine nutrient level,2009. 49 50 416 [19] ICCIDD Current IDD Status Database,2003. 51 417 [http://www.who.int/vmnis/database/iodine/countries/en/index.html]. 52 53 54 55 56 57 58 59 60 19 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 46 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from

1 2 3 4 Figure legend 5 6 Figure 1 Flow chart 7 8 Figure 2 The answer status of the sampled telephone numbers 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 30, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 20 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 46 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 Title page 7 8 Should both iodised and non-iodised salt be made available in 9 10 Chinese cities? A cross-sectional survey 11 † † 12 Jun Y , Peng L , Ying L, Shou-jun L, Dian-jun S* 13 14 Corresponding author 15 Name: Dian-junFor SUN peer review only 16 17 Degree: M.D. 18 19 Position: Chief of Center for Endemic Disease Control 20 21 Affiliation: Center for Endemic Disease Control, Chinese Center for 22 Disease Control and Prevention, Harbin Medical University, Harbin 23 24 150081, People’s Republic of China 25 26 Full address: No.157 Baojian Road, Nan’gang District, Center for 27 Endemic Disease Control, Harbin Medical University, Harbin, China 28 29 Postal code: 150081 30 31 Telephone: 86-451-86612695 32 Fax: 86-451-86657674 33 http://bmjopen.bmj.com/ 34 Email: [email protected] 35 36 37 38 First Author 39 † † 40 Name: Jun YU , Peng LIU † 41 have made the same contribution on September 30, 2021 by guest. Protected copyright. 42 43 Other Author Name:Ying LIU, Shou-jun LIU 44 45 Affiliation: Center for Endemic Disease Control, Chinese Center for 46 Disease Control and Prevention, Harbin Medical University, Harbin 47 48 150081, People’s Republic of China 49 50 Keywords: telephone interview, iodized salt, non-iodized salt, 51 52 knowledge-attitude-practice, iodine deficiency disorders 53 Word count: 4366 words 54 55 56 57 1 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 46 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 Should both iodised and non-iodised salt be made available in 7 8 Chinese cities? A cross-sectional survey 9 10 ABSTRACT 11 12 Objective: To contribute evidence relevant to the policy of supplying 13 iodised salt (IS), non-iodised salt (NIS) or both in Chinese cities. 14 15 Design: Sub-nationalFor telephone peer interview survey.review only 16 17 Setting: China. 18 19 Participants: Totally, 24557 telephone numbers were dialled and The 20 4833 citizens accepted the telephone interview. The telephone numbers 21 22 were randomly selected by random digit dialling and a 23 24 Mitofsky-Waksberg two-stage sampling method in 17 capital cities and 25 six coastal cities from 17 iodine deficiency disorder (IDD)-eliminated 26 27 provinces (municipalities). 28 29 Results: The 4833 citizens finished the telephone interview. Among them, 30 31 3738 (77.3%) citizens chose IS, 481 (10.0%) citizens chose NIS, and the 32 others chose both IS and NIS. The citizens’ awareness rates of IDD and

33 http://bmjopen.bmj.com/ 34 IDD preventive measures were 68.7% and 62.5%, respectively. 35 36 Conclusions: It is not a suitable time to supply IS and NIS simultaneously 37 38 in the developed cities of China, but a pilot project may be conducted in 39 the cities where IDD has been sustainably eliminated, there is strong 40

41 awareness of IDD and the population can make informed decisions on September 30, 2021 by guest. Protected copyright. 42 43 regarding IS. IDD health education should be further strengthened, 44 45 especially regarding the potential for IQ damage. 46 47 48 Keywords: telephone interview, iodized salt, non-iodized salt, 49 50 knowledge-attitude-practice, iodine deficiency disorders 51 52 53 54 55 56 57 2 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 46 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 7 8 9 10 Strengths and limitations of this study 11 12 The vast majority (77.3%) of the population of China living in 13 14 developed cities chose iodized rather than non-iodized salt. 15 For peer review only 16 Over half (68.7%) of the population of China living in developed 17 18 19 cities knew about IDD, and 65.2% knew measures to prevent IDD. 20 21 It is not an appropriate time to supply IS and NIS simultaneously in 22 23 the developed cities of China, but a pilot project may be carried out in 24 25 some cities. 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 30, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 3 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 46 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 1 Should both iodised and non-iodised salt be made available 7 8 2 in Chinese cities? A cross-sectional survey 9 10 3 INTRODUCTION 11 12 4 Approximately 20% of the total global population resides in China, which 13 5 has one of the most severe iodine deficiency epidemics, in part due to the 14 15 6 external environment.For Iodine peer Deficiency review Disorders (IDD) wereonly once 16 17 7 widespread and caused great damages. Universal Salt Iodization (USI) 18 19 8 has been adopted for IDD prevention and control and has proven to be an 20 9 effective measure since 1994. By 2000, 17 provinces (municipalities) in 21 22 10 China achieved the goal of eliminating IDD and have sustained this 23 [1] 24 11 outcome for 10 years based on the WHO criteria for IDD elimination. 25 [2-4] 12 However, an excess of iodine has been discovered, and new issues , 26 27 13 including the increasing rate of thyroid disease and the elevated health 28 29 14 awareness of the Chinese population, have started a debate about the 30 31 15 ability of the population to make informed choices to use iodised salt(IS) 32 16 or non-iodised salt(NIS). As we know, some developed European

33 http://bmjopen.bmj.com/ 34 17 countries, such as Belgium, Finland, France, Germany, Greece, Ireland, 35 36 18 Italy and the Netherlands, supply IS and NIS simultaneously in the 37 [5] 38 19 market. Residents of these countries can purchase IS or NIS voluntarily. 39 20 In developed cities in provinces that have demonstrated the sustainable 40

41 21 elimination of IDD, is it the right time to supply IS and NIS on September 30, 2021 by guest. Protected copyright. 42 43 22 simultaneously? If so, can these populations remain free of IDD? Finally, 44 45 23 is there a proper method to demarcate cities to which IS and NIS is 46 24 supplied simultaneously or to which only IS is supplied? To answer these 47 48 25 questions, the subjective preferences of the residents about choosing IS or 49 50 26 NIS and the relevant influencing factors should be examined. 51 27 Door-to-door interviewing is usually used to conduct such investigations. 52 53 28 Considering the associated time and money expenditures, a 54 55 29 cross-sectional telephone interview is a feasible alternative and is a 56 57 4 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 46 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 30 widespread method that was adopted earlier and used by different 7 [6] 8 31 countries all over the world. Successful cases of the use of 9 10 32 cross-sectional telephone interviewing in China include KAP surveys on 11 [7-10] 33 influenza, smoking and pollinosis. 12 13 34 The application of cross-sectional telephone interviewing is based on 14 15 35 educational andFor economic peer levels, in review addition to the telephoneonly 16 17 36 popularization rate. In China, by the end of 2009, the national average 18 37 Urban Per Capita Disposable Income was 17,175 RMBY, and some 19 20 38 developed cities (municipalities) in the IDD-eliminated provinces were 21 [11] 22 39 all at or beyond the national average (table 1). Additionally, the 23 40 national fixed line penetration had increased to 314 million people (23.5 24 25 41 telephones/100 persons) and 212 million people in the cities (34.1 26

27 42 telephones/100 persons). These facts made telephone interview a 28 29 43 representative and feasible way to research the residents’ KAP toward 30 44 IDD. 31 32 45 As far as we know, there were no such records about the residents’

33 http://bmjopen.bmj.com/ 34 46 subjective preferences between IS and NIS, and no research on the 35 36 47 residents’ awareness of IDD in developed cities (municipalities) has been 37 48 conducted in recent years. Hence, the National Centre for Endemic 38 39 49 Disease Control (CEDC) launched this project in April, 2010, with the 40 50 aim of obtaining data on the above aspects to decide whether supplying

41 on September 30, 2021 by guest. Protected copyright. 42 43 51 IS and NIS simultaneously is feasible in developed cities in 44 52 IDD-eliminated provinces. The methods and results may provide 45 46 53 evidence or references for similar regions in the world. 47 48 54 49 55 METHODS 50 51 56 52 53 57 City sites to be investigated 54 55 58 The 17 capital cities of the 17 provinces (3 municipalities, 6 coastal 56 57 5 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 46 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 59 provinces and 8 inland provinces) that had achieved the goal of IDD 7 8 60 elimination in 2000 were selected (table 1). One coastal city was also 9 10 61 selected in each of the above coastal provinces. A total of 23 cities were 11 62 included in the study. 12 13 63 Table 1 Provinces and cities surveyed 14 15 Provincial capital Coastal cities Category ProvincesFor peer review only 16 Name Tel* Income** Name Tel * Income ** 17 Tangshan 24.58 18,053 18 Hebei Shijiazhuang - 16,607 19 Shandong Ji’nan 39.28 22,272 Yantai 28.27 21,125 20 Jiangsu Nanjing 38.91 25,504 Nantong 35.80 21,001 21 Coastal province Zhoushan 53.45 24,082 22 Zhejiang Hangzhou 48.23 26,864 23 Guangdong Guangzhou 61.28 27,610 Shenzhen 52.64 29,244 24 Guangxi Nanning 16.61 16,254 Beihai - 15,134 25 26 Heilongjiang Harbin 21.66 15,887 27 Jilin Changchun - 16,277 28

29 Shanxi Taiyuan 44.31 15,607 30 Inland He’nan Zhengzhou 29.28 17,117 province 31 Anhui Hefei 32.47 17,158 32 Hubei Wuhan 40.57 18,385 33 http://bmjopen.bmj.com/ 34 Jiangxi Nanchang 32.37 16,472 35 Hu’nan Changsha 33.28 20,238 36 Beijing Beijing 50.89 26,738 37 38 Municipalities Shanghai Shanghai 48.70 28,838 39 Tianjin Tianjin 31.36 21,430 40 64 All data marked with “*” and “**”were obtained from the Statistics Bulletin of the Cities’

41 on September 30, 2021 by guest. Protected copyright. 42 65 Economic and Social Development of 2009: “*” fix telephones/100 persons, “**” Urban per 43 66 capita disposable income (RMBY). The national fixed line penetration was 23.50 telephones/100 44 45 67 persons, and the urban per capita disposable income was 17,175 RMBY in 2009. 46 68 47 48 69 Target population 49 50 70 Citizens who had fixed telephones in the 23 selected cities. 51 71 The main contents of the questionnaire: 52 53 72 If IS and NIS are supplied simultaneously, which will you choose? 54 55 73 And why? 56 57 6 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 27 of 46 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 74 Have you heard of IDD? If so, what are the effects? How is IDD 7 8 75 prevented? 9 10 76 Can you tell us your occupation and age group? 11 77 Sampling method and sampling size 12 13 78 Random Digit Dialling and the Mitofsky-Waksberg two-stage sampling 14 [12] 15 79 method were usedFor in this study. peerIn each city,review only the main urban only zones 16 17 80 were investigated, and two districts were selected randomly from each 18 81 main urban zone. At least 200 effective samples were sampled in each 19 20 82 city included in the survey. The sample quota was divided evenly 21 22 83 between each district. The sample size was calculated according to 23 84 formula 1; the positive response rate ( π ) was approximately 70% 24 25 85 according to the pilot survey, and the allowable error (δ ) was 0.5. Thus, 26 27 86 the calculated sample size was 170 responses per city; therefore, 200 28 29 87 responses per city were considered adequate. 30  2 (1−× ππ ) 31 88 n =  α  π Positive rate δ Allowable error 32  δ 

33 http://bmjopen.bmj.com/ 89 Formula 1 34 35 90 Telephone number acquisition 36 37 91 If the fixed telephone of the neighbourhood offices in the selected district 38 92 was 7- or 8-digit numbers, the first 3 or 4 digits were deemed to be the 39 40 93 direction code. The last 4 digits were obtained by the Random function

41 on September 30, 2021 by guest. Protected copyright. ® 42 94 (=Rand ()×10000) in Microsoft Office Excel . If the function produced 43 44 95 randomised numbers less than 4 digits, and “0” was supplemented in 45 96 front of the generated numbers. The number of telephone numbers 46 47 97 created randomly should be at least 5 times the effective sample size, i.e., 48 49 98 at least 1000 telephone numbers should be created per city. 50 51 99 Quality control 52 100 The research was approved by the review board of the Harbin Medical 53 54 101 University. Participant consent was acquired when they answered the 55 56 57 7 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 28 of 46 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 102 telephone. The survey program and survey result database were designed 7 8 103 and delivered to the project cities by the national CEDC. The survey 9 10 104 program was improved based on a pilot survey, and the questionnaire had 11 105 clear questions that prompted simple answers and avoided providing hints 12 13 106 to the residents. Sensitive questions, such as age, were designed as 14 15 107 “Which age groupFor do you belongpeer to? <20, 20-30,review 30-40... ”. only 16 17 108 In the program, the investigation was carried out in a unified manner, 18 109 including the order and pattern of the questions, the dialling time (18:30 - 19 20 110 21:00), and instructions for making notes of the answers and filling the 21 22 111 database. 23 112 To avoid investigation and report biases and to improve the research 24 25 113 subjects’ cooperation, the interviewers were trained on their attitude, 26 27 114 mood, and countermeasures in case of emergency by the project 28 29 115 provinces themselves according to the rules of the survey program. For 30 116 example, some related questions put forward by residents could be 31 32 117 answered only after the interview was finished to avoid bias.

33 http://bmjopen.bmj.com/ 34 118 To guarantee the reliability of the research information, a return visit was 35 36 119 conducted after a project city finished the telephone interviews. The 37 120 quality control officer of the provincial CDC or CEDC visited the citizens 38 39 121 who had accepted the interview. At least 10% of the respondents were 40 122 randomly sampled to complete the return visit questionnaire. When the

41 on September 30, 2021 by guest. Protected copyright. 42 43 123 coincidence rate of survey information was greater than 80%, the 44 124 investigation results of the city can be regarded as qualified, and 45 46 125 otherwise, the city should be reinvestigated. 47 48 126 Definition of telephone number 49 127 Success: respondent finished the entire questionnaire; Rejection: 50 51 128 respondent refused to answer or did not finish the questionnaire; No 52 53 129 answer: the phone was busy both the first time and when re-dialling after 54 55 130 5 minutes; Invalid number: the phone number was vacant, fax, 56 57 8 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 29 of 46 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 131 switchboard or phone quit. 7 8 132 Data processing and statistical analysis 9 ® 10 133 Epi info 3.5.1 (CDC, Atlanta, GA USA) was used for the database 11 ® 134 design, and SPSS 17.0 was utilized for analysis. The level of 12 13 135 significance was set at 0.05. 14 15 136 For peer review only 16 17 137 RESULTS 18 138 Response situation 19 20 139 The flow chart is presented in figure 1. The survey included 235 districts 21 22 140 of 120 main urban zones in 23 cities. In total, 446 direction numbers were 23 141 chosen, and 24,557 telephone numbers were dialled, of which 13,215 24 25 142 (53.8%) were valid and 11,342 were invalid numbers. An invalid number 26 27 143 was a number that was not used as a telephone number. Among the 28 29 144 13,215 valid telephone numbers, 4,833 successfully finished the 30 145 questionnaire, 1,949 answered the phone but refused to answer the 31 32 146 questions, and 6,433 telephones were not answered. The answered

33 http://bmjopen.bmj.com/ 34 147 numbers (successes and rejections) were 6,782 and accounted for 51.3% 35 36 148 of the valid numbers. The total response rate (4833/(4833+1949)*100%) 37 149 was 71.3% (figure 2).. 38 39 150 Basic information about the interviewee 40 151 The age and occupation questions were answered by 4,861 and 4,857

41 on September 30, 2021 by guest. Protected copyright. 42 43 152 residents, respectively. The age and occupation proportion were similar 44 153 between the capital city (including Beijing) and the coastal city (including 45 46 154 Tianjin and Shanghai) residents. The residents’ age were mainly 47 48 155 distributed between 20 and 60 years, and this range accounted for 80% of 49 156 the respondents. However, compared with the population composition 50 [13] 51 157 from the ‘2009 China Health Statistical Yearbook’ and the Chinese 52 [14] 53 158 population pyramid, the 10-20 age group had a smaller constituent 54 55 159 ratio. This may lead to a non-representative sample of the Chinese 56 57 9 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 30 of 46 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 160 population, although this may be acceptable in this study because 7 8 161 children rarely buy salt. The residents were evenly distributed in each 9 10 162 profession (accounting for approximately 10%), except for the 11 163 professions of servicemen, students and irregular professions, which had 12 13 164 small sample sizes. 14 15 165 For peer review only 16 17 166 Proportion of respondents choosing IS or NIS 18 167 Of the residents investigated, 4,865 answered the question ‘If IS and NIS 19 20 168 are supplied simultaneously, which will you choose?’ The total 21 22 169 percentage of respondents ‘choosing IS’ was 76.8%, ‘choosing NIS’ was 23 170 9.9%, ‘both’ was 3.9% and ‘does not matter’ was 9.3%. The respondents 24 25 171 from the capital cities chose IS at a higher rate (79.5%) than those from 26 27 172 the coastal cities (72.0%). The percentages of respondents choosing NIS 28 29 173 was opposite between the city and coastal residents (6.7% and 15.7%, 30 174 respectively), and the ratio of respondents who chose “both” (<5%) and 31 32 175 ‘does not matter’ (<10%) were nearly the same in the two areas. Of the 23

33 http://bmjopen.bmj.com/ 34 176 cities, there were 12 cities whose IS selection rate was higher than 80%, 8 35 36 177 cities in which the IS selection rate was between 60-80%, and 3 cities 37 178 (Shanghai: 32.8%, Zhoushan: 42.8% and Hangzhou: 49.3%) in which the 38 39 179 IS selection rate was lower than 50% (table 2, table 3). 40 180 Reasons for “choosing IS”

41 on September 30, 2021 by guest. Protected copyright. 42 43 181 Of the 3,738 residents who chose IS, their reasons were as follows: 44 182 61.9% of them accurately understood the benefits of IS, i.e., preventing 45 46 183 IDD and increasing children’s Intelligence Quotient (IQ), 14.7% of them 47 48 184 merely chose IS for superior-quality reasons and knew little about the 49 185 relationship between IS and IDD prevention, and 27.7% had no clear 50 51 186 reason (most explained that it was the custom) (table 2). 52 53 187 54 55 188 56 57 10 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 31 of 46 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 189 Table 2 Respondents’ reasons for buying IS (%) 7 8 Reasons Number 9 IS of Benefit Province City choosing 10 choosing Prevent for Good rate Others 11 IS IDD children’s quality 12 IQ 13 Heilongjiang Harbin 183 91.5 36.6 3.8 1.6 60.1 14 Jilin Changchun 175 86.6 48.6 9.7 6.3 0 Beijing - 202 86.7 61.4 12.9 9.4 28.2 15 Hebei ShijiazhuangFor 174peer 87.0 review 67.8 7.5 12.1only 14.4 16 Shanxi Taiyuan 189 95.0 13.8 0 14.3 70.9 17 Shandong Ji’nan 187 93.5 73.3 27.3 20.3 11.2 18 He’nan Zhengzhou 177 88.5 54.8 15.3 26.6 27.7 19 Jiangsu Nanjing 180 72.6 64.4 25.0 23.9 7.8 20 Anhui Hefei 178 84.8 32.0 0.6 5.6 61.8 21 Hubei Wuhan 163 75.1 41.7 8.0 24.5 29.4 Zhejiang Hangzhou 66 32.8 57.6 16.7 21.2 0 22 Jiangxi Nanchang 184 86.8 52.2 7.6 13.6 8.7 23 Hu’nan Changsha 146 72.3 56.2 6.8 5.5 39.7 24 Guangdong Guangzhou 124 61.7 39.5 1.6 46.0 3.2 25 Guangxi Nanning 170 78.0 38.8 4.1 6.5 46.5 26 Capital city total 2498 79.5 79.5 9.8 15.0 29.0 27 Hebei Tangshan 208 94.1 81.7 17.3 9.6 5.8 28 Tianjin - 200 83.7 75.5 28.5 27.0 6.5 Shandong Yantai 186 93.0 90.3 3.2 8.1 4.8 29 Jiangsu Nantong 153 73.9 46.4 5.9 2.6 47.7 30 Shanghai - 108 49.3 25.9 6.5 29.6 48.1 31 Zhejiang Zhoushan 86 42.8 68.6 12.8 18.6 0 32 Guangdong Shenzhen 144 71.6 22.9 0.7 3.5 42.4 Guangxi Beihai 155 66.2 21.9 2.6 18.7 58.1 33 http://bmjopen.bmj.com/ 34 Coastal city total 1240 72.0 72.0 10.6 14.1 25.0 Total 3738 76.8 51.9 10.0 14.7 27.7 35 190 Note: The reason for choosing IS was a multiple choice question. 36 37 191 38 39 192 Reasons for “choosing NIS” 40 193 Most of the 481 residents who chose NIS lived in Shanghai, Hangzhou,

41 on September 30, 2021 by guest. Protected copyright. 42 194 Zhoushan and Beihai. The majority thought that they had received an 43 44 195 adequate amount of iodine from food or that they did not live in an IDD 45 46 196 area. Unexpectedly, 10% of the respondents who chose NIS thought IS 47 197 was harmful. Nearly 27.0% chose NIS because of reasons that cannot be 48 49 198 classified, such as habit, parent’s decision, or changing taste (table 3). 50 51 199 52 53 200 54 201 55 56 57 11 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 32 of 46 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 202 Table 3 Respondents’ reasons for choosing NIS（%） 7 8 Reasons 9 Not IS 10 City n % Iodine IS is IDD Patient* bad Cheap Others 11 sufficient harmful 12 endemia taste 13 Capital city Total 211 40.0 6.7 13.3 9.0 10.5 2.9 1.4 23.3 14 Coastal city Total 270 27.7 15.7 17.7 5.2 8.1 4.0 1.5 29.9 15 Total 481For 33.1 9.9 peer16.6 7.3review 9.1 3.5 1.5only 27.0 16 203 Note: The “*” meant patients with thyroid disease for whom it is not feasible to take IS. Some provinces had small 17 204 sample sizes; the proportions refer to large ones; n: number that chose NIS; %: NIS choosing rate, the 18 205 reason for choosing NIS was a multiple choice question. 19 20 206 Reasons for “choosing both” 21 22 207 Most of the 192 residents who chose “both IS and NIS” were from 23 208 Hangzhou, Guangzhou and Shanghai. The two major reasons were to 24 25 209 maintain an appropriate iodine intake and to avoid the harmful effect of 26 27 210 excess iodine, accounting for 27.6% and 30.7% of the responses, 28 29 211 respectively. 30 212 Reasons for “does not matter” 31 32 213 The 454 residents who chose “does not matter” mainly resided in Nanjing,

33 http://bmjopen.bmj.com/ 34 214 Wuhan, Hangzhou, Changsha, Guangzhou, Nanning, Tianjin, Shanghai 35 36 215 and Zhoushan. They tended to choose IS or NIS based on which was 37 216 most convenient to buy (52.0% of the respondents who had no 38 39 217 preference). 40 218 Awareness rates of IDD

41 on September 30, 2021 by guest. Protected copyright. 42 219 Of the residents who responded to our phone calls, 4,865 answered the 43 44 220 question “Have you heard of Iodine Deficiency Disorders?”. Among 45 46 221 those, 68.7% (3,344) had heard and 31.3% (1,521) had never heard of 47 48 222 IDD. There was little difference in having heard of IDD between the 49 223 capital cities (68.2%) and the coastal cities (69.8%). The awareness rate 50 51 224 in Yantai (94.5%) was the highest, and in four cities (Shijiazhuang, Jinan, 52 53 225 Hangzhou and Tangshan), the awareness rates were greater than 80%. 54 55 226 Beihai and Taiyuan had the lowest awareness rates (46.2% and 44.7%, 56 57 12 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 33 of 46 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 227 respectively). There were no significant differences in the levels of 7 8 228 awareness between age groups or occupation groups. 9 10 229 Among the 3,344 residents who had heard of IDD, up to 77.1% of them 11 230 knew that iodine deficiency could lead to goitre, but only 11.8% knew 12 13 231 that it could also cause intelligence damage to children, and 16.9% of 14 15 232 them knew nothingFor about itspeer hazard. review only 16 17 233 Awareness rates of IDD preventive measures 18 234 Of the 3,344 residents who had heard of IDD, 62.5% of them knew IS 19 20 235 could prevent IDD, 43.2% of them knew that kelp and purple seaweed 21 22 236 could prevent IDD, and nearly 18.7% of the residents knew nothing about 23 237 how to prevent IDD. 24 25 238 The IS preference rates and IDD awareness rates in each city 26 27 239 All cities could be divided into four groups according to their relationship 28 29 240 between the IS preference rate and IDD awareness rate: Uhigh IS 30 241 preference rate and high IDD awareness rate, such as Yantai, 31 32 242 Shijiazhuang, Beijing, Tianjin, Nanjing, Changsha and Guangzhou;

33 http://bmjopen.bmj.com/ 34 243 Uhigh IS preference rate and low IDD awareness rate, such as Taiyuan, 35 36 244 Harbin, Nanchang, Hefei, Wuhan and Beihai; Uapproximate equal the 37 245 choosing rate and the awareness rate, such as Tangshan, Ji’nan, 38 39 246 Zhengzhou, Changchun, Nanning, Nantong and Shenzhen; and ④low IS 40

41 247 choosing rate and high IDD awareness rate, such as Shanghai, Zhoushan on September 30, 2021 by guest. Protected copyright. 42 43 248 and Hangzhou. Only in Yantai and Nanjing were the IDD awareness rates 44 249 higher than the IS preference rates. The IS choosing rates in Harbin and 45 46 250 Taiyuan were greater than 90%, but their IDD awareness rates were only 47 48 251 57.0% and 44.7%, respectively. There were only 4 cities whose IS rates 49 50 252 and awareness rates were all greater than 80% (figure 3table 4). 51 253 52 53 254 54 55 255 56 57 13 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 34 of 46 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 256 Table 4 The IDD awareness rates and IS preference rates（%） 7 8 9 City IS rate(%) Acknowledge rate(%) 10 Taiyuan 95.0 44.7 11 Tangshan 94.1 85.1 12 13 Jinan 93.5 81.0 14 Yantai 93.0 94.5 15 HarbinFor peer91.5 review57.0 only 16 Zhengzhou 88.5 74.0 17 18 Shijiazhuang 87.0 85.0 19 Nanchang 20 86.8 67.5 21 Beijing 86.7 79.8 22 Changchun 86.6 65.8 23 Hefei 84.8 52.9 24 Tianjin 83.7 76.6 25 26 Nanning 78.0 63.8 27 Wuhan 75.1 59.0 28 Nantong 73.9 62.8 29 Nanjing 72.6 74.2 30 31 Changsha 72.3 69.8 32 Shenzhen 71.6 60.7

33 Beihai 66.2 46.2 http://bmjopen.bmj.com/ 34 Guangzhou 61.7 65.2 Formatted Table 35 36 Shanghai 49.3 57.5 37 Zhoushan 42.8 77.6 38 Hangzhou 32.8 81.1 39 40 257

41 on September 30, 2021 by guest. Protected copyright. 258 The chi square test was used to analyse the influence of IDD awareness 42 43 259 on the IS preference rate. The results showed no differences between the 44 45 260 capital cities’ IS and NIS groups but significant differences between the 46 47 261 coastal cities’ IS and NIS groups. This could explain the weak relations 48 262 between the IS preference rates and the IDD awareness rates in the cities, 49 50 263 whereas for coastal cities, there were associations between the IS 51 52 264 preferences and IDD awareness; specifically, most coastal residents who 53 265 chose NIS had lower IDD awareness rates. 54 55 56 57 14 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 35 of 46 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 266 Results of the return visits 7 8 267 The total coherence rate of all questions was 85.1%. The Kappa values of 9 10 268 the coherence analysis for choosing IS, choosing NIS and IDD awareness 11 269 were 0.752, 0.693 and 0.76, respectively, and were all above the required 12 [15] 13 270 0.4, describing good coherence. 14 15 271 For peer review only 16 17 272 DISCUSSIONS 18 273 Telephone interviews are a widely used method all over the world that 19 [16] 20 274 has the advantages of wide coverage, low costs and high efficiency. 21 22 275 However, The survey subjects could be units or persons with fixed 23 276 telephones, and the selection, supervision and training of investigators are 24 25 277 convenient. These features made it feasible to conduct a large-scale, 26 27 278 easily acceptable survey with clear content. However, telephone surveys 28 [16] 29 279 do have biases and limitations. Only after a detailed design to decrease 30 280 error as much as possible are the results convincible. For some interviews, 31 32 281 such as street intercept or telephone interviews, iif most of the

33 http://bmjopen.bmj.com/ 34 282 interviewees reject the survey and the rejection rate is greater than 70% 35 [17] 36 283 (i.e., the response rate <30%),, the representation of the random 37 284 samples will be questioned; hence, the interviewees will be resampled. In 38 39 285 our study, the telephone survey was conducted with unified quality 40 286 control, the response rate was 71.3%, and the coherence rate of re-survey

41 on September 30, 2021 by guest. Protected copyright. 42 43 287 was above 80% (85.1%). One limitation was There were no significant 44 288 differences in terms of age and occupation. Thus, the results of this 45 46 289 telephone survey could basically reflect the subjective preferences about 47 48 290 IS or NIS of the urban residents in China. Athats it was the first telephone 49 291 survey on the KAP toward IDD in China,China; the traditional PAPI 50 51 292 method was taken, and the results offer references for future surveys 52 53 293 rather than CATI. The other, some “other reason” should be recorded Formatted: Font color: Auto 54 Formatted: Font color: Auto 55 294 more carefully. 56 57 15 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 36 of 46 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 295 The residents’ total IS preference rate was 76.8% in the 6 coastal cities 7 8 296 and 14 capital cities in 14 provinces and 3 municipalities (i.e., Beijing, 9 10 297 Tianjin and Shanghai) wherehad IDD elimination has been sustained for 11 298 nearly ten years. The reasons for choosing IS were mainly ‘IDD 12 13 299 prevention (51.9%)’ and ‘improving children’s IQ (10%)’, the remaining 14 15 300 residents (38%)For had no idea peer about the relationship review between IS onlyand IDD 16 17 301 prevention and their choice of IS was only because IS is of good quality 18 302 or simply a custom. Another phenomenon that emerged in this study 19 20 303 was that most residents had lower IDD awareness rates than IS selection 21 22 304 rates, as was observed in Taiyuan, Hefei, Harbin and Changchun. 23 305 Although the residents’ IS preference rates were greater than 80%, their 24 25 306 awareness of IDD was low. An additional 33.1% of residents chose NIS, 26 27 307 27% of whom had no clear reason, which reflected that aAlthough IDD 28 29 308 health education activities, such as the 5.15 IDD Day, have been carried 30 309 out widely and continuously by the relevant departments, it is still not 31 32 310 enough or exhaustive, and the current propaganda strength and methods

33 http://bmjopen.bmj.com/ 34 311 need further improving. 35 36 312 Considering that the supply system of iodised salt is divided by province 37 313 in China, the KAP toward IDD of the city residents is more important. 38 39 314 Although these provinces all attained the goal of IDD elimination and 40 315 their capital cities and coastal cities are all developed cities, their IS

41 on September 30, 2021 by guest. Protected copyright. 42 43 316 choosing and IDD awareness rates were different, and they should be 44 317 treated differently according to these results. In this study, 12 cities had 45 46 318 an IS preference rate of greater than 80%, and these were all northern 47 48 319 cities with the exception of Nanchang. Five cities had a rate greater than 49 320 90%. The other 11 cities with rates lower than 80% were all southern 50 51 321 cities. The IS preference rates of Hangzhou, Zhoushan and Shanghai were 52 53 322 32.8%, 42.8% and 49.3%, respectively, and were all lower than 50%. 54 55 323 Thise result shows that the IS preference rates of southern city residents 56 57 16 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 37 of 46 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 324 were lower than those of northern city residents. The main reason was 7 8 325 that the people in the southern coastal areas think that they have received 9 10 326 adequate iodine from food other than iodised salt, and another reason was 11 327 that they thought they did not live in an iodine-deficient area, as was the 12 13 328 case with the residents of Hangzhou, Zhoushan and Shanghai. Their IDD 14 15 329 awareness ratesFor were greater peer than their ISreview preference rates, andonly some 16 17 330 residents chose NIS despite knowing about the potential damage caused 18 331 by IDD. In fact, these districts had no apparent IDD prevalence, which is 19 20 332 likely attributable to the widespread IS supply. The survey “Iodine 21 22 333 nutritional status of coastal districts residents in 2009” conducted by the 23 [18] 334 MOH found that under the condition of consumption rates of qualified 24 25 335 IS greater than 90%, the Median Urinary Iodine (MUI) of Shanghai 26 27 336 residents was merely between 100 µg/L and 200 µg/L, i.e., just at the 28 29 337 iodine intake appropriate level, and the MUI of pregnant women was 30 338 lower than 150 µg/L, which corresponds to an insufficient iodine intake 31 32 339 level. Among the residents of the Xiacheng district of Hangzhou city,

33 http://bmjopen.bmj.com/ 34 340 Zhejiang province, the consumption rate of qualified IS was 96.7%, and 35 36 341 the MUI of 8-10-year-old children was between 100 and 200 µg/L, which [19] 37 342 is at the iodine intake appropriate level. If allowing these cities’ 38 39 343 residents to freely choose IS or NIS, the residents may choose NIS 40 344 blindly, and the achievement of controlled IDD that has been obtained

41 on September 30, 2021 by guest. Protected copyright. 42 43 345 will be destroyed. 44 346 From the society and economic development aspects, supplying IS and 45 46 347 NIS simultaneously to some cities or persons is an inevitable trend. In the 47 48 348 public health domain, if a disease has a prevalence rate below 5%, it will 49 349 not become a public health problem. Hence, we deem that if the IDD 50 51 350 awareness rates and the correct selection rates are above 95%, iodised and 52 53 351 non-iodised salt can both be made available. According to this telephone 54 55 352 interview study, it is not a suitable time to supply IS and NIS 56 57 17 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 38 of 46 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 353 simultaneously, even in the developed cities of China. The intake of 7 8 354 non-iodised salt should be performed with caution in China. The next step 9 10 355 is to conduct a trial in which IS and NIS is simultaneously supplied in 11 356 appropriate cities. The selected experimental city should be a developed 12 13 357 city with no IDD prevalence in recent years, and the residents’ awareness 14 15 358 rate of IDD andFor selection peer rate of IS should review be high. According only to our 16 17 359 study, Yantai, Tangshan and Ji’nan may be appropriate cities for this 18 360 experiment. Before the experiment, emergency planning should be 19 20 361 performed, and extensive IDD education should be conducted. If 21 22 362 abnormal situations emerge, the emergency program will be started, or 23 363 the trial will even be stopped. Currently, Universal salt iodization was 24 25 364 carried out as a mandatory effective strategy to prevent IDD in China 26 27 365 since 1996. However, no documentation was found about the residents’ 28 29 366 preferences on salt type (IS or NIS) based on their awareness of IDD and 30 367 its prevention measure knowledge. Hence, the study tried to find out the 31 32 368 reason why people choose iodised or non-iodised salt. Some possible

33 http://bmjopen.bmj.com/ 34 369 reasons were enlisted including preventing IDD, improve children’s IQ, 35 36 370 good quality and convenience, etc, were main reasons people would 37 371 purchase iodised salt, while, regarding to the potential harm and having 38 39 372 thyroid disorders were the reasons for choosing non-iodised salt. These 40 373 results were important for further research and prevention measures and

41 on September 30, 2021 by guest. Protected copyright. 42 43 374 policy decision. 44 375 45 46 376 47 48 377 49 378 Author Affiliation 50 51 379 Jun Y, Peng L, Ying L and Shou-jun L are affiliated with the Institute for 52 53 380 iodine deficiency disorders, the Centre for Endemic Disease Control, the 54 55 381 Chinese Centre for Disease Control and Prevention, and Harbin Medical 56 57 18 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 39 of 46 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 382 University, Harbin, People’s Republic of China. 7 8 383 Dian-jun S is affiliated with the Centre for Endemic Disease Control, the 9 10 384 Chinese Centre for Disease Control and Prevention, and Harbin Medical 11 385 University, Harbin, People’s Republic of China. 12 13 386 14 15 387 AcknowledgmentsFor We thank peer the organisations review that participated only in this 16 17 388 survey: the National Centre for Endemic Disease Control (CEDC), 16 18 389 provincial (municipality) Centres for Disease Control (CDCs) or CEDC, 19 20 390 including those of the Beijing, Tianjin, Shanghai, and Jilin Provinces, 21 22 391 Hebei Province, Shandong Province, He’nan Province, Jiangsu Province, 23 392 Anhui Province, Hubei Province, Zhejiang Province, Jiangxi Province, 24 25 393 Hu’nan Province, Guangdong Province and Guangxi province, and six 26 27 394 coastal city CDCs, including those of Tangshan, Yantai, Nantong, 28 29 395 Zhoushan, Shenzhen and Beihai. The authors of this article thank those 30 396 colleagues from the above organisations who took part in the survey for 31 32 397 their hard work. We also thank the Elsevier for their English editing.

33 http://bmjopen.bmj.com/ 34 398 35 36 399 Contributorship statement Jun Y wrote the project report, Peng L 37 400 analysed the data and wrote the manuscript, Ying L and Shou-jun L 38 39 401 designed the study and managed the project, and Dian-jun S was 40 402 responsible for the project and the publication.

41 on September 30, 2021 by guest. Protected copyright. 42 43 44 403 Competing interests None. 45 46 404 Funding The project was funded by the Ministry of Health of China. 47 48 405 Data sharing No additional data are available. 49 50 406 Ethics approval The review board of the Harbin Medical University. 51 407 52 53 408 54 409 References 55 56 57 19 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 40 of 46 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 410 [1] WHO/UNICEF/ICCIDD. Assessment of iodine deficiency disorders and monitoring their 7 411 elimination. A guide for programme managers. Third edition, France, 2007. 8 9 412 [2] Todd CH, Allain T, Gomo ZA, et al. Increase in thyrotoxicosis associated with iodine 10 413 supplements in Zimbabwe. Lancet. 1995,346:1563-4. 11 414 [3] Stanbury JB, Ermans AE, Bourdoux P, et al. Iodine-induced hyperthyroidism: occurrence 12 415 and epidemiology. Thyroid 1998, 8:83-100. 13 14 416 [4] Parveen S, Latif SA, Kamal MM, et al. Iodized salt induced thyrotoxicosis: Bangladesh 15 417 perspective, MymensinghFor Med J. 2009,18:165-8.peer review only 16 418 [5] ICCIDD. International Council for the Control of Iodine Deficiency Disorder. CIDDS 17 419 Database. Current IDD Status Database.2004. 18 420 [6] Groves RM, Biemer PP, Lyberg LE, et al. Telephone survey methodology [J] John Wiley 19 20 421 &Sons Inc.1989, 2-10. 21 422 [7] Jianhua L, Hanwu M , Yongsheng W, et al. Analysis of know ledge, attitude and practice 22 423 of the influenza A /H1N1of general and immigration population in Shenzhen, Chinese Health 23 424 Education. 2009,25:915-8. 24 25 425 [8] Jin M, Shunxiang Z, Hanwu M, et al. Telephone survey on behavior risk factors of 26 426 Shenzhen residents among six districts in 200. Journal of Disease Control. 2005,9:561-5 27 427 [9] Tao W, Su-jun L, Liang Z, et al. Investigation of Pollinosis in Beijing Residents over the 28 428 Age of 15. J Environ Health, 2008,25:403-4. 29 429 [10] Blair G. Obstetricians' Receptiveness to Teen Prenatal Patients Who Are Medicaid 30 31 430 Recipients, Health Services Research.1997,32:265-82. 32 431 [11] National Bureau of Statistics of China.Statistics Bulletin of the Cities’ Economic and

33 432 Social Development of 2009. http://bmjopen.bmj.com/ 34 433 [12] Potthoff RF. Some Generazations of the Mitofsky-Waksberg techniques for random 35 434 digital dialing. Journal of the American Statistical Association, 1987, 82:409-41. 36 37 435 [13] Yi MING. http://geo.cersp.com/sJxzy/sc/200706/2658.html. 38 436 [14] http://www.moh.gov.cn/publicfiles/business/htmlfiles/zwgkzt/ptjnj/200908/42635.htm. 39 437 [15] Jiezhen W, Jing H, Yanxun L, et al. Kappa statistic value and its application. [J], Chinese 40 438 Health Statistical, 1995,12:46-50.

41 on September 30, 2021 by guest. Protected copyright. 42 439 [16] Ping W, Weijing D. Telephone Survey and Its Application in Public Health. Health 43 440 education of China. 2004,20(6):549-50. 44 441 [17] Guohua D. A powerful tool of the statistic survey—the computer aid telephone survey 45 442 system. Newspaper of Theory Guidance, 2008,12:51-2. 46 443 [18] Chinese CDC, CEDC. Report of coastal residents’ iodine nutrient level,2009. 47 48 444 [19] ICCIDD Current IDD Status Database,2003. 49 445 [http://www.who.int/vmnis/database/iodine/countries/en/index.html]. 50 51 52 53 54 55 56 57 20 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 41 of 46 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 7 Figure legend 8 9 Figure 1 Flow chart 10 11 Figure 2 The answer status of the sampled telephone numbers 12 Figure 3 The relationship between IDD awareness rates and IS preference 13 14 15 rates For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 30, 2021 by guest. Protected copyright. 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 21 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 42 of 46 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

41 on September 30, 2021 by guest. Protected copyright. 42 43 44 45 46 47 Figure 1 134x218mm (300 x 300 DPI) 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 43 of 46 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 For peer review only 16 17 18 19 20 21 22 23 24 Figure 2 25 77x38mm (300 x 300 DPI) 26 27 28 29 30 31 32

33 http://bmjopen.bmj.com/ 34 35 36 37 38 39 40

2 2 8 8 8 8 5-7 5-7

5 5 4-5 4-5

8 8 Reported on page # Reported on page 2 2 7 7

cross-sectional studies cross-sectional

BMJ Open http://bmjopen.bmj.com/ on September 30, 2021 by guest. Protected copyright. For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml

9 9 8 8 8 8

9 9 9 9 9 9

8 8 8 8 7 7

9-14 9-14 8 8

Main resultsMain Outcome Outcome data 15* numbers Report of outcome events or summary measures

Descriptive data 14* characteristicsGive (a) participantsof study (eg demographic, clinical, social) and information on exposures potential and

Participants Results

Statistical Statistical methods 12 ( Quantitative variables 11 how Explain quantitative were variables handled in analyses.the If applicable, describe which groupingschosen were and Study size Page 45 of 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2014-005397 on 11 July 2014. Downloaded from Page 46 of 15-17 15-17

15 15 15 15 17-19 17-19

18 18 14 14

BMJ Open http://bmjopen.bmj.com/ on September 30, 2021 by guest. Protected copyright. For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml