| Objective: There are special methodologies about telephone survey which has been used for many years in North America and Western Europe. But the criterion of conducting telephone survey has not been laid down in China and this method is restrictive to be applied. The purpose of this study was to use external telephone survey method for reference; discuss the method of using telephone survey to collect medicine and public health information in Chongqing area; attempt how to apply telephone survey on large range health survey; explore appropriate sampling methods, population, cooperation of people, design of questionnaire, quality control, etc. and eveluate the reliability and consistency of telephone survey.Methods: A population-based telephone survey of the inhabitants in 8 Chongqing urban areas was conducted in September 2002. Based on Random Digit Dialing, a 4 stages mixed sampling method including stratified, PPS, equal-probability was used. The content was the frequency, duration, severity, and impact of diarrhea during the previous two weeks. Some of respondents in the first interview were assigned to a second interview for recontacting after 15 days.Results: Atotal of 3815 numbers were dialed, 1334 numbers were connected, 1114 subjects were interviewed. The average response rate was 83.51%. 85 ofrespondents were reinterviewed, reinterview rate was 7.6%.There were 4 methods for sampling telephone bureaus. The lowest inexist-number rate owned to the Block PPS Sampling. The Telephone Bureaus Fixed Weight Sampling was not so good, but it covered broader area and put up better randomicity. 57.5% of respondents were women, 96.37% were 15 years of age, 25.07% had more than a college education, and househole income was RMB 750 / person.month, all higher than the information in . The reliability of this study was good, kappa = 0.6805, 95% confidence interval is ( 0.6421, 0.7189 ). Spearman's rank correlation coefficient was 0.7223 which showed that there was excellent agreement between the first and second interviews. The whole survey consumed about 11 work days ( caculated according to one person) and RMB 1,400.Conclusion: Telephone survey is more timely and less expensive than door to door interviews. The reliability of this method is good and people are easy to cooperate. The content of telephone survey should be concise. It is best that survey is conducted in the evening. Telephone survey is appropriate to be used as routine survey and follow-up method for collecting health information and an assist method when large-scale door to door interview is paused. A high cost-benefit ratio will be attained. A multi-stage mixed sampling method is advised. The Block PPS Sampling and The Telephone Bureaus Fixed Weight Sampling are considered to be better for sampling telephone bureaus. At present, it is more appropriate for telephone survey to interview people with middle or high education and income level. With the increasing of the telephone popularization in China, telephone survey will be used more broadly. |
