Demographic and Health Survey and Malaria Indicator Survey 2015-2016 - Tanzania

Abstract
---------------------------

The primary objective of the 2015-16 Tanzania Demographic and Health Survey and Malaria Indicator Survey (TDHS-MIS) is to provide up-to-date estimates of basic demographic and health indicators. This survey collected information on fertility levels, marriage, sexual activity, fertility preferences, awareness and use of family planning methods, breastfeeding practices, nutrition, childhood and maternal mortality, maternal and child health, malaria, and other health-related issues. In addition, the 2015-16 TDHS-MIS provided estimates of anaemia prevalence among children age 6-59 months and women age 15-49 years, estimates of malaria prevalence among children age 6-59 months, and estimates of iodine concentration in household salt and women’s urine.

The information collected through the 2015-16 TDHS-MIS is intended to assist policy makers and programme managers in evaluating and designing programmes and strategies to improve the health of the country’s population.

Geographic coverage
---------------------------

National

Analysis unit
---------------------------

- Household
- Children age 0-5
- Women age 15-49
- Men age 15-59

Kind of data
---------------------------

Sample survey data [ssd]

Sampling procedure
---------------------------

Sample Design
The sample design for the 2015-16 TDHS-MIS was done in two stages and was intended to provide estimates for the entire country, for urban and rural areas in Tanzania Mainland, and for Zanzibar. For specific indicators such as contraceptive use, the sample design allowed the estimation of indicators for each of the 30 regions (25 regions from Tanzania Mainland and 5 regions from Zanzibar). The first stage involved selecting sample points (clusters), consisting of enumeration areas (EAs) delineated for the 2012 Tanzania Population and Housing Census. A total of 608 clusters were selected.

In the second stage, a systematic selection of households was involved. A complete households listing was carried out for all 608 selected clusters prior to the fieldwork. From the list, 22 households were then systematically selected from each cluster, yielding a representative probability sample of 13,376 households for the 2015-16 TDHS-MIS. To estimate geographic differentials for certain demographic indicators, Tanzania was divided into nine geographic zones. Although these zones are not official administrative areas, this classification system is also used by the Reproductive and Child Health Section of the MoHCDGEC. Grouping the regions into zones allowed a relatively large number of people in the denominator and a reduced sampling error. Note that the zones, defined below, differ slightly from the zones used in previous DHS surveys. Therefore, comparisons across the zones and from survey to survey should be made with caution. The zones are as follows:
Western Zone: Tabora, Kigoma
Northern Zone: Kilimanjaro, Tanga, Arusha
Central Zone: Dodoma, Singida, Manyara
Southern Highlands Zone: Iringa, Njombe, Ruvuma
Southern Zone: Lindi, Mtwara
South West Highlands Zone: Mbeya, Rukwa, Katavi
Lake Zone: Kagera, Mwanza, Geita, Mara, Simiyu, Shinyanga
Eastern Zone: Dar es Salaam, Pwani, Morogoro
Zanzibar: Kaskazini Unguja, Kusini Unguja, Mjini Magharibi, Kaskazini Pemba, Kusini Pemba

All women age 15-49 who were either usual residents or visitors in the household on the night before the survey were included in the 2015-16 TDHS-MIS and were eligible to be interviewed. In a subsample of one-third of all the households selected for the survey, all men age 15-49 were eligible to be interviewed if they were either usual residents or visitors in the household on the night before the survey. In all households, with the parent's or guardian's consent, children age 6-59 months were tested for anaemia and malaria. All interviewed women were tested for anaemia. In the households selected for interviews with men, interviewed women were asked to provide a urine sample and a sample of household salt for laboratory testing to detect the presence of iodine.

For further details of sample design and implementation, see Appendix A of the final report.

Mode of data collection
---------------------------

Face-to-face [f2f]

Research instrument
---------------------------

Four questionnaires were used for the 2015-16 TDHS-MIS: the Household Questionnaire, the Woman’s Questionnaire, the Man’s Questionnaire, and the Biomarker Questionnaire. These questionnaires were based on the DHS Program’s standard Demographic and Health Survey (DHS) questionnaires. They were adapted to reflect the population and health issues relevant to Tanzania. Inputs were solicited from various stakeholders representing government ministries, departments, and agencies; non-governmental organizations; and development partners. After the preparation of the definitive questionnaires in English, the questionnaires were translated into Kiswahili.

Cleaning operations
---------------------------

In the 2015-16 TDHS-MIS the first data entry was done concurrently with data collection in the field. After the paper questionnaires were completed, edited, and checked by both the field editor and the supervisor, the data was entered into a tablet equipped with a data entry programme. This was done by the editor. Completed questionnaires were then sent to NBS headquarters, where they were entered for the second time and edited by data processing personnel who were given special training for this task. ICF International provided technical assistance during the entire data processing period.

Processing the data concurrently with data collection allowed for regular monitoring of team performance and data quality. Field check tables were generated regularly during data processing to check various data quality parameters. As a result, feedback was given on a regular basis, encouraging teams to continue in areas of good performance and to correct areas in need of improvement. Feedback was individually tailored to each team. Data entry, which included 100% double entry to minimise keying errors, and data editing, were completed on March 21, 2016. Data cleaning and finalization were completed on April 22, 2016.

Response rate
---------------------------

A total of 13,360 households were selected for the survey, of which 12,767 were occupied. Of the occupied households, 12,563 were successfully interviewed, yielding a response rate of 98%.

In the interviewed households, 13,634 eligible women were identified for individual interviews; interviews were completed with 13,266 women, yielding a response rate of 97%. In the subsample of households selected for the male survey, 3,822 eligible men were identified and 3,514 were successfully interviewed, yielding a response rate of 92%. There is little variation in household response rates between rural and urban residences.

Sampling error estimates
---------------------------

The estimates from a sample survey are affected by two types of errors: (1) nonsampling errors, and (2) sampling errors. Nonsampling errors are the results of mistakes made in implementing data collection and data processing, such as failure to locate and interview the correct household, misunderstanding the questions on the part of either the interviewer or the respondent, and data entry errors. Although numerous efforts were made during the implementation of the 2015 Tanzania Demographic and Health Survey (TDHS) to minimize this type of error, nonsampling errors are impossible to avoid and difficult to evaluate statistically.

Sampling errors, on the other hand, can be evaluated statistically. The sample of respondents selected in the 2015 TDHS is only one of many samples that could have been selected from the same population, using the same design and identical size. Each of these samples would yield results that differ somewhat from the results of the actual sample selected. Sampling errors are a measure of the variability between all possible samples. Although the degree of variability is not known exactly, it can be estimated from the survey results.

A sampling error is usually measured in terms of the standard error for a particular statistic (mean, percentage, etc.), which is the square root of the variance. The standard error can be used to calculate confidence intervals within which the true value for the population can reasonably be assumed to fall. For example, for any given statistic calculated from a sample survey, the value of that statistic will fall within a range of plus or minus two times the standard error of that statistic in 95 percent of all possible samples of identical size and design.

If the sample of respondents had been selected as a simple random sample, it would have been possible to use straightforward formulas for calculating sampling errors. However, the 2015 TDHS sample is the result of a multi-stage stratified design, and, consequently, it was necessary to use more complex formulas. The computer software used to calculate sampling errors for the 2015 TDHS is a SAS program. This program used the Taylor linearization method for variance estimation for survey estimates that are means or proportions. The Jackknife repeated replication method was used for variance estimation of more complex statistics such as fertility and mortality rates.

The Taylor linearization method treats any percentage or average as a ratio estimate, r = y/x, where y represents the total sample value for variable y, and x represents the total number of cases in the group or subgroup under consideration.

For further details on sampling error calculations see Appendix B of the final report.

Data appraisal
---------------------------

Data quality tables were produced to review the quality of the data:
- Household age distribution
- Age distribution of eligible and interviewed women
- Age distribution of eligible and interviewed men
- Completeness of reporting
- Births by calendar years
- Reporting of age at death in days
- Reporting of age at death in months

Note: The tables are presented in Appendix C of the final report.

摘要
---------------------------
2015-16年坦桑尼亚人口与健康调查及疟疾指标调查（TDHS-MIS）的主要目标是提供最新的人口与健康基本指标估计值。该调查收集了关于生育水平、婚姻、性行为、生育偏好、计划生育方法的认知与使用、母乳喂养实践、营养、儿童与孕产妇死亡率、孕产妇及儿童健康、疟疾以及其他与健康相关问题的信息。此外，2015-16年TDHS-MIS还提供了6-59个月龄儿童和15-49岁女性贫血患病率的估计值，6-59个月龄儿童疟疾患病率的估计值，以及家庭食盐和女性尿液中碘浓度的估计值。

通过2015-16年TDHS-MIS收集的信息旨在协助政策制定者和项目管理者评估和设计旨在改善国家人口健康状况的方案和策略。

地理覆盖范围
---------------------------
全国

分析单元
---------------------------
- 家庭
- 0-5岁儿童
- 15-49岁女性
- 15-59岁男性

数据类型
---------------------------
样本调查数据 [ssd]

抽样程序
---------------------------
样本设计

2015-16年TDHS-MIS的样本设计分为两个阶段，旨在为整个国家、坦桑尼亚大陆的城乡地区以及桑给巴尔提供估计值。对于如避孕药使用等特定指标，样本设计允许对30个地区（25个来自坦桑尼亚大陆，5个来自桑给巴尔）的每个地区进行指标估计。第一阶段涉及选择样本点（聚类），包括为2012年坦桑尼亚人口与住房普查划定的统计区域（EAs）。共选择了608个聚类。

在第二阶段，进行了系统的家庭选择。在实地工作之前，对所有608个选定的聚类进行了完整的家庭清单编制。然后，从清单中系统地选择了每个聚类的22个家庭，从而为2015-16年TDHS-MIS提供了13,376个家庭的代表性概率样本。为了估计某些人口指标的地理差异，坦桑尼亚被划分为九个地理区域。尽管这些区域不是官方的行政区域，但该分类系统也被MoHCDGEC的生殖与儿童健康部门所采用。将地区划分为区域允许在分母中有较大的人群数量，并减少了抽样误差。请注意，以下定义的区域与之前DHS调查中使用的区域略有不同。因此，跨区域和跨调查的比较应谨慎进行。区域如下：
西部区域：塔波拉，基戈马
北部区域：基里曼贾罗，塔安加，阿鲁沙
中部区域：多多马，辛吉达，马尼亚拉
南部高地区域：伊林加，琼贝，鲁武马
南部区域：林迪，姆特瓦拉
西南高地区域：姆贝亚，鲁夸，卡塔维
湖区：卡盖拉，姆万扎，吉塔，马拉，希尼扬加，希米尤
东部区域：达累斯萨拉姆，庞尼，莫罗戈罗
桑给巴尔：卡斯基齐尼翁古贾，卡斯基齐尼佩姆巴，姆吉尼马哈拉比，卡斯基齐尼佩姆巴

所有在调查前一个晚上在家庭中居住或访问的15-49岁女性都被纳入2015-16年TDHS-MIS，并有权接受访谈。在所有选定的家庭样本的三分之一中，如果他们在调查前一个晚上在家庭中居住或访问，则所有15-49岁男性都有资格接受访谈。在所有家庭中，在家长或监护人的同意下，对6-59个月龄的儿童进行了贫血和疟疾的检测。所有接受访谈的女性都进行了贫血检测。在选定与男性访谈的家庭中，接受访谈的女性被要求提供尿液样本和家庭食盐样本供实验室检测碘。

有关样本设计和实施的详细情况，请参阅最终报告附录A。

数据收集方式
---------------------------
面对面 [f2f]

研究工具
---------------------------
2015-16年TDHS-MIS使用了四个问卷：家庭问卷、女性问卷、男性问卷和生物标志物问卷。这些问卷基于DHS计划的标准人口与健康调查（DHS）问卷。它们被改编以反映与坦桑尼亚相关的人口与健康问题。征求了代表政府部、委和机构、非政府组织以及发展伙伴的各利益相关者的意见。在准备最终问卷的英文版本后，问卷被翻译成斯瓦希里语。

数据清理操作
---------------------------
在2015-16年TDHS-MIS中，首次数据录入是在实地数据收集的同时进行的。在完成、编辑和检查纸质问卷后，数据被输入到配备数据录入程序的平板电脑中，由编辑完成。完成后的问卷随后被发送到NBS总部，在那里它们被第二次录入并由经过特别培训的数据处理人员编辑。ICF国际在整个数据处理期间提供了技术援助。与数据收集同时处理数据允许定期监控团队表现和数据质量。在数据处理期间，定期生成现场检查表以检查各种数据质量参数。因此，定期提供了反馈，鼓励团队在表现良好的领域继续努力，并在需要改进的领域进行纠正。反馈针对每个团队量身定制。数据录入（包括100%的双录入以最小化按键错误）和数据编辑于2016年3月21日完成。数据清理和最终确定于2016年4月22日完成。

响应率
---------------------------
共选择了13,360个家庭进行调查，其中12,767个被占用。在占用的家庭中，12,563个家庭成功接受了访谈，响应率为98%。在访谈的家庭中，确定了13,634名有资格接受个别访谈的合格女性；完成了13,266名女性的访谈，响应率为97%。在选定进行男性调查的样本家庭中，确定了3,822名有资格的男性，其中3,514名成功接受了访谈，响应率为92%。农村和城市住宅之间的家庭响应率变化很小。

抽样误差估计
---------------------------
样本调查的估计值受两种类型的误差影响：（1）非抽样误差，和（2）抽样误差。非抽样误差是实施数据收集和数据处理过程中所犯错误的结果，例如未能找到和访谈正确的家庭，访谈员或受访者对问题的误解，以及数据录入错误。尽管在实施2015年坦桑尼亚人口与健康调查（TDHS）期间做出了众多努力以最大限度地减少此类错误，但非抽样误差是无法避免且难以从统计上进行评估的。

另一方面，抽样误差可以统计评估。2015年TDHS中选定的受访者样本只是从同一人口中，使用相同的设计和相同规模可以选出的许多样本之一。这些样本中的每一个都会产生与实际选定的样本结果有所不同的结果。抽样误差是衡量所有可能样本之间变异性的指标。虽然变异程度无法确切知晓，但可以从调查结果中估计出来。

抽样误差通常用特定统计量（平均数、百分比等）的标准误差来衡量，它是方差的平方根。标准误差可用于计算置信区间，在此区间内可以合理地假设总体真实值。

例如，对于从样本调查中计算出的任何给定统计量，该统计量的值将在95%的所有可能样本（相同规模和设计）的标准误差的两倍范围内。

如果受访者样本被选为简单随机样本，则可以使用简单的公式来计算抽样误差。然而，2015年TDHS样本是多层次分层设计的产物，因此有必要使用更复杂的公式。用于计算2015年TDHS抽样误差的计算机软件是SAS程序。该程序使用泰勒线性化方法对调查估计值（平均数或比例）进行方差估计。对于如生育率和死亡率等更复杂的统计量，使用了Jackknife重复复制方法进行方差估计。

泰勒线性化方法将任何百分比或平均数视为比率估计，r = y/x，其中y代表变量y的总样本值，x代表考虑的群体或子群体中的总案例数。

有关抽样误差计算的详细情况，请参阅最终报告附录B。

数据评估
---------------------------
生成了数据质量表以审查数据质量：
- 家庭年龄分布
- 合格和接受访谈的女性年龄分布
- 合格和接受访谈的男性年龄分布
- 报告的完整性
- 按日历年出生
- 死亡年龄报告
- 死亡年龄按月报告

注意：表格在最终报告附录C中呈现。