Malaria Indicator Survey 2020 - Kenya

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

The 2020 Kenya Malaria Indicator Survey (2020 KMIS) was a cross-sectional household-based survey with a nationally representative sample of conventional households. The survey targeted women age 15-49 and children age 6 months to age 14 living within conventional households in Kenya. All women age 15-49 who were usual members of the selected households or who spent the night before the survey in the selected households were eligible for individual interviews. In all sampled households, children age 6 months to age 14 were tested for anaemia and malaria.

The sample for the 2020 KMIS was designed to produce reliable estimates for key malaria indicators at the national level, for urban and rural areas separately, and for each of the five malaria endemic zones.

The 2020 KMIS was designed to provide information on the implementation of core malaria control interventions and serve as a follow-up to the previous malaria indicator surveys. The specific objectives of the 2020 KMIS were as follows:
- To measure the extent of ownership of, access to, and use of mosquito nets
- To assess coverage of intermittent preventive treatment of malaria during pregnancy
- To examine fever management among children under age 5
- To measure the prevalence of malaria and anaemia among children age 6 months to age 14
- To assess knowledge, attitudes, and practices regarding malaria control
- To determine the Plasmodium species most prevalent in Kenya

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

National coverage

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

- Household
- Individual
- Children age 0-14
- Woman age 15-49

Universe
---------------------------

The survey covered all de jure household members (usual residents), women age 15-49 years and children age 0-14 years resident in the household.

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

Sample survey data [ssd]

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

The 2020 KMIS followed a two-stage stratified cluster sample design and was intended to provide estimates of key malaria indicators for the country as a whole, for urban and rural areas, and for the five malaria-endemic zones (Highland epidemic prone, Lake endemic, Coast endemic, Seasonal, and Low risk).

The five malaria-endemic zones fully cover the country, and each of the 47 counties in the country falls into one or two of the five zones as follows:
1. Highland epidemic prone: Kisii, Nyamira, West Pokot, Trans-Nzoia, Uasin Gishu, Nandi, Narok, Kericho, Bomet, Bungoma, Kakamega, and Elgeyo Marakwet
2. Lake endemic: Siaya, Kisumu, Migori, Homa Bay, Kakamega, Vihiga, Bungoma, and Busia
3. Coast endemic: Mombasa, Kwale, Kilifi, Lamu, and Taita Taveta
4. Seasonal: Tana River, Marsabit, Isiolo, Meru, Tharaka-Nithi, Embu, Kitui, Garissa, Wajir, Mandera, Turkana, Samburu, Baringo, Elgeyo Marakwet, Kajiado, and West Pokot
5. Low risk: Nairobi, Nyandarua, Nyeri, Kirinyaga, Murang’a, Kiambu, Machakos, Makueni, Laikipia, Nakuru, Meru, Tharaka-Nithi, and Embu.

The survey utilised the fifth National Sample Survey and Evaluation Programme (NASSEP V) household master sample frame, the same frame used for the 2015 KMIS. The frame was used by KNBS from 2012 to 2020 to conduct household-based sample surveys in Kenya. It was based on the 2009 Kenya Population and Housing Census, and the primary sampling units were clusters developed from enumeration areas (EAs). EAs are the smallest geographical areas created for purposes of census enumeration; a cluster can be an EA or part of an EA. The frame had a total of 5,360 clusters and was stratified into urban and rural areas within each of 47 counties, resulting into 92 sampling strata with Nairobi and Mombasa counties being wholly urban.

The survey employed a two-stage stratified cluster sampling design in which, in the first stage of selection, 301 clusters (134 urban and 167 rural) were randomly selected from the NASSEP V master sample frame using an equal probability selection method with independent selection in each sampling stratum. The second stage involved random selection of a fixed number of 30 households per cluster from a roster of households in the sampled clusters using systematic random sampling.

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

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

Computer Assisted Personal Interview [capi]

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

Four types of questionnaires were used for the 2020 KMIS: the Household Questionnaire, the Woman’s Questionnaire, the Biomarker Questionnaire, and the Fieldworker Questionnaire. The questionnaires were adapted to reflect issues relevant to Kenya. Modifications were determined after a series of meetings with various stakeholders from DNMP and other government ministries and agencies, nongovernmental organisations, and international partners. The Household and Woman’s Questionnaires in English and Kiswahili were programmed into Android tablets, which enabled the use of computer-assisted personal interviewing (CAPI) for data collection. The Biomarker Questionnaire, in English and Kiswahili, was filled out on hard copy and then entered into the CAPI system.

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

The 2020 KMIS questionnaires were programmed using Census and Survey Processing (CSPro) software. The program was then uploaded into Android-based tablets that were used to collect data via CAPI. The CAPI applications, including the supporting applications and the applications for the Household, Biomarker, and Woman’s Questionnaires, were programmed by ICF. The field supervisors transferred data daily to the CSWeb server, developed by the U.S. Census Bureau and located in Nairobi, for data processing on the central office computer at the KNBS office in Nairobi.

Data received from the field teams were registered and checked for any inconsistencies and outliers on the central office computer at KNBS. Data editing and cleaning included an extensive range of structural and internal consistency checks. All anomalies were communicated to field teams, which resolved data discrepancies. The corrected results were maintained in the central office computer at KNBS head office. The central office held data files which was used to produce final report tables and final data sets. CSPro software was used for data editing, cleaning, weighting, and tabulation.

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

A total of 8,845 households were selected for the survey, of which 8,185 were occupied at the time of fieldwork. Among the occupied households, 7,952 were successfully interviewed, yielding a response rate of 97%. In the interviewed households, 7,035 eligible women were identified for individual interviews and 6,771 were successfully interviewed, yielding a response rate of 96%.

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

The estimates from a sample survey are affected by two types of errors: non-sampling errors and sampling errors. Non-sampling 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 of 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 2020 Kenya Malaria Indicator Survey (KMIS) to minimise this type of error, non-sampling 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 2020 KMIS is only one of many samples that could have been selected from the same population, using the same design and expected 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.

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% 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 2020 KMIS sample is the result of a multi-stage stratified design, and, consequently, it was necessary to use more complex formulas. Sampling errors are computed in SAS, using programs developed by ICF. These programs use the Taylor linearisation method of variance estimation for survey estimates that are means, proportions, or ratios.

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

Data Quality Tables
- Household age distribution
- Age distribution of eligible and interviewed women
- Completeness of reporting
- Births by calendar years
- Number of enumeration areas completed, by month and malaria endemicity
- Positive rapid diagnostic test (RDT) results, by month and malaria endemicity
- Concordance and discordance between RDT and microscopy results
- Concordance and discordance between national and external quality control laboratories

See details of the data quality tables in Appendix C of the final report.

摘要
---------------------------
2020年肯尼亚疟疾指标调查（2020 KMIS）是一项基于横断面家庭调查，具有全国代表性常规家庭的样本。该调查针对生活在肯尼亚常规家庭中的15-49岁女性和6个月至14岁的儿童。所有15-49岁的女性，无论是选定家庭的常驻成员，还是调查前一天在选定家庭过夜的，均有资格接受个别访谈。在所有抽样家庭中，6个月至14岁的儿童均接受了贫血和疟疾的检测。
2020 KMIS的样本设计旨在为全国、城市和农村地区以及五个疟疾流行区（高原易感、湖区、海岸、季节性和低风险）的国家关键疟疾指标提供可靠的估计。
2020 KMIS旨在提供关于核心疟疾控制干预措施实施的信息，并作为之前疟疾指标调查的后续调查。2020 KMIS的具体目标如下：
- 测量蚊帐拥有率、可及性和使用率
- 评估孕期间歇性预防治疗疟疾的覆盖率
- 检查5岁以下儿童发热管理
- 测量6个月至14岁儿童疟疾和贫血的患病率
- 评估关于疟疾控制的认知、态度和实践
- 确定在肯尼亚最普遍的疟原虫种类

地理覆盖范围
---------------------------
全国覆盖

分析单位
---------------------------
- 家庭
- 个人
- 0-14岁儿童
- 15-49岁女性

总体
---------------------------
调查涵盖了所有法定家庭成员（常驻居民），以及居住在家庭中的15-49岁女性和0-14岁儿童。

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

抽样程序
---------------------------
2020 KMIS遵循了两阶段分层聚类样本设计，旨在为整个国家的关键疟疾指标提供估计，包括城市和农村地区，以及五个疟疾流行区（高原易感、湖区、海岸、季节性和低风险）。

抽样误差估计
---------------------------
样本调查的估计受到两种类型误差的影响：非抽样误差和抽样误差。非抽样误差是数据收集和处理过程中出现的错误的结果，例如未能找到和访谈正确的家庭，访谈者或受访者对问题的误解，以及数据输入错误。尽管在实施2020年肯尼亚疟疾指标调查（KMIS）期间做出了大量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进行统计评估的。
抽样误差另一方面，可以通过统计方法进行评估。2020 KMIS中选定的受访者样本只是从同一总体中选出的许多样本之一，使用相同的设计和预期规模。这些样本中的每一个都会产生与实际选定样本结果略有不同的结果。抽样误差是衡量所有可能样本之间差异的指标。尽管变异程度并不完全清楚，但它可以从调查结果中估计。
抽样误差通常以特定统计量（均值、百分比等）的标准误差来衡量，这是方差的平方根。标准误差可用于计算置信区间，其中可以合理地假设真实人口值的范围。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%的所有可能样本的相同大小和设计中，以加减两倍标准误差的范围内。

数据评估
---------------------------
数据质量表
- 家庭年龄分布
- 有资格和接受访谈的女性年龄分布
- 报告的完整性
- 日历年度出生人数
- 按月份和疟疾流行性完成的枚举区域数量
- 按月份和疟疾流行性快速诊断测试（RDT）结果
- RDT和显微镜结果之间的符合性和不一致性
- 国家和外部质量控制实验室之间的符合性和不一致性

详见最终报告附录C中的数据质量表。