Kenya National Health Account 2007 - Kenya

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

National Health Accounts (NHA) is an internationally recognised method used to track expenditures in a health system for a specified period of time. Specifically, NHA details the flow of funding from financial sources (e.g., donors, Ministry of Finance), to financing agents (i.e., those who manage the funds, such as
the Ministry of Health [MoH] or nongovernmental organisations [NGOs]), to providers (e.g., public and private facilities) and finally to end users (e.g., inpatient and outpatient care, pharmaceuticals).

Actual expenditures, rather than budget inputs, are used to fill a series of tables that show the flow of funding through the health sector. NHA also provides detailed breakdowns of disease-specific expenditures such as those for HIV/AIDS and reproductive health (RH). NHA is designed to be used as a policy tool to facilitate the implementation of health system goals.
This report describes findings from the third round of NHA in Kenya.

The first two estimations covered financial years (FYs) 1994/95 and 2001/02, respectively. This third round, undertaken in 2007 and covering 2005/06 was implemented by the MoH and Kenya National Bureau of Statistics (KNBS) with financial support from the United States Agency for International Development (USAID). USAID’s
Health Systems 20/20 Project, led by Abt Associates Inc., provided technical support. The findings will be used as a platform for informing policy decisions concerning resource allocation and will also be used by stakeholders in the sector.

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

The whole country

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

households and institutions

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

Household health expenditure covered all households in the country whereas the institutional surveys covered firms selected under the review.

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

Administrative records data [adm]

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

Kenya is divided into eight administrative provinces. The provinces are in turn subdivided into 70
districts. Each district is subdivided into divisions while the divisions are split into locations and finally
each location into sublocations.

During the 1999 population census, each sublocation was subdivided into smaller units called enumeration areas (EAs). Kenya has about 62,000 EAs. The EAs provided census information on households and population. This information was used in the design of the National Sample Survey Evaluation Programme (NASSEP) IV master sample with 1,800 selected EAs.

The cartographic records for each EA in the master sample were updated in the field, one year preceding the NHA survey. The 1,800 clusters were distributed into 540 urban and 1,260 rural clusters.

The province provided a natural stratification of the population. The six major urban centres Nairobi, Mombasa, Kisumu, Nakuru, Eldoret, and Thika were further substratified into five socioeconomic classes based on incomes to circumvent the extensive socioeconomic diversity inherent in them as follows: upper, lower upper, middle, lower middle and lower; this improved the precision of estimates due to reduced sampling variation.

It was estimated that 8,844 households would be sufficient to provide estimates both at provincial and national levels as well as disaggregation to urban and rural components of the country. This sample was to yield 6,060 interviews in the rural and 2,784 in the urban clusters (Table 2.2). This was to be achieved through coverage of 737 clusters (505 rural and 232 urban clusters).

Twelve households were to be covered in each cluster. The method of proportional allocation was used in assigning the sample households to the provinces and districts. The count of the households was transformed to the square
root of the census households to avoid under-representing the smaller districts.

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

Face-to-face [f2f]

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

To expedite data entry and monitor data quality, all completed questionnaires were sent to a data
management unit at the MoH Planning Department, which was the designated secretariat for the activity.

This approach helped in standardizing and speeding up data entry and reducing errors. Questionnaires were also checked for completeness before entry.

Data were entered in a Census and Survey Processing System (CSPro) by a team of data entry clerks under the supervision of data entry supervisors. The data were reentered for validation.
The data files were then converted into SPSS, the
software used for data analysis. Much of the analysis was replicated using Stata, to confirm that weighted estimates were correct. Stata was also used to perform analysis that could not be undertaken using SPSS.

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

A total of 8,844 households were selected for the survey. Of these, 8,453 were successfully interviewed, giving a response rate of 95.6
percent, and the survey reported observations on 38,235 individuals living in these households.

摘要
---------------------------
国家卫生账户（NHA）是一种国际认可的方法，用于追踪在一定时期内卫生系统中资金支出的流向。具体而言，NHA详细说明了资金从金融来源（例如，捐助者、财政部）流向融资代理（即管理资金者，如卫生部[MoH]或非政府组织[NGOs]），再流向提供者（例如，公立和私立设施）以及最终用户（例如，住院和门诊护理、药品）的过程。
实际支出而非预算输入被用来填充一系列表格，以展示资金在卫生部门的流动情况。NHA还提供了针对特定疾病的支出详细分解，如HIV/AIDS和生殖健康（RH）。NHA旨在作为政策工具，以促进卫生系统目标的实施。本报告描述了肯尼亚第三次国家卫生账户调查的结果。
第一次和第二次估算分别涵盖了1994/95财年和2001/02财年。这次第三次调查于2007年进行，涵盖了2005/06财年，由卫生部和肯尼亚国家统计局（KNBS）实施，并得到美国国际开发署（USAID）的财政支持。由Abt Associates Inc.领导的USAID卫生系统20/20项目提供了技术支持。这些发现将被用作制定有关资源分配的政策决策的平台，并将由该领域的利益相关者使用。
地理覆盖范围
---------------------------
全国
分析单元
---------------------------
家庭和机构
总体
---------------------------
家庭卫生支出覆盖了全国所有家庭，而机构调查涵盖了审查下的企业。
数据类型
---------------------------
行政记录数据[adm]
抽样程序
---------------------------
肯尼亚被划分为八个行政省。这些省份随后被划分为70个区。每个区被划分为分区，而分区被划分为地点，最后每个地点被划分为次地点。
在1999年人口普查期间，每个次地点被划分为更小的单位，称为统计区域（EA）。肯尼亚大约有62,000个统计区域。这些统计区域提供了关于家庭和人口普查的信息。这些信息被用于设计国家样本调查评估计划（NASSEP）IV主样本，其中选择了1,800个统计区域。
主样本中每个统计区域的地图记录在NHA调查前一年在实地更新。1,800个集群被分配到540个城市集群和1,260个农村集群。
省份为人口提供了自然的分层。六个主要城市中心内罗毕、蒙巴萨、基苏木、纳库鲁、埃尔多雷特和提卡被进一步根据收入划分为五个社会经济阶层，以规避其中固有的广泛社会经济多样性，如下所示：上层、中上层、中层、中下层和下层；这通过减少抽样误差提高了估计的精度。
估计有8,844个家庭足以提供省级和国家层面的估计，以及对国家和城市成分的细分。这个样本预计将在农村产生6,060次访谈，在城市产生2,784次访谈（见表2.2）。这是通过覆盖737个集群（505个农村集群和232个城市集群）来实现的。
每个集群应覆盖12个家庭。在将样本家庭分配到省份和地区时使用了比例分配法。将家庭数量转换为普查家庭数量的平方根，以避免小地区代表性不足。
数据收集方式
---------------------------
面对面[f2f]
数据清理操作
---------------------------
为了加快数据录入并监控数据质量，所有完成后的问卷都被发送到卫生部规划部门的数据管理单位，该单位是活动的指定秘书处。
这种做法有助于标准化和加快数据录入，并减少错误。在录入之前也会检查问卷的完整性。
数据由数据录入员团队在数据录入监督员的监督下输入到人口普查和调查处理系统（CSPro）中。数据被重新输入以进行验证。
然后，数据文件被转换为SPSS，这是数据分析使用的软件。大部分分析都使用Stata进行复制，以确认加权估计是正确的。Stata还用于执行无法使用SPSS进行的分析。
响应率
---------------------------
总共选择了8,844个家庭进行调查。其中，8,453个家庭成功接受了访谈，响应率为95.6%，调查报告了对这些家庭中38,235个个体的观察。