dhis2-devs team mailing list archive
-
dhis2-devs team
-
Mailing list archive
-
Message #12648
[Branch ~dhis2-documenters/dhis2/dhis2-docbook-docs] Rev 337: Added chapter on conceptual design principles and on integration
------------------------------------------------------------
revno: 337
committer: Lars Helge Overland <larshelge@xxxxxxxxx>
branch nick: dhis2-docbook-docs
timestamp: Sat 2011-06-18 18:37:00 +0200
message:
Added chapter on conceptual design principles and on integration
added:
src/docbkx/en/resources/images/implementation_guide/
src/docbkx/en/resources/images/implementation_guide/data_input_output.png
src/docbkx/en/resources/images/implementation_guide/indicator_report.png
src/docbkx/en/resources/images/implementation_guide/interoperable_his.png
renamed:
src/docbkx/en/dhis2_implementation_guide_system_design.xml => src/docbkx/en/dhis2_implementation_guide_conceptual_design_principles.xml
modified:
src/docbkx/en/dhis2_implementation_guide_en.xml
src/docbkx/en/dhis2_implementation_guide_integration.xml
src/docbkx/en/dhis2_implementation_guide_conceptual_design_principles.xml
--
lp:~dhis2-documenters/dhis2/dhis2-docbook-docs
https://code.launchpad.net/~dhis2-documenters/dhis2/dhis2-docbook-docs
Your team DHIS 2 developers is subscribed to branch lp:~dhis2-documenters/dhis2/dhis2-docbook-docs.
To unsubscribe from this branch go to https://code.launchpad.net/~dhis2-documenters/dhis2/dhis2-docbook-docs/+edit-subscription
=== renamed file 'src/docbkx/en/dhis2_implementation_guide_system_design.xml' => 'src/docbkx/en/dhis2_implementation_guide_conceptual_design_principles.xml'
--- src/docbkx/en/dhis2_implementation_guide_system_design.xml 2011-02-12 16:14:18 +0000
+++ src/docbkx/en/dhis2_implementation_guide_conceptual_design_principles.xml 2011-06-18 16:37:00 +0000
@@ -1,5 +1,61 @@
<?xml version='1.0' encoding='UTF-8'?>
+<!-- This document was created with Syntext Serna Free. -->
<!DOCTYPE chapter PUBLIC "-//OASIS//DTD DocBook XML V4.4//EN" "http://www.oasis-open.org/docbook/xml/4.4/docbookx.dtd"; []>
<chapter>
- <title>System Design</title>
+ <title>Conceptual Design Principles</title>
+ <para>This chapter provides a introduction to some of the key conceptual design principles behind the DHIS 2 software. Understanding and being aware of these principles will help the implementer to make better use of the software when customising a local database. While this chapter introduces the principles, the following chapters will detail out how these are reflected in the database design process. </para>
+ <para>The following conceptual design principles will be presented in this chapter:</para>
+ <itemizedlist>
+ <listitem>
+ <para>All meta data can be added and modified through the user interface</para>
+ </listitem>
+ <listitem>
+ <para>A flexible data model supports different data sources to be integrated in one single data repository</para>
+ </listitem>
+ <listitem>
+ <para>Data Input != Data Output</para>
+ </listitem>
+ <listitem>
+ <para>Indicator-driven data analysis and reporting</para>
+ </listitem>
+ <listitem>
+ <para>Maintain disaggregated facility-data in the database</para>
+ </listitem>
+ <listitem>
+ <para>Support data analysis at any level in the health system</para>
+ </listitem>
+ </itemizedlist>
+ <para>In the following section each principle is described in more detail.</para>
+ <section>
+ <title>All meta data can be added and modified through the user interface</title>
+ <para>The DHIS 2 application comes with a set of generic tools for data collection, validation, reporting and analysis, but the contents of the database, e.g. what data to collect, where the data comes from, and on what format, will depend on the context of use. This meta data need to be populated into the application before it can be used, and this can be done through the user interface and requires no programming. This allows for more direct involvement of the domain experts that understand the details of the HIS that the software will support. </para>
+ <para>The software separates the key meta data that describes the raw data being stored in the database, which is the critical meta data that should not change much over time (to avoid corrupting the data), and the higher level meta like indicator formulas, validation rules, and groups for aggregation as well as the various layouts for collection forms and reports, which are not that critical and can be changed over time without interfering with the raw data. As this higher level meta data can be added and modified over time without interfering with the raw data, a continuous customisation process is supported. Typically new features are added over time as the local implementation team learn to master more functionality, and the users are gradually pushing for more advanced data analysis and reporting outputs. </para>
+ </section>
+ <section>
+ <title>A flexible data model supports different data sources to be integrated in one single data repository</title>
+ <para>The DHIS 2 design follows an integrated approach to HIS, and supports integration of many different data sources into one single database, sometime referred to as an integrated data repository or a data warehouse.</para>
+ <para>The fact that DHIS 2 is a skeleton like tool without predefined forms or reports means that it can support a lot of different aggregate data sources. There is nothing really that limits the use to the health domain either, although use in other sectors are still very limited. As long as the data is collected by and orgunit, described as a data element (possibly with some disaggregation categories), and can be represented by a predefined period frequency, it can be collected and processed in DHIS 2. This flexibility makes DHIS 2 a powerful tool to set up integrated systems that bring together collection tools, indicators, and reports from multiple health programs, departments or initiatives. Once the data is defined and then collected or imported into a DHIS 2 database, it can be analysed in correlation to any other data in the same database, no matter how and by whom it was collected. In addition to supporting integrated data analysis and reporting, this integrated approach also helps to rationalise data collection and reduce duplication.</para>
+ </section>
+ <section>
+ <title>Data input != Data output</title>
+ <para>In DHIS 2 there are three dimensions that describe the aggregated data being collected and stored in the database; the where - organisation unit, the what - data element, and the when - period. The organisation unit, data element and period make up the three core dimensions that are needed to describe any data value in the DHIS 2, whether it is a in a data collection form, a chart, on a map, or in an aggregated summary report. When data is collected in an electronic data entry form, sometimes through a mirror image of the paper forms used at facility level, each entry field in the form can be described using these three dimensions. The form itself is just a tool to organise the data collection and is not describing the individual data values being collected and stored in the database. Being able to describe each data value independently through a Data Element definition (e.g. ‘Measles doses given <1 year’) provides important flexibility when processing, validating, and analysing the data, and allows for comparison of data across collection forms and health programs. </para>
+ <para>This design or data model approach separates DHIS from many of the traditional HIS software applications which threat the data collection forms as the key unit of analysis. This is typical for systems tailored to vertical programs’ needs and the traditional conceptualisation of the collection form as also being the report or the analysis output. The figure below illustrates how the more fine-grained DHIS design built around the concept of Data Elements is different and how the input (data collection) is separated from the output (data analysis), supporting more flexible and varied data analysis and dissemination. The data element ‘Measles doses given <1 y’ is collected as part of a Child Immunisation collection form, but can be used individually to build up an Indicator (a formula) called ‘Measles coverage <1y’ where it is combined with the data element called ‘Population <1y’, being collected through another collection form. This calculated Indicator value can then be used in data analysis in various reporting tools in DHIS 2, e.g. custom designed reports with charts, pivot tables, or on a map in the GIS module.</para>
+ <graphic fileref="resources/images/implementation_guide/data_input_output.png" align="center"/>
+ </section>
+ <section>
+ <title>Indicator-driven data analysis and reporting</title>
+ <para>What is referred to as a Data Element above, the key dimension that describes what is being collected, is sometimes referred to as an indicator in other settings. In DHIS 2 we distinguish between Data Elements who describe the the raw data, e.g. the counts being collected, and Indicators, which are formula-based and describe calculated values, e.g. coverage or incidence rates that are used for data analysis. Indicator values are not collected like the data (element) values, but instead calculated by the application based on formulas defined by the users. These formulas are made up of a factor (e.g. 1, 100, 100, 100 000), a numerator and a denominator, the two latter are both expressions based on one or more data elements. E.g. the indicator "Measles coverage <1 year" is defined a formula with a factor 100, a numerator ("Measles doses given to children under 1 year") and a denominator ("Target population under 1 year"). The indicator "DPT1 to DPT3 drop out rate" is a formula of 100 % x ("DPT1 doses given"- "DPT3doses given") / ("DPT1 doses given"). These formulas can be added and edited through the user interface by a user with limited training, as they are quite easy to set up and do not interfere with the data values stored in the database (so adding or modifying an indicator is not a critical operation).</para>
+ <para>Indicators represent perhaps the most powerful data analysis feature of the DHIS 2, and all reporting tools support the use of indicators, e.g. as displayed in the custom report in the figure above. Being able to use population data in the denominator enables comparisons of health performance across geographical areas with different target populations, which is more useful than only looking at the raw numbers. The table below uses both the raw data values (Doses) and indicator values (Cov) for the different vaccines. Comparing e.g. the two first orgunits in the list, Taita Taveta County and Kilifi County, on DPT-1 immunisation, we can see that while the raw numbers (659 vs 2088) indicate many more doses are given in Kilifi, the coverage rates (92.2 % vs 47.5 %) show that Taita Taveta are doing a better job immunising their target population under 1 year. Looking at the final column (Immuniz. Compl. %) which indicates the completeness of reporting of the immunisation form for the same period, we can see that the numbers are more or less the same in the two counties we compared, which tells us that the coverage rates are comparable across the two counties. </para>
+ <graphic fileref="resources/images/implementation_guide/indicator_report.png" align="center"/>
+ </section>
+ <section>
+ <title>Maintain disaggregated facility-data in the database</title>
+ <para>When data is collected and stored in DHIS 2 it will remain disaggregated in the database with the same level of detail as it was collected. This is a major advantage of having a database system for HIS as supposed to a paper-based or even spreadsheet based system. The system is designed to store large amounts of data and always allow drill-downs to the finest level of detail possible, which is only limited by how the data was collected or imported into the DHIS 2 database. In a perspective of a national HIS it is desired to keep the data disaggregated by health facility level, which is often the lowest level in the orgunit hierarchy. This can be done even without computerising this level, through a hybrid system of paper and computer. The data can be submitted from health facilities to e.g. district offices by paper (e.g. on monthly summary forms for one specific facility), and then at the district office they enter all the facility data into the DHIS 2 through the electronic data collection forms, one facility at a time. This will enable the districts health management teams to perform facility-wise data analysis and to e.g. provide print-outs of feedback reports generated by the DHIS 2, incl. facility comparisons, to the facility in-charges in their district.</para>
+ </section>
+ <section>
+ <title>Support data analysis at any level in the health system</title>
+ <para>While the name DHIS indicates a focus on the District, the application provides the same tools and functionality to all levels in the health system. In all the reporting tools the users can select which orgunit or orgunit level to analyse and the data displayed will be automatically aggregated up to the selected level. The DHIS 2 uses the orgunit hierarchy in aggregating data upwards and provides data by any orgunit in this hierarchy. Most of the reports are run in such a way that the users will be prompted to select an orgunit and thereby enable reuse the same report layouts for all levels. Or of desired, the report layouts can be tailored to any specific level in the health system if the needs differ between the levels.</para>
+ <para>In the GIS module the users can analyse data on e.g. the sub-national level and then by clicking on the map (on e.g. a region or province) drill down to the next level, and continue like this all the way down to the source of the data at facility level. Similar drill-down functionality is provided in the Excel Pivot Tables that are linked to the DHIS 2 database. </para>
+ <para>To speed up performance and reduce the response-time when providing aggregated data outputs, which may include many calculations (e.g. adding together 8000 facilities), DHIS 2 pre-calculates all the possible aggregate values and stores these in what is called a data mart. This data mart can be scheduled to run (re-built) at a given time interval, e.g. every night. </para>
+ </section>
</chapter>
=== modified file 'src/docbkx/en/dhis2_implementation_guide_en.xml'
--- src/docbkx/en/dhis2_implementation_guide_en.xml 2011-06-18 13:38:33 +0000
+++ src/docbkx/en/dhis2_implementation_guide_en.xml 2011-06-18 16:37:00 +0000
@@ -42,6 +42,8 @@
</bookinfo>
<xi:include xmlns:xi="http://www.w3.org/2001/XInclude"; href="dhis2_implementation_guide_users_and_user_roles.xml" encoding="UTF-8"/>
<xi:include xmlns:xi="http://www.w3.org/2001/XInclude"; href="dhis2_implementation_guide_data_elements_and_custom_dimensions.xml" encoding="UTF-8"/>
+ <xi:include xmlns:xi="http://www.w3.org/2001/XInclude"; href="dhis2_implementation_guide_conceptual_design_principles.xml" encoding="UTF-8"/>
+ <xi:include xmlns:xi="http://www.w3.org/2001/XInclude"; href="dhis2_implementation_guide_integration.xml" encoding="UTF-8"/>
<!-- <xi:include xmlns:xi="http://www.w3.org/2001/XInclude"; href="dhis2_implementation_guide_system_design.xml" encoding="UTF-8"/>
<xi:include xmlns:xi="http://www.w3.org/2001/XInclude"; href="dhis2_implementation_guide_database_development.xml" encoding="UTF-8"/>
<xi:include xmlns:xi="http://www.w3.org/2001/XInclude"; href="dhis2_implementation_guide_harmonisation.xml" encoding="UTF-8"/>
@@ -55,6 +57,5 @@
<xi:include xmlns:xi="http://www.w3.org/2001/XInclude"; href="dhis2_implementation_guide_human_resource.xml" encoding="UTF-8"/>
<xi:include xmlns:xi="http://www.w3.org/2001/XInclude"; href="dhis2_implementation_guide_support.xml" encoding="UTF-8"/>
<xi:include xmlns:xi="http://www.w3.org/2001/XInclude"; href="dhis2_implementation_guide_user_training.xml" encoding="UTF-8"/>
- <xi:include xmlns:xi="http://www.w3.org/2001/XInclude"; href="dhis2_implementation_guide_integration.xml" encoding="UTF-8"/>
<xi:include xmlns:xi="http://www.w3.org/2001/XInclude"; href="dhis2_implementation_startup_strategies.xml" encoding="UTF-8"/> -->
</book>
=== modified file 'src/docbkx/en/dhis2_implementation_guide_integration.xml'
--- src/docbkx/en/dhis2_implementation_guide_integration.xml 2011-02-11 15:11:12 +0000
+++ src/docbkx/en/dhis2_implementation_guide_integration.xml 2011-06-18 16:37:00 +0000
@@ -1,5 +1,75 @@
<?xml version='1.0' encoding='UTF-8'?>
+<!-- This document was created with Syntext Serna Free. -->
<!DOCTYPE chapter PUBLIC "-//OASIS//DTD DocBook XML V4.4//EN" "http://www.oasis-open.org/docbook/xml/4.4/docbookx.dtd"; []>
<chapter>
<title>Integration</title>
+ <para>This chapter will discuss the following topics:</para>
+ <itemizedlist>
+ <listitem>
+ <para>Integration and interoperability</para>
+ </listitem>
+ <listitem>
+ <para>Benefits of integration</para>
+ </listitem>
+ <listitem>
+ <para>What facilitates integration and interoperability</para>
+ </listitem>
+ <listitem>
+ <para>Architecture of interoperable HIS</para>
+ </listitem>
+ </itemizedlist>
+ <para>In the following each topic will be discussed in greater detail.</para>
+ <section>
+ <title>Integration and interoperability</title>
+ <para>In a country there will usually be many different, isolated health information systems. The reasons for this are many, both technical and organizational. Here the focus will be on what benefits integration of these systems will bring, and why it should be a priority. First, a couple of clarifications:</para>
+ <itemizedlist>
+ <listitem>
+ <para>When talking about integration, we think about the process of making different information systems appear as one, i.e. data from them to be available to all relevant users as well as the harmonization of definitions and dimensions so that it is possible to combine the data in useful ways.</para>
+ </listitem>
+ <listitem>
+ <para>A related concept is interoperability, which is one strategy to achieve integration. For purposes related to DHIS2, we say that it is interoperable with other software applications if it is able to share data with this. For example, DHIS2 and OpenMRS are interoperable, because there is support in both to share data definitions and data with each other.</para>
+ </listitem>
+ </itemizedlist>
+ <para>To say that something is integrated, then, means that they share something, and that they are available from one place, while interoperability usually means that they are able to do this sharing electronically. DHIS2 is often used as an integrated data warehouse, since it contains (aggregate) data from various sources, such as Mother and Child health, Malaria program, census data, and data on stocks and human resources. These data sources share the same platform, DHIS2, and are available all from the same place. These subsystems are thus considered integrated into one system. Interoperability will then be a useful way to integrate also those data sources available on also other software applications. For example, if census data is stored in some other database, interoperability between this database and DHIS2 would mean census data would be accessible in both (but only stored one place).</para>
+ </section>
+ <section>
+ <title>Benefits of integration</title>
+ <para>There are several potential benefits related to integration of systems. The most important are:i</para>
+ <itemizedlist>
+ <listitem>
+ <para>Calculation of indicators: many indicators are based on numerators and denominators from different data sources. Examples include mortality rates, including some mortality data as numerator and population data as denominator, staff coverage and staff workload rates (human resource data, and population and headcount data), immunization rates, and the like. For these to be calculated, you need both the numerator and denominator data, and they should thus be integrated into a single data warehouse. The more data sources that are integrated, the more indicators can be generated from the central repository.</para>
+ </listitem>
+ <listitem>
+ <para>Reduce manual processing and entering of data: with different data at the same place, there is no need to manually extract and process indicators, or re-enter data into the data warehouse. Especially interoperability between systems of different data types (such as patient registers and aggregate data warehouse) allows software for subsystems to both calculate and share data electronically. This reduces the amount of manual steps involved in data processing, which increases data quality.</para>
+ </listitem>
+ <listitem>
+ <para>There are organizational reasons for integration. If all data can be handled by one unit in the ministry of health, instead of in various subsystems maintained by the health programs, this one unit can be professionalized. With staff which sole responsibility is data management, processing, and analysis, more specialized skills can be developed and the information handling be rationalized.</para>
+ </listitem>
+ </itemizedlist>
+ </section>
+ <section>
+ <title>What facilitates integration and interoperability</title>
+ <para>There are three levels that need to be addressed in this regard:</para>
+ <itemizedlist>
+ <listitem>
+ <para>The motivation and will to integrate (organizational level)</para>
+ </listitem>
+ <listitem>
+ <para>Standard definitions (language level)</para>
+ </listitem>
+ <listitem>
+ <para>Standard for electronic storage and exchange (technical level)</para>
+ </listitem>
+ </itemizedlist>
+ <para>The first level is less of a topic in this guide, which takes as a point of departure that a decision has been taken about integration of data. However, it is an important issue and usually the most complex to solve given the range of actors involved in the health sector. Clear national policies on data integration, data ownership, routines for data collection, processing, and sharing, should be in place to address this issue. Often some period of disturbance to the normal data flow will take place during integration, so for many the long-term prospects of a more efficient system will have to be judged against the short-term disturbance. Integration is thus often a step-wise process, where measures need to be taken for this to happen as smoothly as possible.</para>
+ <para>At the language level, there is a need to be consistent about definitions. If you have two data sources for the same data, they need to be comparable. For example, if you collect malaria data from both standard clinics and from hospitals, this data need to describe the same thing if they need to be combined for totals and indicators. If a hospital is reporting malaria cases by sex but not age group, and other clinics are reporting by age group but not sex, this data cannot be analyzed according to either of these dimensions, though a total amount of cases will be possible to calculate. There is thus a need to agree on uniform definitions.</para>
+ <para>In addition to uniform definitions across the various sub-systems, data exchange standards must be adopted if data is to be shared electronically. The various software applications would need this to be able to understand each other. DHIS2 is supporting several data formats for import and export, but one standard format now supported by WHO is called SDMX-HD (Statistical Data and Metadata Exchange - Health Domain). Other software applications are also supporting this, and it allows the sharing of data definitions and aggregate data between them. For DHIS2, this means it supports import of aggregate data that are supplied by other applications, such as OpenMRS (for patient management), iHRIS (for human resources management)</para>
+ </section>
+ <section>
+ <title>Architecture of interoperable HIS</title>
+ <para>Since there are many different use-cases for health information, such as monitoring and evaluation, budgeting, patient management and tracking, logistics management, insurance, human resource management, etc, there will be many different types of software applications functioning within the health sector. Above the issue of interoperability has been addressed, and a plan or overview of the various interoperable software applications and their specific uses, along with what data should be shared between them, is termed an architecture for health information. </para>
+ <para>The role of the architecture is to function as a plan to coordinate the development and interoperability of various sub-systems within the larger health information system. It is advisable to develop a plan for the various components even if they are not currently running any software, to be able to adequately see the requirements in terms of data sharing. These requirements should then be part of any specification for the software when such is developed or procured.</para>
+ <para>Below is a simple illustration of an architecture, with a focus on the data warehouse for aggregate data. The various boxes represent use cases, such as managing logistics, tracking TB patients, general patient management, etc. All of these will share aggregate data with DHIS2. Note that the arrows are two-way, because there is also a synchronization of meta-data (definitions) involved, to make sure that the right data is shared. Also, an example of the logistics and financial data applications sharing data is also shown, as there are strong links between procuring drugs and handling the budget for this. There will be many such instances of sharing data; the architecture helps us to plan better for this being implemented as the ecosystem of software applications grow.</para>
+ <graphic fileref="resources/images/implementation_guide/interoperable_his.png" align="center"/>
+ </section>
</chapter>
=== added directory 'src/docbkx/en/resources/images/implementation_guide'
=== added file 'src/docbkx/en/resources/images/implementation_guide/data_input_output.png'
Binary files src/docbkx/en/resources/images/implementation_guide/data_input_output.png 1970-01-01 00:00:00 +0000 and src/docbkx/en/resources/images/implementation_guide/data_input_output.png 2011-06-18 16:37:00 +0000 differ
=== added file 'src/docbkx/en/resources/images/implementation_guide/indicator_report.png'
Binary files src/docbkx/en/resources/images/implementation_guide/indicator_report.png 1970-01-01 00:00:00 +0000 and src/docbkx/en/resources/images/implementation_guide/indicator_report.png 2011-06-18 16:37:00 +0000 differ
=== added file 'src/docbkx/en/resources/images/implementation_guide/interoperable_his.png'
Binary files src/docbkx/en/resources/images/implementation_guide/interoperable_his.png 1970-01-01 00:00:00 +0000 and src/docbkx/en/resources/images/implementation_guide/interoperable_his.png 2011-06-18 16:37:00 +0000 differ
