USE OF HMIS-DHIS 2 HMIS M-4. 1 Module-4 Data Quality and Validation Learning objectives: After reading this module you will be able to understand: 1. What is data quality and its importance for HMIS. 2. How to do data quality check at point of data entry. 3. How to create data validation rules. 4. How to carry out data triangulation. 5. How to analyze data status.
21
Embed
Data Quality and Validation - · PDF fileData Quality and Validation ... HMIS software should be built in with different tools to do data quality ... desired formula
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
USE OF HMIS-DHIS 2 HMIS
M-4. 1
Module-4
Data Quality and
Validation
Learning objectives:
After reading this module you will be able to understand:
1. What is data quality and its importance for
HMIS.
2. How to do data quality check at point of data
entry.
3. How to create data validation rules.
4. How to carry out data triangulation.
5. How to analyze data status.
USE OF HMIS-DHIS 2 HMIS
M-4. 2
4.1 Overview of data quality check
Ensuring data quality is a key concern in building an effective HMIS. Data
quality has different dimensions including:
If we have poor quality data, we will have “garbage in and garbage out”
situations. Use of poor quality data leads to ill informed decisions. So, the
HMIS software should be built in with different tools to do data quality
checks and validation.
4.1.1 Data quality checks
Data quality checking can be done through various means, including:
1. At point of data entry, the software can check the data entered to
see if it falls within the min-max ranges of that data element over
the last six months or as defined by the user.
• Correctness: Data should be within the normal range for data collected at that facility. There should be no gross discrepancies when compared with data from related data elements.
• Completeness: Data for all data elements for all health
facilities/blocks/Taluka/districts should have been submitted.
• Consistency: Data should be consistent with data entered during earlier months and years while allowing for changes with reorganization, increased work load, etc. and consistent with other similar facilities.
• Timeliness: All data from all health facilities/blocks/Taluka/districts
should be submitted at the appointed time.
USE OF HMIS-DHIS 2 HMIS
M-4. 3
2. Defining various validation rules, which can be run once the user
has finished data entry. The user can also check the entered data
for a particular period and Organization Unit(s) against the
validation rules, and display the violations for these validation rules.
3. Analysis of data sets, ie, examining gaps in data.
4. Data triangulation which is comparing the same data or indicator
from different sources.
4.2 Data quality check at the point of data entry
Data quality can be checked at the point of data entry in the following two
ways:
a) By setting the minimum and maximum value range for each element
manually. Or
b) Generating the min-max values using the DHIS 2 if there is historical
data available for that data element.
a) Setting the minimum and maximum value range manually
If you are using the default entry screen click on the element for which you
want to set the min-max value, as shown below.
USE OF HMIS-DHIS 2 HMIS
M-4. 4
A pop up window will appear as shown below. Here you can enter the
min-max values.
On subsequent data entry if the value entered does not fall within the set
min-max range the text box will change colour to red. The user will also
USE OF HMIS-DHIS 2 HMIS
M-4. 5
get a popup as shown below. This change in colour is a prompt to check
the data entered and make necessary correction.
On the data entry screen the users also have the option to add a comment
on how the discrepant figure might be explained (if required). This you can
do by using the drop down menu of the ‘comment’ box.
In case you are using the custom data entry screen which is displayed
when you deselect the ‘default data entry form’ option on the top right
USE OF HMIS-DHIS 2 HMIS
M-4. 6
corner of the screen. In this case the minimum and maximum values can
be added by double-clicking on the data entry box instead of the data
element.
b) Generated min-max values
If you have a minimum of six months of your data entered in the DHIS2 it
is possible to generate the min-max value, element-wise, using the
DHIS2. In such case you merely need to click on the ‘Generate min-max’
tab as shown below.
In case of default data entry screen the min and max values, when
generated, will appear on the left and right side of the data entry box.
In case you deselect the default data entry form the generated values will
appear on the top right end of the screen as shown in the following
screenshot.
USE OF HMIS-DHIS 2 HMIS
M-4. 7
USE OF HMIS-DHIS 2 HMIS
M-4. 8
4.3 Defining Validation Rules
Validation rules are data quality check mechanism based on verification of
the logic of relation between related data elements. Validation rules are
relational expressions comprising of related data elements and an
operator that states the expected / logical relation between the elements.
For example number of infant deaths cannot be greater than the number
of deliveries. As can be seen from the example a validation rule comprises
of a left and a right side. On the left side of the expression, there must be
a data element or a combination of data elements, and the same on the
right side. The left and right hand sides of the expression are separated
with a validation operator which states the realtion between the elements.
As validation rules have a relational property there must be atleast two
data elements for which the validation rules may be applied.
4.3.1 Types of validation operators (equal to, less than,
greater than):
Following are some validation operators used for data quality analysis in
DHIS.
• Equal to: It will validate the validation rule only if both sides are
equal.
• Not Equal to: It will validate the rule if both the sides Not Equal
• Greater Than: It will validate the rule if the left side is greater than
the right side.
• Greater Than Equal to: It will validate the rule, if the left side is
Greater or Equal to the right hand side.
USE OF HMIS-DHIS 2 HMIS
M-4. 9
• Less Than: It will validate the rule if the left side is smaller than the
right hand side.
• Less Than Equal to: It will validate the rule if the left side is either
smaller or equal to the right side.
4.3.2 Adding new validation rule
Follow the steps below to add a new valiadtion rule.
First select the Data Quality module from the drop down menu of the
Services module located on the main tool bar.
USE OF HMIS-DHIS 2 HMIS
M-4. 10
In the screen that is displayed – Validation Rule Management screen, click
on ‘Add new’
The following screen will appear.
USE OF HMIS-DHIS 2 HMIS
M-4. 11
Enter the first three fields specifically validation rule name, description of
the validation rule and select the particular operator that forms the
validation rule. Next click on the Edit left side button to enter the ‘left side’
details of the concerned validation rule.
The following steps can be used by the user:
1. Add Description
2. Select data element from the Available Data Elements options
shown on the right side.
3. Add Operators in between the data elements to generate the
desired formula.
USE OF HMIS-DHIS 2 HMIS
M-4. 12
4. When you have entered the required fields click on ‘update’. This
will return you to the previous window. Here click on the ‘Edit right
side’ and follow the steps that you followed for the ‘Edit left’.
4.4 Validation Checks
When you open the Data Quality module you will see a menu on the left
side that lists different options related to validation rules. For purposes of
validation checks you will need to use the ‘Run validation’ option. This is
described below.
4.4.1 Run validation:
1. If you select the run validation option the following screen will be
displayed.
USE OF HMIS-DHIS 2 HMIS
M-4. 13
2. You will be required to specify the period for which you want to run the
validation check by selecting the start and end dates. This you can do by
using the drop down calender provided for the date fields.
3. Next select the particular organisation unit (s) for which you want to run
the validation.
4. Finally click on the ‘validate’ tab.
5. When you click on ‘Validate’ (Number 5 on the screenshot) button the
following popup will be displayed which will list the validation rules that are
violated with data values of the elements constituting the particular
validation rule.
USE OF HMIS-DHIS 2 HMIS
M-4. 14
4.4 Diagnosing the source of the validation
violation:
This you can accomplish by selecting the ‘Run validation by avergae’
option. You can run this validation after entering the required fields of the
‘run validation’ screen which includes the organisation units(s) and the
period for which you want to run the validation. The result of ‘Run
validation by average’ is a pop up screen (see screenshot below) that
displays the percentage of validation rules violated by the selected
orgnaisation unit(s).
Amogst the Orgunits showing violation, select the one which has
maximum violation percentange, and drill down to get its detailed
USE OF HMIS-DHIS 2 HMIS
M-4. 15
validation list. You could do the same for any other organisation unit as
well.
The screen that gets displayed presents the list of validation rules that
have been violated by the specific orgnaisation unit (see screenshot
below).
USE OF HMIS-DHIS 2 HMIS
M-4. 16
To get drilldown for one validation, click on any validation rule, it will give
you the detailed validation analysis for the selected orgunit and its
immediate children.
USE OF HMIS-DHIS 2 HMIS
M-4. 17
If you click on any orgunit you can drilldown to its children.
4.5 Analysis of data status
The purpose of analysis of data status is to see what is the percentage of
missing or unreported data either by data elements or by facilities.
4.5.1 Types of missing data
Missing data can be listed by facilities and or data elements. Missing data
creates different kinds of problems, such as:
1. Incomplete reports.
2. Indicator calculations will be misleading as there will be some
numerator or denominator that is missing.
3. Effective decisions cannot be based on incomplete data.
USE OF HMIS-DHIS 2 HMIS
M-4. 18
4. Probably, the facility which is not reporting data is the one which
needs more care and support.
4.5.2 Generating missing data reports by facilities, data
elements and periods
‘Data Status’ option provides us the tool to analyze how much data is
entered. You can find this option in Dashboard Module which is inturn
displayed in the drop down menu of the services module.
Clicking on the dashboard module will lead to the following screen where
you can find the ‘Data status’ option.
USE OF HMIS-DHIS 2 HMIS
M-4. 19
Once you click on ‘Data Status’ option you will get the following screen
where you can select orgunit, dataset, period for which you want to
generate the data status. Once you have done this click on ‘View data
status’ as shown below.
USE OF HMIS-DHIS 2 HMIS
M-4. 20
You will get the following output.
From here you can go drill down to its immediate children by clicking any
orgunit to obtain more detailed sub facility wise data status.
USE OF HMIS-DHIS 2 HMIS
M-4. 21
4.6 Data Triangulation
Data (for example on institutional deliveries) is collected from different
sources such as routine health data and NFHS surveys. By plotting data
on this data element across the three surveys and juxtaposing it with
routine data, we can have a method of data triangulation.
In the boxes below, the NFHS trends in institutional delivery are compared with
trends of monthly figures from the state routine HMIS.