Example: HRIS Data Quality Guidelines

Objective

Institute procedures to reduce data entry errors.

Use the example guidelines from Uganda as a model for preparing a data quality plan, with the goal of minimizing errors during data entry.

Data Quality: Concepts

When entering data from paper forms, maintaining the quality, or accuracy, of the data is the most important consideration. Introduce procedures to minimize data entry errors, even if those procedures slow the data entry process.

Dual data entry, in which a paper record is entered into an electronic database at two separate times by two staff members, is the most effective way to reduce errors; any discrepancies between the two entries can be caught at the point of entry, compared against the original document, and corrected.

Should dual data entry be prohibitively expensive or time-intensive, a system of spot-checking, in which a randomly selected list of electronic records is checked against the original data collection forms, can be substituted. Log data-entry errors and review errors frequently to identify needed improvements in training methods, data collection forms, and software modifications.

Data entrants should be contracted to transfer data from paper to electronic form.  To encourage quality of work, it is recommended that the data entrants should be paid on a per-record basis, and only on the records that have passed the validation test.   To determine the per-record reimbursement, time tests should be done on a sample of the paper records to determine a reasonable number of records that can be performed in a normal working day.  The rate should be reflective of this number of records and will incentivize increased efforts by the data entrants.

Well-defined roles are useful for enforcing accountability in updating staff data. To distinguish roles and responsibilities in the chain of command, user levels and permissions can be set within iHRIS.  These roles can also be connected to the facility, department, district, or region that they manage.  Each level has explicit accountability for data accuracy and completeness.

Tool filename (offline version): Tool3-E.pdf; download tool (Internet connection required)