Interesting post. Information Quality/Data Quality is a challenging area. Only today, the UK Audit Commission slated the NHS Data Quality programme ( http://bit.ly/KL9q) for lack of rigour.
Controlling access to data and rights to create/update/delete data is one way to help ensure quality. Buying in tools to help implement governance, oversight and checking is also an option. However, if you think of your database as a lake, this amounts to not much more than blocking input from certain streams you know to be dirty and having filtering and cleaning processes to clean the lake water (even if pollution is still coming in from elsewhere).
However, the key lessons from other industries (and from manufacturing quality) is that the optimum solution is to move your quality metrics and controls as close to the point of information creation as possible. As part of that you need to understand the process, ensure your people are trained in the process, and ensure that your people understand the WHY of the process as well as the HOW.
By understanding your process, and how that links to the objectives of your organization, you can better identify what metrics matter and make better decisions about tool investments and targeted actions to tackle root causes in process, technology or people issues.
The IAIDQ held a webinar back in March on the topic of Data Quality in Healthcare. Problems on the audio recording meant we couldn’t put up a podcast, but the slides can be found here: http://bit.ly/COG2L, with detailed speaker notes here:http://bit.ly/viYO
[Disclosure: I’m the Publicity Director of the IAIDQ, the only professional association focused solely on the challenges of Information/Data Quality.]
Tim,
Interesting post. Information Quality/Data Quality is a challenging area. Only today, the UK Audit Commission slated the NHS Data Quality programme ( http://bit.ly/KL9q) for lack of rigour.
Controlling access to data and rights to create/update/delete data is one way to help ensure quality. Buying in tools to help implement governance, oversight and checking is also an option. However, if you think of your database as a lake, this amounts to not much more than blocking input from certain streams you know to be dirty and having filtering and cleaning processes to clean the lake water (even if pollution is still coming in from elsewhere).
However, the key lessons from other industries (and from manufacturing quality) is that the optimum solution is to move your quality metrics and controls as close to the point of information creation as possible. As part of that you need to understand the process, ensure your people are trained in the process, and ensure that your people understand the WHY of the process as well as the HOW.
By understanding your process, and how that links to the objectives of your organization, you can better identify what metrics matter and make better decisions about tool investments and targeted actions to tackle root causes in process, technology or people issues.
The IAIDQ held a webinar back in March on the topic of Data Quality in Healthcare. Problems on the audio recording meant we couldn’t put up a podcast, but the slides can be found here: http://bit.ly/COG2L, with detailed speaker notes here:http://bit.ly/viYO
[Disclosure: I’m the Publicity Director of the IAIDQ, the only professional association focused solely on the challenges of Information/Data Quality.]