Getting To Knowledge and Insight
“We are drowning in information, but starved for knowledge.”
–Futurist John Naisbitt
Does this ring true for your organization? Twitter alone generates more than 20 terabytes of data every day. Specific to healthcare, clinical and administrative information systems are generating terabytes of data that are begging to be used to measure and improve the care we provide across the continuum.
“The goal is to transform data into information, and information into insight,” said former Hewlett-Packard’s CEO Carly Fiorina. At LVHN, we are on the way to achieving that goal—implementing new data warehouse and business intelligence (BI) tools. In a best-of-breed environment, this is quite a challenge. In addition to creating the new warehouse and mapping the data elements, we are employing state-of-the-art normalization tools to ensure, for example, that a Hgb from one system is matched up with a Hb from another system; and an enterprise master patient index to ensure that the data regarding Don Levick in one system are matched up with the data on Donald Levick from another system. The leading-edge BI tool will allow clinicians to perform ad hoc reporting and dashboard creation, both in retrospective mode for population health management and at the point of care.
The biggest hurdles we have identified to achieve this very cool vision are two-fold: ensuring data are entered into the system accurately and that the information is entered in the correct place. The new data profiling tools have uncovered the unsettling statistic that a significant percentage of our registrations are incomplete or incorrect (i.e., lacking employer information or including an incorrect date of birth). It has become clear in working with our physicians that there are inconsistencies in how clinical information is captured in the ambulatory EMR. Does “smoking cessation counseling” get entered as free text in the Plan section, or checked off on the Preventive Care template? Needless to say, the free text entry is very difficult to capture for reporting purposes. Natural language processing can’t come too soon.
To address these issues, and to build a culture that truly values data as a corporate asset, we have instituted a data governance program. The goals include: identifying sources of inaccurate data and implementing technical and process solutions, working with physicians (both at the group and individual levels) to standardize the input process in our clinical systems, and seeking a “single source of truth” for data collection and analysis. The use of Lean technology and other “standard work” tools is facilitating this work.
With any major change process, culture and behavior change is just as important as the technology, if not more so. Everyone in the organization must understand the value of accurate and timely data, and must develop ownership in the process. I’ll let you know how we’re doing in a future column.
–Futurist John Naisbitt
Does this ring true for your organization? Twitter alone generates more than 20 terabytes of data every day. Specific to healthcare, clinical and administrative information systems are generating terabytes of data that are begging to be used to measure and improve the care we provide across the continuum.
“The goal is to transform data into information, and information into insight,” said former Hewlett-Packard’s CEO Carly Fiorina. At LVHN, we are on the way to achieving that goal—implementing new data warehouse and business intelligence (BI) tools. In a best-of-breed environment, this is quite a challenge. In addition to creating the new warehouse and mapping the data elements, we are employing state-of-the-art normalization tools to ensure, for example, that a Hgb from one system is matched up with a Hb from another system; and an enterprise master patient index to ensure that the data regarding Don Levick in one system are matched up with the data on Donald Levick from another system. The leading-edge BI tool will allow clinicians to perform ad hoc reporting and dashboard creation, both in retrospective mode for population health management and at the point of care.
The biggest hurdles we have identified to achieve this very cool vision are two-fold: ensuring data are entered into the system accurately and that the information is entered in the correct place. The new data profiling tools have uncovered the unsettling statistic that a significant percentage of our registrations are incomplete or incorrect (i.e., lacking employer information or including an incorrect date of birth). It has become clear in working with our physicians that there are inconsistencies in how clinical information is captured in the ambulatory EMR. Does “smoking cessation counseling” get entered as free text in the Plan section, or checked off on the Preventive Care template? Needless to say, the free text entry is very difficult to capture for reporting purposes. Natural language processing can’t come too soon.
To address these issues, and to build a culture that truly values data as a corporate asset, we have instituted a data governance program. The goals include: identifying sources of inaccurate data and implementing technical and process solutions, working with physicians (both at the group and individual levels) to standardize the input process in our clinical systems, and seeking a “single source of truth” for data collection and analysis. The use of Lean technology and other “standard work” tools is facilitating this work.
With any major change process, culture and behavior change is just as important as the technology, if not more so. Everyone in the organization must understand the value of accurate and timely data, and must develop ownership in the process. I’ll let you know how we’re doing in a future column.