Article Type
Changed
Fri, 09/14/2018 - 12:16
Display Headline
Learn How Best To Avoid Some of Data Mining’s Potential Pitfalls

Ensuring data quality and equivalency can present major challenges in data analytics, especially given the field’s dearth of uniform standards.

“The joke is that the great thing about health-care data standards is that there’s so many to choose from,” says Brett Davis, general manager of Deloitte Health Informatics. If data integration remains a big challenge, however, Davis says the cost and complexity of the technology is dropping rapidly.

A lack of electronic health records (EHR) can limit more advanced data-mining functions. But that’s no excuse for not exploring the technology, says Steven Deitelzweig, MD, SFHM, system chairman for hospital medicine at Ochsner Health System in New Orleans and chair of SHM’s Practice Analysis Committee.

Deployment of that partial prerequisite also seems to be happening quickly around the country. The Office of the National Coordinator for Health IT (ONC) estimates that hospital adoption of at least a basic EHR system more than tripled between 2009 and 2012, to 44% from 12%. Meanwhile, an estimated 85% of hospitals were at least in possession of certified EHR technology by 2012.

Despite the falling barriers, Davis cautions that users should have clear goals in mind when setting up a new system. “There is the risk of building bridges to nowhere, where you just integrate data for the sake of integrating data but not knowing what questions and insights you want to glean from it,” he says.

ONC spokesman Peter Ashkenaz agrees, citing governance within a hospital or health center and education of all participants as important elements of any data-analytics plan. Among the questions that must be addressed, he says, are these: “Have we collected the right information? Are we doing so efficiently and securely with respect to privacy requirements? Are we sharing the data with the appropriate parties? Are we doing so in a way that is easily understood? Are we asking the right questions about how to use the information?”

The most fundamental question, Dr. Deitelzweig says, may be whether a hospitalist group, hospital, or health system is truly committed to using the technology. “If you’re going to make the investment in such things, then you really better be dedicated to understanding them and how best to utilize them. And give it some time,” he says. “I think people want solutions fast, and often they don’t take the time to individualize it or customize it.” TH

Bryn Nelson is a freelance medical writer in Seattle.

Issue
The Hospitalist - 2013(10)
Publications
Sections

Ensuring data quality and equivalency can present major challenges in data analytics, especially given the field’s dearth of uniform standards.

“The joke is that the great thing about health-care data standards is that there’s so many to choose from,” says Brett Davis, general manager of Deloitte Health Informatics. If data integration remains a big challenge, however, Davis says the cost and complexity of the technology is dropping rapidly.

A lack of electronic health records (EHR) can limit more advanced data-mining functions. But that’s no excuse for not exploring the technology, says Steven Deitelzweig, MD, SFHM, system chairman for hospital medicine at Ochsner Health System in New Orleans and chair of SHM’s Practice Analysis Committee.

Deployment of that partial prerequisite also seems to be happening quickly around the country. The Office of the National Coordinator for Health IT (ONC) estimates that hospital adoption of at least a basic EHR system more than tripled between 2009 and 2012, to 44% from 12%. Meanwhile, an estimated 85% of hospitals were at least in possession of certified EHR technology by 2012.

Despite the falling barriers, Davis cautions that users should have clear goals in mind when setting up a new system. “There is the risk of building bridges to nowhere, where you just integrate data for the sake of integrating data but not knowing what questions and insights you want to glean from it,” he says.

ONC spokesman Peter Ashkenaz agrees, citing governance within a hospital or health center and education of all participants as important elements of any data-analytics plan. Among the questions that must be addressed, he says, are these: “Have we collected the right information? Are we doing so efficiently and securely with respect to privacy requirements? Are we sharing the data with the appropriate parties? Are we doing so in a way that is easily understood? Are we asking the right questions about how to use the information?”

The most fundamental question, Dr. Deitelzweig says, may be whether a hospitalist group, hospital, or health system is truly committed to using the technology. “If you’re going to make the investment in such things, then you really better be dedicated to understanding them and how best to utilize them. And give it some time,” he says. “I think people want solutions fast, and often they don’t take the time to individualize it or customize it.” TH

Bryn Nelson is a freelance medical writer in Seattle.

Ensuring data quality and equivalency can present major challenges in data analytics, especially given the field’s dearth of uniform standards.

“The joke is that the great thing about health-care data standards is that there’s so many to choose from,” says Brett Davis, general manager of Deloitte Health Informatics. If data integration remains a big challenge, however, Davis says the cost and complexity of the technology is dropping rapidly.

A lack of electronic health records (EHR) can limit more advanced data-mining functions. But that’s no excuse for not exploring the technology, says Steven Deitelzweig, MD, SFHM, system chairman for hospital medicine at Ochsner Health System in New Orleans and chair of SHM’s Practice Analysis Committee.

Deployment of that partial prerequisite also seems to be happening quickly around the country. The Office of the National Coordinator for Health IT (ONC) estimates that hospital adoption of at least a basic EHR system more than tripled between 2009 and 2012, to 44% from 12%. Meanwhile, an estimated 85% of hospitals were at least in possession of certified EHR technology by 2012.

Despite the falling barriers, Davis cautions that users should have clear goals in mind when setting up a new system. “There is the risk of building bridges to nowhere, where you just integrate data for the sake of integrating data but not knowing what questions and insights you want to glean from it,” he says.

ONC spokesman Peter Ashkenaz agrees, citing governance within a hospital or health center and education of all participants as important elements of any data-analytics plan. Among the questions that must be addressed, he says, are these: “Have we collected the right information? Are we doing so efficiently and securely with respect to privacy requirements? Are we sharing the data with the appropriate parties? Are we doing so in a way that is easily understood? Are we asking the right questions about how to use the information?”

The most fundamental question, Dr. Deitelzweig says, may be whether a hospitalist group, hospital, or health system is truly committed to using the technology. “If you’re going to make the investment in such things, then you really better be dedicated to understanding them and how best to utilize them. And give it some time,” he says. “I think people want solutions fast, and often they don’t take the time to individualize it or customize it.” TH

Bryn Nelson is a freelance medical writer in Seattle.

Issue
The Hospitalist - 2013(10)
Issue
The Hospitalist - 2013(10)
Publications
Publications
Article Type
Display Headline
Learn How Best To Avoid Some of Data Mining’s Potential Pitfalls
Display Headline
Learn How Best To Avoid Some of Data Mining’s Potential Pitfalls
Sections
Disallow All Ads
Content Gating
No Gating (article Unlocked/Free)