When it comes to big data, the majority of CIOs say their organizations aren’t there yet — and that vendors are overpromising on what they can deliver. According to the June healthsystemCIO.com Snap Survey, just 24 percent of CIOs say their organizations are doing high-level, predictive analytics, while about half (52 percent) are performing some analytics, but not at a sophisticated level.
It’s a far cry from the picture some vendors are trying to paint, according to the survey, which found that 76 percent of CIOs believe vendors are selling a concept that isn’t fully baked. “Vendors know it’s the key word, so they are selling it and hoping to find the right customer to help them develop what customers are really wanting,” one CIO noted.
Although there are a number of obstacles to achieving big data, from high costs to a lack of standards, gaining organizational buy-in was identified as the biggest hurdle. “Many view it as an added expense for another reporting system that might not be widely used,” one CIO noted.
And beyond selling organizations leaders on the investment, it’s also about helping them grasp what big data entails, and what it will require — from both an IT and a staffing standpoint. “It’s not buy-in so much as understanding,” said one CIO. “Getting managers to reevaluate their existing workflows and expectations by asking what-if questions of the data is not a common skillset.” Another noted that although his organization supports the initiative, “I don’t think they understand how to leverage the tools they keep telling me to evaluate/buy/build. They want analytics and understand their power, but have trouble figuring how to get there.”
In terms of staffing, two-thirds of CIOs say their organization lacks the manpower and skillset needed to achieve a high level of analytics. “We have some great report writers, but we struggle to find the right people to understand and analyze the results,” said one respondent.
(SnapSurveys are answered by the healthsystemCIO.com CIO Advisory Panel. To see a full-size version of all charts, click here. To go directly to a full-size version of any individual chart, click on that chart.)
1. What is your organization’s level of engagement when it comes to business intelligence/analytics?
Doing high-level, predictive analytics
- We’re probably somewhere between doing some and doing “high-level.”
Doing some analytics, but not at a sophisticated level
- Most of our analytics are around costs.
- This has been a struggle for our organization.
- We have formed a committee to evaluate current status of data analytics and develop a plan.
- It will take another year to get to high-level.
We’re in the planning phase
- We wasted two years trying to convince the organization we need to do this and now they finally are going to move on it.
We have no plans at this time
2. From a staffing perspective, does your organization have the manpower and skillsets needed to achieve a high level of analytics?
- But we’re only in our infancy on the skill set. I believe we are well positioned, but it is an ongoing challenge.
- Once the EMR build levels off, we’ll have the manpower to do analytics.
- We have staffed up over the past 18 months.
- This is a huge challenge. We have some great report writers, but we struggle to find the right people to understand and analyze the results.
- We are missing the skillset.
- No, not enough for BI or data governance at all.
- We do have a dedicated staff but are short in support for clinical analytics in particular.
I’m not sure
- We have a great core staff that requires augmentation to focus on higher analytics.
3. If not, how do you plan to address this?
Hiring more staff and/or investing in training
I don’t know
- We have not defined our analytics needs yet.
- Our CFO was all for the project until he got the price tag. So we do not know what the approach will be yet.
- We’re making a plan. We’re working to show the value, then will consider staffing.
- Using both current staff and consultants.
- All of the above being considered.
- We’re making do. But we really need more focus on this. Pressure to report ambulatory quality metrics (ACO, ARRA, PQRS, HEDIS, payers, etc.) will force this issue in the next year or so.
- We’ll use a combination of repurposing existing staff and hiring consultants.
4. What is the biggest obstacle to achieving a high level of analytics?
Lack of data quality (patient matching, etc)
- It is really about the data; the back-end stuff is really not the barrier. If you think about the variation in clinical documentation and the lack of structure (never mind predictable structure), we have a long way to go.
Lack of industry-wide data standards
Getting organizational buy-in
- Getting folks to understand how the data “can” be used and getting them to “trust” the data are the biggest issues.
- It’s not buy-in so much as understanding. Getting managers to reevaluate their existing workflows and expectations by asking what-if questions of the data is not a common skill set.
- Many view it as an added expense for another reporting system that might not be widely used.
- Without the right understanding of what it will take to meet the expectations, the project will fail. And I expect it to do so here.
High cost of resources
All of the above
- These resources aren’t cheap, and the data lacks integrity because it isn’t really adequately monitored on the clinical side.
- The data can’t analyze itself. Institutionally, we don’t know what is important to analyze and we don’t have the discipline to act on the analytics.
- We need to have a global, strategic view of enterprise-wide analytics. Currently everything is in silos.
- I think the organization is bought in, but I don’t think they understand how to leverage the tools they keep telling me to evaluate/buy/build. They want analytics and understand their power, but have trouble figuring how to get there.
- It’s a journey. Some hospitals, like mine, are still focused on getting to Stage 1 Meaningful Use. You have to get the building blocks in place before you can have a data warehouse to even do analytics. But we are moving in that direction and plan to be “there” by the time we’re at Stage 2.
5. Are vendors overpromising when it comes to what they can deliver to help your organization achieve ‘big data’?
- To be honest, I’m not sure either way, but checked yes because they always have great use cases, but don’t take into consideration the internal processes and culture. It’s like Microsoft — they sell you Office with all the bells and expect you to seek out the training and expertise to use it to its fullest extent. And one size always fits all.
- I believe that vendors know it’s the key word, so they are selling it and hoping to find the right customer to help them develop what customers are really wanting.
- They cannot ensure the integrity of the data because there simply are not adequate automatic monitoring tools at the time of input on the clinical data. Financial data is a lot more standard, but clinical is all over the map.
- Lots of promise with little that is deliverable.
- This market will not mature for another couple of years.
- It’s not about the back-end.
- No, I don’t think they are. However, I don’t believe they understand how behind the curve we as an industry are in our understanding what BI is and how to do it.