From Clinfowiki
Jump to: navigation, search

With the advent of the Health Information Technology for Economic and Clinical Health (HITECH) Act, healthcare organizations are creating and collecting electronic data at an ever-increasing rate. This data can help identify areas of improvement, find solutions, incentives to change behaviors, and then monitor the effect of these changes.

Healthcare Data Collection, Aggregation, and Reporting in the Era of HIT

Historically, the only consistent source of data was generated from the billing and claims process of reimbursing the provider and the system. This provided information on the end point of care, procedures, and some areas of resource utilization, however, it provided little insight into the workflow process that make up the steps in the delivery of care.

With claims data, the payors are able to identify variance in costs, but have never been able to drill down and identify the relevant steps to decrease costs and improve the care delivery system. As a result, many change initiatives have been founded upon educated guesses as to what was causing increased claims and attempts to curb cost were simply aimed at the end point and were not strategically targeted.

As electronic data began to amass, it produced historically uninvestigated stores of clinically relevant data. Much of this data was available before HITECH and widespread adoption of electronic medical records (EMR), however, to get at the paper chart data and aggregate it into a useful purpose was extremely time consuming and costly. There was some new data that was generated, for instance EMRs provided time stamps and intervals of care data that was often not available, but the majority of the data was always there; it was just nearly impossible to get and even more difficult to utilize.

While the first few years of this era of health IT has been primarily focused on EMR adoption and meeting the first stage of Meaningful Use, the realized potential of the secondary gains of this data is beginning to come to the forefront of healthcare. Accountable care organizations and proposed quality based reimbursement models have also furthered the interest of payors, policy makers, administrators, employers, and providers in these secondary gains.

Definition Problems

As with many new fields, making decisions in this new area is complicated by poorly defined terms and concepts. While much of this is expected in a nascent field, there are certainly some vendors that are exploiting the lack of universally accepted and understood definitions to support what their technologies are capable of achieving. For instance, many of these new vendor solutions are aimed at helping organizations with the development of their accountable care organizations (ACO). The problem is that at the beginning of 2011, there are very few operational ACOs and most in existence have varied structures and goals.

Some that are now considered an ACO were in existence before the term was even coined and in fact the most widely agreed up definition of an ACO did not surface until the Federal government released its proposed rules for ACOs in the spring of 2011. This was long after many companies were marketing comprehensive “ACO” solutions.

Another term that is often misrepresented in this field is the concept of a Health Information Exchange (HIE). While there are set requirements and definitions under the government’s criteria for HIEs, many vendors exploit the lack of familiarity with these rules and represent simple interfaces between disparate systems as a fully functional HIE.

Data warehouses, data marts, psychographic profiles, attribution and many more poorly understood terms are now becoming increasingly used to represent often different concepts.

Categories of Data

When availability of new data presented itself and new uses began to emerge, there were also new descriptions of the categories of data that evolved. While the classification of the data is often at the discretion of those using it, there are several broad categories that are becoming relevant to healthcare costs, quality, and operations. As noted, claims data has been the most utilized repository of data, but now that EMRs are being adopted there is a new broadly defined area considered “clinical data.”

This type can be functionally broken down into provider data and patient data. In short, this can provide information about the patient’s care and provide details about the provider’s behaviors in caring for these patients. There is also another category of data that has been even used even less in redesigning healthcare; this can be considered “financial data.”

Loosely this relates to the cost of operating the hospital, supply costs, personnel costs, consumer costs, etc. Most hospitals have been collecting this information in their accounting and finance offices, but have never been able to directly link this information to the other data generating processes of a healthcare organization.

Population data is another category that encompasses many of the data elements of the above categories, but is filtered to a defined set of individuals.

An example of this would be hospital employee data. Although, organizations are in the business of providing care, they often are the largest employer in their area and therefore the cost of providing care to their employee population is as important as the revenue generated from providing care to others. This category of population data is rapidly increasing in its importance to individual organizations and beyond.

Data Storage

Even though masses of useful electronic data are being captured, much of this information is siloed in disparate systems or behind proprietary code. Ironically, much of the same revenue generating technology that allows for collection of electronic data serves as a major impediment to the actual aggregation of data. Further complicating data integration, is that the organizational and administrative structures of many large organizations have created redundant, yet separate systems, processes, and personnel that form a political barrier for aggregation.

As discussed above, the finance and accounting department of most organizations has been collecting and analyzing data for years around financial activity, revenue, and costs.

However, convincing them to share this data can be quite challenging and can generate significant intra-facility tension and turmoil.

Those in healthcare that do see the benefit in consolidating data are then faced with the enormous burden of defining a useful architecture for the data so that it could be functionally harvested for future purposes. Many systems not wanting to invest in data warehouses focused on developing interfaces that allows for transmission of data as it is requested, but does not always allow large scale aggregation and reporting that is needed for actions such as, process improvement and population health analysis.

Collection, Aggregation, and Reporting Vendors

If the vendor market for health IT is a newborn industry, then the niche industry focused on collection, aggregation, and reporting electronic health IT data is at the stage of a morula. Dozens if not hundreds of companies are developing their own twist on the solutions for achieving the secondary gains that providers and organizations crave.

Many large IT companies have merged onto the scene with their history and experience of collecting data that was not developed from within the healthcare industry.

Other solutions were developed from within large EMR vendor companies to fill this growing need, while still others were small start-up companies who’s sole business objective was to pull together disparate data, analyze it, and report the results. Some company’s entire business model is generated around the concept of driving behavior change through the powerful presentation of quality and efficiency metric data.

In analyzing these varied backgrounds, there appears to be three rough categories of vendors entering the market. The first is the established health IT company that has been supplying solutions prior to the development of these tools. Whether it be a specific EMR vendor or simply a health IT company that previously focused on a data niche, the experience managing and manipulating data appears to entice these types of organizations to develop solutions in this new realm. The second and quite prevalent type is that of the start up niche vendor. Many of these appear to have developed from individuals with a data background and a connection to homegrown organizational solution. Some of these started in an office or hospital IT department and then broke off to start up their own companies. The last broad classification would be that of the crossover data company. For example, there are some companies that have been dealing with non-healthcare related data for years and have now entered this arena attempting to capitalize on their experience in data management and analysis. A data based marketing company would be an example of a vendor of this type.


Another issue with this field is that it is very difficult to understand what many of the vendors are actually offering. They will note rhetoric like “we will provide your organization the solution to drive changes in physician behavior to improve quality and reduce cost.” In the end, their product is only a reporting tool that requires you to provide them the collected, aggregated, and validated data so that they can put it into a visually appealing dashboard. In fact, several products in the market truly are no more than Excel charts on steroids.

There are also varied philosophical vendor approaches to presenting the data. Some are focused on the administrative needs of quality reporting and hone in on the requirements of organizations to oversee the compliance of quality metrics through detailed and automated reports. However, other vendors neglect this need of the organization and concentrate their product on physician’s desires to see their own personal data. In reality, both ends of the spectrum will be needed to promote acceptance and buy-in of the metrics, while providing the oversight reports that the payors and employers require.


One of the positive outcomes of successful physician EMR adoption is that providers are now taking ownership of their data and looking for ways that they can use the data that they generate to improve their performance and even to identify areas of cost savings. Physicians are quite competitive and will seek out personal performance improvement, whether it is in quality or efficiency, when supplied with data that they trust. For this reason, more and more physicians are getting involved in areas that once were solely dominated by those with IT backgrounds. In fact, one of the founding theories of ACOs is that they be physician led and directed. While this likely will provide unprecedented movement and acceptance in quality initiatives, it also thrusts providers into decision trees where they have had little experience. For example, many physician leaders of the recent quality movements and ACOs have had little training in the formal process of vendor evaluation and selection through the Request for Information and Request For Proposal cycles. Combine this with physicians’ desire to make their own decisions and their overwhelming hatred for delays and there is a significant risk for hasty, poorly vetted decisions.

Future Possibilities

This new field of data collection, aggregation, and reporting in health IT holds tremendous potential for improving efficiency and quality, but can also reveal previously unrealized rewards for providers and for organizations. Providers seem to forget all of the struggles of EMR adoption when they realize the potential for personal improvement through these tools. Conversely, they are also quite skeptical of all of this data if they have not had the opportunity to validate it on some scale. This is especially true when data is being used for metrics to determine reimbursement.

Organizations are realizing that the aggregation of this data with existing stores allows for planning and strategies that were never considered possible before these tools existed. For instance, there are some companies that now combining provider data to generate psychographic profiles of physicians to help understand how a physician thinks and behaves. They can further utilize this information to identify practice patterns of who is practicing at just one hospital or is splitting their time between competitors. Referral patterns, resource utilization, and identifying high value physicians all become a matter of clicking a button once these tools are reliably in place.

While these tools and vendors will certainly spread throughout the healthcare industry, the landscape will change, as smaller companies will be acquired for their successes, while others will be simply driven out by competition. The next frontier in the development of these solutions will be in the form of predictive modeling. As companies begin to collect massive amounts of reference data, they will then be able to utilize an organization’s data not only for regional or national benchmarking, but will be able to predict behaviors and trends in your organization.

There are now several companies that will run predictive modeling formulas to highlight the patients in your population that are likely to get sick in a selected time period. This has huge potential for care management, but also targeted marketing. Instead of mass mailers and billboards, organizations and physicians will market directly to those that are likely to need their services. Some have even developed philanthropic models from the aggregation of massive data stores to help organizations identify and focus on those that are likely to donate to their foundations.


In conclusion, this industry has the potential to revolutionize the power of the electronic data the HITECH era has spawned. It is a very immature field with a definition problem and a number of barriers to overcome. However, it will undoubtedly continue to grow into areas that once were considered impossible. Simply aggregating consumer, patient, and provider data holds the potential to revolutionize how our healthcare organizations approach marketing, quality, incentives, and many more areas of operation.

Tripp Jennings, MD, FACEP