Measuring Quality Improvement in Healthcare

At one point or another, all of us become uses of healthcare, whether it is a routines check-up, on slot from illness or injury. Without much thought or research, it is very easy to understand and explain the quality of care that you received; medical professionals assessed your situation, medicine or medical services are prescribed, and follow-up care is explained. We can see the results, sometimes immediate; other instances require time for internal healing. In worst-case scenarios, quality is measured by helping life to end with less distress.

photo courtesy of imagerymajestic/
photo courtesy of imagerymajestic/

The question then arises: Is quality in healthcare a black or white picture? The answer is unequivocally no. There are many straightforward and easily identifiable situations where quality in care is obvious. When recuperation is recognized or health is restored, and the patient is satisfied with the road to recovery or end result, quality can be quantified. However, is all care linked only with renewal and contentment?

From the point of view of those working within the healthcare system, there are many other facets that attribute to the improvement of quality and the ability to measure such advances. Without droning through a list of all possible departments and sectors along with the areas where improvements are sought, a more overall structure has been defined that can carry across many divisions. Three fields to measurement are:

  • Structural (equipment, facilities, etc.)
  • Processes (how the system functions)
  • Outcome (final product or result)

As helpful as this information can be, these are only the things to measure, but not the steps to take in order to produce results. Duke University Medical Center has formed a very organized illustration called the FADE Model of quality improvement. This outline identifies the steps that need to taken and how to explore the most effectual processes to improve standards. The FADE Model includes:

  • Focus – define/verify what needs to be improved
  • Analyze – collect and analyze data to find the root cause of problems and possible solutions
  • Develop – use analyzed data to develop a plan to:
    • improve
    • implement
    • communicate
    • measure or monitor
  • Execute – implement the development stage
  • Evaluate – install monitoring processes to validate improvements

Understanding where improvements need to be made and analyzing data are both very complicated steps. Most any business would consider a reduction in costs and errors to be stating the obvious. However, finding the inefficiencies may not be as obvious. Outside and unbiased recognition comes in the form of dedicated software, which eliminates guess work by taking collected data from as many sources as possible to produce a complete picture of procedures and events.

image courtesy of cooldesign/
image courtesy of cooldesign/

This picture defines the areas in need of improvement, and may also provide the answers to questions of root-causes. As much as we might like to believe that problems should be apparent, the truth is some improvements are complex with several outlying factors that contribute to the performance. With analytical data in hand, more quality information is present and the ability to attain better results is achievable.

Why We Need Healthcare Data Analysts

There will always be careers in health care for persons with a desire to help and care for others. Aside from abilities relative to basic patient care, anyone beginning a health care career now and in the future will need more than basic computer skills. The increasing use of technology in health data management adds a new dimension to the skill set required just a few decades ago.

Healthcare Data Analysts
photo courtesy of imagerymajestic/

The next new high demand specialty in the health care field may be the Healthcare Data Analyst. As more information about patient care is entered into digital databases, more complex processes will be needed to retrieve the information and use it in ways to improve hospital efficiency and patient care.

Knowing how to query the database to pull out the relevant information and use it effectively will require skill with statistical analysis. This is in addition to knowing what relevant information to look for.

There are a number of software platforms available for health care institutions to choose from for their clinical data storage and retrieval. Designers of those systems are often physicians themselves, or have strong ties to the medical field. They often have a problem to solve, and data management is the fist step in finding a solution.

Armed with a repository full of data, hospital decision makers can direct resources to specific health treatments and programs that will have the greatest impact on their patient population.

A clinical analytics application with completely integrated data from a large health care enterprise can identify and reduce gaps in patient care and resource management. By “aggregating data from the health system’s Electronic Health Record (EHR) and other critical IT systems,” a skilled healthcare data analyst can tailor a query to evaluate readmission rates for instance. Then she can work with physicians and hospital staff to develop a strategy to lower the numbers. Along with saving the hospital valuable resources, insurance providers and the patients will also see savings.

Saving Money with Analytics
photo courtesy of Baitong333/

Development of software that allows the patient to take a more active role in his own health care is gaining in popularity. An approach known as telemedicine is expected to save the health care industry and its clients up to six billion dollars per year.

With telemedicine, patients and physicians can communicate via telephone or video and can often resolve issues without needing an office visit. In addition, nurses and other hospital staff are often able to answer questions that don’t involve diagnosing conditions or prescribing medications.

Research is an area where electronic health records can have a tremendous impact. Studies can be designed to analyze disease trends within the hospital’s own patient population or on a global scale. There can be problems, though, when clinicians ask information technology workers for sets of data.

Health Care Data Analysts
photo courtesy of hin255/

Applying clinical analytics to disordered data sets may be more time consuming than a busy physician would like to deal with. IT workers may not understand all the ramifications of the data they generate in response to requests from researchers. Healthcare data analysts can bridge many of these communications gaps by having a foot in each world.

Healthcare has a compelling need to use more information, better. Software engineers who can help medical enterprises achieve that goal will not lack for work for the foreseeable future.


Healthcare Data in Action

When you are making a big decision in your life, you usually don’t turn to the Magic 8 Ball or local newspaper’s horoscope predictions.  More often than not, research, other people’s experiences, and common available knowledge are gathered together and weighed against doubts or other insights.  This is not much different than the information that comes out when visiting a doctor or hospital.  The collected data and experience from years of research, trial and error, and patient’s encounters are weighed against current signs and symptoms.  This total gathering of healthcare data makes it possible to receive more precise treatment on an individual basis, eliminate unnecessary testing and wait times, and also prevent future diseases and outbreaks.

Healthcare data saves lives
photo courtesy of Sura Nualpradid/

In this day and age, we are not satisfied with simply taking a pill for everything that ails us; we want and expect that cures are being pursued, and that our communities and nation are being protected from epidemics.  These leaps in expectations are warranted, however don’t come without years of exploration into healthcare data.  Found within all that data that has been generated, and is still being generated, are patterns that contain answers.

However, there are billions and billions of bytes of data already created, with more being added each moment.  In fact, the healthcare industry is one of the top producers of data each year, and thus serves to benefit the most from all that stockpiled information.  The thought that someone or a group of people could sit down and begin to decipher any significant patterns is impossible.  That is why the business of healthcare has turned to big data solutions.

When comparing all the information manufactured in the realm of healthcare, not many industries stack up to the complexity and privacy issues.  Not only do treating doctors and nurses need access, but billing departments and insurance representative and even government offices need some of the information stored within the data warehouses.  Thus the need for efficiency and effectual extraction of information is in high demand.

The gathering and storing of such significant and immense amount of information is not a project that is tackled with something as straightforward and simple as an Excel spreadsheet.  The need to have a model and progressive goals to eliminate errors, produce more meaningful outputs and ultimately to reduce expenses for everyone involved is crucial.  This healthcare analytics adoption model lays out the steps that lead down a path of better utilization of the collected healthcare data.  As the steps are perfected, and as technology also provides further aid to analyze stored information, goals for producing advances in treatment and reduction of costs will truly be met.

Data collection in healthcare
photo courtesy of stockimages/

As much as I am glad that my doctor doesn’t turn to a Magic 8 Ball to determine the best course of action with my healthcare, I see the benefits that combining the processing of such large amounts of health data and the analytical tools to understand and find significant information held within can open the doors to more cures, better care and an overall superior knowledge to such a vast and necessary industry.  And, having a plan of action will make it possible to reach the highest potential.

8 Levels of the Healthcare Analytics Adoption Model

Accountable Care Organizations (ACO’s) across the country are discovering every day just how useful and effective healthcare analytics can be. In today’s marketplace, data management is becoming an integral part of providing quality service to the general public.

According to an article published by Health Catalyst, there are 8 levels of analytics adoption in which an ACO can find itself. As you read through this summary, see if you can figure out how your own personal provider is doing in adopting this technology into their practice.

Level 0: Little to no analytics integration

What is an ACO?
(photo courtesy of photostock/

- This provider has a self-contained information system that doesn’t share with or draw from a larger pool of data
- Inconsistent and incomplete data sets cause confusion and wasted time and resources
- Reporting is labor-intensive and less helpful
- There is virtually no data governance

Level 1: Integrating the Enterprise Data Warehouse (EDW)
- This provider has begun to use a data warehouse to collect data on HIMSS EMR Stage 3 data, the revenue cycle, financial information, supply chains, and the patient experience
- There is a searchable metadata repository that can be accessed across the enterprise
- The EDW is updated within one month of source system changes
- The CIO of the organization receives reports from the EDW

Level 2: Standardized Vocabulary & Patient Registries
- Various, disparate source system content is identified and standardized in the EDW
- Local standards define the naming, definition and types of data
- Patient registries are defined by ICD billing data
- The governance of data forms around the definition and evolution of patient registries

Level 3: Automated Internal Reporting for consistent and efficient production
- Analytics are consistent and efficient, and are focused on supporting basic management and operation of the ACO
- KPI’s are accessible from the executive level to front-line management
- Data analysts are regularly involved in order to collaborate and guide the EDW
- The governance of data is being managed in a way to steer the ACO towards constant improvement

Level 4: External reporting integration
- Analytics are focused on consistency, efficient reporting and accreditation requirements, payer incentives and specialty society databases
- Industry-standard vocabularies are adhered to
- Clinical data content is searchable with simple key word targeting
- Data governance is centralized  for review and approval of externally released data

Level 5: Reduction of waste and care variability
- Analytics are focused on measuring clinical best practices, waste reduction, and eliminating variability
- Data is used to support teams focusing on improving the health of patient populations, not just individuals
- Population-based analytics are implemented in order to improve individual-level care
- Teams are in place that continuously and regularly look for ways to improve care quality while reducing risks and costs
- Data is improved by incorporating information from labs, pharmacies and clinical observations
- Content from the EDW is organized into evidence-based and standardized data streams that simultaneously manage clinical and cost data associated with patient registries
- Insurance claims and HIE data feeds are included in data content
- EDW updates happen within one week of source system changes

health analytics and data management
(photo courtesy of Ambro/

Level 6: Population health management
- The ACO shares in the financial risks as well as the rewards tied to clinical outcomes
- Acute care cases are managed under bundled payments at a level of 50% or higher
- Data content is expanded to incorporate bedside devices, home monitoring data, external pharmacies and activity based costs
- Data governance is factored into the metrics of quality-based compensation plans for clinicians and executives
- EDW updates happen within one day of source system changes
- C-level executives accountable for balancing cost of care with quality of care receive EDW reports

Level 7: Risk prevention and predictive analysis
- Diagnosis-based, fixed-fee per capita reimbursement models are addressed by the analytic motives
- Predictive modeling, forecasting and risk stratification are used to support outreach, triage, escalation and referrals, expanding focus from management of cases to collaboration with clinician and payer partners for management of episodes of care
- Risks and rewards are shared through collaboration between physicians, hospitals, employers, payers, and members/patients
- Patients who are unwilling or unable to participate in care protocols are flagged in the registries
- Data is expanded to incorporate home monitoring data, long term facility care data, and protocol-specific patient reported outcomes
- EDW updates occur within one hour of source system changes

Level 8: Personalized medicine & prescriptive analysis
- Data management is used to influence wellness management, physical and behavioral health and mass customization of care
- Analytics are expanded to include NLP of text, prescriptive analytics, and intervention support
- Patient-specific outcomes are improved at the point of care, based upon population outcomes
- Data content is expanded to incorporate biometric data, genomic data and familial data
- Updates in the EDW occur within a few minutes of changes in the source systems

What type of provider would you prefer? Healthcare analytics can transform our current model into a data-driven process, which serves the population as a whole as well as individuals. As more and more ACO’s integrate the EDW into their processes, the better the service will be across the spectrum of providers. As consumers, we can aid in the process by learning about the process and participating in the discussion.


Application of Business Healthcare Intelligence

Business intelligence (BI) is a general term used across many different industries of businesses that helps to explain utilization of stored data.  A more specific definition is the extracting, identifying and analyzing of computerized business data to provide current and predictive view of operations.

When integrating BI into the demands of healthcare the possibilities are almost limitless.  However, the need for a data warehouse coupled with BI is integral to uncovering the potential of the stored data.  How these work together is that the data warehouse is the efficient and effective way to store the collected information, which can then be accessed by all the necessary personnel.

The BI tools look at all that data and are able to dive down to find problem or red-flagged areas that should be looked over. Imagine sitting in a very important financial meeting where everyone has their own statistical data to backup their opinion for changes.  One department shows there is a surplus and demands more spending power, but according to your numbers, that same department is showing a significant loss and should be cutting back severely.  So, who is right?  Both parties have evidence that supports their point of view, and both consider the other to be incorrect.

Situations like this happen all too often because two or more groups are extracting their numbers differently, possibly from different sources, and basing decisions with far reaching consequences on that data.  If everyone was on the same page, with the same numbers, the same conclusions and all of it being accurate, more appropriate and effectual decisions could be made.

This is the importance of the data warehouse: everyone is pulling from the same information, and interpreting it the same way.  Thus better assessments can be made with everyone in agreement. Another typical situation that comes up can be that everyone sees that a department is losing money, everyone has settled on that fact, but the no one can explain how, where or why it is happening.  Sort of like looking at an Excel worksheet with all the numbers plugged in, the totals summed and yet no better understand of what is going on than before you looked at it.

How can you make changes if you don’t know where to start?  What if the changes you believe you should make don’t aid in the health of the company? All the stored data is in one location, the necessary people can access it to help draw current and future decisions, but what does this do if there is no analysis tool that makes meaningful use and presentable layouts that let you know what all the numbers really mean?

BI might be divided up just a little further to seen for the ultimate usability and also the end strategy that you want to meet.  You can have all the graphs, reports and summaries you can handle, but if you don’t know what your end-game is, than you are left with a very expensive data system. Obviously, more often than not, cutting costs and inefficiencies is at the top of everyone’s list.  There is nothing wrong with that in the field of healthcare, and in most cases, this would help with patient safety, treatment and overall cost.

To be more specific though: what if you wanted to cut down on the number of habitual patients seeking pain medicine.  This would mean understanding the fundamentals that go into a patient coming to a hospital or clinic on a regular basis.  This isn’t found within the vast array of numbers, but is found within notations and other information that can be housed in a data warehouse.  By having a strategy, you can see where you are headed and why. There are other pitfalls to be aware of that can bring down any giant of the business world.

In one word, let me say Segway.  Just because you build it, you hype it, you cover it in mystery and you launch to a national audience doesn’t mean that everyone will flock to buy it.  The same goes with business intelligence: just because you install it, instruct everyone and toot your own horn doesn’t mean everybody will support the decision or utilize it to the fullest potential.  Change is required and you need to understand that all your problems won’t be solved just because you instituted new software.

Additionally, no matter how much of a perfect fit BI software may be for you, if data quality isn’t an enforced concept for everyone involved, the old adage, “GIGO (garbage in, garbage out)” will always ring true.  A simple example of this could be if a health professional consistently entered a wrong spelling for medicine administered to patients, the information wouldn’t be pulled out for statistical reporting or other pertinent accounting.  This simple action could lead to eventually falling behind in having that medicine purchased or slightly skewing expenditures.  Without consistent checks to make sure that all information being entered is reliable, the conclusions made could be biased in one direction or another, but never completely accurate.

Healthcare business intelligence may be of the most complex of any known industry, but also may yield more long-lasting answers no matter your part in the healthcare system.  However, be ready to know the inner workings of the business and the strategies that will help to accomplish whatever goals may be requested or required of you.

Is the Clinical Data Warehouse still relevant?

I was doing some reading on data warehouses when I came across this article. It’s an interesting read on what the Texas Children’s Hospital did to turn its data warehouse into a very valuable tool. This is a great analysis on what modern health care practitioners face in regards to technology, and what solutions might currently be available. Studies like this are one example of how population heath management will be leading the way to the future of health care.

Clinical Data Warehouse

Data warehousing, on the surface, may seem like a straightforward idea to grasp.  However, there are so many functions that must happen in the right order for a user trying to access information from that warehouse to receive what he or she wants.  In most cases, we all take having and getting access to data for granted; that it is something easy and we should just be able to type or click a few things, and voila, instant answers in the form of reports or electronic records.

WarehouseAll the work that goes into making easy for you and me to pull out information from some unknown location down the wire from our computers comes from the hard work that software programmers put into the organization, maintenance and then retrieval of all the inputted data.

I can almost hear the groaning happening right now.  Who can even understand when some IT guy (or gal) tries to explain the huge complexities that goes into their job, and, on top of that, includes all the jargon that only programmers understand, anyway?  Honestly, more than anything, it takes someone that understands the programming world and can explain it in more relevant terms to breach this gap.

Let’s imagine that you want to store a huge amount of information on all your clients, vendors and any other miscellaneous details that may need to be accessed later.  More specifically, you are a doctor within a group of other medical professionals. This data could be written down and stored on thousands upon thousands of papers, files and tabs that would have to be gone through one by one to extract the exact data you needed.  This obviously is not efficient and is prone to human error.  So, the next obvious place to store information is on a computer.  However, if everyone in an office is vying for information at the same time, that one computer is not sufficient.  Most of you are probably saying, “Network the computers!”

By setting up and having a network, many different computers can be utilized at one time, and the information is available to all.  But, what happens when the hub of the networking computers no longer has space for any more data to be entered?  It must be time to upgrade to something bigger, and depending upon your needs, this could be a larger hard drive for a computer or moving to server(s).  This allows for significant growth of data, and everyone still has access.  Access to what, though?  Do you have everything loaded onto Word or Excel sheets?  Or did you spring for software that allows production of reports, receipts and any statements that may be required?

Now, we are stepping into the big boy’s game of clinical data warehousing.  The warehousing portion of this refers to where massive amounts of information are stored for access by more than just a few people.  And, just in case you might be wondering, the data is not on Excel sheets but is carefully filed into categories within the software so that it can be easily distinguished, sorted and managed.

At this point, it could be very easy to throw IT lingo at you and expect you to keep up, but that is not my intent.  Data, especially clinical or medical data, is only helpful when you are able to use it for making decisions, extracting for specific purposes or finding patters and trends.  This is where the software you chose becomes essential.  But, not only the software as a name but as a functioning entity you can employ.  If you wanted a list of all your clients in alphabetical order with addresses and phone numbers, most software will function quite well.  On the other hand, what if you needed more dynamic information that was up-to-the-minute and contained client details that requires privacy protections.  That Excel sheet isn’t going to be of any help!

Because you now have made a serious investment for your medical practice with a software program that allows for you to keep on top of everything, you don’t necessarily want to spend more money every so often if there is are changes on your network or within your system.  Or, worse yet, what if there are new requirements for reporting that the corporate offices or government need?  Instead, what if you were able to just ask for new parameters to be added, changed or removed in order to comply with the higher-ups?   With software that processes the data with late-binding procedures, the flexibility that you are afforded allows you to keep the same software without significant upgrades, updates, or changes on each and every computer in your office that might shut you down for a while.

I know, I just threw IT wording at you: late-binding.  Don’t stop reading; I want to explain the benefits.  There are always updates that occur, but instead of having to buy the newest version and licensing for all computers on your system, you may only need to buy the newest version of the software for the newest computer to the system or if the boss (you) wants the latest edition.  Everyone else can continue to work with what they are used to, and you have the latest and greatest.  Because of the late-binding function within the software, flexibility is at a maximum and your company isn’t forced to make unnecessary upgrades that could be costly and inefficient.

It seems as if the healthcare industry is just a few steps behind many of the other industries and professions out there. However, the benefits that could be seen and felt within healthcare could surpass just about anyone else’s needs out there. Whether you are a doctor, other health professional or work at some other business industry, you can understand the need for flexibility and up-to-the-minute information that great software provides at a doctor’s office or hospital.

Image credit: