Driving Out Waste: A Framework to Enhance Value in Clinical Care (HFMA)

Having the right infrastructure and the buy-in of clinicians are the two key requirements for a successful initiative focused on reducing variation in the ordering and delivery of care to improve the value of care delivery processes.

[Written by Dr. David Burton. View original hfma article here]

At a Glance

  • Many healthcare providers today are seeking to improve the value of the care they deliver by implementing standardized clinical practice guidelines aimed at reducing variations in care, avoiding complications, and lowering costs.
  • To succeed, such an initiative requires the full support and participation of the clinicians who will use the guidelines.
  • Providers also should have a fully developed infrastructure consisting of a clinical content system, an analytics system, and a deployment system.

Value is rapidly becoming the linchpin of success in our volatile healthcare environment. With payers, regulators, and consumers increasingly demanding proof of high–quality, safe and cost-effective care, it is incumbent on hospitals to find a way to reduce costs by improving clinical performance.

One way to accomplish this is to focus on eliminating waste throughout the U.S. healthcare system, including waste caused by:

  • Variation in what care is ordered
  • Variation in how care is delivered
  • Preventable complications resulting from care delivery—termed defects in Sigma/Lean terminology—that can harm or injure patients (e.g., hospital-acquired conditions)

Creating standardized clinical effectiveness guidelines that reduce variation can definitely improve clinical performance. However, to be truly effective, such an undertaking requires the participation and support of the clinicians who are expected to implement the guidelines, including the physicians and other clinicians and stakeholders who can “make or break” these initiatives.

To generate enthusiasm among time-challenged and often skeptical physicians, clinical transformation projects must quickly produce meaningful, actionable data that demonstrate the value of these efforts for producing measurable improvements. Replacing the common fragmented approach to quality and process improvement with comprehensive, standardized processes that address waste throughout organizations is equally important.

Critical Ingredient No. 1: A Three-Part Infrastructure

Success in improving both clinical and financial outcomes depends, first, on having the ability to harness the flood of data pouring into hospitals from disparate sources. Electronic health records (EHRs) are essential, but they are just one component needed to provide reliable data about quality, safety, service, and resource management.

Arriving at value-based measures that physicians will embrace also involves more than simply importing commercially available professional guidelines and imposing them from the top down—although evaluating available resources usually starts the process.

Before embarking on any comprehensive initiative aimed at reducing variations in care, avoiding complications, and lowering costs, an organization should have in place a fully developed infrastructure consisting of three parts:

  • A clinical content system
  • An analytics system
  • A deployment system

Clinical content. Organizations will require a clinical content system to begin standardizing the care process. The assets contained in this system should be developed from best practices drawn from the scientific literature and expert opinion, including specialty society guidelines.

These assets should include three types of starter sets—that is, sets of essential information and intelligence drawn from the scientific literature and expert opinion that identify the primary considerations for managing a patient population:

  • Clinical effectiveness starter sets for key categories of care based on best practices(e.g., health maintenance and preventive guidelines, diagnostic algorithms; criteria for triage to treatment venue; treatment and monitoring algorithms; indications for referral, admission, and intervention; substance selection criteria; order sets)
  • Value stream map and process improvement tool starter sets to optimize workflows for each care unit and type of care
  • Patient safety risk assessment and injury prevention protocol starter sets, as well as tracking systems to detect failures in patient safety processes

Analytics. Data from transactional systems (e.g., clinical [EHR], financial, administrative, departmental, patient satisfaction) should be integrated into an enterprise data warehouse to allow key processes to be measured and analyzed. Developing such a resource also allows the healthcare system to implement refined versions of the clinical content starter sets to:

  • Automate regulatory/compliance reporting
  • Automate measures of clinical effectiveness (e.g., Healthcare Effectiveness Data and Information Set and pay for performance)
  • Identify, track, and measure the effectiveness of care for specific patient populations, types of care, and care units
  • Measure workflow efficiency

Deployment. The organization should create and train multidisciplinary teams of experts—including physicians, nurses, and administrators—to prioritize, refine, and deploy clinical content assets and measures of processes and outcomes across the enterprise. By overlooking this important means of putting value-based measures into operation, hospitals could be left with only isolated pockets of excellence. As former major league baseball manager Casey Stengel once said, “Finding good players is easy. Getting them to play as a team is another story.”

To successfully improve outcomes, hospitals also must understand how best to apply infrastructure in driving out waste and establishing standardized guidelines for best care delivery practices. The following examples offer insight into how best to approach this challenge.

Variations in the Ordering of Care

How can hospitals use data generated by this kind of infrastructure to eliminate the waste that drags down clinical and financial outcomes? To illustrate, let’s consider what is typically involved in the treatment of heart failure.

Heart failure is a common diagnosis in hospitals, but the tests used to arrive at this diagnosis vary widely from physician to physician, facility to facility, and region to region. Not all of these tests are of equal value in establishing the diagnosis and monitoring the care provided. Sorting these tests according to their clinical- and cost-effectiveness can improve patient care and lower costs for hospitals and patients.

Broadly, there are three types of tests currently being ordered for diagnosing and managing heart failure (and other disease conditions).

Tests that are deemed essential for every patient suspected of heart failure. Two tests are consistently warranted: left ventricular ejection fraction (LVEF), which measures how well the left ventricle pumps out blood (an ejection fraction below 40 percent signals significant heart failure), and brain natriuretic peptide (BNP), which can essentially rule out heart failure when levels are less than 100 pg/ml.

Tests that can help diagnose heart failure, but that are not specific to the diagnosis. A chest X-ray, for example, can show an enlarged heart, fluid outside the lung, or congestion in the lung–each of which could result from other conditions—or a set of blood gases that shows decreased oxygen saturation, which could be indicative either of pneumonia or heart failure. Depending on the circumstances, each of these tests may be useful in confirming the diagnosis of heart failure or managing the condition effectively.

Tests that do not add value. A ventriculogram is performed as part of a cardiac catheterization to visualize how much blood is ejected each time the heart pumps would be wasteful. The procedure costs multiple times more than an echocardiogram and does not add significant diagnostic value.

In this way, hospitals can move through their model of clinical care—from diagnosis to triage to care provided throughout the continuum—to identify which tests are essential, contributory, or wasteful. This approach can be applied to all diseases, not just heart disease. The findings provide a basis for quantifying costs that can be wrung out of the system by standardizing care processes, leading to enhanced clinical

Variation in the Delivery of Care

Value stream maps are a useful tool for pinpointing workflow process improvement potential. By identifying and standardizing the most effective method for carrying out each step in a process and eliminating time wasted between steps, an organization can reduce variation, which improves efficiency and, therefore, cost-effectiveness.

For example, mapping each required step for an inpatient surgery, from scheduling through recovery, can pinpoint bottlenecks that slow turnover, impacting patient care as well as costs. (For an example of a value stream map, go to hfma.org/hfm.)

By applying time stamps from its EHR to each individual task, one hospital determined that its ORs were standing empty for 10 minutes after each case before the cleaning crew arrived. Investigation identified the root cause: The crew was not being notified immediately when a patient was wheeled out of the operating room (OR). The solution was to identify exactly who was responsible for notifying the crew and what means of communication would be most efficient for this purpose: cell phone, beeper, alert, or text message. This newly standardized workflow now ensures automatic, immediate notification, and the analytic system measures how well it is working. Since implementing the process improvement protocol, the time has decreased, on average, by 10 minutes per case.

Improving the cycle time in this and other ways opens up the possibility of adding another case each day; for a hospital with five ORs, that means the capacity for 25 additional cases each week.

Defects in Care

Consider the case of a person who is admitted to a hospital with a collapsed lung caused by pulmonary fibrosis and then develops a pressure ulcer within 12 days of the admission, requiring treatment that adds two days to the length of stay. Now that Medicare no longer pays for such extra care or extra days, every case like this one cuts into a hospital’s margin.

To improve patient safety by reducing pressure ulcers, one hospital first set out to define the set of patients who always should be screened to determine whether they are at risk for this kind of injury. The hospital concluded that patients who require screening are those in critical care and medical-surgical units who are confined to bed, sometimes for extended periods.

The next step was to develop criteria to identify the subgroup of at-risk patients to target with preventive interventions. In this case, the clinical asset database pointed toward a score of 14 or less on the Braden Scale, which predicts the risk of developing pressure sores.

For patients who score below the cut-off point, the hospital designed a multipart intervention, again drawing on evidence-based clinical effectiveness guidelines. Steps constituting the protocol include using a special mattress, turning the patient frequently, and limiting the types of substances that come in contact with the skin of pressure areas. The hospital also implemented an ongoing tracking system to measure its effectiveness in preventing pressure sores.

This protocol led to a significant reduction in the number of inpatients who developed stage III or stage IV pressure ulcers. By adopting this approach to patient safety analytics, hospitals can avoid defects that undermine patient safety, many of which also would have significant financial implications if payers have included them on their lists of “never events” for which they decline to provide coverage.

Value Measures in Practice

To see how value measures produced by the enterprise data warehouse and integrated into clinical protocols can work to improve medical management and outcomes, let’s consider diabetes, which consumes an increasing share of healthcare resources every year.

A key indicator of diabetes is hemoglobin A1c, which measures blood glucose values over a three-month period. Many physicians sincerely believe they follow best practices for A1c testing, but in reality, physicians all too often do not test as many as half of their patients for A1c at appropriate intervals. Moreover, these physicians are often unaware that their care falls short of best practice standards. The use of standardized metrics across the board to uncover such variances can provide a framework for constructive engagement and performance improvement for all types of care, including diabetes management.

A system that can convey useful information in a timely manner by converting analytic data into easy-to-understand graphics can be a profoundly useful asset for reducing waste. Consider the sheer amount of information required for clinicians and hospitals to achieve the highest standards for diabetes care. For example, for an initiative focused on improving diabetes care, an organization would require ready access to the following information:

  • The number of diabetic patients tested within the specified time frame
  • The proportion of those tested with A1c levels above and below 7 percent
  • The percentage of patients with A1c readings of 9 percent or higher (indicating that they face a 100-fold greater risk of complication than do patients with a value of 7 percent)

Moreover, because patients with diabetes have a much greater risk of heart attacks, strokes, nerve damage, blindness, amputations and other complications, controlling their cholesterol (especially LDL cholesterol), blood pressure, and other measures is also essential. Therefore, a holistic approach is required that incorporates similar monitoring and reporting for other key indicators. As an example, for LDL cholesterol alone, the hospital would require the following information:

  • The percentage of patients for whom their physician follows the recommended timetable for LDL cholesterol testing
  • The proportion of patients whose LDL levels are within the acceptable target range of less than 100 mg/dl, and the breakdown among the group whose levels exceed that level
  • The extent to which protocols for blood pressure monitoring are being followed and how successful patients are at maintaining readings consistent with the recommended guideline

Similar types of information would also be required with respect to kidney function and eye exams.

Clearly, having ready access to accurate, complete and timely data is critical to ensuring the success of such an initiative. Progress would be slow, and would likely grind to a stop, if the initiative required wading through pages of numbers to extract useful information.

Moreover, even the most accurate, complete, and timely data alone cannot guarantee that a health system will be able to reduce waste significantly. Achieving meaningful results also depends on the organization’s ability to translate the data into actionable intelligence to create the foundation for ongoing, enterprisewide patient quality and safety initiatives. For this purpose, the organization requires automated drill-down analytics and reporting capabilities that can enable it and its physicians to identify and easily contact outlier patients to schedule appointments for needed tests and/or to adjust treatment regimens. Such tools are also valuable for identifying variability and waste in ordering and workflow by helping to benchmark the facility and individual clinicians to clinical effectiveness and workflow guidelines.

Critical Ingredient No. 2: Physician Buy-In

Again, it should be stressed that achieving physician buy-in and participation from the beginning is vital for success. The initiative will succeed only if physicians and other clinicians are convinced that the process is clinically driven, transparent, and focused on improving both the quality and cost-effectiveness of care, and not simply an imposed cost-cutting mandate. The process should be designed to build consensus and to educate and continuously update all stakeholders.

Physicians also should have a guiding role in the initiative. The very first step should be to identify physicians to lead the initiative. Often these “early adopter” physicians have begun improvement initiatives on their own at one institution within a system, but lack a robust infrastructure to support their efforts.

The next step is to introduce these physician leaders to the analytics system through the starter sets. It is then up to the physicians (enterprisewide leaders and front-line) to refine the criteria and metrics in the analytics system and the guidelines and protocols in the content system based on their knowledge of the scientific evidence and experience. The guidelines and protocols will continue to evolve as needed based on new insights, evidence, guidelines, regulations, and lessons learned.

Simply put, physicians must take ownership of the clinical effectiveness guidelines for improvement initiatives to succeed.

From Data to Waste Removal to Payer Collaboration

Hospitals and health systems that are able to standardize best practices into value-based care and drive out all three types of waste will be in a prime position to negotiate favorable relationships with payers. In fact, the transformation of the healthcare system and the shared desire among providers and payers to reduce waste make it more likely that payers will agree to allow their claims data to be integrated into providers’ enterprise data warehouses. This level of integration would enable payers and providers to develop collaborative utilization management programs using shared views of entire episodes of both ambulatory and acute care, with quantitative tracking of results. These are among the essential ingredients of effective population health management.

In this way, savings realized through improved processes and outcomes can be used to advance value-based purchasing and new finance and care models that encompass pay for performance, meaningful use, and risk assumption by providers for bundled payments, comprehensive disease and population management, capitation, and other innovative payment approaches.

It all starts with a data infrastructure that allows organizations to shrink waste by reducing variation and defects in clinical care.

[Written by Dr. David Burton. View original hfma article here]