Section 2 (continued)

Cost of Poor Quality or Waste in Integrated Delivery System Settings

2.5.2. Reductions in Waste that Damage Health Care Providers' Bottom-Line Financial Performance

A health care provider's long-term financial viability rests on net operating income (NOI, or margin), calculated by subtracting total operating expenses (cost of operations) from total revenues received. Most care providers explicitly track NOI by individual types of service (bundled episodes). If sustained NOI margins fall below about 2 to 3 percent of total operating costs over time, then the health care provider will fail financially. A minimum level of retained earnings is essential to replace aging buildings and equipment and for growth into new areas of care delivery as the science and technology of medicine advances over time. For-profit health care providers, including private practice physicians, must also build a profit margin (return on investment) and tax payments into their NOI margins.

Health care providers divide their costs of operations into two broad categories: fixed costs and variable costs. Fixed costs are expenses that a physician, clinic, hospital, or delivery system must pay regardless of the number of patients they treat. Fixed costs include such elements as payments for buildings, equipment, licensing and regulatory fees, malpractice insurance, and minimum levels of staffing required by regulation. Variable costs are those expenses that fluctuate directly with patient volumes, such as medications, disposable equipment, other supplies, and short-term staffing levels. The mix of fixed versus variable costs varies with the type of care being given; however, across most care delivery settings, 60 to 75 percent of all expenses fall in the fixed cost category (Roberts et al., 1999). Thus, only about 25 to 40 percent of the savings generated by quality-based waste reduction—representing the variable cost proportion—immediately appears on a health care deliverer's financial balance sheet as NOI margin. Sixty to 75 percent of the potential savings, representing the fixed cost proportion, appears as unused ("excess") capacity.

If a care provider can recruit additional, clinically appropriate patients to use the newly freed capacity, then fixed cost savings arising from waste reduction efforts will quickly drop to the care provider's bottom line, especially if the service episodes represented by the new patients have larger NOI margins than the episodes that were eliminated by waste reduction (Maureen Bisognano at IHI calls this "converting light green dollars into dark green dollars"). With higher patient volumes, the care provider's fixed costs will spread across a larger patient population (all of the former patients, plus the additional new ones) so that the effective fixed cost per patient falls, producing a larger NOI margin for each patient treated. If a care provider cannot increase patient volumes, then it must "manage out" the excess fixed capacity to recover the associated costs—an activity that can easily take several years; involve significant transition costs; and involve wrenching human decisions, such as eliminating jobs and releasing long-term employees. These management costs usually make it impossible to recover the entire excess capacity and realize related potential fixed-cost savings.

Different care delivery episode types have different levels of NOI margins. In particular, in the current U.S. health care system, some episode types have negative margins—providing a needed service means that the care provider will lose money, on average, for each patient treated. To survive financially, care providers who provide such services must counterbalance negative margin services with positive margin services. "Contribution to margin" (or "contribution") combines the concept of episode-level NOI margin with fixed versus variable cost accounting. A health care provider will insist that an episode type generate at least enough revenue to cover its associated variable costs. Any revenues in excess of variable costs then "contribute" to paying the care provider's fixed costs. The task of administrative financial management then becomes finding enough "contribution" to cover all of a provider's fixed costs, plus some level of retained earnings for capital replacement, growth, and return on investment (for for-profit providers).

"Contribution" also explains a set of business strategies that some health care providers bring to the health marketplace. "Cherry picking" describes a strategy of selectively providing only services with a high NOI margin. When unit costs are bundled to a case level (i.e., the DRG case-level payment system), "skimming" refers to the practice of selecting individual cases that have a high NOI margin, while refusing negative margin cases of the same episode type. The most common form of skimming involves physicians with practice privileges at two competing organizations, who have a direct or indirect financial return at one of the organizations. In such a circumstance, a physician will selectively route positive margin cases to the organization that generates higher personal income, while routing negative margin cases to the other organization.

Exhibit 5 diagrams the interaction between waste recovery levels and payment mechanisms currently used in the United States, from the NOI margin perspective of a health care provider (the level where waste recovery efforts necessarily reside). Only waste recovery at the Patient Level uniformly results in financial benefit under current payment mechanisms. As a result, in the past, almost all administrative efforts for cost control have focused at the Patient Level, where all successful activities provide financial benefit to the administrative team that initiates action to save costs. Given their long history and a high degree of understanding of these mechanisms within health care delivery, within this project we chose to ignore many traditional cost reduction strategies, such as negotiating better prices for supplies, replacing amortized equipment with more efficient models, or shifting staff mix to reduce seniority levels (more senior staff means higher pay for the same position) or staffing mix. We instead focused our Patient-Level efforts on TPS tools, an area largely unexplored in current health system management. 

Exhibit 5. Effect of cost reduction strategies on a health care delivery organization's net operating income (NOI), based on common payment mechanisms

Cost Reduction StrategyPayment Mechanism
Discounted Fee For ServicePer CasePer DiemShared Risk
1. Decrease the cost of a unit of care (Patient Level)
2. Decrease the number of units of (Episode Level) care per case:
a. decrease the number of nursing hours per case
b. decrease the number of any other unit of case per case
c. avoid complications in ways that decrease units, but reclassify the case
3. Decrease the number of cases (Population Level)

Notes: A "unit of care" is any single diagnostic, treatment, or support activity performed on a patient's behalf, such as any single laboratory test, an imaging examination, a dose of a drug, or an hour of nursing care. For example, if a hospital is paid discounted fee-for-service for a particular case and successfully reduces the number of hours of nursing care services required to treat that case, then the hospital will lose reimbursement for those hours of service (column 1, row 2 above). If the hospital's administration had built a small margin of NOI into each hour of nursing care services, then the hospital will lose that NOI. Successful waste elimination always reduces costs, but some payment mechanisms may deliver those savings exclusively to a health care payer (e.g., the Federal Medicare program, an HMO, or a self-insured employer), leaving the health care delivery group to cover the costs of the waste elimination activity and, at the same time, find other funds to cover the lost NOI.

The fact that waste elimination at the Episode and Population Levels often damages care providers' financial viability, while care providers are the only group that can change care delivery practices to eliminate the waste, is the foundation for the recent Pay for Performance (P4P) movement. P4P aims to redesign health care delivery payment so that care providers share in the savings generated by waste reduction, both as a financial incentive to eliminate waste and to provide the financial resources necessary to undertake waste reduction efforts. In particular, under current payment mechanisms and given health care's high proportion of fixed expenses, successful waste reduction efforts can greatly increase care providers' incentives to increase their "unit of care" frequency.

As item 2(c) in Exhibit 5 notes, successful waste reduction efforts at the Episode Level can also damage care providers' NOI margin by changing the episode type. To illustrate, in 1995, Dr. Kim Bateman, a physician medical director managing nine rural facilities within Intermountain Healthcare, successfully introduced an evidence-based best practice guideline for community-acquired pneumonia (CAP). Intermountain tracked the impact of the guideline by comparing its effect in the 9 rural facilities with CAP results in 12 other adult hospitals within the Intermountain Healthcare system where the practice guideline was not introduced (a prospective nonrandomized controlled trial, or quasi-experiment). There were 1,179 CAP cases treated from 1994 through 1996 in the 9 rural hospitals using the new CAP protocol, compared with 3,455 CAP cases treated without the protocol in those same 9 hospitals (preprotocol period) plus the 12 Intermountain control hospitals (both pre- and postprotocol periods, to establish secular trend). All cases were fully risk-adjusted based on 3M APR-DRG severity of illness (a commercial risk adjustment system that predicts cost based on patient presentation), patient age, gender, location (urban versus rural), specific hospital, and calendar month of treatment (to pick up medical inflation over time, as well as general changes in pneumonia care that affected the entire community).

One major effect of the Care Process Model (CPM) was to change physicians' initial choice of antibiotics. Use of recommended "ideal" antibiotics increased from 22 percent to 40 percent in test hospitals following introduction of the practice guideline. As appropriate choice of ideal antibiotics increased, complication rates decreased (from 15.3 percent to 11.6 percent of patients suffering a major complication while hospitalized) and mortality rates decreased (from 7.2 percent to 5.3 percent, representing about 70 lives per year). With fewer complications to treat, costs under the CPM fell by an average of $572 per case (from a baseline average of $4,851 per case), an 11.9 percent drop in cost per case. Across the 1,179 patients in the test group, costs fell by a total of $674,329.

At the same time, payments dropped by $894 (16.9%) per case against straight-line projections of expected revenues (initial average across all payers: $5,306). The total loss in expected payments totaled $948,317. Intermountain suffered a net loss of $273,988, or about 5 percent of the total cost of care (Exhibit 6). In other words, all of the savings flowed back to payers and pulled additional NOI operating margin as well. The reason for the payment reduction was changes in DRG categories that resulted from fewer complications. For example, pneumonia-associated complications often result in a patient requiring support from a mechanical ventilator. Such ventilator support shifts the patient's DRG assignment from DRG 89—CAP to DRG 475—Long-Term Ventilator Support. At the time of this study, DRG 475 paid about $16,400 and provided a positive NOI margin of about $800. DRG 89, on the other hand, had a negative NOI margin of about $400 (Intermountain received about $400 less per case than the true costs of operations). 

Exhibit 6. Financial results of quality waste reduction in CAP

Exhibit 6. Line graph displays the financial results of quality waste reduction in CAP. For details, go to Notes below.

Month relative to protocol introduction


  • Actual costs (blue dashed line) versus expected costs (green line) for the cases managed under the guideline. The guideline was introduced at time point 0 on the x-axis (at the vertical line). Note that average actual costs tracked very closely to expected costs in the "before" period but that there is a net accumulation of lower costs after the intervention.
  • The upper and lower dotted red lines are statistical process control limits—a method of showing statistical significance (p = 0.001) graphically, for single points.
  • The black line is reimbursement, which fell further than costs.
  • All of the data are fully risk adjusted.
  • Further to the right in the graph, case numbers decline, as reflected in the widening control limits (the red lines). That is because hospitals joined the study one at a time, usually one per month ("staggered implementation," which strengthens the statistical validity of the study design).
  • The green line (expected costs) rises over time, reflecting actual medical inflation over the life of the study.

2.5.3. Intermediate Financial Conclusions that Affect Waste Estimation and Reduction

One man's waste is another man's profit. Whether an expenditure appears to be waste or profit depends on one's level in the Chain of Effect for Quality, as reflected through the four listed payment mechanisms. For example, when confronted with cost controls that reduce the amount a physician receives per patient visit, or that a hospital receives per case, physicians and hospitals routinely respond by increasing the number of visits or the number of hospitalizations.

Paid for waste. Another anomaly of current payment policies is that health care providers are paid for quality waste. "Dr. R. Burney (2004) observed that the health care system tolerates poor quality and pays the same for poor and high quality care". Current reimbursement policies actually provide financial incentives for poor quality care. Examples include payment for treatments for complications due to adverse events, which are technically defects of the system. The industry has clearly recognized this problem.

Margins matter most. NOI margins, not total cost or savings, are the key measure for return on waste reduction investments at the care provider level.

Throughput. CFOs are interested in throughput because of inefficient use of physical and human capital. In one sense, throughput rests on the idea of continuum of care; by breaking up the continuum of care, health care providers can increase the number of episodes while reducing their internal cost per episode. Alternatively, throughput is the key to more efficient use of facilities and labor. Throughput, or its opposite (delay), has immediate and second-order impacts. The primary impact is on the inefficient use of capital—human and material. Each additional day in the hospital requires additional nursing hours, blood draws, meals, medical consults, and so on. Typically, these costs are in proportion to the additional time spent. Second-order impact is that each additional day involves new staff who must be oriented to the patient, with attendant handoff problems and costs. Also, more days means additional chances for infections or other adverse events. Third-order delay is administrative (e.g., more complex staffing and scheduling plans, added facility space for holding areas), including the need to house patients sometimes in inappropriate beds (e.g., use of an ICU bed versus a telemetry bed, a psych bed, or an ED holding area). Complex staffing and scheduling can also increase patient transport and reconciliation costs. Managing throughput is a more efficient way of continuously managing fixed cost inputs, by spreading them over larger numbers of cases, rather than conducting a waste reduction project, creating excess capacity, then reacting to "manage out" the excess capacity.

Page last reviewed September 2008
Internet Citation: Section 2 (continued): Cost of Poor Quality or Waste in Integrated Delivery System Settings. September 2008. Agency for Healthcare Research and Quality, Rockville, MD.