New Strategies Needed to Monitor Blood Usage

The US blood system is a vast network comprised of approximately 190regional blood centers, which collect 90% of the nation's blood, and 621hospital blood centers, which collect the remaining 10%.[1] Many of theregional blood centers are operated by the American Red Cross, which collectsapproximately 45% of the blood in the United States.[2]

The United States is almost completely self-sufficient in its bloodsupply, with only 2% of the supply being imported from Western Europe eachyear.[3] Safety protocols for the collection, processing, and distributionof blood and blood products are regulated by the FDA through the Centerfor Biologics Evaluation and Research.

Whole blood, packed red blood cells, platelets, and frozen plasma productsaccount for 0.29% of total hospital expenditures in the United States,including blood storage, processing, and transportation.[4]

Hospital administrators, looking to improve their bottom line and remaincompetitive in a managed care environment, are holding physicians moreaccountable for blood and blood product usage. However, most of the effortsto assess physician performance have focused on utilization review techniques.

Recently, "report cards" have been developed to more formallyassess physician performance in the use of blood products and other importantmedical resources.

These efforts at accountability have inherent cultural and operationalobstacles, as described below, and are not generally well accepted by manyphysicians. Newer approaches have recently been developed and implemented,based on prospective evaluations that include assessments of quality ofcare, patterns of care, and costs of care.

The Report Card Approach

At the policy level, report cards were developed in response to demandsfrom health care regulators, contractors, purchasers, and consumers foran effective means by which to evaluate and compare health care providers.Many large managed care organizations have developed report cards on providers'quality of care, while policy makers and researchers at the state and federallevel have generated surgical procedure report cards for physicians andhospitals.[5-8]

Report card implementation has resulted in the definition of sets ofhealth plan performance measures, the most prominent of which is the HealthPlan Employer Data and Information Set (HEDIS), developed by the NationalCommittee for Quality Assurance (NCQA).[9]

Validity Questioned

The use of report cards in comparing providers has caused much scientificcontroversy. Questions of validity have been raised by physicians who doubtthe fairness and consistency of the measures applied in different healthcare settings.

Physicians argue that report cards fail to control adequately for condition-specific factors such as severity of illness, comorbidity, and coincidentconditions, as well as for patient acceptance of effective care, all ofwhich play major roles in determining patient outcomes.[10]

At hospitals throughout the country, report cards have been implementedto monitor the use of transfused blood products. For example, at one hospital,software programs have been developed that identify all transfusions topatients with hemoglobin concentrations above a certain level. Each caseso identified is reviewed by a pathologist for assessment of clinical need,based on documentation in the medical record.

For those cases in which documentation is insufficient, the attendingphysician is asked to provide support for the decision to use transfusion.If upon further review, the transfusion is still felt to be clinicallyunnecessary, an additional assessment by the utilization review committeeis carried out. If this committee agrees with the determination, formaldocumentation about the incident is attached to the physician's file.

The entire review process is based on retrospective assessments andrequires multiple levels of communication and documentation. The inadequacyof this process has been substantially documented. Furthermore, the reportcard process runs counter to today's advanced management thinking concerningmotivation and performance.[11]

In addition to concerns over the validity of report cards, many feelthat their widespread use may have significant adverse health care implications.Feedback from report cards based on evaluation of overuse may cause physiciansto alter their practice patterns to err on the side of underuse of importantblood products so as to minimize their chances of being identified as anoutlier.

Another problem with report cards is that they are inherently statictools, and although they may be able to determine certain performance deficits,they do not clarify which of many underlying processes contribute to thosedeficits. It has also been documented that poor outcomes are due largelyto failures in the processes of care rather than to individual error.[11]

Instrument Panel Approach

The numerous concerns surrounding report cards have caused many in thehealth care industry to explore other potentially more valid and more acceptableperformance measures. The trend toward the use of real-time performanceimprovement tools in many business settings has prompted the use of similarmeasures, namely the "instrument panel," in the health care arena.

Potential Impact of Instrument Panel Approach on Physician Performance

The use of instrument panels to help develop clinical practice guidelines,or monitor physician compliance with them, is especially applicable tothe blood banking industry, since blood usage rates can vary widely amongphysicians.

Hospital administrators, faced with the task of improving cost-effectiveness,have traditionally taken a reactive, report card type approach to standardizingtreatment protocols among their physicians, rather than a proactive approach.

By singling out and reprimanding physicians with questionable bloodusage patterns, the hospital is sending a message: That physicians needto be constantly looking over their shoulder, and if and when their practicesare questioned, they may find themselves in an uncomfortable, defensivesituation that is not conducive to education or change.

Performance improvement efforts using instrument panels would likelyhave a positive impact in blood banking by making better, timelier informationreadily available to physicians and their patients.

In this system, selected indicators are measured at frequent intervalsto generate statistical process control charts. These charts provide qualityassurance personnel access to essential, current information that can beused to facilitate decision making and needed change. When using an instrumentpanel in blood banking, potential indicators might include quality of care,patient satisfaction, quality of life, length of stay, functional status,return to work, and cost of care.[12]

One advantage of the provider-initiated, continuous quality improvementapproach relative to static report cards is that it is broader in its organizationalreach, thereby reducing certain systematic and potentially confoundingsources of variation. Providers are also capable of capturing data pertinentto selectively chosen measurement indicators, a large advantage over themore rigid, externally based report cards.

The practical benefit of the instrument panel in the clinical settingcomes when it is integrated with clinical practice guidelines. This performancemeasurement methodology is extremely beneficial in the effort to standardizeclinical practice through guidelines, thereby contributing greatly to qualityand efficiency of care.

A Real-Life Example

Citrus Valley Health Partners (CVHP), comprised of the Queen of theValley and Inter-Community campuses of Citrus Valley Medical Center, aswell as Foothill Presbyterian Hospital and Citrus Valley Hospice, servesa suburb of approximately 720,000 northeast of Los Angeles.

It is an example of a large health-care provider that has recently initiateda capitated approach to blood product usage. As a part of this capitationprogram, CVHP is ideally suited to develop prospective performance measuresin its blood banks using a model similar to the instrument panel describedhere.

This provider has adopted a universal computer system for its threehospital networks that facilitates its ability to collect the necessaryclinical data required to monitor prospective performance improvement measures.

A program to establish a comprehensive database to track and monitorblood and blood product overusage and underusage throughout CVHP is currentlyunderway. The information systems are designed to identify clinical episodesthat are associated with overuse, under-use, or misuse of blood productsin a prospective fashion.

One of CVHP's goals is the development of a total blood management systemthat will determine appropriate blood and blood product usage based onsuccessful treatment regimens within the CVHP system. Target ranges arecalculated according to previous yearly usage rates and projected ratesas determined by revised clinical practice guidelines.

CVHP takes a proactive approach to ensure physician compliance withits guidelines, namely educating physicians beforehand rather than reprimandingthem after the fact. The company openly communicates its expectations forits physicians and explains its position that standard treatment protocolsare beneficial not only for CVHP but also for the patients it serves.

When variant usage patterns are determined through computer monitoring,the company can assess whether the patterns are consistently variant orrepresent isolated incidents.

If usage patterns are variant throughout the system, then target rangesand clinical practice guidelines will be reevaluated. If usage patternsare isolated as coming from a single physician, then that physician willbe reoriented as to the established clinical practice guidelines.

An important aspect of this approach is the cultural fit of instrumentpanels. Physicians at CVHP have been involved in other capitated paymentprograms with successful implementation in the past few years.[12]

In one of the early examples of instrument panel benefits, three physicianoutliers were identified based on adverse outcome rates. After reorientationof the physicians, adverse outcome rates decreased from 30% to 4% withinsix month.

In addition to instrument panels, CVHP has devised other strategiesto succeed in a capitated payment environment. For example, they have enteredinto a blood usage risk-share agreement with their main blood service provider,HemaCare Corporation.

HemaCare works with CVHP to develop target ranges for each individualblood product, and if CVHP usage falls below those ranges, the cost savingsfrom lowered utilization are split between CVHP and HemaCare, creatinganother incentive for CVHP's proactive approach to its blood usage.

CVHP is also attempting to implement widespread use of newer technologiessuch as leuko-reduced platelet products using the COBE LRS system, whichHemaCare provides at no additional charge. The use of leuko-reduced productseliminates the need for bedside filtration units and cytomegalovirus testing,as well as lowering required nursing administration time, a cost benefitthat would be maintained even if health care providers were charged morefor leuko-reduced products.

Recommendations

The shift from externally based report card programs to provider-initiatedintegration of guidelines and instrument panels could be very beneficialto the blood banking industry as a whole, especially as medical paymentstrategies move toward capitation.

The teamwork and analysis that are required to implement such a performanceimprovement strategy are likely to enable providers to thrive in a rapidlychanging health care environment, a secondary effect of performance improvementprograms that has been previously documented.[13]

Health care providers who use performance measures to focus on individualcomponents within the health care system, such as blood banking, whileconsidering their use as part of a larger integrated delivery system arein a better position to determine and implement effective performance measures.

Use of effective performance measures brings health care providers onestep closer to improving patient treatment outcomes while at the same timeachieving a better cost profile. These goals are even more attainable whenadvanced performance measures integrating instrument panels with clinicalpractice guidelines are implemented.

References:

1. Wallace EL, Churchill WH, Surgenor DM, et al: Collection and transfusionof blood and blood components in the United States, 1992. Transfusion 35:802-812,1995.

2. Leveton LB, Sox HC, Stoto MA: HIV and the Blood Supply: An Analysisof Crisis Decisionmaking. Washington, DC, National Academy Press, p. 6,1995.

3. Wallace EL, Surgenor DM, Hao HS, et al: Collection and transfusionof blood and blood components in the United States, 1989. Transfusion 33:139-144,1993.

4. Woolhandler S, Himmelstein DU, Lewontin JP: Administrative costsin U.S. hospitals. N Engl J Med 329:400-403, 1993.

5. Pennsylvania Health Care Cost Containment Council: A Consumer Guideto Coronary Artery Bypass Graft Surgery. Harrisburg, PA, 1995.

6. Welch HG, Miller ME, Welch WP: Physician profiling: An analysis ofinpatient practice patterns in Florida and Oregon. N Engl J Med 330:607-612,1994.

7. NY State Department of Health: Coronary Artery Bypass Graft Surgeryin New York State. Albany, NY, New York State Department of Health, 1992.

8. Sawyer D, Donaldson S: Surgical Scorecards. Prime Time Live Transcript#248, 1992.

9. Corrigan JM, Nielson DM: Toward the development of uniform reportingstandards for managed care organizations: The Health Plan Employer Dataand Information Set (Version 2.0). Jt Comm J Qual Improv 19:566-575, 1993.

10. Berwick DM, Wald DL: Hospital leaders' opinions of the HCFA mortalitydata. JAMA 263:247-249, 1990.

11. Bader B: Rediscovering Quality. Rockville, Md, Bader and Associates,1992.

12. Ullman M, Metzger CK, Kuzel T, et al: Performance measurement inprostate cancer care: Beyond report cards. Urology 47:356-365, 1996.

13. Gottlieb LK, Margolis CZ, Schoenbaum SC: Clinical practice guidelinesat an HMO: Development and implementation in a quality improvement model.Qual Rev Bull 16:80-86, 1990.