The Challenge and Potential for Assuring Quality Health Care for the 21st Century
From Quality Measures to Quality Care: Examples of Quality Improvement at Work
Federal and State governments, private purchasers, physicians, nurses, insurers, labor unions, health plans, hospitals, accreditation organizations, and others have begun to address some of the significant quality problems in the United States health care system. One approach they have taken is to improve the ability to measure and report on the quality of care being delivered. The reporting of quality measures prompts a closer look at provider or health plan practices both as feedback for clinicians or as publicly available scorecards for consumers and purchasers to evaluate. Several of these endeavors have led to improvements in the quality of care and in the health outcomes for consumers, while often saving costs. Among these are:
The Prevention of Heart Attacks with Beta-Blocker Treatment
Despite the clear medical evidence that the use of beta blockers can significantly reduce the odds that a heart attack patient will have a second, and often fatal, attack, researchers have shown that too often these patients are not prescribed this life-saving therapy. Faced with this problem, the Medicare Peer Review Organizations (PROs) in Alabama, Connecticut, Iowa, and Wisconsin have launched a concerted campaign to increase use of beta blockers. They monitored beta-blocker use after heart attacks in Medicare patients and provided feedback to all practitioners in their States. As a result, beta-blocker use climbed from 31.8 percent of cases to 49.7 percent of cases (Marciniak, et al., 1998). Based on these promising findings, Medicare expanded this campaign nationwide among its PROs beginning in 1995.
With support from the Agency for Health Care Policy Research (AHCPR), the Minnesota Clinical Comparison and Assessment Program worked to generate consensus and commitment from providers to increase use of life-saving drugs like beta-blockers (Soumerai, et al., 1998). Using local medical opinion leaders at 20 hospitals, they focused on the evidence, identified barriers to change, and offered feedback of comparative performance. Their efforts yielded a 63-percent increase in the use of beta-blockers.
Using another life-saving treatment for heart attack patients, the Middletown Regional Hospital in Ohio reported decreased mortality rates by reducing the time from hospital admission to the administration of thrombolytic therapy. Mortality rates decreased from 10.9 percent to 6.5 percent in less than 1 year. For their efforts, this hospital received the Joint Commission on Accreditation of Healthcare Organization's 1997 Codman Award, recognizing excellence in the use of outcome measurement to improve quality.
The National Committee for Quality Assurance has begun requiring health plans to submit data on the use of beta blockers after a heart attack as part of its Health Plan Employer Data Information Set, version 3.0 (HEDIS 3.0), which is a standardized set of measures that assesses the performance of managed care organizations. This information assures that purchasers and consumers have comparative information about the use of beta blockers by the health professionals in a particular health plan. In 1998, the Health Care Financing Administration, the Federal agency that administers and oversees the Medicare and Medicaid programs, began requiring health plans participating in Medicare to submit HEDIS data, including information on beta blocker use.
Improving the Quality of Care for Heart Bypass Surgery
More than half a million Americans undergo coronary artery bypass surgery each year, but the quality of that surgery varies from hospital to hospital. To help consumers choose the best care, the State of New York and a consortium of hospitals have been using quality measurement and reporting techniques to flag problem areas and to improve the quality of care for bypass patients. For nearly a decade, the New York State Department of Health has collected and released hospital-specific data on coronary artery bypass surgery. Using these data, New York targets quality improvement efforts to those hospitals with quality problems, including site visits, comprehensive consultations, and probationary action taken against facilities until they agree to implement necessary changes. As a result of these efforts, between 1989 and 1992, the risk-adjusted mortality rate for bypass patients in New York State declined 41 percent (Hannan, et al., 1994). By 1993, the cardiac surgery mortality rate had declined by 52 percent (Longo, et al., 1997; Hannan, et al., 1994a).
The New England Cardiovascular Project, involving five hospitals in northern New England, was begun in 1990 to reduce mortality rates among coronary bypass patients. Similar to the New York program, this effort includes site visits, data reporting, and training for health care professionals. At the end of the project, researchers reported a 24-percent reduction in coronary artery bypass surgery mortality among the five hospitals (O'Connor, et al., 1996).
Other groups have undertaken quality measurement and reporting projects aimed at improving survival rates among heart bypass patients. General Motors, First Chicago NBD, and others initiated the Southeast Michigan Health Care Consortium, which is collecting outcomes data for all health care centers in the region. They plan to publish data on angioplasty and coronary artery bypass surgery in the fall of 1998 (The Business Roundtable, 1997). Similarly, the California Office of Statewide Health Planning and Development and the Pacific Business Group on Health have developed the California Coronary Artery Bypass Graft Mortality Reporting Program, which collects and reports risk-adjusted, hospital-level mortality data for California hospitals that perform bypass surgery. Both of these groups hope to stimulate quality improvement by enabling comparisons among participating hospitals and health care providers (Meyer, et al., 1997).
Reducing Unnecessary Cesarean Deliveries
In 1995, nearly 4 million women entered hospitals to give birth, and 785,000 women underwent cesarean section. Research into the variations in the rates of cesarean section procedures has raised questions about whether cesarean sections are being used when they are not necessary. While clinically indicated for some deliveries, this surgical procedure carries risks as well as morbidity when used as an alternative to vaginal delivery. Studies show that women living in Southern States had the highest cesarean rates while women living in Western States had the lowest rates (Clarke and Taffel, 1996). Even within States, wide variations exist. For example, cesarean section rates in Washington State hospitals ranged from 0 percent to 43 percent (McKenzie and Stephenson, 1993).
The Greater Cleveland Health Quality Choice Coalition produces a biannual report card documenting patient outcomes in 27 local hospitals. In this report, the Coalition has tracked the number of cesarean section deliveries as a measure of performance in those hospitals. As a result, several Cleveland hospitals have developed and implemented practice guidelines and have fostered collaborative efforts with physicians to improve care. Over three successive reporting periods, the aggregate trend of total cesarean section rates in local hospitals has declined (Cleveland Health Quality Choice Program, 1997).
The Missouri Department of Health also developed a consumer report on obstetrical services. The ShowMe Buyer's Guide: Obstetrical Services, issued in 1993, covered all 90 Missouri hospitals (Longo, et al., 1997). The guide looked at the length of stay, number of births, level of perinatal care, and the availability of labor, delivery, and recovery beds at each hospital. It reported the cesarean delivery rate, high-risk infant transfer rate, ultrasound rate, vaginal birth after cesarean, very low birth weight of each facility along with the rates of neonatal mortality, and patient satisfaction. Within 1 year of the report, approximately 50 percent of Missouri hospitals had taken action to address some of the issues raised by the guide: instituting or planning formal transfer agreements with tertiary care centers for referral of high-risk infants, developing car seat programs to improve safety, and training nurse educators for breast-feeding. On follow-up, hospitals previously performing a high level of cesarean deliveries also had reviewed their procedures and significantly lowered their rates for this procedure.
Various firms and group purchasers have embarked on similar efforts to improve obstetrical care. The Massachusetts Healthcare Purchaser Group, a coalition of 27 corporate and government health care purchasers, released a 1994 report card to coalition members on obstetric care and held a "best-practice forum" on cesarean sections (Jordan, et al., 1995). The Dallas-Ft. Worth Business Coalition has initiated a pilot study to develop best practices for pregnancy and childbirth through quality measurement and reporting. General Motors, Chrysler, and the United Auto Workers union have also joined together to develop best practices for cesarean section in Flint, Michigan (AFL-CIO, 1997).
Reducing Asthma-Related Deaths
Despite advances in treatment, mortality from asthma has risen 58 percent from 1979 to 1992 (NAEPP, 1998). Over the same period, both the incidence of asthma cases and hospitalizations related to it climbed. In 1993, asthma led to 468,000 hospitalizations, 100 million days of restricted activity annually, and a yearly cost of $6.2 billion. However, there are a number of examples of quality improvement efforts that have been effective at improving health outcomes and reducing costs for people with asthma. These steps have resulted in fewer emergency room visits and hospital stays.
In 1989, the National Institutes of Health initiated the National Asthma Education and Prevention Program (NAEPP). One of its landmark accomplishments was the 1991 release of the Expert Panel Report: Guidelines for the Diagnosis and Management of Asthma. In this report, the NIH recommends the use of anti-inflammatory therapy (e.g., corticosteroids) for long-term control of persistent asthma, along with patient/provider education and other measures (NAEPP, 1991). The NAEPP disseminated the report widely to health care professionals, medical schools, health care organizations, professional societies, and patients and updated the guidelines in 1997 (NAEPP, 1997). Prior to the report's release, only 21 percent of emergency physicians reported prescribing early corticosteroid administration in asthma. In 1994, 82 percent of physicians reported that they followed this practice recommended in the guidelines (Lantner, et al., 1995).
Parkland Memorial Hospital in Dallas, TX, has implemented a plan of increased education for asthma patients and providers, coupled with more intensive treatment (NAEPP, 1996). The hospital recorded a 53-percent reduction in emergency department visits by asthma patients during the first 2 years. In an inner-city program in Boston, children with asthma experienced an 86 percent reduction in hospital admissions and a 79-percent decline in emergency department visits in response to interventions, including one-on-one consultations with an asthma outreach nurse, patient education in self-management, and regular monitoring of patient progress (Greineder, 1995).
The Mayo Clinic of Rochester, MN, reduced asthma-related emergency department and urgent care center visits by 22 percent (Weiss, 1997). Hospitalization rates for Family Medicine asthma patients under age 65 also fell, by 47 percent for those patients between the ages of 45 and 64. Their approach involved improved monitoring of asthma severity through the use of peak flow meter readings and site visits to physician offices to promote prescribing of anti-inflammatory drugs for asthma. By creating slots in the Family Medicine scheduling system, the hospital ensured that asthmatics could make follow-up appointments before leaving the Center.
The National Committee for Quality Assurance has introduced a measure of appropriate medications for people with asthma in its HEDIS 3.0 testing set. The Office of Personnel Management and the Foundation for Accountability (FACCT) have also embarked on a collaborative effort to improve asthma disease management (AFL-CIO, 1997).
Employers like AT&T, First Chicago NBD, and Deere and Company have also made efforts to improve the care of employees with asthma and reduce costs for such care (The Business Roundtable, 1997). In the first year of its program, Deere and Company, an Illinois-based manufacturer of farm equipment, recorded a 12-percent reduction in asthma-related treatment costs (Strickland, 1996). The company has achieved this by inviting workers and dependents with asthma to participate in an asthma education program, and facilitating the development of personal action plans with their physicians to guide employees' self-management of this chronic condition.
Health plans have also recorded success with disease management programs in asthma. The Family Health Plan Cooperative of Milwaukee, Wisconsin, has developed an Asthma Treatment Plan based on the NIH guidelines. They have identified the population at high risk, developed an education plan for asthmatics focusing on self-management, and encouraged physicians to work with their patients to create effective, customized asthma management plans. From 1993 to 1994, the Family Health Plan Cooperative reported a 22-percent reduction in asthma hospital admissions for patients between the ages of 2 and 19 (GHAA, 1995).
Increasing Influenza Vaccination Among Adults
Influenza and pneumonia is the sixth leading cause of death in the United States (NCHS, 1997). For the elderly, influenza and its complications are particularly problematic, with 95 percent of the resulting deaths occurring in Americans over the age of 60 (Govaert, et al., 1994). Studies of influenza vaccine demonstrate saved lives and lower costs. In one clinical trial, influenza vaccination halved the incidence of the flu among patients over the age of 60 (Govaert, et al., 1994). Yet only 52 percent of people age 65 and over received the vaccine in 1993 (CDC, 1995).
The National Committee for Quality Assurance now requires health plans to report on the number of older adults in the plans who receive flu shots each year. The Department of Veterans Affairs has encouraged local VA hospitals to immunize 100 percent of its patients against the flu (Department of Veterans Affairs, 1996). Under a 4-year program, the Centers for Disease Control and Prevention (CDC) and the Health Care Financing Administration (HCFA) have implemented the Medicare Influenza Vaccine Demonstration project to promote use of this preventive measure (CDC, 1992). HCFA distributed letters to Medicare beneficiaries, provided physician reminders, trained nurses to recognize high-risk patients, and piggybacked vaccination messages on telephone company mailers. Over a 3-year period, overall flu vaccination rates climbed from 26 percent to 48 percent. In a few of the intervention sites, vaccination rates surpassed 60 percent.
Building on the success of this demonstration project, influenza vaccine became a covered benefit under the Medicare program, and HCFA has undertaken a national effort to increase the use of the influenza vaccine. HCFA is working with its Peer Review Organizations (PROs), the Centers for Disease Control and Prevention, and the National Coalition for Adult Immunization to determine how best to improve immunization rates. It is also working with States, community-based organizations, and senior advocacy groups to promote this practice. The overall goal of the HCFA effort is to achieve a 60-percent influenza immunization rate among Medicare beneficiaries age 65 and older by the year 2000.
Reducing the Incidence of Pressure Ulcers
For the hospitalized and home care patient, pressure ulcers, or bed sores, can be a serious and chronic problem. While bedbound or immobile, unrelieved pressure on the skin can lead to its breakdown and damage to the underlying tissue. Patients in critical care, with hip fractures or with spinal injuries, are at particular risk. Studies have found that as many as one in four nursing home patients experience pressure ulcers and the cost of treating them is more than $1.3 billion annually (Bergstrom, et al., 1994).
The Agency for Health Care Policy and Research has published two clinical practice guidelines addressing the prediction, prevention, and treatment of pressure ulcers (Bergstrom, 1992; Bergstrom, 1994). Various groups have implemented quality improvement programs to prevent and treat pressure ulcers based on these guidelines (Suntken, et al., 1996).
In its 140 long term care facilities, the Department of Veterans Affairs has undertaken a major quality improvement program to reduce the rate of pressure ulcers. In 1991, the VA Office of Quality Management began a program of calculating facility-specific incidence rates of pressure ulcers and disseminating the results to all long term care facilities in the VA system (Berlowitz, 1997). The VA viewed pressure ulcer development as an ideal indicator because they are common, required the coordinated efforts and attention of multiple health care providers, and could serve as a sentinel for more systemic problems in a facility. A VA facility in Memphis, TN, employed a team-based approach to monitor patients and provide skin care. The rate of pressure ulcer development dropped from 11.5 percent to zero. Overall, the VA has experienced a decline in the incidence of pressure ulcers from 4.9 percent in 1990 to 3.1 percent in 1993. This offers another example where the feedback of quality measures prompts improvements in the quality of care.
Reducing Low-Birthweight Births
Though infant mortality rates in the United States have declined, the incidence of low birthweight babies has not. This trend is troubling since 70 percent of infant mortality traces to low birthweight newborns (Goldenberg, et al., 1998). Also, low birthweight babies often require extensive and costly treatments. An AHCPR-sponsored Patient Outcome Research Team (PORT) studied the use of drugs called corticosteroids as one treatment to reduce morbidity and mortality in these newborns. For these premature newborns, the PORT found that corticosteroids prevented many of the complications associated with prematurity—respiratory distress syndrome, brain hemorrhage, and death.
Based on the PORT's work and other findings, an NIH consensus panel issued a guideline recommending corticosteroid treatment of women at risk of very pre-term deliveries (NIH, 1994). The NIH disseminated this guideline to Federal health agencies, health care organizations, continuing medical education directors at hospitals, deans of medical schools, and directors of State and county medical societies. The media and major medical journals also covered the release of the guideline. Prior to the release of the guideline, only 30 percent of women who delivered prematurely received corticosteroids. By 1996, this had risen to 70 percent (Goldenberg et al., 1998). Another study found that nearly 87 percent of obstetricians surveyed after the release of the NIH consensus statement believed that corticosteroids could reduce infant mortality. Consistent with the NIH guidelines, 91 percent of obstetricians reported that they would prescribe this drug therapy to prevent the complications of pre-term deliveries (Wright, 1996). An NIH-commissioned study concludes that if corticosteroids were used in 60 percent of cases threatening preterm delivery, $157 million could be saved annually (NIH, 1994).
Preventing Adverse Drug Reactions
Preventable adverse drug events cause significant mortality each year (Leape, 1994). Responding to this problem, providers have led efforts to make significant improvements in the rate of adverse drug events.
LDS Hospital in Salt Lake City, Utah, has taken steps to improve its rate of adverse drug events. From 1990 to 1993, they discovered that 2.43 percent of hospital patients experienced an adverse drug event (Classen, 1997). The hospital projected that if 50 percent of these reactions were prevented, cost savings would total $500,000, and hospitalization would be reduced by 450 days annually. The hospital took a multi-pronged approach that involved providing timely feedback to physicians to prevent severe problems; more effective tracking of patient drug allergies; a program that monitors kidney function while patients receive certain antibiotics with toxic effects on the kidneys; and a computerized disease management program for antibiotic use that integrates all of these components. These efforts led to a 75-percent decline in adverse drug reactions related to antibiotics.
Through the Institute for Healthcare Improvement's Breakthrough series, Michigan's Pontiac Osteopathic Hospital Medical Center achieved an 80-percent reduction in adverse drug events in its cardiac monitoring unit in 1 year (Leape, 1998). The program included standardization of physician order forms, especially for common conditions covered by standardized treatment regimens; thorough notation of patient drug allergies through computerized systems and wristbands; education on and easy access to drug incompatibility information; increased availability of needed medications; and routine monitoring of drug distribution. In total, 43 health care organizations participated in the Breakthrough series, setting a goal of reducing adverse drug reactions by 30 percent or more. After 1 year, one-third of the organizations had exceeded those targets.
The American Medical Association recently launched a National Patient Safety Foundation with broad support from accreditation organizations, academic institutions, health plans, and other partners. Its mission is to "enhance the safety of the U.S. health care system" by (1) promoting research on human and organizational error and its prevention; (2) increasing awareness and communication in patient safety and errors; and (3) encouraging the application of knowledge in this field (National Patient Safety Foundation, 1998).
The Joint Commission on the Accreditation of Healthcare Organizations has established a Sentinel Event Policy, which encourages accredited health care organizations to voluntarily report unexpected events or errors that lead to death, or serious physical or psychological injury (JCAHO, 1998). If an organization does not conduct an acceptable analysis of the root cause of the error, the JCAHO will place that hospital on Accreditation Watch, a fact that may be publicly disclosed as part of the organization's existing accreditation status. This designation signifies that the organization is under close monitoring by the Joint Commission.
The Department of Veterans Affairs has also designed an error reduction system in health care for use throughout its delivery system. This VA effort has taken lessons from the Aviation Safety Reporting System. Under this system, the Federal Aviation Administration has increased the reporting of errors in aviation and improved safety with steps ranging from better air traffic control procedures to a recognition of problems with pilot fatigue (President's Advisory Commission, 1998).