AHRQ Annual Highlights, 2010 (continued)
Crosscutting Activities include a variety of research projects and activities related to quality, effectiveness, and efficiency, which support the five research portfolios. These activities include data collection and measurement, dissemination, rapid cycle research, training, and intramural and extramural research sponsored by multiple portfolios.
Healthcare Cost and Utilization Project (HCUP)
HCUP is a family of health care databases and related software tools and products developed through a Federal-State-industry partnership and sponsored by AHRQ. HCUP databases bring together the data collection efforts of 43 State data organizations, hospital associations, private data organizations, and the Federal Government to create a national information resource of patient-level health care data. HCUP includes the largest collection of all-payer encounter-level longitudinal hospital care data in the United States, beginning in 1988. These databases enable research on a broad range of health policy issues, including cost and quality of health services, medical practice patterns, access to health care programs, and outcomes of treatments at the national, State, and local market levels.
HCUP Data Revisit Analyses
Made available for public use, the newly developed HCUP Supplemental Files for Revisit Analyses are discharge-level files designed to facilitate analyses to track patients across time and hospital settings in the HCUP inpatient, ambulatory surgery, and emergency department databases. Each record in an HCUP database represents one discharge abstract from a hospital setting (inpatient, emergency department, or ambulatory surgery)—meaning if an individual visited the hospital three times in a given year, the HCUP databases would include three separate records in the respective database. After being combined with the corresponding databases, the HCUP Supplemental Files for Revisit Analyses allows researchers to uniformly identify sequential visits for an individual and use the available clinical information to study if the visits are unrelated, complications from a previous treatment, or an unexpected revisit or re-hospitalization. Total charge information from each visit can be combined to determine the total charge or cost for an episode of hospital care.
New Data Added to HCUP
In FY10, AHRQ added 2008 data to its HCUP national database marking the 20th anniversary of providing powerful data to researchers and policy makers. The Nationwide Inpatient Sample (NIS) for 2008 includes 1,056 hospitals from 42 States. Data from NIS are available from 1988 to 2008, allowing analysis of trends over time. The NIS is nationally representative of all short-term, non- Federal hospitals in the United States. It approximates a 20-percent stratified sample of hospitals in the United States and is drawn from the HCUP State Inpatient Databases, which include 90 percent of all discharges in the United States. The NIS includes all patients from each sampled hospital, regardless of payer—including persons covered by Medicare, Medicaid, and private insurance, as well as uninsured persons. It encompasses all discharge data from more than 1,000 hospitals in 40 States.
HCUP Statistical Briefs
In FY10, AHRQ continued to issue HCUP Statistical Briefs, a series of Web-based publications containing information from HCUP. These publications provide concise, easy-to-read information on hospital care, costs, quality, utilization, access, and trends for all payers (including Medicare, Medicaid, private insurance, and uninsured). Each Statistical Brief covers an important health care issue. For example:
- Nearly 25 percent of patients in public hospitals were covered by Medicaid, compared with 17.3 percent in private not-for-profit hospitals. Public hospitals cared for over 75 percent more uninsured patients than did private not-for-profit hospitals (8.3 percent versus 4.7 percent). Nearly 11 percent of all patients in metropolitan public hospitals were uninsured and 27.7 percent were covered by Medicaid. (Go to Public Hospitals in the United States, 2008, http://www.hcup-us.ahrq.gov/reports/statbriefs/sb95.jsp.)
- The rate of infections among medical and surgical discharges peaked in 2004 and 2005 at 2.30 per 1,000 stays, then declined to 2.03, a rate similar to year 2000. (Go to Adult Hospital Stays with Infections Due to Medical Care, 2007, http://www.hcup-us.ahrq.gov/reports/statbriefs/sb94.jsp.)
- In 2008, nearly one in five hospitalizations were related to patients with diabetes, totaling over 7.7 million stays and $83 billion in hospital costs. Hospital stays for patients with diabetes were longer, more costly, and more likely to originate in the emergency department than stays for patients without diabetes. (Go to Hospital Stays for Patients with Diabetes, 2008, http://www.hcup-us.ahrq.gov/reports/statbriefs/sb93.jsp.)
- In 2008, approximately 18 percent of Medicare beneficiaries were dually eligible for Medicaid. Dually eligible patients accounted for about one-third of all Medicare hospital stays with a principal diagnosis of pressure ulcers (36 percent), asthma (32 percent), and diabetes (32 percent); and roughly one-quarter of stays for urinary tract infection, chronic obstructive pulmonary disease, and bacterial pneumonia. (Go to Hospitalizations for Potentially Preventable Conditions among Medicare-Medicaid Dual Eligibles, 2008, http://www.hcup-us.ahrq.gov/reports/statbriefs/sb96.jsp.)
- Hospital costs for bone marrow transplants shot up 85 percent from $694 million to $1.3 billion between 2004 and 2007. Ten procedures experienced rapid cost increases between 2004 and 2007. About 75 percent of the rise was due to increases in the number of patients who underwent these procedures and 25 percent resulted from higher costs per case treated. In addition to bone marrow transplantation, the procedures with the most rapid increases in hospital costs included:
- Open surgery for noncancerous enlarged prostate: Up 69 percent to $1 billion.
- Aortic valve resection or replacement: Up 38.5 percent to $1.9 billion.
- Cancer chemotherapy: Up 33 percent to $2.6 billion.
- Spinal fusion: Up 29.5 percent to $8.9 billion.
- Lobectomy (a type of lung cancer surgery): Up 29 percent to $1.8 billion.
- Incision and drainage of skin and other tissues: Up 29 percent to $1 billion.
- Knee surgery: Up 27.5 percent to $9.2 billion.
- Nephrostomy (surgery to allow urine to pass through the kidneys): Up 25 percent to $683 million.
- Mastectomy (breast removal because of cancer): Up 24 percent to $660 million.
- Nearly 8 million patients were admitted to U.S. hospitals on weekends in 2007 and received 36 percent of major procedures on the day of admission. In comparison, patients who were admitted on weekdays received 65 percent of all major procedures on their first day in the hospital. For example:
- Patients who were admitted on weekends were more likely to be there due to emergencies, such as heart attack, stomach bleeding, fractures, or internal injuries, than patients hospitalized on a weekday (65 percent versus 44 percent).
- A smaller share of weekend than weekday admissions was elective (11 percent weekend and 28 percent weekday).
- Sixty-four percent of heart attack patients admitted on a weekend had a major cardiac procedure, such as angioplasty or heart bypass surgery, performed by the second day of their hospitalization, compared with 76 percent of heart attack patients admitted on a weekday. A smaller share of weekend than weekday admissions received treatment on the day of admission for back surgery (35 percent versus 90 percent); angina (23 percent versus 37 percent); gallbladder removal (23 percent versus 32 percent); and hernia repair (54 percent versus 68 percent).
- Weekday admissions were often planned in advance. For example, 99 percent of admissions for osteoarthritis and 93 percent of those for back problems occurred on weekdays.
- About 2.4 percent of patients admitted on a weekend died in the hospital, compared with 1.8 percent of patients admitted on a weekday.
- Hospital charges for uninsured stays grew by 88 percent between 1998 and 2007. The average charge for an uninsured hospital stay grew from $11,400 in 1998 to $21,400 in 2007 after adjusting for inflation. In addition:
- From 1998 to 2007, the number of uninsured hospital stays increased by 31 percent, which far exceeds the 13 percent overall increase in hospital stays during the period.
- The percentage of uninsured hospital stays increased the most in the South, rising from 5.8 percent to 7.5 percent. In contrast, in the Midwest, the percentage of uninsured hospital stays declined from 4.7 percent to 4.0 percent.
- The top reason that uninsured patients were hospitalized was for childbirth. In 2007, roughly a quarter of a million uninsured women gave birth in hospitals. This reason was followed by mood disorders (94,300); chest pain with no observed cause (77,000); skin infections, which more than doubled from 31,000 to 73,300; and alcohol-related disorders (66,600).
For more information about HCUP, go to http://www.ahrq.gov/data/hcup.
Knowledge Transfer project results in two States adding their data to HCUPnet
After attending workshops featuring HCUPnet offered as part of AHRQ's Knowledge Transfer Quality Diagnostic Tools for States project, two States—Maine and Tennessee—agreed to share their data publicly through HCUPnet (http://hcupnet.ahrq.gov).
Representatives from Maine attended the AHRQ Knowledge Transfer Program's "State Healthcare Quality Improvement Workshop." This workshop featured several AHRQ products and tools to diagnose quality issues and improve quality of care. The Maine participants were impressed by the various capabilities of the HCUPnet tool. After the workshop, they urged Al Prysunka, Executive Director of the Maine Health Data Organization, to share data in HCUPnet.
Prysunka, who also serves on the Healthcare Cost and Utilization Project Advisory Committee at AHRQ, notes that several participants from the December 2007 workshop called him to discuss sharing Maine's data as a result of the HCUPnet presentations and information gained at the meeting. Upon deciding to share Maine's data, Prysunka says that "the relationship between AHRQ and Maine is quite positive. AHRQ has a staff that does a lot of analytical and methodological work that is wonderful."
Another AHRQ workshop was influential in encouraging Tennessee's participation in HCUPnet. Tennessee officials attended a technical assistance meeting, "Using Administrative Data To Answer State Policy Questions," which focused on several AHRQ tools—including HCUP and HCUPnet—that support a State's analysis of administrative data. Attendees included staff from the Tennessee Hospital Association, which holds the State contract to collect hospital discharge data, as well as the Tennessee Department of Health, which regulates the collection of data. The workshop included a live demonstration of HCUPnet by HCUP staff. The demonstration impressed the Tennessee delegation and convinced its members that the State's information should be included in the online tool.
Brooks Daverman, Health Quality Analyst, Tennessee Department of Finance and Administration's Division of Health Planning, notes, "After the presentation, our delegation decided independently that Tennessee's data should be on HCUPnet, and we agreed upon our return to Nashville to seek the necessary approvals from our various groups. Those approvals were quickly obtained, and we were able to instruct AHRQ to put Tennessee's information on the AHRQ Web site soon after." Tennessee has continued to increase the amount of data the State shares in HCUP, recently adding emergency room data.
AHRQ Quality Indicators
In FY10, AHRQ released an updated version of its Quality Indicators (QIs) software 4.1a and 4.1b (for both Windows and SAS®). The AHRQ QIs are used to highlight potential quality concerns, identify areas that need further study and investigation, and track changes over time. A number of the measures are also used in comparative hospital reporting and basing payment on quality. The AHRQ QIs are organized into four modules, each of which measures quality associated with the delivery of care occurring in either an outpatient or an inpatient setting:
- Prevention Quality Indicators (PQIs) are ambulatory care-sensitive conditions that identify adult hospital admissions that evidence suggests could have been avoided, at least in part, through high-quality outpatient care.
- Inpatient Quality Indicators (IQIs) reflect quality of care for adults inside hospitals and include: inpatient mortality for medical conditions; inpatient mortality for surgical procedures; utilization of procedures for which there are questions of overuse, underuse, or misuse; and volume of procedures for which there is evidence that a higher volume of procedures may be associated with lower mortality.
- Patient Safety Indicators (PSIs) also reflect quality of care for adults inside hospitals, but focus on potentially avoidable complications and patient safety events.
- Pediatric Quality Indicators (PDIs) are indicators of children's health care that can be used with inpatient discharge data. They are designed to examine both the quality of inpatient care and the quality of outpatient care that can be inferred from inpatient data, such as potentially preventable hospitalizations.
Currently, 48 AHRQ QIs and AHRQ QI composites are endorsed by the National Quality Forum. In addition, 21 States use the QIs, giving the public access to hospital quality information. For more information, see http://www.qualityindicators.ahrq.gov.
The 2009 National Healthcare Quality and Disparities Reports
In April 2010, AHRQ released the 2009 National Healthcare Quality Report (NHQR) and the 2009 National Healthcare Disparities Report (NHDR) the seventh consecutive, annual editions of the congressionally-mandated reports. Overall, the reports found that quality is improving, but the pace is slow. In addition, many disparities are not decreasing, with cancer, heart failure, and pneumonia needing particular attention.
The 2009 reports also include a new section on lifestyle modification, such as preventing or reducing obesity, which is crucial for the health of many Americans and an important task for health care providers. For example:
- More than 40 percent of obese adults have never received advice from their doctor about exercise.
- Obese adults who are black, Hispanic, or poor or have less than a high school education are less likely to receive diet advice from their doctor.
- Most overweight children and one-third of obese adults report that they have not been told by their doctor that they are overweight.
- Most American children have never received counseling from their health care provider about exercise, and almost half have never received counseling about healthy eating.
National Healthcare Quality Report (NHQR)
Sample of Findings: Deaths Due to Heart Attack
The 2009 NHQR found that the overall proportion of hospital patients who died in the hospital after a heart attack fell by 37 percent between 2000 and 2007, from 106 per 1,000 patients to 67 per 1,000 patients during the period. Medicare patients experienced the largest decline (37 percent) and Medicaid patients, the smallest (27 percent). While the privately insured and uninsured rates fell at a similar pace (32 percent), the uninsured were much more likely to die from a heart attack (93 versus 67 deaths per 1,000 heart attack admissions).
From 2000-2007, Midwestern hospitals went from having the highest heart attack death rate in the country to the lowest (from 112 to 63 deaths per 1,000 heart attack admissions). Western hospitals had the highest rate of heart attack deaths than any other region in 2007 (71 deaths per 1,000 heart attack admissions). In 2000, they had the second highest. The death rate from heart attacks fell the most in the largest hospitals (500 beds or more), and by 2007 the heart attack death rate was almost 1.5 times lower than that of smaller hospitals with fewer than 100 beds (60 versus 87 deaths per 1,000 heart attack admissions).
Sample of Findings: Deaths Due to Complications
The 2009 NHQR also found that fewer hospital patients died from complications in their health care between 2001 and 2006, but Asians/Pacific Islanders and Hispanics were less likely to survive than either whites or blacks. The overall death rate for patients ages 18 to 74 who developed a complication such as pneumonia, blood clots, or blood infections during their hospitalization decreased 23 percent (from 152 deaths to 117 deaths for every 1,000 patients with complications) from 2001 to 2006.
Although the death rate for Asians and Pacific Islanders fell 24 percent during the period, they had the highest death rate of any group in both 2001 and 2006. The death rate for Hispanic patients declined by 21 percent, yet by 2006, their rate was the second highest of any group (122 deaths per 1,000 patients). The largest decrease in death rate was for black patients, which declined by 30 percent. In 2001, blacks had higher rates than whites, but by 2006, the black death rate was the lowest of any of the four groups of patients (111 deaths per 1,000 patients).
2010 NHQR Theme: The Effects of Insurance Status on Quality of Care
The overall theme for the 2009 NHQR was insurance status. In summary, this increased scrutiny on the effect of insurance status on quality of health care found that uninsured patients had worse outcomes and received worse care than patients with private insurance. Several key areas were identified as most pronounced:
- The overall proportion of hospital patients who died in the hospital after a heart attack fell between 2000 and 2006 for all insurance groups (the privately insured, Medicare and Medicaid recipients, and the uninsured). However, death rates among uninsured patients were higher than among patients with private insurance.
- The biggest difference in health care services was in receipt of mammograms by women ages 40 to 64, in which 74 percent of those with private insurance had a mammogram, while less than 40 percent of those with no insurance had been screened.
National Healthcare Disparities Report (NHDR)
One of the basic tenets of a high quality health care system is that all groups within the population should receive equally high care quality. Getting into the health care system (access to care) and receiving appropriate health care in time for the services to be effective (quality care) are key factors in ensuring good health outcomes for all groups.
Consistent with extensive research and findings in previous NHDRs, the 2009 report finds that disparities related to race, ethnicity, and socioeconomic status are still pervasive in the American health care system. Although varying in magnitude by condition and population, disparities are observed in almost all aspects of health care. Members of racial and ethnic minority groups, poor people, less educated people, and people without health insurance face significant barriers to care and experience worse health outcomes.
Sample of Findings: The Five Largest Observed Disparities
The NHDR identifies five measures for which the disparities are largest and are worsening over time for racial and ethnic minorities compared with whites:
- Deaths per 100,000 population per year for colorectal cancer among blacks.
- Asian adults age 65 and over who did not ever receive a pneumococcal vaccination.
- American Indian and Alaska Native adults age 50 and over who did not receive colorectal cancer screening.
- American Indian and Alaska Native hospital patients with heart failure who did not receive recommended hospital care.
- Hispanic adults age 65 and over who did not ever receive a pneumococcal vaccination.
Sample of Findings: Areas Where Multiple Groups Experience Disparities
In some cases, such as respiratory diseases, disparities are seen in multiple groups. The gap between the proportion of black and white Americans with asthma who took an inhaled or oral medicine daily to prevent attacks grew wider between 2003 and 2006. There was no significant difference in the use of daily asthma medicine between the two groups in 2003 (29 percent of blacks compared with 30 percent of whites. By 2006, the proportion of blacks who reported taking daily asthma medicine had fallen to 25 percent, while 34 percent of whites reported taking it.
The gap between Hispanic and white asthma sufferers who reported daily use of medicine also widened from 2003 to 2006. Specifically, roughly 28 percent of Hispanics and 31 percent of whites reported taking medicine daily for asthma in 2003, and in 2006, the number of Hispanics taking the drugs decreased to 23 percent, while the number of whites taking them increased to 35 percent.
Sample of Findings: In Some Areas, Disparities Are Decreasing
Disparities, however, are not exclusively increasing, as some disparities have been nearly eliminated. From 2003 to 2006, the gap in use of asthma medications closed between high-income people and people at other income levels. During the same period, the gap also closed between people who did not finish high school and those with some college education.
Furthermore, improvements in care are being seen in some areas, with multiple groups experiencing increases in receipt of care. For example, screening women for osteoporosis rose dramatically. The proportion of women age 65 and over on Medicare who said that they had been screened for osteoporosis increased from 33 percent in 2001 to 64 percent in 2006. Additionally, the rate of bone density or bone mass screening in white women nearly doubled during the period (36 percent to 67 percent). Hispanic women reported the most dramatic increase in screening, from 22 percent to 55 percent. The percentage of black women who reported undergoing osteoporosis screening also rose significantly, from 16 percent to 38 percent. While all women reported increases in osteoporosis screening, income was a factor: by 2006, only 46 percent of poor women reported having had a screening test, compared with 80 percent of high-income women.
The number of residents at nursing homes who were kept physically restrained dropped by more than half from 2000 to 2007, from 10.4 percent in 2000 to 5 percent in 2007. Further analyses revealed that from 2000 to 2007, the gap between Asians/Pacific Islanders (APIs) and whites in the percentage of residents who were physically restrained decreased. Nevertheless, in 2007, the percentage of residents who were physically restrained was still higher for APIs than for whites (6 percent compared with 5 percent). From 2000 to 2007, the gap between Hispanics and whites also decreased. However, in 2007, the percentage of residents who were physically restrained was still higher for Hispanics than for whites (7 percent compared with 5 percent).
Sample of Findings: Disparities for Those With Hypertension
Despite improvements, disparities in some areas persist. At nearly 5 times that of whites, the hospital admission rate for blacks with hypertension was 161 per 100,000 people in 2006, (compared to 33 admissions per 100,000). Furthermore, the admission rate for Hispanics with high blood pressure was 61 per 100,000 people, or nearly twice that of whites. Asians and Pacific Islanders had the lowest admission rate for high blood pressure (26 per 100,000). Women were admitted for high blood pressure more often than men (56 versus 40 hospitalizations per 100,000). The poorest Americans were 2.5 times more likely to be admitted for high blood pressure than the wealthiest (83 versus 32 admissions per 100,000).
The 2009 NHQR and NHDR are available online at . In addition to copies of the reports, the Web site also includes an online Data Query System—NHQRDRNet—that provides access to national and State measures of quality, and access to nearly 40 data sources at http://nhqrnet.ahrq.gov/nhqrdr.
For FY10, the Agency's State Snapshots, a State-by-State health care quality and access comparison tool, was expanded to include increased data on health insurance. Much of the data on health care quality can now be categorized by source of payment, including private insurance, Medicare, Medicaid, and no insurance.
As always, the 2009 State Snapshots (available at: http://statesnapshots.ahrq.gov) provide State-specific health care quality information, which can compare strengths, weaknesses, and opportunities for improvement. State-level information used to create the State Snapshots stems from data collected for the 2009 NHQR. Overall, States get mixed reviews for the quality of care they provide. As in previous years, AHRQ's 2009 State Snapshots show that no State does universally well or poorly on quality measures.
Some States do far better or worse than others. For instance, Maine, Maryland, Wyoming, South Carolina, and the District of Columbia showed the greatest improvements. The five States showing the smallest improvements were North Dakota, Texas, West Virginia, Nebraska, and Washington. For each State, it is possible to identify specific clinical conditions that could account for the differing rates of improvement.
Furthermore, the new health insurance section allows users to compare payer-specific quality rates as well as differences among payers. For instance, a State can compare the quality of care received by Medicaid or uninsured patients with that received by these same patients nationally. In addition, a State can assess whether its insurance-related disparities are larger or smaller compared with the Nation as a whole.
The 2009 State Snapshots provide additional ways to analyze the quality of health care for each State compared with all States, or with States in the same region. For example, the Mountain States region of the United States, which includes Montana, Wyoming, Idaho, Utah, Nevada, Colorado, Arizona, and New Mexico, reported the lowest average (best) rate of potentially avoidable hospitalizations for heart failure in the Nation in 2006, at 266 admissions per 100,000 population. Continuing the regional analysis found that the following regions had the next lowest rates of potentially avoidable heart failure hospitalization rates:
- Pacific (which includes California, Oregon, Washington, and Alaska) had the second lowest average rate, at 316.5 admissions per 100,000 population.
- West North Central (which includes North Dakota, South Dakota, Nebraska, Iowa, Missouri, Minnesota, and Kansas), 362 per 100,000.
- New England (which includes Connecticut, Rhode Island, Massachusetts, New Hampshire, Vermont, and Maine), 364 per 100,000.
The regions with the highest rates were:
- East South Central (which includes Alabama, Mississippi, Tennessee, Kentucky), with a rate of 583 admissions per 100,000 population.
- East North Central (which includes Wisconsin, Michigan, Illinois, Indiana, and Ohio), 502 admissions per 100,000 population.
- West South Central (which includes Texas, Oklahoma, Arkansas, Louisiana), 496 admissions per 100,000 population.
- Southeast (which includes Florida, Georgia, North Carolina, South Carolina, Virginia, West Virginia, Maryland, Delaware), 460 admissions per 100,000 population.
- Mid-Atlantic (which includes New Jersey, New York, Pennsylvania), 430 admissions per 100,000 population.
Enhanced sections on asthma care, diabetes care, and health care disparities are also included on the State Snapshots. For example, two States, Oregon and Vermont, reported the Nation's lowest rates of potentially preventable hospitalizations for asthma in children ages 2 to 17 in 2006. Oregon reported the lowest rate of potentially avoidable hospitalizations, at 44 per 100,000 children, with Vermont followed closely with 46 admissions per 100,000 children. States that also reported low rates of potentially avoidable asthma hospitalizations per 100,000 children ages 2 to 17 included New Hampshire (62 per 100,000 children), Iowa (66), Utah (74), Nebraska (75), and Maine (78).