Future Directions for the National Healthcare Quality and Disparities Reports
Chapter 3: Updating the Framework for the NHQR and NHDR (cont.)
Rationale for the Four New Quality of Care Components
The committee concludes that high-quality, equitable health care is facilitated by enhanced access to care, efficiency, care coordination, and a supportive health systems infrastructure. For that reason, the committee has included all four of these quality components in the updated framework. This chapter does not present the rationale for including the pre-existing framework components of effectiveness, safety, timeliness, and patient centeredness as the rationale for each was presented in Envisioning the National Healthcare Quality Report, and AHRQ has responded by reporting on these topics.
The IOM defines access as "the timely use of personal health services to achieve the best possible health outcomes" (IOM, 1993, p. 4). Access to care remains a central challenge for the U.S. health care system (Ginsburg et al., 2008; IOM, 1993, 1998, 2009), and this topic has been highlighted in the NHDR as a component of health care quality that exhibits disparities. The committee finds that improving access is a fundamental aspect of quality for the entire population. Therefore, access should be addressed in both reports.
With more than 46 million uninsured Americans as of 2008 (U.S. Census Bureau, 2009a) and large numbers of Americans reporting they have gone without needed care (Cunningham and Felland, 2008; IOM, 2009), access is a critical issue for the nation. Uninsurance affects all population groups, not only low-income or minority groups. For example, as of 2008, people with household incomes greater than $50,000 per year (middle and higher income families) constituted 22.2 percent of the uninsured population (U.S. Census Bureau, 2009b), and non-Hispanic Whites made up nearly half of the uninsured individuals in the United States (U.S. Census Bureau, 2009c). Although the availability of health insurance is significant when measuring access and utilization—insurance is an entryway into the health care system and is often linked with health status (DeVoe et al., 2003; Hadley, 2002; Ross et al., 2006)—other aspects are also barriers to receiving appropriate medical care. For example, even if more people obtain insurance coverage, problems will likely persist in access to care, including affordability (Cummingham et al., 2008; IOM, 2009), access to a usual or ongoing primary care provider (Goldman and McGlynn, 2005; Sack, 2008), and the ability to see those physicians (Ahmed et al., 2001; Hall et al., 2008).
Affordability of health care is a major concern for Americans (Blendon et al., 2004; Gallup Consulting, 2009)—medical causes and related costs (in the form of medical bills, or lost wages due to days unable to work) were behind 62 percent of all personal bankruptcies in the United States in 2007 (Himmelstein et al., 2009). Additionally, 35 percent of adults with health insurance still experience access problems due to cost (Collins et al., 2008; Cummingham et al., 2008; Gabel et al., 2009; Wang et al., 2009). Deemed the underinsured, the number of individuals who fall into this category rose from 16 to 25 million from 2003 to 2007 (Schoen et al., 2008). High deductibles and copayments, exclusion by condition or by service, the Medicare Part D donut hole, and caps on coverage all contribute to lack of affordability (Briesacher et al., 2009; IOM, 2009).
Primary care represents the entry point and foundation for successful health care systems (Grumbach and Mold, 2009; Starfield and Shi, 2002, 2007). Individuals who report having usual and continuous sources of care (particularly primary care) are associated with overall better health regardless of other factors (including general health status, insurance status, greater utilization of health care services, fewer delays in getting care, and better preventive care) (Doescher et al., 2001; RWJF, 2002; Starfield and Shi, 2007; Starfield et al., 2003). Ensuring access to care for other specialties, such as mental and oral health care, is also important for overall health and availability to needed care (Chapin, 2009; Edelstein and Chinn, 2009; Pomerantz et al., 2008). Regular access to health care services has been shown to correlate with reduced hospital use while preserving quality because ongoing clinical preventive services allow for the prevention of disease or detection of asymptomatic disease or risk factors at early, treatable stages (Bodenheimer, 2005). If someone is not seeking ongoing care for his or her conditions, it is possible that the illness will not be managed effectively (Collins et al., 2008), increasing one's risk of a worsened condition that ultimately may be costlier to treat.
The combination of insurance coverage, affordability, and access to ongoing sources of care illustrates that access is a broad topic with multiple dimensions, and that it can be assessed by a variety of measures (AHRQ, 2009b; Cantor et al., 2007; The Commonwealth Fund Commission on a High Performance Health System, 2008; NCQA, 2009). The printed version of the 2008 NHDR reported on 10 access measures (AHRQ, 2009b). AHRQ breaks down the access section in the NHDR into two categories: "facilitators and barriers to care"and "health care utilization"(which includes measures of dental, emergency, and mental health care). By organizing the access section in this way, AHRQ attempts to capture the discrete variables that affect access. The committee suggests that AHRQ begin, at a minimum, to include those same or related measures in the NHQR as indicators of how well the structure of the nation's health care system responds to the various needs of patients.4 Exploratory methods for measuring some other aspects of access to care are presented in Appendix D.
Efficient care is defined in the IOM's Quality Chasm report as "avoiding waste, including waste of equipment, supplies, ideas, and energy"(IOM, 2001a, p. 6). While this definition captures the concept of efficiency, to better convey the technical aspects involved with measuring efficiency, the committee also uses a definition put forth by Safavi (2006) and McGlynn (2008): maximizing performance (health care outcomes) by producing the best possible outputs from a given set of resources or inputs. Efficiency measurement includes optimal management of resources (such as administrative, operational, and clinical policies and practices) in order to maximize health care value (Bentley et al., 2008).
Efficiency is only one aspect of value-based care. Assessing efficiency is a distinct process that focuses solely on the objective use of resources (e.g., human labor, supplies, devices, money) relative to producing health care outcomes (e.g., hospital discharge, clinical examinations). For instance, measuring efficient performance could mean assessing the number of health care professionals required to properly execute a surgical procedure. Unlike value, it does not include aspects of patient-centeredness or valued patient experiences (for which there are limited metrics) nor is it always equated with comparative-effective analysis, which is another way to assess value.
Efficiency was previously omitted from the original framework recommended in Envisioning the National Healthcare Quality Report on the grounds that it was "outside the scope of the Quality Report and will be better addressed by specific efforts designed to face the considerable methodological and measurement challenges involved"(IOM, 2001b, p. 66). Nine years later, growing costs and purchaser concern with value have created an increased level of interest in measures of efficiency for the health system. As of 2008, the United States spent 16.2 percent of its gross domestic product (GDP) ($2.3 trillion) on health care (CMS, 2010), a total that is projected to reach nearly 20 percent of the GDP (an estimated $4.3 trillion) by 2017 (Keehan et al., 2008). The committee, therefore, agrees that the contribution of efficiency to health care value and quality cannot be ignored and that this component must be more comprehensively addressed in the NHQR and NHDR. AHRQ first attempted to address the component of efficiency in the 2007 NHQR, yet efficiency remains an underdeveloped aspect of the report for which AHRQ has specifically requested guidance.
In assessing efficiency, it is important to note that much of the research on health care efficiency suggests that cost and quality are not necessarily correlated (Fisher et al., 2003; Roski et al., 2008; Scholle et al., 2005; Solberg et al., 2002): the amount of money or resources spent on health care is not always indicative (or predictive) of the quality of services received or outcomes achieved (Weinstein and Skinner, 2010). Examples show that some of the most cost-efficient delivery of health care services is occurring in settings with the highest quality care, providing models for others of how to attain efficient and high-quality care that offers high value (Cantor et al., 2007; The Commonwealth Fund Commission on a High Performance Health System, 2008). Yet there are examples that demonstrate the contrary, where higher total per capita state spending on health care is correlated with better quality care (Cooper, 2009a). The complexity implied in these results illustrates the challenges in providing national measures of efficiency.
To better understand how efficiency can be measured, it is useful to refer to Bentley and colleagues' description of the different types of waste in the U.S. health care system. Equating waste with inefficiency, they break down the different aspects of efficiency in the system into three main components: administrative, operational, and clinical:
Administrative waste is the excess administrative overhead that stems primarily from the complexity of the U.S. insurance and provider payment systems (e.g., billing/claims processing, sales/marketing practices, compliance procedures, benefits design), operational waste refers to other aspects of inefficient production process (e.g., unnecessary or duplicative procedures, use of defective devices that cause errors, or wasted time transporting people or materials), and clinical waste is created by the production of low-value outputs (e.g., overuse of certain procedures). (Bentley et al., 2008, p. 632)
Because outputs are always considered when evaluating efficiency measures, there is a distinction to be made between efficiency measures and measures of cost. Cost measures consider resource consumption (the inputs used) relative to costs without consideration, or in isolation, of the results produced (i.e., resources used by unit price) (AQA, 2009; Krumholz et al., 2008).
Efforts to incorporate quality outcomes (whether a patient's health outcomes or a provider's performance outcomes) in the construction of efficiency measures are underdeveloped, and significant questions have been raised regarding the use of these measures for public reporting, tiered network design, or pay-for-performance (McGlynn, 2008). Nonetheless, some cost and efficiency measures being used may help suggest opportunities for development of this area in the NHQR and NHDR (go to Appendix D).
This committee encourages the development of efficiency measures that determine health outcomes as an output. McGlynn's systematic review identifies two different types of outputs, or products, of the health care system: health services (e.g., visits, drugs, admissions) and health outcomes (e.g., preventable deaths, functional status, blood pressure control) (McGlynn, 2008). The review notes that the vast majority of efficiency measures from the examined literature focused on health services as the output, and that only 4 (out of 250) used health outcomes as the desired end. The committee agrees that ideal assessments of efficiency would use health outcomes as the outputs of interest, as the goal of high-value care is not merely to provide inexpensive care. The dearth of such efficiency measures deserves attention, and their development in the future could be an area that AHRQ plays a role in supporting. Involvement in this task will be important because a number of unresolved methodological issues persist regarding the creation of credible and reliable efficiency measures (e.g., how to incorporate quality outcomes, ensuring reliability of measurement, attribution of providers, and validating risk-adjustment methods) (Hussey and McGlynn, 2009).
Considerations for Reporting Efficiency Measures
The committee acknowledges the inevitable trade-off in reporting measures that capture information at a national level rather than at a more local or organizational level; the more national level a measure is, the more useful it might be to policy makers and the less useful it might be to health care providers or community-level decision-makers. Likewise, the more local or organizationally specific measures are, the more the reverse is true. However, the committee agrees that as national reports, the NHQR and NHDR should continue to primarily focus on reporting system-level (state and national) efficiency measures that reflect a broader perspective. Additional data relevant to other stakeholders (e.g., providers, payers), or reported at a more local level, could be presented via online mechanisms, derivative publications (e.g., fact sheets), or links to other source data (e.g., CMS).
AHRQ should begin presenting cost and efficiency measures in both the NHDR and the NHQR and stratify these measures in ways that illustrate the variation in care for different populations. Such stratification will be useful because significant differences have been found regarding the use of health care services among different populations. For example, non-Hispanic Whites have high rates of overuse for many procedures including coronary revascularization (Epstein et al., 2003), typanostomy tubes (Kogan et al., 2000), and use of antibiotics (Gonzales et al., 1997). In comparison, African Americans, and in some cases Latinos, have higher rates for theoretically avoidable procedures (e.g., treatment of late stage cancer, limb amputations) and inappropriate use of emergency department visits and hospitalizations for avoidable conditions (potentially due to neglected prevention screenings and disease management) (Fiscella, 2007; Shavers et al., 2009). Reporting such findings is informative for promoting more targeted quality and disparities interventions.
The IOM has previously identified care coordination as 1 of its 20 national priorities for improving quality and as a primary area for performance measurement (IOM, 2003a, 2006). Other organizations, including CMS, WHO, The Commonwealth Fund, NQF, and the NPP, have also identified care coordination as a valuable component for enhancing health care delivery and patient experiences (CMS, 2008b; The Commonwealth Fund Commission on a High Performance Health System, 2006; NPP, 2008; NQF, 2009; WHO, 2008). Increasing evidence shows that fragmented or uncoordinated care often hinders optimal patient care. Suboptimal care coordination can refer to poor transitions at hospital discharge (Coleman et al., 2007), inadequate reconciliation of medications (NPP, 2008), and inadequate communication between primary care physicians, specialists, and other health care providers that can lead to contradictory messages or instructions for patient care. These gaps contribute to errors, adverse events, and avoidable costs including avoidable hospitalizations and unnecessary duplication of tests and procedures (Bodenheimer, 2008; Epstein, 2009; Wolff et al., 2002). Consequently, the committee feels that care coordination is sufficiently important for providing quality care, and highlights it as a separate framework component.
In a well-coordinated system, information for decision-making and care provision is shared across providers and settings so that integrated and well-communicated care occurs seamlessly throughout a patient's care experience (AHRQ, 2007). Efforts to coordinate care occur within a variety of health care environments (including across public and private sectors) and aim to improve patient outcomes and reduce health care spending (AHRQ, 2007). Care coordination programs have been found to reduce readmissions in hospitals, increase length of time between discharge and readmission, improve patient and caregiver communication (AHRQ, 2007; Naylor et al., 2004), improve patient satisfaction with care received (Neumeyer-Gromen et al., 2004), and improve health outcomes (Foy et al., 2010; Peikes et al., 2009; Wadhwa and Lavizzo-Mourey, 1999). Most of these programs address complex chronic diseases and aim to reduce the costs associated with these conditions (CMS, 2009a). That said, there is reason to believe that care coordination can benefit all populations and individuals (Starfield and Shi, 2004).
Involving the patient in information exchanges and decision-making is another aspect:
Care coordination is a function that helps ensure that the patient's needs and preferences for health services and information sharing across people, functions, and sites are met over time. Coordination maximizes the value of services delivered to patients by facilitating beneficial, efficient, safe, and high-quality patient experiences and improved healthcare outcomes. (NQF, 2006)
Care coordination's importance as an emerging area for measurement is further supported by its potential to reduce costs to the health care system. While evidence of cost reduction is mixed, in some instances, increasing the integration of services or coordination of care among multiple settings has been demonstrated to be cost-effective (AHRQ, 2007; Choudhry et al., 2007; Neumeyer-Gromen et al., 2004; Peikes et al., 2009; Smith et al., 2007). Such findings are particularly significant given that 10 percent of individuals in the U.S. account for 70 percent of total health care expenditures (Monheit, 2003). In other instances, however, care coordination programs have not been shown to provide any cost savings (Peikes et al., 2009; Wadhwa and Lavizzo-Mourey, 1999).
In spite of the mixed empirical findings, the committee believes that care coordination, because of the potential to improve health outcomes and patient experiences and lower costs, is an important foundational element of quality across the spectrum of care and contributes to each of the other care components (e.g., effectiveness, safety, patient-centeredness). Thus, care coordination should be monitored through reporting in the NHQR and NHDR.
Although AHRQ is expected to report on a number of care coordination measures in the 2009 NHQR and NHDR, some measures are intended to appear only once due to limitations in AHRQ's data sources. Among those measures being planned for reporting include: integration of information (receipt of test results, doctor with information about care from specialists, and other providers in practice with enough information about an individual to provide care), transitions of care (complete written discharge instructions, inadequate discharge information), and perception of care coordination.5 Reporting of these measures is an improvement, and the committee believes that AHRQ should continue to report care coordination measures in future reports, giving the topic appropriate attention in a separate chapter. Additional suggestions for reporting care coordination measures can be found in Appendix D.
Capabilities of Health Systems Infrastructure
Ensuring well-coordinated, high-quality health care requires supportive systems infrastructure. Such an infrastructure means having information systems in place for data collection, quality improvement analysis, and clinical communication support. Additionally, systems infrastructure includes having an adequate and well-distributed workforce in place, and the organizational capacity to support emerging models of care, cultural competence services, and ongoing improvement efforts. Adequate systems infrastructure for various care models helps promote and sustain performance improvement and has the potential to increase system efficiency by streamlining administrative, operational, and clinical processes, and reducing duplication of work (Bodenheimer and Grumbach, 2003; Bodenheimer et al., 2002, 2009; Grumbach, 2003; Grumbach and Bodenheimer, 2004). Conversely, a lack of system capabilities can disadvantage specific populations (e.g., rural populations with fewer available health care professionals, minority populations served by providers without health information technology [HIT] support).
Because many of the performance measures for infrastructure capabilities are still developing, the committee encourages further investigation and evaluation of measures in this area. Among the infrastructure capabilities that could be further evaluated for reporting in the national health care reports are care management processes, the adoption and use of Health IT, workforce distribution, and the relevance of these capabilities to disparity populations.
Integrated Delivery Systems
Growing evidence highlights the benefits of integrated delivery systems on system efficiency and patient outcomes (Bradley et al, 2005; Coleman et al., 2009; Enthoven, 2009), including integrated systems that promote cultures of safety and team-based practices (Shortell et al., 2004; Singer et al., 2009). Examples of effective integrative models of care include the patient centered medical home (PCMH) and Wagner's chronic care model (CCM). Each promotes the collaboration of various health care professionals, within and across settings, to provide continuous, patient-centered care. A PCMH is defined as "a team-based model of care led by a personal [primary care] physician who provides continuous and coordinated care throughout a patient's lifetime to maximize health outcomes"(American College of Physicians, 2010). Medical homes enhance access to care through "open scheduling, expanded hours, and new options for communication between patients, their personal physician, and practice staff"(The Patient-Centered Primary Care Collaborative, 2007). One study has shown that disparities in health care quality can be diminished or eliminated through clinical practices (e.g., ready telephone access, availability of medical care or advice on weekends/evenings, organized and timely office visits, provider continuity) (Beal et al., 2007). Similarly, the CCM promotes health systems whose structure enables community-based services, support for self-management of care, information support systems, and delivery system design (Improving Chronic Care Illness, 2010). The element of information support systems is particularly important for these models of care, as it provides continuity in patient records and clinician communication.
As the future of health care becomes more electronically driven, adequate Health IT systems will increasingly provide a foundation for tracking quality improvement in care delivery and patient outcomes. Although the adoption of an Health IT system is no guarantee of better health care quality outcomes, and more work is needed to determine its impact on quality improvement, the increased and more efficient use of Health IT can make available additional sources of valuable data on clinical outcomes (Arrow et al., 2009). Appropriate Health IT represents a supportive foundation for new health care models (e.g., the CCM and PCMH) and payment reforms (e.g., pay-for-performance and value-based purchasing) (Bodenheimer et al., 2002; The Patient-Centered Primary Care Collaborative, 2007). The adoption and use of Health IT as a tool to manage costs and improve the quality of care delivered (Balfour et al., 2009) has been shown to help reduce medical errors and adverse events, enable better documentation and file organization, provide patients with information that assists their adherence to medication regimens and scheduled appointments, and assist doctors in tracking their treatment protocol (Balfour et al., 2009; Herzer and Seshamani, 2009; Keenan et al., 2006; Keyhani et al., 2008; O'Connell et al., 2004). The committee recognizes, though, that not all aspects of Health IT adoption have resulted in positive effects. Systems that integrate poorly with other information systems may be more time-consuming to use or may unnecessarily duplicate efforts (Campbell et al., 2006). Many electronic health record (EHR) systems that providers currently use have little, if any, interoperability with one another, creating an inability to share information between providers (Improving Chronic Care Illness, 2010). Furthermore, heavy reliance on these systems may affect general provider communication skills and the occurrence of face-to-face interactions among clinicians or with their patients (Ash et al., 2007).
The proposed requirements for receiving incentive payments under the HITECH Act include the collection and reporting of race, ethnicity, and language data for at least 80 percent of Medicare or Medicaid patients seen by that hospital or provider (CMS, 2009b, pp. 50, 55, 69, 77-78). Each hospital or provider seeking a HITECH incentive payment will have to provide patient quality data stratified by race, ethnicity, and language (CMS, 2009b, pp. 52, 56, 83). As these hospitals and providers implement Health IT systems, and as states build health information exchanges to share these data, the nation's overall capacity for quality data collection and reporting by race, ethnicity and language will dramatically increase. These additional data will provide a stronger basis for identifying cultural competence needs and other disparity gaps. Such endeavors highlight the advantage of having solid infrastructure capabilities from which the national healthcare reports will likely benefit.
The adoption of Health IT in the United States is relatively low. Evidence suggests that only 17 percent of physicians in ambulatory care environments have EHR access (RWJF et al., 2008), and a study of acute care hospitals shows that only 1.5 percent of those surveyed have a comprehensive EHR system (i.e., present in all clinical units) (Jha et al., 2009). Nevertheless, the need to establish such systems has gathered momentum from the HITECH portion of the American Recovery and Reinvestment Act of 2009 (ARRA).6 In 2009, the federal government invested $49 billion for Health IT, most specifically for EHRs, but also for e-prescribing, quality reporting, and health information exchange (Chang, 2009). HITECH, which focuses on quality, promotes Health IT as a means to improve health outcomes and efficiency of health care systems (Blumenthal, 2009).
Health Care Workforce
The health care workforce is another aspect of systems infrastructure on which the NHQR and NHDR should report. Considered the backbone of the health care system, the workforce is comprised of all health care providers, from physicians, dentists, and nurses, to laboratory and pharmacy technicians, to nursing home staff. Ensuring a sufficient number of providers is important for the health care delivery system and can be an indicator of the quality of care delivered. For example, Cooper examined the supply of physicians in various states relative to reported state rankings of quality care and found that the total supply of physicians (both specialists and primary care) was associated with the quality of care delivered (Cooper, 2009b). Other studies demonstrate that the ratio between nurses and patients in a given organization can also impact the quality of care delivered (Gordon et al., 2008; Kane et al., 2007; Needleman et al., 2002). Currently, staffing shortages are a concern for several physician specialties (AMA, 2009; IOM, 2008), nurses (Gerson et al., 2005), and other health care professionals (HRSA, 2009). Ensuring a large enough and appropriately distributed workforce to respond to expected increases in patient demand (IOM, 2008) will be an important task.
The implications of shortages are illustrated by recent data that indicate access to primary care has been declining, in part due to an emerging primary care workforce shortage (Bodenheimer et al., 2007). A significant number of primary care physicians cannot and will not accept new patients (CDC, 2007). Combined with the aging of the baby boomer population, shortages of primary care and other health care professionals are expected to remain in many areas of the country (IOM, 2008).
The ratio of providers per 100,000 has been widely used to estimate provider shortages in geographic areas (HRSA, 2009). For example, there are numerous parts of the United States that are designated as medically underserved areas or populations (MUA/Ps) as well as areas designated as health professional shortage areas (HPSAs) (HRSA, 2009), where the distribution of health care professionals or sites available to serve populations are lower than what is recommended. Southern and mid-western states tend to have the highest number of HPSAs compared to other regions of the country (HRSA, 2010). Reporting some data on these designated underserved areas, perhaps at the state level, may help inform where additional action could be taken to improve delivery of or access to care. Furthermore, analyzing these data in conjunction with information on receipt of health care services and patient outcomes would be the type of informative analyses that the Future Directions committee would like to see provided in the NHQR and NHDR. AHRQ could provide other assessments of availability for various types of health care professionals to better inform this issue, including but not limited to information on primary care and specialist physicians, nurses, mental health, and dental care professionals.
Infrastructure to Support Access and Utilization
The significance of health infrastructure capabilities can be of particular importance for underserved areas and priority populations. Appropriate information systems and a well-trained workforce are key elements for providing access to needed care in the form of enabling services, such as patient outreach, patient navigation services, and training in cultural competence (Fiscella, 2007; HRSA, 2007; Ro et al., 2003).
Knowledge of a patient or group's language and cultural needs better equips providers to deliver high-quality care and communicate effectively with patients. Currently, there are limited national data regarding linguistic competency (among providers or patients) or the use of various interpreter services (e.g., in-person translation, telephonic and video health care interpretation, translation of documents). Yet, the presence of these services in health care settings will be increasingly important as the population of the United States increases in diversity and potentially includes more individuals with limited English proficiency (Betancourt et al., 2005; Moreno et al., 2009). For example, with the U.S. Hispanic population projected to comprise 30 percent of the population by 2050 (U.S. Census Bureau, 2008), reporting on measures that capture the persistence of linguistic barriers will be important.
AHRQ currently reports data on a measure of workforce diversity that reflects the racial and ethnic make-up of reported registered nurses, licensed practical nurses, and licensed vocational nurses in the United States, and another measure on the availability of language assistance at the usual source of care for limited English-proficient adults (AHRQ, 2009a). Striving to report these data at the state level would be even more informative. AHRQ might also further analyze the data reported for the workforce diversity and language assistance measures to determine whether the individuals who receive culturally or linguistically competent care have better outcomes. Since the purpose of performance reporting is to inform how to improve quality care and patient outcomes, analysis that can illuminate these findings will be beneficial.
4 This committee has been informed by AHRQ staff that the 2009 NHQR will include data on insurance and underinsurance status.
5 Personal communication, Ernest Moy, Agency for Healthcare Research and Quality, October 13, 2009.
6 American Recovery and Reinvestment Act of 2009, Public Law 111-5 §4101, 111th Cong., 1st sess. (February 17, 2009).