EXECUTIVE SUMMARY
Before the mid-1990s, Italians relied upon acute care hospitals to solve a wide range of problems related to the assistance of the elderly. As a result, in 1992 the average length of stay in Italian internal medicine wards was about 18 days, mostly due to the elderly not being discharged.
In the early 1990s, the Italian National Health Service (NHS) underwent a series of reforms. One reform divided the nation into approximately 250 local Health Agencies, each further subdivided into districts; the Health Agencies' general directors (CEOs) received the mandate to adopt a "corporate mentality" even though the enterprise was state-owned. Another reform led to the introduction of Diagnosis-Related Groups (DRGs), which dramatically changed the approach to the long-term care of frail elderly patients.
The aging of Italy's population and the rising number of Health Agency clients, along with a new prospective payment system and the NHS's corporate philosophy, pushed the local Health Agency CEOs toward redesigning the long-term care system and shifting resources from the hospital to the community.
In this context, the Institute of Internal Medicine and Geriatrics of the Catholic University of Rome, a well-known research institution, demonstrated that community-based care could be improved by training personnel in multidimensional geriatric assessment and by integrating services via case management.
At the same time, Pfizer Italy was updating its mission, deciding to target local health authorities and Health Agency CEOs. A new partnership arose between Pfizer and Catholic University that allowed the pinpointing of the Health Agencies' primary needs and promoted a community care model of services based on case management and a second-generation assessment tool, the Resident Assessment Instrument-Home Care (RAI-HC). The Catholic University group was responsible for the scientific and educational leadership of the project, dubbed the Silver Network, over its whole duration. (The project has been going on since 1997.)
Thirty-two Health Agencies began to work according to the agreed model; to date, 25 (10 percent of the total number of Health Agencies) continue to do so.
The intervention produced a change in the relationships between professionals, who had been accustomed to working by themselves without sharing information, as well as a change in the flow of information among the professionals, making it more logical. A second outcome was a reduction in the number of admissions to acute care hospitals among clients followed by case management, thus reducing the costs of care.
In general, Health Agency CEOs and policymakers immediately understood that the RAI-HC/case management system was effective and could provide data useful for epidemiological, administrative, and management purposes. They were easily convinced by the evidence-based data produced by intervention research carried out using RAI-HC as the assessment tool. There were some difficulties, however. Because people working in geriatric care were, before the RAI-HC's introduction, almost entirely ignorant of multidimensional assessment techniques, various professionals had a hard time accepting the rationale of the new system. Once the initial difficulties were overcome, however, the professionals responded extremely positively to the assessment tool and the case management system.
The future course of the RAI-HC in Italy has three aspects: legislative/regulatory, administrative, and research-related.
On the legislative/regulatory front, the regions of Italy will in the future have a tool to use for accrediting home care programs run either by the Health Agencies or by private cooperatives. On the administrative level, the Health Agencies can justify and standardize the costs of home care services through the use of an MDS-HC database now under development. That database will serve as the launch site for exploring the RAI-HC's research potential. As shown by experiences elsewhere, databases generated by comprehensive geriatric assessment instruments can provide valuable information about a population usually excluded from mainstream research.
RAI-HC is, beyond doubt, a good assessment tool. Nevertheless, nothing is sold without sellers. Pfizer's people did what the Italian representatives of interRAI and their coworkers could never have done: they went to the Health Agencies and directly promoted the product. Without their efforts, very few Health Agencies would currently be using the RAI-HC.
A general problem remains, however, regarding how to disseminate tools, procedures, and processes that can improve present policies and behaviors on such huge issues as frailty and long-term care. First of all, policies need to be evidence-based, and in this field evidence is very difficult to obtain. Even when the evidence is available, as is the case with the RAI-HC, a tool's success among potential users is not assured. Presenting this story of RAI-HC implementation in Italy is one way of addressing the problem.
BACKGROUND
Italy has one of the world's fastest-growing elderly populations, and, as in all western countries, growth is most rapid among the oldest old. Curiously, however, in the early to mid-1990s, Italy ranked lowest among western nations in the number of residential beds for the elderly: about 22 beds per 1,000 people over 65. (No other western country had fewer than 60.) The numbers of elderly receiving professional care services at home were even lower, accounting for less than 1 percent of those over 65 years of age, with higher rates in a few northern regions (ISTAT 1997; Lopez 1987; Lori, Golini, and Cantalini 1995).
The low number of nursing home beds and the shortage of home care services were due both to the typical Italian family structure and to the widespread use of acute care hospitals for the care of the sick elderly. Italian families, whether because of choice or of lack of opportunities, traditionally take responsibility for the care of their elderly members, even when the elderly person is very frail. In southern Italy, this tradition is central to a culture that views care of the elderly as a natural responsibility of the family.
Moreover, prior to the mid-1990s, Italians relied upon acute care hospitals to solve a wide range of problems related to the assistance of elderly people. As a result, the length of stay in Italian internal medicine wards in 1992 averaged around 18 days, mostly due to the elderly not being discharged. This high use of inpatient beds (which in those years numbered 7 per 1,000 inhabitants) was not consistent throughout the country, however. There were marked differences between the north and the south. A study sponsored by the National Research Council in the late 1980s showed that in-hospital mortality was six times higher in hospitals located in the north than in hospitals in the south, even adjusting for disease and clinical severity. Southern Italian families (which tend to be larger, with more relatives available for elder care, and with a higher level of unemployment, so therefore also more available for caregiving) have more time and emotional resources to manage the death of a relative, and they often ask that a relative be discharged when the prognosis is terminal. In contrast, in the north, families often leave the management of a terminal patient and the subsequent death to the hospital.
Before the mid-1990s, the only state directives on the care of the elderly were those included in the so-called Progetto Obiettivo Anziani (Targeted Program on Aging), which invited all local health authorities to provide the elderly with nursing homes and home care programs supervised by a Geriatric Evaluation and Management Unit (GEMU). Not many authorities followed those suggestions.
In the early 1990s, the National Health Service (NHS) underwent a series of reforms. One reform divided the nation into approximately 250 local Health Agencies, each further subdivided into districts; the Health Agencies' general directors (CEOs) received the mandate to adopt a "corporate mentality" even though the enterprise was state-owned. In the 1990s, the key word of the reform process was aziendalizzazione, which denoted the process of transforming local health care providers into private enterprises that, like private companies, would follow rules of corporate management and engage in competition. Another reform led to the introduction of something that dramatically changed the approach to frail elderly patients in need of long-term care: in 1993, hospital reimbursement shifted from a fee-for-service system to a prospective system based on the Diagnosis-Related Groups (DRGs).
The aging of the Italian population, the rising number of "clients" of the Health Agencies,1 this new prospective payment system, and the corporate philosophy were the driving forces that pushed the CEOs of the local Health Agencies toward redesigning the long-term care system and shifting resources from the hospital to the community.
The Health Agencies and their CEOs immediately had to attempt to rectify the poor geriatric training of the professionals who were to be involved in this redesign of long-term care (LTC). At the end of the 1980s, medical schools (in particular, these schools' residency programs in geriatrics) were the only Italian institutions offering appropriate training in the management of geriatric patients and the care of the frail elderly. While these programs were able to provide good preparation in the use of geriatric assessment and drugs and in the treatment of syndromes related to the aging process, they were not ready to help design the new health care models needed to change the structure of the NHS.
Other schools training professionals who participate in the care of the elderly (e.g., nurses, rehabilitation therapists, social workers) were not prepared even to give the minimal training provided by the medical schools. For example, the concepts of geriatric assessment accepted by the medical schools in the mid-1980s were not yet being taught in nursing schools ten years later.
At that time, only partial and "local" solutions for managing the care of the frail elderly were available, because there was no law indicating the direction of LTC or the practical terms for its reorganization. But no one was worried about this situation, and some probably preferred it, feeling that the vagueness of the situation allowed them a personal, "creative" approach in experimenting with new care models and strategies. This "individualistic" approach applied even to gigantic problems, such as the redesign of LTC. Some Health Agencies opened home care programs (some in cooperation with the local municipalities). Others opened nursing homes under the direction of general practitioners (GPs), while others hired geriatricians to run such facilities. Some Health Agencies created a GEMU with a GP as chief; still others put a pediatrician in place as chief; and yet others even chose a lawyer. No Health Agency, however, instituted anything that might properly be called a network of care.
In this context, the Institute of Internal Medicine and Geriatrics of the Catholic University of Rome, a well-known geriatrics research institution, sought to demonstrate that community-based care could be improved through the following two activities:
The Catholic University research team performed a two-part test of its proposal in Rovereto, a town in northern Italy with a population of 35,000 (the size of a typical health district). The local Health Agency had considerable experience with community-based care, although it was not using a comprehensive multidimensional assessment tool and did not integrate its services with the social services delivered by the municipality. In 1995, the Ministry of Scientific Research provided funding for the research team to work with the Rovereto Health Agency and town government to accomplish the following:
The educational section of the program was based on the Resident Assessment Instrument (RAI) for nursing homes. Although this instrument was designed for nursing home residents, the research team felt that it, because of its comprehensiveness, would best acquaint personnel with geriatric assessment. Also, the care-planning facilitating tools embedded in the RAI (the Resident Assessment Protocols, or RAPs) were particularly useful in highlighting and summarizing problems that are common in the frail elderly (incontinence, pressure ulcers, cognitive and physical impairment, etc.).
The results of this project were impressive. The GEMU personnel in Rovereto—a total of 25 nurses, physical therapists, and social workers—easily understood the technique as well as the philosophy underlying it. The two nurses who became case managers perceived their new role as a professional advancement—a step above their traditional role as physicians' assistants.
Furthermore, training in the RAI's multidimensional approach to assessment provided the nurses with a method for identifying problems that was much better than the vague, not fully justified process used before. The study's results, including the participating professionals' satisfaction with this educational intervention, were published in a series of papers on gerontological education (Cipriani et al. 1995). Finally, the RAI system, with its comprehensiveness, triggering process, and RAP-based procedures for designing a care plan, promoted a team approach to care, unusual in such an individualistic country as Italy.
After appointing the case managers, the research team proceeded to randomize clients into two groups of 100 each. The case managers and the GEMU followed one group; the other—the controls—continued to receive the traditional kind of home care. The research team chose an existing comprehensive, multidimensional assessment form—one developed to determine eligibility for long-term care services in British Columbia, Canada—as the basis for assessment of all participants during the baseline and one-year evaluations.
The study revealed that, after one year, clients being served by the case-management approach showed greater improvement than the controls in physical and cognitive function and mood status and were admitted to institutions (acute care hospitals or nursing homes) much less frequently than their counterparts.
The integrated care model also achieved savings of approximately US$1,000 per year per client. Keeping in mind that this was the amount of money given each year by the NHS to each Health Agency for the comprehensive health care needs of each citizen, this saving was significant (Bernabei, Landi, et al. 1998).
This study generated two important pieces of information useful for Health Agencies throughout Italy. First, the personnel who deliver care to the elderly can be trained to understand and use geriatrics "skills"—that is, the multidimensional assessment and the team approach. Second, a new, validated model of integrated care—the case management approach—was available for organizing LTC for elderly people in the community.
OUTREACH
In 1997, the interRAI group developed a new instrument, the RAI-Home Care (RAI-HC) (Mor et al. 1997; Morris, Fries, Bernabei et al. 1996; Morris et al. 1997; Phillips et al. 1997). One of the authors of the present paper, Roberto Bernabei, served on the RAI-HC overview committee and was the leader of the Catholic University geriatric research team. The availability of the new instrument, along with the positive results of the Rovereto study, allowed the Catholic University research team to ask the Italian Research Council (Consiglio Nazionale delle Ricerche, or CNR) for an extension of their research project into the use of case management, now to include the use of the RAI-HC. The CNR funded this proposal with a dedicated grant in 1997.
The research experimentation initially proceeded slowly, due to scant resources (the total amount of the grants for the Rovereto study and from the CNR did not exceed US$100,000) and the lack of any political mandate to implement the new system in any other Health Agency. Even in Rovereto, the new system faced difficulty: a new political coalition took over the government of the municipality and immediately decided to stop the case management-based home care program and return to the previous model. The old custom of finding local solutions for elderly care, even against the evidence, was clearly winning in Rovereto and around the country.
The change in this situation came from an unforeseen source, and the consequences were far-reaching. At the time, Pfizer Italy was updating its mission in the pharmaceutical industry, which had traditionally focused on research and development and the marketing of drugs. Pfizer had always been concerned with the community and the general practitioner, and the company had been the first to use the logic of customer care as the basis for a partnership with, rather than a sponsorship of, patients. Against this background, Pfizer decided to create a department devoted to the implementation of innovative projects in health care. These projects would alleviate both the financial and the organizational burdens of new models of care and would introduce technology to support the integration of care. Traditionally, the pharmaceutical industry's main customers had been physicians, but in the new health care environment they would not remain the only decision makers, because access to health care services would depend more and more on economics. For this reason, Pfizer decided to find ways to start targeting local health care authorities and Health Agency CEOs.
The company thus initiated an intervention to support an improvement of services delivered at the community level, using the logic of quality, effectiveness, and efficacy that are typical in managed care. For this intervention, Pfizer chose the services directed at the elderly population—specifically, home care services—because this segment of the population was increasing at a very fast rate, making Italy the "oldest" nation in the world. As a first step, Pfizer Italy, in collaboration with social researchers from the Center for Social Studies and Policies (Centro Studi Investimiento Sociale, or CENSIS) and epidemiologists from the University of Rome-Tor Vergata, analyzed the realities of home care in Italy. The results were surprising. Only 15 percent of those receiving a service in the community were satisfied with it. Only 118 out of 228 Health Agencies had started experimental home care programs, and most of these were not integrated with social services (as had been seen in Rovereto). Further, medical/nursing and social service home care programs were scarce, and there were no common standards to determine eligibility, oversee management, or evaluate outcomes in the care of elderly patients.
In addition, there had been a decline in the number of inappropriate hospital admissions and a dramatic reduction in lengths of stay, both consequences of the introduction of the prospective payment system based on the DRGs. The effects of the prospective payment system on elderly care were particularly pronounced because there had been no increase in rehabilitation, postacute, and home care services needed to guarantee continuity of care after hospital discharge, even as hospitals could no longer provide a buffer for the paucity of the long-term and community services for the elderly. Moreover, such services as did exist were not satisfactory, either qualitatively or quantitatively. The implementation of the DRG-based prospective payment system had, in fact, worsened the inadequacy of the entire elderly care system.
The Pfizer/CENSIS/Tor Vergata analysis then proceeded to map the Geriatric Evaluation and Management Units in Italy. The goal was to identify those Health Agencies that were in particular need of integrated home care services but lacked the scientific, technical, and educational support to implement such programs. Marina Panfilo, Pfizer's national key account manager and one of the authors of the present paper, contacted many research groups to find existing initiatives that could provide models for the needed changes. Based on its successful experience on the Rovereto study, the Catholic University group was chosen to be Pfizer's partner in this initiative.
The partnership between Pfizer and Catholic University pinpointed the primary needs of the Health Agencies: a model for community care services and an assessment tool. The model was the case management approach already tested in Rovereto. The assessment tool would be the RAI-HC, because it could provide a complete and multidimensional assessment for the identification of patients' problems and serve as the basis for creating individualized care plans. This validated tool was already available, but it was necessary to translate it, to create data-entry software to store the information, and to provide a teaching program on its use. Pfizer obtained an interRAI license to use the RAI-HC in Italy and financed its translation, computerization, and implementation. The implementation was made possible through training courses given by the Catholic University team and, as explained below, through Pfizer's network of people. By mutual agreement between Pfizer and the Catholic University, this tool would also to be used to assess resources utilization, so that the cost-effectiveness of the intervention could be evaluated (Morris et al. 1996).
The Catholic University group was responsible for the scientific and educational leadership of the project (by that time dubbed the Silver Network) for the entire duration of the project, which has been going on since 1997 (Landi, Lattanzio, et al. 1997). The university developed and provided a standardized program for a six-day training course (Landi et al. 1996). The program was organized as follows:
This outline is reported in full because it explains how it has been possible to effectively communicate the core elements of the geriatric assessment tool to those involved in providing care to the elderly. The Catholic University team considered the RAI-HC instruction manual an outstanding geriatric-gerontological textbook because of its explanation of the CAPs (analogous to the RAPs in the nursing home RAI) and because it provides the capability to identify a problem using just the triggers embedded in the assessment form. This intensive course gives personnel a level of knowledge that they could not achieve so rapidly and efficiently in any other way (Morris et al. 1996).
With Pfizer's support, the team drew up a plan to inspire Italian Health Agencies to train their professionals in the use of the RAI-HC and to implement it as part of an entirely new home care service program that would include including case management. Pfizer's ability to communicate this new model to Health Agency decision makers was the crucial element in this process. A group of 15 persons from Pfizer, specifically trained in public health and health economics, had the mission of creating the best possible environment for its new-drugs market and of disseminating a disease-management culture in which drugs would play a significant role. The Pfizer group contacted Health Agency CEOs and tried to create a bridge between specialists in geriatrics, general practitioners, nurses, social workers, and health care managers. This bridge rested on two pillars: case management and the RAI-HC.
To be a suitable candidate for this plan, a Health Agency had to have—or had to be willing to put in place—a home care program with at least 15 nurses, thus providing a reasonable number of patients and the certainty that the Health Agency would be willing to redesign its LTC services. Pfizer's national key account manager (KAM) contacted those CEOs who expressed a willingness and ability to implement the new service, and a formal deal was worked out. Pfizer, using the Catholic University team, committed itself to training the personnel who would be involved in the Health Agency's home care program and to providing all the participants in the course with the RAI-HC instruction manual and the software to store the information; a new section on costs was added to the tool as well. In turn, Pfizer asked the Health Agency to commit itself to organizing its home care service according to the case management approach, using the training course as the occasion to select and train case managers. Finally, Pfizer and the Health Agency had to coordinate their work with local GPs, who provide the bulk of medical care to the elderly.
This last part of the deal was the most problematic because it touched on the GPs' role and prerogatives. In Italy as in many other countries, GPs are the cornerstone of the National Health Service (NHS). According to Italian law, the GP is solely responsible for the citizen's health, and in an individualistic country like Italy, he or she doesn't forget it! Italy's 55,000 registered GPs represent what is probably the country's single most powerful professional guild. Understandably, GPs do not easily accept what they consider intrusions into their domain.
The most recent contract between the GPs and the NHS specifies that an extra amount of money is to be paid for what are called "planned home visits" to elderly patients. According to many GPs, this additional remuneration allows them to meet the needs of the homebound elderly. This clause brought the GPs into direct conflict with the Health Agencies over the new approach being pushed by the Pfizer-Catholic University partnership. The new system allowed case managers to take care of their patients without (apparently) the direct involvement of the GP, and many general practitioners believed this to constitute a clear threat to their relationship with their patients.
The Catholic University team leader, Pfizer's national KAM, and the local KAM therefore met multiple times with the local GPs in each Health Agency to explain how the Silver Network would, in fact, be very useful to them. The Pfizer-Catholic University representatives emphasized the following points:
These conversations were only successful, however, in cases where the Health Agency CEOs were very committed to the effort and used all the leverage they had to persuade the reluctant GPs, by renegotiating the number of funded annual planned visits, for example, or by promising to pay any GP the money budgeted for one planned visit in exchange for that GP's participation in the GEMU meeting at which the care plan of his or her patients was discussed.
In the end, the team contacted 80 Health Agency CEOs, 35 of whom agreed to initiate a home care service program—or transform an existing home care program—according to the case management model, using RAI-HC as the assessment system. These 80 Health Agencies were the only ones with home care programs large enough to be worth reorganizing. The approximately 170 other Health Agencies totally lacked such programs or provided home care services to very few patients. To date, 32 Health Agencies have begun implementing the new model, of which 25 are still in operation. We estimate that, at present, about 10 percent of Italian Health Agencies are using the same model for caring for the frail elderly—a model that is based on case management and that uses the RAI-HC.
OUTCOMES
The potential effects of the implementation of the RAI-HC system on care planning are summarized in tables 1 and 2; these outcomes were initially conceptualized by one of the case managers trained by the Catholic University team, who worked full-time in the Vittorio Veneto Health Agency (Landi, Onder, Russo, et al. 2001).
The tables show the methodological revolution that follows the implementation of a comprehensive, care planning-oriented tool such as the RAI-HC. This revolution is similar in concept to the modifications seen in industry when a new technology is introduced and the traditional process of production is changed. It is not by chance that some have defined comprehensive geriatric assessment (CGA) as the "technology" of geriatrics (Rubenstein, Wieland, and Bernabei 1995). Because it is more complex than traditional tools, and probably also because, in Italy, the assessment instrument was being used by people who had no previous knowledge of this kind of technology, the RAI-HC produced true innovation.
The first set of outcomes concerned the relationships between professionals in the field of elderly care. Before the RAI-HC's implementation, these professionals had been accustomed to working by themselves, without sharing information. Now they worked in teams. The flow of information among professionals also changed, becoming more logical. The implementation also altered the role of the nurse. The new case manager who prepared the two tables reproduced here was taking on a role previously unknown among Italian nurses: that of following a screening and diagnostic process to identify the causes of the problems reported by the client.
Until the introduction of the RAI-HC, too much was left to the physician, while other health care professionals played only a marginal role, at most that of executing the physician's orders. Today, the case managers are able, for instance, to precisely detect nutritional problems in their clients because the assessment lets them know if a client has lost more than 5 percent of body weight over the last 30 days (item L1a=1, which triggers the CAP on nutrition). Further, the nutrition CAP itself helps them rule out causal hypotheses of, say, multiple drug regimen, or terminal illness, or depression. Once this diagnostic process has been accomplished, the case manager can interact with the physician to discuss and together develop an appropriate care plan that can include both medical and social interventions. Before the introduction of RAI-HC, all the nurse could do was to relate a subjective impression of a weight loss, without any validation or standardization. As a consequence, the physician was less likely to take the communication into serious consideration, and might just order some examination or diet that the nurse would arrange without any involvement in the decision making process.
The second set of outcomes of the RAI-HC/case management system's implementation has to do with the client. Two papers describe the client-related outcomes: one describes what occurred in Rovereto, which was the first Health Agency to implement the new system (Landi, Gambassi, et al. 1999); the other documents the activities of the four Health Agencies that were next to initiate home care programs based on this approach (Landi, Onder, Russo, et al. 2001). In all these experiences, admissions to acute care hospitals fell for clients followed by case management. In addition, when clients were admitted to the hospital, they were discharged more quickly. All this was accomplished without a change in overall mortality. When clients were able to stay at home, there was a substantial improvement in their quality of life.
Another outcome relates not just to the reduction in hospitalization rates and lengths of stay but to the associated savings realized by the Health Agencies. One of the authors of this paper, Giuseppe Panio, was general director of the Health Agency of Venosa (in the south of Italy) at the time of the home care implementation there. As CEO, he had seen the financial problems caused when too many older citizens belonging to his Health Agency were admitted to hospitals of other Health Agencies. His agency was responsible for paying for these admissions, and the data generated by the intervention convinced him that the case management approach could reduce these costs. He decided to implement a Silver Network site. After a single year, hospital admissions dropped from 15 to 5 percent, with corresponding money savings.
The Venosa experience is common to all the Health Agencies: hospital admission is very costly, and reducing costs related to hospital admissions is necessary to contain the budget. The dramatic worldwide increase in the size of the elderly population along with the concurrent pressure for cost containment efforts in the health sector necessitate a shift in resources from the hospital to the community, where costs are lower and problems related to frailty are also better managed. In Italy, this change was coincident with a reduction in the number of acute hospital beds in the country (now 6 per 1,000 population) and an increase in the number of people in home care. The Silver Network helps by producing original data that support Health Agency CEOs in making choices that permit this shift in resources.
SUSTAINABILITY
The future course of the RAI-HC in Italy has three aspects: legislative/regulatory, administrative, and research-related.
On the legislative/regulatory front, the regions of Italy will in the future have a tool to use for accrediting home care programs run either by the Health Agencies or by private cooperatives (the accreditation process is the regions' responsibility).
The RAI-HC system has been tested in one of the districts of the Health Agency of Bergamo, a medium-size, wealthy city in the north of Italy that is participating in the Silver Network. The study performed in Bergamo compared the results of the RAI-HC assessment (and the care plan and overall management that followed) with those obtained by other well-established tools, including the Barthel and the Lawton scales, which measure physical function; the Mini-Mental State Examination, which measures cognitive function; and the Geriatric Depression Scale, which measures mood status (Landi et al. 2000).
The comparison was made possible because another district of the same Health Agency had a case management-based home care program, but one that used these "traditional" scales. At the beginning of the study, all the clients involved had similar clinical and functional characteristics, but one year later, the clients assessed with the RAI-HC had higher levels of physical and cognitive functioning and had been hospitalized less frequently. This is probably due to the increased confidence of personnel using the RAI-HC, which better enables them to detect and treat clients' problems. Further, care plans triggered by the RAI-HC are more precise in targeting the problems of individual clients. This more accurate care results, in turn, in clients being able to delay or even avoid hospitalization. The paper presenting these results (Landi, Onder, Tua, et al. 2001) provided partial evidence regarding some of the characteristics that a home care program should have to perform at the best possible level. An evidence-based home care program should use RAI-HC as its assessment tool, and the published data now available support the region in requiring that the RAI-HC be implemented as a requisite for accreditation.
On the administrative level, the Health Agencies must justify and standardize the costs of home care services. This is problematic because these services are increasingly often provided by private cooperatives of nurses and physical therapists, which are reimbursed by the Health Agency. Under a grant from the Italian Ministry of Health, a group from the Istituto Nazionale Ricovero e Cura Anziani (INRCA) and the Catholic University's geriatric research team was able to validate the Resource Utilization Groups, version III (RUG-III), case-mix system in the context of the Italian NHS. This system measures the burden posed by the individual patient, in terms of minutes of care, on the nursing home and its staff and makes it possible to obtain the case-mix of a facility by averaging the complexity of the needs of all the patients in that facility. The RAI-HC includes the Home-care Utilization Groups (HUGs) algorithm, which is similar to RUG-III and allows the same kind of case-mix determination to be made about the clients of a home care program. We expect that, in the future, Health Agency administrators will use this case-mix measure to reimburse those who provide home care services. In addition, the HUGs can be used to define clients' characteristics in terms of resource consumption. By using either the HUGs or the scales embedded in RAI-HC it is also possible to "classify" the client and to make the delivery of services more rational with regard to personnel needed and procedures used. It is clear that this can only occur, however, if databases are created that permit analysis of the information collected with the multidimensional assessment tool in multiple settings throughout Italy.
Such databases would serve as the launch site for exploring the research potential of the RAI-HC. As experience elsewhere has shown, databases generated by comprehensive geriatric assessment instruments can provide valuable information on a population usually excluded from mainstream research. It is, for example, useful to recall the experience of the Systematic Assessment of Geriatric drug use via Epidemiology (SAGE). SAGE, which was assembled by Roberto Bernabei, working with Vincent Mor of Brown University, contains the nearly two million assessments performed over a five-year period in nursing homes in five U.S. states (Kansas, Maine, Mississippi, New York, and South Dakota), encompassing a total of nearly 750,000 residents (Landi, Onder, Tua, et al. 2001). The database also contains data from the U.S. Centers for Medicare & Medicaid Services (Medicare part B) on hospital admission and data from the OSCAR (Online Survey Certification and Reporting) survey on facilities' characteristics. This data set has made it possible to study a population with a mean age of 83 years—a group that is seldom included in traditional studies. Evidence-based medicine is, in fact, usually based on randomized clinical trials that exclude persons older than 75 years, either for logistical reasons or because of concerns about the presence of potential confounders due to comorbidity. Research conducted using the SAGE database revealed, for example, that pain suffered by nursing home residents with cancer often went untreated (Bernabei, Gambassi, et al. 1998), that ACE-inhibitors are more effective than digoxin in treating heart failure in people older than 80 (Gambassi et al. 2000), and that the first drug shown to be effective in the treatment of Alzheimer's disease, tacrine, was often incorrectly prescribed (Gifford et al. 1999). While the nursing home residents are the "extremes," the clients of a home care program are closer to the general elderly population. To gain information on this population, the Catholic University geriatric research team, thanks to a grant from the Ministry of Health, is collecting all the assessments performed at the sites participating to the Silver Network.
The first information forthcoming from this database will address measures of the quality of care delivered, using the indicators contained in the RAI-HC (use of psychotropic drugs, prevalence of incontinence, prevalence of mood disturbances, prevalence of pain, and so on). Afterward, it will be possible to start a continuing education effort for the professionals of the Silver Network home care program—one based on the actual problems detected by analyzing the data they provide. For example, if a given home care program shows a prevalence of incontinence much higher than that of other programs with similar case-mixes, and if the comparative data are presented to the people working in that program, it is likely that they will be better motivated to learn how to improve the management of incontinence.
It should be stressed that the database and continuing education effort just described are only possible when comparable data are obtained from the multiple programs, centers, and agencies providing integrated home care in different Health Agencies. This, in turn, requires that all work be performed in a standardized fashion, using the same assessment tool.
Following the same logic, it becomes possible to envision comparisons of very large scope, conducted at the European level. The European Commission's Fifth Framework Programme, "Quality of Life and Management of Living Resources," Key Action 6, "The Ageing Population and Disabilities," has just funded (2001—03) a 1.3 million euro grant coordinated by Roberto Bernabei and involving 11 countries (the Czech Republic, Denmark, Finland, France, Germany, Iceland, Italy, the Netherlands, Norway, Sweden, and the United Kingdom). The project, which goes by the acronym AD HOC (AgeD in HOme Care), aims to identify and propose a model of home care for the elderly through the analysis of the structural and organizational characteristics of multiple nations' home care services and the clinical and functional characteristics of their clients. The identification of the factors that can be correlated with positive patient outcomes at one year will contribute to the design of the first evidence-based home care service model. The project also aims to achieve several intermediate objectives: The AD HOC Project requires a considerable organizational and data-collecting effort—an effort that is made possible by the availability of a common assessment tool, the RAI-HC.
Finally, the Italian Society of Gerontology and Geriatrics (SIGG) has decided to support the adoption of the MDS family of instruments. This formal sponsorship has produced an agreement with the Association of Italian Directors of Nursing Homes (ANSDIPP) and Pfizer, Inc., which holds the RAI 2.0 license in Italy, to implement RAI 2.0 in 40 nursing homes throughout Italy. More than 500 nursing homes are members of the ANSDIPP, and the implementation in 40 nursing homes is an experiment to understand whether RAI 2.0 would be the appropriate common assessment tool for all the nursing homes that are members of the association. In the summer of 2002, the Ministry of Health, together with Pfizer, funded the construction of a database based on the RAI instruments for monitoring the quality of care for the elderly in acute and long-term care settings. Use of the RAI instruments in Italy will increase as additional ANSDIPP nursing homes participate, as the number of Silver Network home care agencies expands, and as the acute sector, including the Gruppo Italiano di Farmacoepidemiologia nell' Anziano (GIFA), which represents more than 50 internal medicine and geriatric wards located throughout the country, move into the RAI fold.
CONCLUDING OBSERVATIONS
The use of RAI-HC generated a major difficulty, mostly due to the almost total ignorance of the multidimensional assessment technique by the people working in geriatric care. Various professionals had a hard time accepting the rationale of such an assessment tool and system. The authors of this paper usually opened Silver Network preliminary training courses by asking if anyone in the audience was knowledgeable about the activities of daily living, which are the cornerstone of geriatric care. Many times, no one in the audience was able to respond positively. It is thus understandable that it was not easy for this new instrument or its cultural underpinnings to be easily accepted and disseminated. The instrument's complexity and the administrative time it consumes also motivated initial resistance to the implementation.
Once the initial difficulties were overcome, however, the professionals responded extremely positively to the new assessment tool and system. This enthusiasm helped personnel endure the long training courses and to complete the very long field assessments. (The necessity of a long training program and the time required for the assessments are the very factors that could hamper the use of this tool in countries such as the United States.)
Italian professionals have come to understand that the RAI-HC's usefulness does not stop at the assessment level and that, in fact, the assessment tool allows General practitioners' acceptance of the RAI-HC system was facilitated by their understanding that the nurse's role, though upgraded, would still be subordinate to the physician's. Formal documents from the Health Agencies were not sufficiently convincing, and so it was necessary to hold many meetings with GPs during which they could discuss their reticence regarding case management and the RAI-HC. This direct communication, as well as the CEOs' willingness to participate, fostered the GPs' acceptance.
Italian geriatricians' persistent refusal of the RAI system was probably due to their suspicion of the agreement between the Catholic University geriatric team and Pfizer. The geriatricians, and probably many other health care policymakers and providers, as well, felt that there may have been "dirty dealing" between the two, and that only the Catholic University team stood to gain from the arrangement. Aside from this, however, it remains the case that Italian geriatricians' acceptance of any common, standardized assessment system would be extraordinary. As we have said, the Italian mentality is very individualistic, and trained specialists use their own judgment and follow their personal intuition, often excluding what might be objectively desirable. (And, to be frank, the geriatrics field worldwide seems to be infected with this "individualistic virus.")
At the level of the CEOs of the Health Agencies and of policymakers in general, however, everything ran smoothly. The CEOs immediately understood that the RAI-HC was an effective system and could provide data useful for epidemiological, administrative, and management purposes. They were easily convinced by the data produced by the intervention research done with the RAI. Interestingly, they behaved as the MDs should behave, quickly coming to rely on a tool that produces evidence-based data. The RAI and case management system's full potential has not yet been exploited, however, because the construction of a database proceeds slowly, and CEOs are not yet able to envision how it might be used. Local policies have therefore not changed since the implementation of RAI-HC and the associated data collection. On this score, the Catholic University research team made a mistake by failing to supervise the local users' teams in the provision of accurate data collection within the specified reporting times.
The RAI-HC is, beyond doubt, a good assessment tool, and we believe that its value, as compared with older, more traditional assessments is equivalent, say, to the value of a CT scan over the traditional x-ray. Moreover, not only does it produce a much higher "image resolution" and much more sensitive and specific diagnosis, but it also reduces costs. Nevertheless, nothing is sold without sellers. The support of Pfizer, which provided a network that "sold" the RAI-HC, has been essential to the assessment tool's success in Italy. Pfizer's KAMs did what the Italian representatives of interRAI and their coworkers could never have done: they went to the Health Agencies and directly promoted the product. Without their efforts, only three or four Health Agencies would currently be using the RAI-HC, and then only as an experimental tool.
Still, the Italian system of elderly care faces a critical paradox: even though there is a great and widespread need for instruments such as the RAI-HC and the other RAI tools, implementation is so far much lower than expected and has encountered considerable resistance. Reasons for this resistance can be found in Italian care providers' tradition of individualism, in the autonomy of each region and the lack of a centralized mandate for RAI utilization, and in the suspicion provoked by the agreement between the Catholic University research team and Pfizer.
A general problem thus remains regarding how to disseminate tools, procedures, and processes that can improve present policies and behaviors on such huge issues as frailty and long-term care. First of all, policies need to be evidence-based, and in this field evidence is very difficult to obtain. Even when the evidence is available, as is the case with the RAI-HC, a tool's success among potential users is not assured. We believe that the creation of the database on elderly quality of care funded by the Italian Ministry of Health will help in promoting the RAI instruments' diffusion. In addition, this presentation of the story of RAI-HC implementation in Italy is one way of addressing the problem.
NOTES
1 In 1993, the number of Italians over 64 exceeded those under 15 for the first time (16 percent 64 years and older versus 15.2 percent from 0 to 14 years of age).
REFERENCES
Bernabei, R., G. Gambassi, K. Lapane, F. Landi, C. Gatsonis, R. Dunlop, L. Lipsitz, K. Steel, and V. Mor (SAGE Study Group). 1998. Management of Pain in Elderly Patients with Cancer: Systematic Assessment of Geriatric Drug Use via Epidemiology. Journal of the American Medical Association 279(23):1877–82.
Bernabei, R., F. Landi, G. Gambassi, A. Sgadari, G. Zuccalà, and V. Mor. 1998. Randomised Trial of Impact of Model of Integrated Care and Case Management for Older People Living in Community. British Medical Journal 316:1348–51.
Gambassi, G., J. Lapane, A. Sgadari, P.U. Carbonin, C. Gatsonis, L.A. Lipsitz, V. Mor, and R. Bernabei (SAGE Study Group). 2000. Effects of ACE-Inhibitors and Digoxin on Health Outcomes of Very Old Patients with Heart Failure. Archives of Internal Medicine 160:53–6.
Gifford, D.R., K. Lapane, G. Gambassi, F. Landi, V. Mor, R. Bernabei, and SAGE Study Group. 1999. Tacrine Use in Nursing Homes: Implications for Prescribing New Cholinesterase Inhibitors. Neurology 52:238–44.
Istituto Nazionale di Statistica (ISTAT). 1997. Some Demographic and Social Features of the Italian Elderly Population (trans. of Italian title). Press release, January.
Landi, F., A. Sgadari, L. Cipriani, M. Pahor, P.U. Carbonin, and R. Bernabei. 1996. A Simple Program to Train Case Managers in Community Elderly Care. Aging—Clinical and Experimental Research 8:211–8.
Landi, F., G. Gambassi, R. Pola, S. Tabaccanti, T. Cavinato, P.U. Carbonin, and R. Bernabei. 1999. Impact of Integrated Home Care Services on Hospital Use. Journal of the American Geriatrics Society 47(12):1430–4.
Landi, F., F. Lattanzio, G. Gambassi, G. Zuccalà, G. Sgadari, M. Panfilo, M.P. Ruffilli, and R. Bernabei. 1999. A Model for Integrated Home Care of Frail Older Patients: The Silver Network Project. Aging—Clinical and Experimental Research 11:262–72.
Landi, F., E. Tua, G. Onder, B. Carrara, A. Sgadari, C. Rinaldi, G. Gambassi, F. Lattanzio, and R. Bernabei (Silvernet-HC Study Group of Bergamo). 2000. Minimum Data Set for Home Care: A Valid Instrument to Assess Frail Older People Living in the Community. Medical Care 38(12):1184–90.
Landi, F., G. Onder, A. Russo, S. Tabaccanti, R. Rollo, S. Federici, E. Tua, M. Cesari, and R. Bernabei. 2001. A New Model of Integrated Home Care for the Elderly: Impact on Hospital Use. Journal of Clinical Epidemiology 54(9):968–70.
Landi, F., G. Onder, E. Tua, B. Carrara, G. Zuccalá, G. Gambassi, P. Carbonin, and R. Bernabei. 2001. Impact of a New Assessment System, the MDS-HC, on Function and Hospitalization of Homebound Older People: A Controlled Clinical Trial. Journal of the American Geriatrics Society 49(10):1288–93.
Lopez, A. 1987. The Impact of Demographic Trends on Health. World Health Statistics Quarterly 40:2–5.
Lori, A., A. Golini, and B. Cantalini. 1995. Atlante dell'invecchiamento della popolazione. Rome: Consiglio Nazionale per le Ricerche.
Mor, V., O. Intrator, B.E. Fries, C. Phillips, J. Teno, J. Hiris, C. Hawes, and J. Morris. 1997. Changes in Hospitalization Associated with Introducing the Resident Assessment Instrument. Journal of the American Geriatrics Society 45:1002–10.
Morris, J.N., B.E. Fries, R. Bernabei, N. Ikegami, R. Gilgen, K. Steel, and I. Carpenter. 1996. RAI-Home Care Assessment Manual. Washington, D.C.: InterRAI Corporation. Italian edition edited by R. Bernabei, F. Landi, F. Lattanzio, M.G. Di Niro, and L. Manigrasso published as Resident Assessment Instrument-VAOR (Valutazione Anziano Ospite di Residenza): Manuale d'istruzione (Rome: Pfizer Italia Spa, 1996).
Morris, J.N., B.E. Fries, K. Steel, N. Ikegami, R. Bernabei, and G.I. Carpenter. 1997. Comprehensive Clinical Assessment in Community Setting: Applicability of the MDS-HC. Journal of the American Geriatrics Society 45:1017–24.
Phillips, C.D., J.N. Morris, C. Hawes, et al. 1997. Association of the Resident Assessment Instrument (RAI) with Changes in Function, Cognition, and Psycho-social Status. Journal of the American Geriatrics Society 45:986–93.
Rubenstein, L.Z., D. Wieland, and R. Bernabei, eds. 1995. Geriatric Assessment Technology: The State of the Art. Milan: Kurtis.
