The general goals of nursing home care are 1) to provide a safe and supportive environment for chronically ill and dependent persons; 2) to maximize individual autonomy, functional capabilities, and quality of life; 3) to stabilize and delay, if possible, the progression of chronic illnesses; and 4) to prevent subacute and acute illnesses and recognize and manage them rapidly when they do occur [1]. Because nursing home residents are heterogeneous, the goals for caring for specific residents vary. This heterogeneity can be illustrated by categorizing nursing home residents into several types. Examples of subgroups and components of their medical evaluation that need particular emphasis are listed in Table 1. Although not all nursing home residents fit neatly into one of these categories, and residents may change from one type to another as their conditions change, this general nosology can help physicians target their evaluations and management for each type of nursing home resident.

As in other comprehensive assessments of geriatric patients in other settings, the timing and purposes for evaluating nursing home residents are important in determining the scope and areas of emphasis for evaluation. Physicians evaluate nursing home residents at the time of admission, at periodic visits every 30 to 90 days, when acute problems occur, and at the time of annual review for residents who stay longer than 1 year. The objectives and elements of evaluation at these different times depend on the clinical status of the resident and goals for care.

Finally, physician evaluation of nursing home residents must be viewed as only one component of a multidisciplinary process that produces an overall care plan for each resident. The goals and context of nursing home care require that a broad range of health professionals participate in care planning and overall management for nursing home residents. In addition, recently implemented federal nursing home regulations mandate comprehensive, multifaceted assessment with interdisciplinary communication and participation. Specific components of the physician evaluation are presented below. However, the relevant federal rules and the role of the interdisciplinary team are reviewed first.

The specific recommendations made in this article are based on literature review and experience, not on a meta-analysis of research. In fact, no research on the most cost-effective strategies to evaluate and care for subgroups of nursing home residents exists. We hope the recommendations described stimulate discussion and research on these issues. These recommendations may be difficult to achieve given the nature of the nursing home environment, the need to spend more time in the nursing home, and inadequate Medicare reimbursement for physician care in nursing homes. Despite recent increases in the relative value of nursing home visit codes achieved by the American Medical Directors Association and the American Geriatric Society, creative and efficient strategies and close cooperation among the interdisciplinary team and physician extenders (where available) are needed to fulfill our recommendations.

Federal Rules and the Role of the Interdisciplinary Team

The Omnibus Budget Reconciliation Act (OBRA) of 1987 contained new federal rules for nursing home care. After considerable public comment, debate, and revision, the new rules became effective in 1991 [2]. Although these rules address a broad range of general and administrative aspects of nursing home care, the process and quality of clinical care are heavily emphasized [3]. With this act, the goal of care is achieving the highest practicable level of functioning (as opposed to custodial care). The act also requires that when a resident's condition deteriorates or complications develop, documentation must show why such situations were "medically unavoidable." Although many physicians, nurses, and other health professionals believe OBRA 1987 represents unnecessary governmental intrusion into the clinical care of nursing home residents, the rules were developed in response to an Institute of Medicine report [4]. When read carefully, the rules provide a sound basic paradigm for improving the process and outcomes of nursing home care. Federal and state nursing home inspectors have interpretive guidelines for these new regulations and will focus increasingly on compliance with OBRA 1987 in the next several years.

One of the central elements of OBRA 1987 is the mandate for a comprehensive, reproducible assessment of all nursing home residents within 14 days of admission, including the "Minimum Data Set," a standard 4-page form composed of 16 sections [5]. Selected areas of the Minimum Data Set must be updated quarterly, and the entire assessment must be updated whenever an important change in patient condition occurs (Table 2).

On a national level, the Minimum Data Set will be used to compile standardized data on nursing home residents, as a tool for quality assurance, and eventually as a component of a prospective reimbursement system. For the individual nursing home, the Minimum Data Set provides health professionals a tool to identify clinical problems and to develop a comprehensive care plan for each resident. Selected items from the Minimum Data Set, called "triggers," are designed to alert the interdisciplinary team that a particular problem or set of problems should be evaluated further. A standard set of assessment protocols (the Resident Assessment Protocols) address 18 common problems in nursing home residents [6] (Table 2). The Resident Assessment Protocols were developed by experts in geriatric medicine, gerontologic nurses, and other gerontologists through a contract with the Health Care Financing Administration. The Resident Assessment Protocols provide recommendations on critical elements of the history, physical examination, and diagnostic testing useful in identifying potentially treatable conditions that may underlie the clinical problem. Most nursing homes have not adequately developed the interdisciplinary cooperation and communication required to make full use of the Minimum Data Set and Resident Assessment Protocols, and the reliability and validity of the data recorded on the Minimum Data Set by typical nursing home staff need further study. In fact, based on our experience and on discussions with other nursing home professionals across the country, the Minimum Data Set is often viewed as just another component of the onerous paperwork required of nursing home staff. Yet, these tools can help physicians and members of the interdisciplinary team to identify important problems in their nursing home residents and to incorporate evaluations from multiple disciplines into the overall care plan. Table 2 lists the disciplines generally responsible for completing specific sections of the Minimum Data Set and the Resident Assessment Protocols. This is usually accomplished at care plan meetings, in which the interdisciplinary team establishes a comprehensive care plan for each resident and updates it quarterly. Because physicians generally do not attend these quarterly meetings, medical aspects of the Minimum Data Set, Resident Assessment Protocols, and care plan must be discussed regularly on routine rounds with the nurses, social service providers, rehabilitation therapists, and other members of the interdisciplinary team. When available, nurse practitioners and physician assistants may play an important role in communicating with the interdisciplinary team and implementing the care plan.

Admission Evaluation

Physicians evaluate nursing home residents within three admission contexts: direct admission from home, admission from an acute care hospital after an acute illness that requires nursing home care, and readmission of a nursing home resident after hospitalization for an acute illness. Often, the primary physician changes during these transitions. Thus, adequate and timely transfer of information critical to patient care among nursing homes, hospitals, and physicians' offices is important. Explicit policies and procedures should be developed to address information exchange, and simplified, comprehensive, and standardized documents should be used during these transfers.

Table 3 lists the key elements of the physician evaluation during nursing home admission. Some elements require greater emphasis than others, depending on the circumstances of the admission. In addition, an interdisciplinary team can conduct several of the recommended assessments because they are requirements of the Minimum Data Set (see Table 2). Standardized assessment instruments besides the Minimum Data Set are available to assist in selected areas of the evaluation, including hearing [7, 8], mobility [9, 10], cognitive function [11, 12], affective status [13-15], and overall function [16, 17]. To screen for hearing impairment, hand-held audioscopes are available that provide frequency sounds at several decibel levels. Detecting hearing loss is an important part of the evaluation and may lead to improved quality of life even among frail nursing home residents [18]. Standardized assessments of mobility involve observing the resident's sitting balance, ability to stand and transfer, stability while standing, and balance and safety while walking. Such assessments are helpful to predict the risk for falls among frail older patients [19] and to identify potentially reversible factors and comorbid conditions that may cause subsequent falls [20, 21]. The Mini-Mental State examination, a 30-point scale that tests orientation, attention, short-term memory, and selected components of higher cognitive function, is the most commonly used standardized screening test of cognitive function. It is reasonably sensitive and specific for detecting dementia [11, 12], is a more objective assessment than is included in the Minimum Data Set, and is sensitive to change. Level of education, primary language, and cultural considerations are important factors to consider when interpreting the results of screening tests for cognitive impairment. Screening questions and scales are available to help detect depression [13, 14], but their reliability and validity for use in cognitively impaired persons have been questioned [22, 23]. Given the high prevalence of untreated depression and its association with death in nursing home residents [24-27] and the frequent coexistence of dementia and depression in this setting [28], identification of depression is an important component of the evaluation process. The Minimum Data Set provides a detailed assessment of basic activities of daily living (such as dressing, grooming, using the toilet, eating) that is similar to the well-established Katz Activities of Daily Living index [29]. For some nursing home residents, particularly those receiving rehabilitation to return to their homes, an assessment of instrumental activities of daily living (such as housekeeping, using the telephone, cooking, and managing finances and medications) may determine their potential for discharge [30].

Relatively few nursing home residents are admitted directly from home. When they are, their need for nursing home care may be based on a broad range of medical, physical, cognitive, affective, and socioeconomic factors. Epidemiologic data suggest that for every nursing home resident, two or three similarly dependent persons live in the community and that most of their care is provided by relatives [31]. Thus, admission of a person to a nursing home from home often is precipitated by a weakening or failure of the support system that kept that person in the community or by the development of behavioral disturbances (such as wandering, night-time agitation, physical or verbal aggression, or incontinence) that family caregivers cannot manage. In addition, many of these persons are treated with several medications, with the potential for adverse effects. Thus, an assessment of socioeconomic status, prescription and nonprescription medications, and behavioral problems should be included in the initial evaluation of persons admitted from home.

Most often, admission to a nursing home occurs after hospitalization for an acute condition. In this case, the acute medical problem is usually the focus of care. But if this is the first nursing home admission, focusing only on the acute problem often leaves many important areas inadequately evaluated. Such areas include concomitant chronic medical conditions, physical and cognitive function, affective status, and socioeconomic status. Thus, when admitting a nursing home resident, whether directly from home or from an acute care hospital, the physician should assess several areas (Table 3).

Acute hospitalization of nursing home residents is common, and infections, fractures, and cardiopulmonary and gastrointestinal conditions are the most frequent causes [32-34]. The readmission evaluation should summarize the status of the acute condition(s) that precipitated hospitalization and update areas outlined in Table 3 that may have changed (for example, overall functional status, mobility, continence, or cognitive function) or been addressed during hospitalization (diagnostic tests related to cognitive function or nutritional status or decisions about intensity of care, for example). Nurse practitioners and physician assistants can be especially helpful in completing these readmission summaries.

Continuing Care

Periodic Visits

Physicians are generally required by federal or state requirements to re-evaluate nursing home residents every 30 to 60 days. When visits for intercurrent subacute problems are used to meet this requirement, important aspects of the resident's overall condition may be unevaluated for several months. Thus, physicians (or nurse practitioners or physician assistants who can make every other required visit after the first one) should use a standardized format for progress notes to ensure that all relevant areas are addressed (Table 4). Software is available to generate notes on a laptop computer and to structure the documentation to include the elements listed in Table 4 [35].

In addition to the periodic evaluation, selected practices are recommended to monitor specific aspects of chronic disease or therapy [1, 36] (Appendix Table 1). Although no data support the efficacy of such practices for preventing morbidity or death, the frailty of nursing home residents, the high prevalence of multiple chronic conditions, and nonspecific evidence of many acute and subacute conditions that could result from complications of therapies warrant a systematic approach to patient monitoring. The monitoring practices we recommend can be accomplished by standing orders, with the results reviewed at periodic visits. Criteria can be set for physician notification of abnormal results [37], and nurse practitioners and physician assistants, when available, can help implement these recommendations.

Acute and Subacute Problems

Many physicians are fortunate to work with nurse practitioners or physician assistants who can evaluate acute and subacute problems at the nursing home. Protocols are available to assist nurse practitioners and physician assistants in these evaluations [1, 38], and both can be reimbursed for their services by Medicare. Involving nurse practitioners and physician assistants gives the nursing staff better access to medical input when problems arise; improves the information physicians receive, which in turn facilitates better decision making; and can prevent some emergency room visits and acute care hospitalizations [39-41].

Most physicians do not, however, have a nurse practitioner or physician assistant who can evaluate nursing home residents in person when problems occur. As a result, physicians usually do the initial evaluation of acute changes in a nursing home resident's status over the telephone. We developed a policy and procedure on "When to Call the Doctor" that delineates specific problems that require immediate or nonimmediate physician notification [1, 37]. This policy helps nursing staff to identify clinically important acute and subacute conditions or laboratory values that require physician evaluation and assists nurses and physicians by eliminating unnecessary telephone calls. When a nurse does call a physician, she or he should have the chart available, have taken the resident's vital signs, and have described the resident's new condition in detail. Physicians on call (who may not be familiar with the resident) should ask, at a minimum, the following questions: What are the resident's vital signs? What medications is the resident receiving? Has the resident's mental status changed? Is the resident eating and drinking fluids? Does the resident have a "No CPR" order or a "No Hospitalization" order? The answers to these questions are critical to understanding the acuity of the situation and the need for hospitalization. Medical directors and nursing directors should develop a policy and procedure regarding nurse-physician communication. A medical fact sheet, which includes a succinct summary of the resident's medical, functional, and psychosocial status on one page, can be used by nurses to give physicians a more comprehensive description of a resident's overall medical and functional status [1, 42]. When laboratory tests are ordered or other orders are requested, communication by facsimile may be an efficient strategy in some settings. Facsimile orders to nursing homes are recognized as legitimate by the Health Care Financing Administration [43].

Annual Review

Although the value of the routine annual history and physical examination for nursing home residents has been questioned [44, 45], long-term nursing home residents should have a comprehensive annual review performed by a physician to summarize relevant findings from 1) the Minimum Data Set [which must be updated annually]; 2) the physician's evaluation of the resident's medical status; and 3) the results of selected screening procedures. Substitution of a history and physical examination done during hospitalization for an intercurrent acute illness for the required annual examination will inadequately address many issues, such as underlying chronic conditions and cognitive and functional status, that should be reviewed at least annually.

Few data exist on which to base recommendations for screening practices that might be included in the annual review of long-term nursing home residents. We recommend selected screening practices (Appendix Table 2, recognizing that these recommendations differ somewhat from those offered by the U.S. Preventive Services Task Force [46] and that they are based on our review of the literature and opinion, not on data from clinical trials. The areas included in functional status should be assessed by members of the interdisciplinary team during the required annual updating of the Minimum Data Set (see Table 2) and summarized briefly in the physician's annual review. Selected laboratory tests [listed in Appendix Table 2 may be useful to identify potentially treatable conditions in nursing home residents [47-50]. Screening for tuberculosis with a PPD (purified protein derivative) should be done on admission and annually if the result is initially negative because of the high risk for conversion and spread of infection in the nursing home [51-54]. A repeated test should be done in 10 to 14 days if the admission PPD test result is negative to rule out the booster phenomenon (a positive test result developed in response to the PPD, which generally does not represent true conversion or reactivation of tuberculosis) [55, 56]. A chest roentgenogram and an electrocardiogram [57] are often useful as baseline examinations for comparison during subsequent illnesses, but annual examinations are probably not cost-effective [49]. Screening for cervical cancer with a Papanicolaou test is probably unnecessary after age 65 years if the resident has had at least two negative results [58, 59]. Screening for breast cancer with an annual mammogram and for prostate cancer with a prostate-specific antigen test is controversial. The guiding principle should be to do these tests only if the results would change the patient's care plan. In this context, a mammogram would make sense for some long-term nursing home residents because curative treatment is available [60]. However, because most breast carcinomas grow slowly, the value of detecting and treating an early cancer in an 85-year-old woman with a short life expectancy is debatable [61]. Discovery of early-stage prostate cancer through prostate-specific antigen screening in a man older than 75 years who has substantial comorbid conditions would probably not prompt initiation of curative therapy, because life expectancy and the natural history of the disease suggest that these men will die of another illness [62-67]. In addition, the number of false-positive results in these patients could lead to substantial morbidity and expense from further evaluation (such as prostate ultrasonography and biopsy). The recommendations outlined above are different from those published for younger persons [68, 69] and for older noninstitutionalized adults [46, 70], for whom some data support the recommendation. Until better data are available, decisions to use these screening tests for nursing home residents must be based on specific patient needs.

Table 5 outlines a format for annual physician evaluation of long-term nursing home residents. Rather than simply dictating a standard history and physical examination and medically oriented problem list, physicians should use the annual review to summarize succinctly data gathered from their medical evaluation, the Minimum Data Set completed by the interdisciplinary team, and the results of selected screening procedures. Establishing overall goals for the resident's medical care and documenting information relevant to advance directives are essential components of this review. This type of annual review is useful for the interdisciplinary team, provides critical documentation should the resident be hospitalized or transferred to another physician or health care facility, and helps set realistic expectations of the resident, family, and nursing home staff for subsequent medical care. Medicare reimburses the cost of one comprehensive nursing home evaluation each year, and nurse practitioners and physician assistants can help complete this annual evaluation.