This Article Abstract Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation

Nursing Facility Compliance With Do-Not-Hospitalize Orders

Correspondence: Address correspondence to Aram Dobalian, PhD, JD, Assistant Professor, Department of Health Services Administration, University of Florida, PO Box 100195, Gainesville, FL 32610-0195. E-mail: adobalia{at}hp.ufl.edu.

	Abstract

TOP Abstract Methods Results Discussion References

 
Purpose: This study sought to determine whether nursing homes comply with residents' do-not-hospitalize (DNH) orders prohibiting inpatient hospitalization. Design and Methods: With the use of data from the nationally representative 1996 Nursing Home Component of the Medical Expenditure Panel Survey, a multivariate logistic regression model was developed. Results: Three percent of residents had DNH orders. These residents were half as likely to be hospitalized. Residents in not-for-profit or public facilities were less likely to be hospitalized than those in for-profit homes. Hospitalization was more likely among men, racial or ethnic minorities, those with more diagnosed health conditions, and those in facilities in the South compared with those in the Midwest. Hospitalized residents with DNH orders had no limitations of activities of daily living, were not located in hospital-based nursing homes, were less likely to be in a for-profit facility, and were sicker than nonhospitalized residents with DNH orders. Implications: Improved education regarding advance directives, particularly DNH orders, is necessary for health care practitioners and patients. More consistent and rigorous policies should be implemented in nursing facilities.

Key Words: Advance directives • End-of-life care • Transfers • Nursing homes • Appropriateness

There are number of reasons why one may choose to complete an advance directive, including concern that future incapacity may limit one's ability to express preferences for medical treatment, or the desire to ensure that family members and health care practitioners provide care that is consistent with one's wishes (Emanuel et al., 1991; Harrison, Molloy, Darzins, & Bédard, 1995). Nonetheless, there is substantial variation in the use of advance directives among different subpopulations. Numerous factors have been posited to explain these differences in the rates of completion of advance directives. Differences in knowledge about advance directives, differential access to health care, and cultural and racial or ethnic differences have been suggested as barriers to completion (Bradley, Wetle, & Horwitz, 1998; Palker & Nettles-Carson, 1995). It has also been suggested that physicians may be hesitant to initiate discussions regarding advance directives with patients, particularly in the early stages of an illness (Christakis & Iwashyna, 1998). Nevertheless, the reasons for these differences in completion rates remain unclear.

Although the use of advance directives has been relatively extensive in certain settings, health care provider awareness and responsiveness to these documents has been less expansive. Prior research has demonstrated that advance directives often do not change interventions at the end of life (SUPPORT Principal Investigators, 1995; Teno, Lynn, et al., 1997). A large, multisite study conducted from 1989 to 1994 demonstrated that 47% of physicians were unaware of their patients' preferences to avoid CPR, at least in hospital inpatient settings (SUPPORT Principal Investigators, 1995). This study failed to find evidence that increasing the rates of advance directives resulted in care that was more consistent with patients' expressed preferences (Teno, Licks, et al., 1997). Because residents often remain in nursing homes for longer periods of time than patients in hospitals, one might suspect that health care providers in nursing homes would be more aware of the existence of residents' advance directives. Nevertheless, a study of nursing home patients and their family members regarding preferences for treatment found that 25% of patients received care that was inconsistent with their previously expressed wishes (Danis et al., 1991).

Transfers from nursing facility to hospital that are not in compliance with resident wishes can lead to unnecessary expense because of the higher cost of care received in hospital settings. One recent study suggests that inappropriate transfers may represent a significant problem (Saliba et al., 2000). The researchers studied 100 hospital records, examining transfers of residents from skilled nursing facilities to hospitals, and found that 45% of admissions were inappropriate as rated by physician reviewers after they considered the residents' advance directives. This study did not examine residents who were not transferred to a hospital.

A do-not-hospitalize (DNH) order is a type of advance directive that indicates that the resident or responsible party (e.g., relative or legal guardian) does not wish the resident to be hospitalized. DNH orders are medical orders written by physicians that apply to the resident's current health status. In this study, I hypothesized that residents with DNH orders would be less likely to be hospitalized than those without these documents. I also examined sociodemographic resident factors and facility characteristics related to whether a resident had a DNH order. The data in this study were derived from a large, nationally representative sample of the U.S. nursing home population, including facilities and residents with a diverse range of demographic characteristics.

	Methods

TOP Abstract Methods Results Discussion References

 
Data Source
This retrospective cohort study uses data from the complete 1996 Nursing Home Component (NHC) of the Medical Expenditure Panel Survey (MEPS), a nationally representative, multistage, stratified probability sample of nursing homes and their residents. The reference population is all persons in the United States who spent one or more nights in calendar year 1996 as a nursing home resident (Bethel, Broene, & Sommers, 1998; Potter, 1998). Data were collected by the Agency for Health Care Policy and Research, now the Agency for Healthcare Research and Quality (AHRQ), and the National Center for Health Statistics. Data are available at the AHRQ's Center for Cost and Financing Studies Data Center in Rockville, MD for use by researchers with approved projects.

MEPS conducted interviews by using a computer-assisted interviewing system. MEPS collected data from administrators and staff at the facilities, and community members (relatives, guardians, friends, community-based caregivers, and residents after discharge). Much of the data were gathered from persons who referenced the resident's Minimum Data Set (MDS), a comprehensive assessment of nursing home residents that is required by the Centers for Medicare and Medicaid Services (CMS) upon the resident's admission and that is updated quarterly as well as when there are significant changes in the resident's condition (Bethel et al., 1998; Potter, 1998). DNH orders are a required item that must be recorded on the MDS. MEPS completed missing data by using medical records or other sources where possible (Potter, 1998). This study was certified as exempt from human subjects review by the University of California, Los Angeles Institutional Review Board.

All Medicare- and Medicaid-certified nursing facilities must periodically conduct standardized, comprehensive assessments of all residents by using the Resident Assessment Instrument (RAI) as mandated by the Omnibus Budget Reconciliation Act of 1987. The care assessment component of the RAI is the MDS, which records information on the functional, medical, cognitive, psychological, and social status of each resident. Each item of the MDS has its own coding conventions, and a manual describes how to ask questions, what to observe, and who to contact for information. The assessor, a trained clinical professional (typically a nurse, social worker, or therapist), interacts with the resident, reviews the medical record, and is responsible for obtaining information regarding functional performance from professional staff. Studies have demonstrated that interrater reliabilities for the MDS items are adequate for research purposes (Hawes et al., 1995; Snowden et al., 1999).

Participants
In the initial sampling phase, facilities were selected and basic information about the nursing homes was gathered; residents were chosen in the second stage. In order to be considered eligible for the survey, a home had to be a facility or distinct unit of a facility that was either certified by Medicare or Medicaid, or licensed as a nursing home with three or more beds and providing on-site 24-hr skilled nursing care (Potter, 1998). Two separate samples of residents were selected within those facilities that chose to cooperate. The first consisted of a cross-sectional sample of residents on January 1, 1996, and the second was a sample of persons admitted during 1996 who had no prior admissions to an eligible facility during 1996 (Potter, 1998). Of the 997 eligible nursing homes that were sampled in the NHC, 82% responded to surveys (Sommers, Bethel, & Broene, 1999). The final sample consisted of 815 responding facilities, as well as 5,899 eligible residents, including 3,209 residents in the facility on January 1 (99% response rate) and 2,690 residents admitted during 1996 (93% response rate).

For each resident, an analytic weight was constructed by the AHRQ to adjust the sample to correspond with the reference population. Each analytic weight indicates the number of residents in the underlying population represented by that resident. Taking the inverse of the probability of selecting a particular facility provided the weights used to produce population estimates for the nursing homes (Sommers et al., 1999). A conditional within-facility weight was then created for each resident to use for the person-level weights. The construction of the analytic weights is described in detail elsewhere (Sommers et al., 1999). The final weighted estimates for the NHC were 16,760 nursing homes, 1,560,003 residents as of January 1, 1996, and 1,536,525 admitted residents who were not in a nursing home on January 1, 1996 (3,096,528 total persons in the United States).

Dependent Variables
The first analysis examined the presence of a DNH order at baseline (January 1, 1996 or upon admission during 1996). The second analysis examined whether the resident had been discharged directly from the resident's sampled nursing facility to a hospital during 1996.

Previous studies have examined the reliability scores (Spearman-Brown intraclass correlation) of the MDS and found that the correlations are adequate for research purposes (Kiely, Mitchell, Marlow, Murphy, & Morris, 2001). Reliability scores for DNH orders in the MDS are not available, but Hawes and colleagues (1995) found that the Identification and Background section, which includes the question regarding DNH orders, had an average reliability score of 0.71 for the 13 items examined in that section. An intraclass correlation of 0.4 or higher is considered adequate for research purposes, whereas those with a correlation of 0.7 or higher are deemed to have excellent reliability (Fleiss, 1986).

Independent Variables
On the basis of prior literature, three groups of covariates were hypothesized to relate to DNH order status and hospitalization: resident demographics and psychosocial variables, health status, and facility characteristics. Resident demographics and psychosocial variables included gender, race or ethnicity (White or non-White), age (>65, 65–84, or 85+ years), educational attainment (up to 8th grade, 9th grade to high school graduate, or some or more college), and largest source of payment to the nursing facility (Medicare, Medicaid, or other, which consisted of private insurance, Social Security, pension, self or family income or assets, Department of Veterans Affairs, HMO contract, other, or none). Health status included whether the resident had been diagnosed with a variety of conditions (43 items), and the resident's need for assistance with activities of daily living (ADLs) or physical functioning (0 or 1–6 ADLs).

Facility characteristics included census region (Northeast, Midwest, South, or West), ownership status of the facility (for profit, not for profit, or public), and type of nursing home (hospital based, facility with multiple levels of care, or freestanding). Facility characteristics were based on the resident's initial nursing home; only 10% of the residents were in more than one home during the year.

Statistical Analysis
In the first analysis, I evaluated the relationship between the binary dependent variable DNH order and the following independent variables: resident demographics and psychosocial variables, health status, and facility characteristics. In the second analysis, I evaluated the relationship between hospitalization (yes or no), the three groups of covariates, and the resident's DNH order status. Significance testing, with statistical significance at p <.05, was performed by regression. I conducted all analyses by using analytic weights to estimate population prevalence, and the standard errors accounted for the multistage sample design (clustering and stratification). I constructed multivariate logistic regression models to evaluate the independent relationship of patient demographics, psychosocial variables, health status, and facility characteristics with the dependent variables. I derived odds ratios (ORs) and 95% confidence intervals (CIs) for all logistic regression analyses. I performed statistical analyses by using Stata (College Station, TX) statistical software (StataCorp, 1999).

	Results

TOP Abstract Methods Results Discussion References

 
Table 1 provides descriptive information about resident characteristics for the sample; 97% answered the questions about DNH orders (n = 5,708). Three percent of the residents in the sample (190) had documentation of DNH orders. Sixty-six percent of the residents were female; the average age was 80 years. Eighty-seven percent of the respondents were White; residents had an average of 3.5 diagnosed conditions; 91% required assistance with ADLs; and 25% had been discharged to a hospital directly from the sampled nursing facility.

	Discussion

TOP Abstract Methods Results Discussion References

 
This study found that nursing homes generally appear to respect residents' rights to individual autonomy and self-determination as expressed through their residents' DNH orders. As hypothesized, residents with DNH orders were less likely to be hospitalized than those without that type of advance directive. Residents with DNH orders were about half as likely to be hospitalized as those without documentation of these orders. Although this result suggests that nursing home compliance with resident preferences for care is superior to that found in hospital inpatient settings (SUPPORT, 1995), 25 residents (13%) with DNH orders were hospitalized directly from their sampled nursing home. Nevertheless, the majority of nursing homes complied with resident preferences to forego hospitalization, indicating that DNH orders are effective devices for promoting patient wishes for medical treatment in nursing facilities. However, the small number of residents with DNH orders is an indication of the limited use of advance directives in the U.S. nursing home population, and it suggests that more effort should be expended to promote their wider use.

Further research is needed to determine why nursing homes do not always comply with resident preferences for medical treatment. For example, it is possible that some of the residents with DNH orders, when confronted with circumstances in which hospitalization was recommended, decided to permit hospitalization despite their previously expressed wishes to forego it. Furthermore, family members, designated surrogates, or other responsible parties may influence the course of a resident's treatment and thus be a factor in whether residents with DNH orders are hospitalized.

To determine what factors differentiated residents with DNH orders who were hospitalized from those residents with DNH orders who were not hospitalized, I performed bivariate analyses with various sociodemographic, clinical, and facility characteristics. Unlike nonhospitalized residents with DNH orders, hospitalized residents with DNH orders had no ADL limitations, were not located in hospital-based nursing homes, and were less likely to be in a for-profit facility. There were no differences associated with particular health conditions (including Alzheimer's disease or other dementias, cancer, emphysema or chronic obstructive pulmonary disease, heart disease, psychiatric diagnosis, stroke or transient ischemic attack, etc.), although hospitalized residents with DNH orders had more health conditions than nonhospitalized residents. Thus, hospitalized residents with DNH orders appear to be "sicker" than nonhospitalized residents, but there does not seem to be a particular pattern to the variation in health status.

Transfers may occur for a variety of reasons. For example, residents may be transferred because the nursing facility lacks the capacity to provide necessary services. Transfers have also been associated with the receipt of poor-quality acute care in skilled nursing facilities (Saliba et al., 2000). Similarly, transfers of residents with DNH orders who do not express a subsequent desire to undergo hospitalization would appear to be inappropriate. Providing inappropriate, expensive services, as in these instances, is an inefficient allocation of limited resources.

With respect to other factors related to hospitalization, I found that racial and ethnic minorities were more likely to be hospitalized than Whites, which may be due to a greater need for acute hospital care and limited access to primary care services for minorities (Employer Health Benefits, 1999). As would be expected, residents in poorer health, as determined by a greater number of diagnosed health conditions, were more likely to be hospitalized. Male residents were more likely to be hospitalized than female residents, which may suggest that providers do not treat women and men with comparable health status similarly. Finally, I found that residents in nursing facilities located in the South were more likely to be hospitalized than residents in homes in the Midwest. This regional difference may be due to variation in treatment preferences among health care providers trained in different parts of the nation and may indicate a need for efforts aimed at decreasing these variations in the education and training of health and social service workers. It may also relate to state-level differences in the regulation of advance directives, or differential enforcement of legislation in some states. Future studies should attempt to disentangle the impact of these factors.

With regard to whether the resident had a DNH order, I found that residents with Medicare as their largest payer were less likely to have DNH orders than those with other payers, although residents with Medicare are typically more acutely ill than long-term patients. Medicare ceases to pay for nursing home care after 100 days, under the assumption that care provided after that time is more custodial and less oriented toward providing clinical interventions. Because length of stay did not differentiate between residents with and without DNH orders, Medicare is not acting as a proxy to differentiate long-term care and subacute residents from more acutely ill residents. The inclusion of length of stay and an interaction term between Medicare and length of stay in (unreported) multivariate analyses did not alter my results with respect to either who has DNH orders or compliance with DNH orders, although residents with shorter stays or with both Medicare and a shorter stay were more likely to be hospitalized.

Although number of diagnosed conditions was not associated with DNH status, I conducted additional bivariate analyses to determine whether particular health conditions were more prevalent among residents with DNH orders. Of more than 20 possible physical and mental conditions, only Alzheimer's disease differed between the two groups, and this effect disappeared in multivariate analysis. In the bivariate analysis, residents with DNH orders were more likely to have Alzheimer's disease (OR = 1.66; 95% CI = 1.09–2.54).

Compared with residents in freestanding homes, those in hospital-based facilities were less likely to have DNH orders. Residents in hospital-based nursing facilities are often admitted to those nursing facilities following discharge from a hospital, and therefore residents with DNH orders may simply be less likely to be in hospital-based nursing homes in the first place. Finally, compared with residents in the Midwest, those in the South were less likely to have DNH orders. This finding may be related to differences in state laws.

This study has several limitations. My findings are limited by the relatively small sample size for those with DNH orders; the low prevalence of DNH orders suggests the possibility that they are being underused or used selectively when providers are less likely to recommend hospitalization for residents. Better data on health status might account for differences in completion of DNH orders and likelihood of hospitalization. In addition, the findings may be subject to an omitted variable bias such as state-level variation in the requirements for completing DNH orders. The data do not include measures of physicians' or other providers' knowledge or beliefs for or against the use of DNH orders; some studies have found that these factors may play a role in determining whether an individual completes an advance directive (Eliasson et al., 1999; Miles, Koepp, & Weber, 1996; Morrison, Morrison, & Glickman, 1994; Paris, Carrion, Capello, & Mulvihill, 1993). The data also do not permit a determination of when the DNH order was completed. In addition, the data do not include specific information regarding each resident's preferences for care as expressed through his or her DNH orders, and thus some variation that is due to content differences might not be evident in the present study. Furthermore, the data do not include information concerning whether residents altered their preferences for care prior to hospitalization. It is possible, for example, that some residents with DNH orders chose to permit hospitalization despite their previous preference to forego it. Finally, because I do not know how many residents wished to have DNH orders, I cannot determine whether documentation of these orders by nursing home staff occurs more frequently for those residents for whom the facility plans to comply with the order prohibiting hospitalization. This limitation suggests that this study may understate the degree to which residents are receiving care that is inappropriate (as measured by facility compliance with the DNH order).

This study found that, in general, nursing homes comply with resident preferences to forego hospitalization as expressed in their DNH orders. This finding is inconsistent with other studies that have examined health care provider awareness and responsiveness to advance directives in hospital inpatient settings. Proponents of the use of advance directives to promote patient autonomy and self-determination should be encouraged by this result as compared with health care provider responsiveness to advance directives in hospital inpatient settings. Nevertheless, some of the residents with DNH orders were hospitalized, suggesting that nursing homes should undertake additional measures, including the implementation of more consistent and rigorous policies, to ensure that patient preferences for medical care are honored. Improvements in the education of health care practitioners and patients regarding advance directives, particularly DNH orders, should also improve compliance.

	Footnotes

 
This work was supported by Grant HS00046 from the Agency for Healthcare Research and Quality. At the time of the data analysis, Dr. Dobalian was a program analyst with the Agency for Healthcare Research and Quality.

¹ Department of Health Services Administration, University of Florida, Gainesville.

Decision Editor: Linda S. Noelker, PhD

Received for publication January 22, 2003. Accepted for publication April 29, 2003.

	References

TOP Abstract Methods Results Discussion References

 

• 42 C.F.R. 489.100 (2004).
• 42 C.F.R. 489.102 (2004).
• Bethel, J., Broene, P., Sommers, J. P., 1998;. Sample design of the 1996 Medical Expenditure Panel Survey Nursing Home Component (MEPS Methodology Report No. 4; AHCPR Publication No. 98-0042). Rockville, MD: Agency for Health Care Policy and Research.
• Bradley, E. H., Wetle, T., Horwitz, S. M., 1998;. The Patient Self-Determination Act and advance directive completion in nursing homes. Archives of Family Medicine,. 7:417-423.[Abstract/Free Full Text]
• Christakis, N., Iwashyna, T., 1998;. Attitude and self-reported practice regarding prognostication in a national sample of internists. Archives of Internal Medicine,. 158:2389-2395.[Abstract/Free Full Text]
• Danis, M., Southerland, L. I., Garrett, J. M., Smith, J. L., Hielema, F., Pickard, C. G., et al 1991;. A prospective study of advance directives for life-sustaining care. New England Journal of Medicine,. 324:882-888.[Abstract]
• Eliasson, A. H., Parker, J. M., Shorr, A. F., Babb, K. A., Harris, R., Aaronson, B. A., et al 1999;. Impediments to writing do-not-resuscitate orders. Archives of Internal Medicine,. 159:2213-2218.[Abstract/Free Full Text]
• Emanuel, L. L., Barry, M. J., Stoeckle, J. D., Ettelson, L. M., Emanuel, E. J., 1991;. Advance directives for medical care—A case for greater use. New England Journal of Medicine,. 324:889-895.[Abstract]
• Employer Health Benefits., 1999;. 1999 annual survey. Menlo Park, CA: Kaiser Family Foundation and Health Research and Educational Trust.
• Fleiss, J., 1986;. The design and analysis of clinical experiments. New York: Wiley.
• Fleiss, J., 1986;. The design and analysis of clinical experiments. New York: Wiley.
• Harrison, C., Molloy, D. W., Darzins, P. J., Bédard, M., 1995;. Should people do unto others as they would not want done unto themselves? Journal of Clinical Ethics,. 6:14-19.[Medline]
• Hawes, C., Morris, J. N., Phillips, C. D., Mor, V., Fries, B. E., Nonemaker, S., 1995;. Reliability estimates for the Minimum Data Set for nursing facility resident assessment and care screening (MDS). The Gerontologist,. 35:172-178.[Abstract]
• Kiely, D., Mitchell, S., Marlow, A., Murphy, K., Morris, J., 2001;. Racial and state differences in the designation of advance directives in nursing home residents. Journal of the American Geriatrics Society,. 49:1346-1068.[Medline]
• Miles, S. H., Koepp, R., Weber, E. P., 1996;. Advance end-of-life treatment planning: A research review. Archives of Internal Medicine,. 156:1062-1068.[Abstract/Free Full Text]
• Morrison, R. S., Morrison, E. W., Glickman, D. F., 1994;. Physician reluctance to discuss advance directives: An empiric investigation of potential barriers. Archives of Internal Medicine,. 154:2311-2318.[Abstract/Free Full Text]
• Palker, N., Nettles-Carson, B., 1995;. The prevalence of advance directives: Lessons from a nursing home. Nurse Practitioner,. 20:13, 17-18, passim.
• Paris, B. E., Carrion, V. G., Meditch, J. S., Jr., Capello, C. F., Mulvihill, M. N., 1993;. Roadblocks to do-not-resuscitate orders: A study in policy implementation. Archives of Internal Medicine,. 153:1689-1695.[Abstract/Free Full Text]
• Potter, D. E. B., 1998;. Design and methods of the 1996 Medical Expenditure Panel Survey Nursing Home Component (MEPS Methodology Report No. 3; AHCPR Publication No. 98-0041). Rockville, MD: Agency for Health Care Policy and Research.
• Saliba, D., Kington, R., Buchanan, J., Bell, R., Wang, M., Lee, M., et al 2000;. Appropriateness of the decision to transfer nursing facility residents to the hospital. Journal of the American Geriatrics Society,. 48:154-163.[Medline]
• Snowden, M., McCormick, W., Russo, J., Srebnik, D., Comtois, K., Bowen, J., et al 1999;. Validity and responsiveness of the Minimum Data Set. Journal of the American Geriatrics Society,. 47:1000-1004.[Medline]
• Sommers, J., Bethel, J., Broene, P., 1999;. Construction of weights for the 1996 Medical Expenditure Panel Survey Nursing Home Component (MEPS Methodology Report No. 7; AHCPR Publication No. 99-0045). Rockville, MD: Agency for Health Care Policy and Research.
• StataCorp., 1999;. Stata statistical software: Release 6.0 [Computer software]. College Station, TX: Stata Corporation.
• SUPPORT Principal Investigators., 1995;. A controlled trial to improve care for seriously ill hospitalized patients: The Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatment (SUPPORT). Journal of the American Medical Association,. 274:1591-1598 (Published correction appears in Journal of the American Medical Association, 1996, 275, 1232).[Abstract/Free Full Text]
• Teno, J. M., Licks, S., Lynn, J., Wenger, N., Connors, A. F., Jr., Phillips, R. S., et al., for the SUPPORT Principal Investigators 1997;. Do advance directives provide instructions that direct care? Journal of the American Geriatrics Society,. 45:508-512.[Medline]
• Teno, J., Lynn, J., Wenger, N., Phillips, R. S., Murphy, D. P., Connors A. F., Jr., et al., for the SUPPORT Principal Investigators 1997;. Advance directives for seriously ill hospitalized patients: Effectiveness with the Patient Self-Determination Act and the SUPPORT intervention. Journal of the American Geriatrics Society,. 45:500-507.[Medline]
• Wenger, N. S., Oye, R. K., Bellamy, P. E., Lynn, J., Phillips, R. S., Desbiens, N. A., et al 1994;. Prior capacity of patients lacking decision making ability early in hospitalization: Implications for advance directive administration. The SUPPORT Investigators. Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments. Journal of General Internal Medicine,. 9:539-543.[Medline]

This article has been cited by other articles:

				P. N. Lanken, P. B. Terry, H. M. DeLisser, B. F. Fahy, J. Hansen-Flaschen, J. E. Heffner, M. Levy, R. A. Mularski, M. L. Osborne, T. J. Prendergast, et al. An Official American Thoracic Society Clinical Policy Statement: Palliative Care for Patients with Respiratory Diseases and Critical Illnesses Am. J. Respir. Crit. Care Med., April 15, 2008; 177(8): 912 - 927. [Full Text] [PDF]

				J. T. van der Steen and M. R. Helton Hospitalization trends for pneumonia among older persons. JAMA, May 10, 2006; 295(18): 2137 - 2138. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation