Outsourcing veterans for long term care: comparison of community and state veterans' nursing homes.
This study compares the characteristics of state veterans' nursing homes and community nursing homes with VA per-diem residentes between 1999-2002. A structure, process, and outcome model was used to examine whether there was any difference in the multi-dimensional quality measures among the three types of community nursing homes (for profit, not-for-profit, and government) and state veterans' nursing homes. For profit community nursing homes were less likely to achieve nurse staffing standards while government facilities were more likely to achieve CNA staffing standards when compared to the state veterans' homes. All community nursing homes had a lower prevelance of tube feeds and catheterization when compared to state veterans' nursing homes. Only government community nursing homes had significantly lower quality of life deficiencies and pressure sore prevelance when compared to state veterans' nursing homes. Vigilant monitoring of all long-term care facilities utilized by veterans is needed.
Veterans with institutional long-term care needs have historically resided in Veterans Administration (VA) nursing homes, community nursing homes, or state veterans' nursing homes. From 1998 to 2005, utilization of state veterans' nursing homes increased from 14,600 to almost 18,000 residents (GAO 2004, 2006). In 2005, the VA also outsourced over 4,400 veterans to community nursing homes (GAO 2006). In 2005, community nursing homes and state veterans' nursing homes provided care for over 60% of all veterans residing in long-term care facilities (GAO 2006).
The VA faces a daunting task of providing institutional long-term care to a rapidly increasing aging veteran population. The number of veterans over the age of 85 is expected to increase by 145% (from 278,400 to 681,400) between the years 2004 and 2012 (US Senate Committee, 2005). People aged 85 years and older are the heaviest users of long-term care with nearly one in five of them living in nursing homes in 2000 (U.S. Bureau of the Census, 2001). The VA is responding to this need by outsourcing veterans to both community nursing homes and state veterans' nursing homes. Therefore, it is of utmost importance for the VA to examine the quality of care that veterans receive in these types of facilities.
There are very few empirical research studies that have evaluated the quality of care provided to veterans in either type of nursing home. Berlowitz et al. (2005) compared the quality of care for veterans in VA nursing homes and those on contract in community nursing homes. Using data for the state of New York from 1997 to 1999, this study found that veterans in VA nursing homes were significantly less likely to develop a pressure ulcer but more likely to experience functional decline than veterans in community nursing homes. On the other hand, residents of community nursing homes with VA contracts were significantly less likely to develop a pressure ulcer but more likely to die than residents in non-contract homes. Few nursing homes were consistently among the best or worst performers on three risk-adjusted quality measures: pressure ulcers, functional decline and mortality. Only 19 of 650 nursing homes, of which four had VA contracts, were in the top quartile on all three performance measures. Fifteen nursing homes, seven of which had VA contracts, were in the bottom quartile on all performance measures. Johnson et al. (2006) compared the quality of care between community nursing homes with veterans and community nursing homes without veterans. Nursing homes participating in the VA Community Nursing Home Program were less likely to meet CNA or RN standards, had more residents being tube fed, newly catheterized, and restrained. These facilities also had higher levels of deficiencies across all of the types examined (quality of care, quality of life, total deficiencies, and actual harm citations) and the residents were more likely to have new pressure sores.
To date, there have been no studies examining quality of care in state veteran nursing homes, nor studies contrasting quality of care for veterans in community nursing homes versus state veterans' nursing homes. The purpose of this study is to compare the characteristics of community nursing homes with VA per-diem residents to the characteristics of state veterans' nursing homes.
VA COMMUNITY NURSING HOME PROGRAM
The Veteran Health Administration (VHA) is divided into 21 Veterans Integrated Services Networks (VISNs). Each VISN is responsible for a specific multi-state geographic region and the care provided to veterans that reside within these areas. There are multiple VA Health Care Facilities (VAHCFs) within each VISN that have primary service areas where they are responsible for providing health care to the veterans in that region. The VAHCF qualifies community nursing homes to participate as a contracting entity and they negotiate the contract rates to be paid to the nursing home through the VA's per-diem system. The per-diem amounts generally reflect a negotiated percentage increase above local Medicaid or Medicare rates for services. A community nursing home that wishes to contract with a VAHCF must be either a Medicare or Medicaid certified facility by the Centers for Medicare and Medicaid Services (CMS), or receive special approval from the VA's Geriatric and Extended Care Strategic Healthcare Group. Each VAHCF has a budget for community nursing home services for veterans seeking care within their primary service area.
VAHCFs have community nursing home review teams at the local level that qualify facilities for participation with VA contracts. The review team analyzes the CMS Nursing Home Compare MDS quality measure profile (http://www.medicare.gov/NHCompare 2006) and the most recent state licensing survey for the facility. A nursing home will be disqualified from participation in the VA contracting program if any four of the following seven factors are present: (1) three level "G" or worse deficiencies are on the survey; (2) the total number of health requirement deficiencies is twice the state average; (3) a level "E" or higher deficiency is given for restraints, abuse, staff treatment of patients, dignity, or licensure; (4) registered nurse (RN) hours per resident day are below the state average; (5) total nurse staffing hours per resident day are below the state average; (6) a level "E" or higher deficiency is given for nursing services, nurse aide training, regular in-service training, proficiency of nursing aides, or staff qualifications; and (7) six or more of the CMS Quality Measures listed in Nursing Home Compare fall above the state average.
There is a 12-month review process once a community nursing home is under contract. The review team revisits the licensing survey and Nursing Home Compare Quality Measures. In addition, there are follow-up visitation requirements for veterans placed in these facilities. The follow-up visit plan is developed before the veteran is placed in the facility and typically there is a visit required by a social worker or RN either every 30 days or every sixth months. Veterans who are long stay residents that have been in the home for over a year without any re-hospitalizations or significant changes in health status typically falls within the six-month visit category as long as there are no family complaints or a facility decline in Nursing Home Compare Quality Measures.
STATE VETERANS' NURSING HOMES
Unlike community nursing homes, all state veterans' nursing homes are owned and operated by their respected state. Historically they were created by states after the civil war. It was necessary for the states to open these homes because the legislation signed by President Lincoln to provide national homes for disabled veterans was only for those who served in the union army. State residency is a common requirement for a veteran to be able to be admitted to a state veterans' nursing home. A veteran has to apply and be accepted to the state veterans' nursing home. This is very different from community nursing homes where the outsourcing transaction is controlled by a contract between the VA and the community nursing home (Smolker et al., 2002).
Like the community nursing homes, the VA uses a daily per diem rate to reimburse the state veterans' homes. However that is where the similarities end. The VA paid a per diem rate of $59.43 in 2006, which is significantly less that the per diem rate paid by Medicaid. Florida Medicaid, for example, paid an average per diem rate of $150 for that same year. The VA also provides construction grants to help states build new facilities or renovate (expand) existing facilities. Other sources of revenue for state veterans' nursing homes include Medicaid, Medicare and other state agencies.
In terms of quality of care, the VA performs annual surveys to certify state veterans' nursing homes. Although the VA is not able to withdraw veterans from state veterans' homes, they are able to withhold the per diem rate. However, unlike the community nursing homes, the VA does not provide 30 day follow up visits (GAO, 2001). In addition, not all state veterans' homes are CMS Medicare or Medicaid-certified. In 2001, only 43 of the 94 state veterans' homes were CMS certified and thus had to comply with the CMS standards (GAO, 2001). These facilities were also included in CMS Nursing Home Compare.
VETERANS IN COMMUNITY NURSING HOMES VERSUS STATE VETERANS' NURSING HOMES
Table 1 compares the demographics of veterans in state veterans' nursing homes versus VA per-diem veterans in community nursing homes. State veterans' nursing homes had a greater proportion of white veterans (90.1% versus 80.4%) and of veterans over the age 75 (65.4% versus 52.6%) than community nursing homes. Table 2 lists the top 15 diagnoses for veterans admitted to state veterans' nursing homes and for VA per-diem residents housed in community nursing homes. Although the order of the top 15 diagnoses is not exactly the same, the proportions for each diagnosis are very similar across both populations. The major difference is that community nursing homes had a greater proportion of veteran residents with hypertension compared to state veterans' nursing homes (43.6% versus 34.9%).
NURSING HOME QUALITY OF CARE
Donabedian's (1988) model of structure, process and outcomes (SPO) is readily used by studies to evaluate the quality of care provided in the health care industry (Davis 1991). Structure reflects the capacity for a provider to provide quality of services, such as nurse staffing levels (Binns et al. 1991). Empirical research shows that nurse staffing levels in nursing homes are related to quality outcomes (Munroe, 1990; Bliesmer, Smayling, Kane, & Shannon, 1998; Harrington et al., 2000).
Processes are what the facility does to provide care. For example, facilities may provide a greater number of tube feeds in an effort to reduce the demand for heavy care activities like toileting or feeding residents. However, high numbers of tube feeding residents could lead to poorer outcomes among nursing home residents (Aronheim et al. 2001).
Outcome indicators are the states that result from care processes, such as decreases in cognitive decline or increases in quality of life (Kane 1998). Within the SPO framework, good structures increase the likelihood of good processes, and good processes increase the likelihood of good outcomes. Good structures may also directly impact outcomes. Nursing home studies examining the impact of organizational characteristics have used Donabedian's SPO framework for quality assessment (Davis 1991).
Using the SPO model, the purpose of this study is to compare the quality of care of community nursing homes with VA per-diem residents with that of CMS-certified state veterans' nursing homes. More specifically, the study has three main research questions:
Question #1: Do state veterans' nursing homes have higher staffing levels than community nursing homes with VA per-diem veterans?
Question #2: Do state veterans' nursing homes have better processes of care than community nursing homes with VA per-diem veterans?
Question #3: Do state veterans' nursing homes have better outcomes than community nursing homes with VA perdiem veterans?
This study uses the Online Survey Certification and Reporting (OSCAR) from 1999-2002 to conduct the facility-level analyses. The OSCAR file is used to obtain data on quality indicators and facility characteristics. The information contained in OSCAR is routinely collected through the Medicare and Medicaid certification process, which is conducted by state licensure and certification agencies. This data is updated annually as part of the recertification survey.
The study is interested in examining quality of care in facilities where veterans are being outsourced. As such, only community nursing homes that had VA per-diem veterans and CMS-certified state veterans' homes are included in this study. Approximately 41% of the community nursing homes housed at least one VA per-diem resident during the four-year period of this study. There were a total of 192 state veterans' nursing homes and 24,360 community nursing homes.
Structure, process, and outcome variables are used as dependent variables to examine if the type of facility where veterans were outsourced to significantly predicts quality differences across facilities. Structure quality related variables were specified at the nursing home level for the direct care nursing staff. OSCAR data was used to capture whether the nursing home met the CMS recommended minimum and/or preferred certified nursing assistant (CNA), licensed staff (total hrs of licensed practical nurse and registered nurse) (LPN/RN), registered nurse (RN) hours per resident day standards. The minimum CMS standard is 2.8 hours per patient per day for CNAs, 1.30 hours per patient per day for licensed staff and 0.75 hours per patient per day for RNs (Abt associates 2001). Empirical research shows that staffing levels in nursing homes is related to quality outcomes (Bliesmer, Smayling, Kane, & Shannon, 1998; Munroe, 1990; Harrington et al., 2000).
Process quality related variables were defined at the facility level as prevalence of tube feeding, indweller catheterizations, and use of physical mobility restraints. The process variables were chosen to examine techniques used during the provision of care as opposed to outcomes of care. High numbers of tube feeding residents could lead to poorer outcomes among nursing home residents (Aronheim et al. 2001). Too many indweller catheterizations has been shown to be related to less quality in facilities (Johnson et al. 2001). The use of physical restraints as part of the care process has received much scrutiny in the literature and these devices should be used with great caution (Karlsson et al. 2001).
Outcomes related quality variables were defined as development of new pressure sores within the facility, quality of care deficiencies, quality of life deficiencies, and actual harm citation on the nursing home licensing survey. Pressure ulcers are injuries to the skin and underlying tissue that result from remaining in the same position for an extended period. Pressure ulcers, if left untreated can become serious, even fatal. Lower pressure ulcer prevalence and incidence is indicative of high-quality care because it suggests that residents are being moved and attempts are being made to alleviate the side effects of being bed-ridden.
State surveyors under contract with the CMS inspect facilities on a yearly basis. From these inspections, facilities may be cited for quality of care and/or quality of life deficiencies. Harrington et al. (2001) describe "Quality of care" deficiencies as those categorized by the federal survey under the following rubrics: resident assessment, quality of care, nursing services, dietary services, physician services, rehabilitative services, dental services pharmacy services, and infection control. "Quality of life" deficiencies are those categorized as followed: residents' rights; admission, transfer, and discharged rights, resident behavior and facility practices; quality of life; and physical environment. An example (from the Nursing Home Compare Website (CMS) of a quality of care deficiency would be if a facility failed to ensure that residents were adequately nourished. An example of a quality of life deficiency would be if a facility failed to prepare food that is nutritional, appetizing, tasty, attractive, well-cooked, and at the right temperature. There is evidence that the higher the total number of deficiencies cited against a nursing home the lower the quality of care (Harrington et al. 2003). Table 3 lists the dependent variables included in the models.
The main independent variable of interest is the type of nursing home: state veterans' versus community nursing home. Community nursing homes vary by ownership type: for- profit, not-for-profit, and public (government) facilities. Prior studies have found quality differences by ownership type, where for-profit nursing homes have lower quality than not-for-profit facilities (Tsai et al. 2003; Hillner et al. 2005, Harrington et al. 2001, Spector et al 1998). Therefore, we categorized the community nursing homes into ownership type: 152 state veteran homes, 4,123 government, 19,066 not-for-profit, and 43,755 for-profit nursing homes
We control for size, member of system or chain, payer-mix (% of Medicaid residents, % Medicare residents), case-mix, geographic location (VISN), state, and year. Larger facilities and chain-affiliated homes usually have greater resources that can result in a better management and staffing profile, and this can ultimately affect quality of care. The amount of financial resources available to the nursing home may impact its ability to provide quality care. Facilities that receive a large portion of their total revenues from Medicaid will be reimbursed at lower levels than private pay and other sources. Higher ratios of Medicare residents could indicate shorter stay residents and lower acuity levels.
We used the acuity index to control for case-mix differences across nursing homes. The acuity index is the sum of OSCAR-based activities of daily living (ADL) index with a special treatments index (Cowles 2002). The ADL index is calculated by adding proportions of residents dependent on eating, toileting, transferring, bedfast, chair bound, and ambulatory within the facility. The special treatments index is the sum of the proportion of residents receiving respiratory care, suctioning, intravenous therapy, tracheotomy care, and parenteral feeding. These two indexes combine to form the acuity index, which has been used in other studies as a case-mix control variable (Shefer et al. 2005; Cowles 2002).
VA budgets and nursing home assignments vary by geographic region; therefore, VISN is used as a set of dummy variables to capture this local variation. VISN 8 (Florida, Southern Georgia, Puerto Rico, and the U.S. Virgin Islands) was omitted from the analysis for comparison purposes. We also used state fixed effects (dummies) to control for potential interstate variations in OSCAR reporting (Angelelli et al. 2003) and industry regulations. Finally, year is entered as a series of dummy variables to control for year-to-year variations in quality of care. Table 4 lists all independent variables included in the analysis.
There are three types of dependent variables in this study: 1) dichotomous variable indicating whether or not a nursing home met a quality standard during a given year; 2) a ratio of a process or outcome relative to the number of residents in the facility; and 3) counts of a quality indicator. Therefore three separate types of analysis were used.
A logistic model was used in cases where the dependent variable was dichotomous (met CNA standard, met RN/LPN standard, met RN standard, actual harm citation). Ordinary least squares regression was used for the ratio variables (tube feeding, catheters, mobility restraints, and pressure sores). Negative binomial regression was used for the count variables (quality of care deficiencies and quality of life deficiencies). The coefficients from the negative binomial regression are converted to incidence rate ratios by exponentiating the coefficients.
Because individual facilities could appear in the data up to four times (once for each year of data collection), it is possible that observations within facilities are correlated over time. To account for potential correlation among observations from the same facility, the standard errors are adjusted using the Huber/White heteroskedastic consistent estimator of the variance/covariance matrix (White 1980) with a cluster correction, as found in STATA (2001).
Table 5 presents the descriptives for all variables used in the analysis by type of nursing home. State veterans' nursing homes had the highest proportion of facilities meeting the CMS recommended nurse staffing levels, while for-profit community nursing homes had the lowest proportion of facilities meeting the recommended staffing levels. State veterans' nursing homes had the highest prevalence of catheters, while government community nursing homes had the highest prevalence for restraints and tube feeds compared to other types of nursing homes. For-profit community nursing homes had the highest number of quality of care and quality of life deficiencies, and had the highest proportion with actual harm citations. State veterans' nursing homes had a higher prevalence of pressure sores. In terms of organizational characteristics, state veterans' and government community nursing homes were larger compared to other types of nursing homes. For-profit community nursing homes had the highest proportion of chain or system affiliated nursing homes. Government and for-profit community nursing homes had the highest proportion of Medicaid residents, while for-profit community nursing homes had the highest proportion of Medicare residents.
Table 6, 7 and 8 present the regression results for the structure, process, and outcome measures of quality. For the structure variables, for-profit community nursing homes were less likely to meet CMS' CNA staffing (p<0.01), LPN/RN staffing (p<0.05), RN staffing (p<0.01) requirements when compared to state veterans' homes (Table 6). Not-for-profit nursing homes also were less likely to meet RN standards (p< 0.10) than state veterans' nursing homes. However government facilities were more likely to meet CMS CNA staffing level (p<0.05) when compared to state veterans' homes. For the process variables, government (p<0.01), for-profit (p<0.01) and not-for-profit (p<0.01) community nursing homes types had a lower prevalence of catheters and tube feeds than state veterans' nursing homes (Table 7). For outcome variables, only government nursing homes had a lower prevalence of pressure sores (p <0.10) and a lower number of quality of life deficiencies (p< 0.01) than state veterans' nursing homes (Table 8).
Some interesting time trends emerged from the analysis. While CNA levels increased between 1999 and 2002, LPN/RN and RN levels decreased during this period. The prevalence of restraints and tube feeds and the probability of actual harm cited also decreased over the same time period.
The VA is faced with the daunting task of providing institutional long-term care for a rapidly increasing number of veterans. With the VA outsourcing many of its veterans to non-VA nursing homes, it is of utmost importance to examine the quality of care of these facilities. Prior research has shown that veterans are being placed in community nursing homes with lower quality of care compared to nursing homes with no veterans (Johnson et al. 2006). However, this research did not include state veterans' nursing homes that provide care to a significant number of veterans.
This is the first study to compare the quality of care for veterans in community nursing homes and state veterans' homes. Our major findings are that for-profit community homes with VA per-diem veterans are significantly less likely to meet CMS staffing standards for CNAs, LPN/RNs and RNs than state veterans' nursing homes. Conversely, government community nursing homes with VA per-diem veterans have higher CNA staffing. This is an important finding since prior studies have indicated that better nurse staffing are associated with better processes and outcomes of care in nursing homes (Bliesmer, Smayling, Kane, & Shannon, 1998; Munroe, 1990; Harrington et al., 2000).
All types of community nursing homes had a lower prevalence of catheters and tube feeds than state veterans' nursing homes. With higher nurse staffing, a nursing home has a higher capacity to provide care. Therefore it is counter intuitive that state veterans' nursing homes would have a greater prevalence of these processes of care when compared to not-for-profit or for-profit nursing homes. Although better staffing levels did not translate into better processes and outcomes for state veterans' homes, government community nursing homes, which had better CNA staffing than state veterans' nursing homes, achieved better processes and outcomes.
Some interesting time trends emerge from the findings of this study. The number of facilities that met CMS CNA recommended staffing levels has increased over time while the RN levels have decreased. These trends could be indicative of shifting in the nurse staffing mix of nursing homes. By having CNAs serve as substitutes for RNs, facilities would be able to lower their costs. Although CNAs and nurses both provide patient care, their roles and many of their functions can be somewhat different. CNAs are primarily responsible for direct care like positioning, bathing and feeding, while RNs will tend to have administrative as well as patient care duties. RNs must provide the right medication, make sure the physician is aware of any changes in the condition of the resident, and direct the care provided by the CNAs. This change in staffing mix may influence the processes and outcomes of nursing homes.
The prevalence of restraints, tube feeds and actual harm cited has decreased over time. Intuitively, one would expect a decrease in restraints and tube feeding with the increase in CNA staffing, since they are the direct care workers in the nursing home industry. Furthermore, there have been a number of initiatives by the federal government to improve all areas of care in nursing homes (CMS, 1998).
Further research is needed to examine why the better staffing patterns do not consistently translate to better quality of care. Although the study did control for case mix differences across facilities, there may still be unmeasured differences between state veterans' and community nursing homes that may explain these findings. For example, while veterans in the VA per-diem program represent a small proportion of residents in community nursing homes (mean of 2% of the facilities that had veterans) (Johnson et al., 2006), state veterans' nursing homes specialize in serving the veteran population and this brings unique challenges to these facilities.
This study presents several limitations. The main limitation of the OSCAR staffing data is that it is self-reported and it is not subject to regular audits. However, a recent study by Feng et al. (2005) found a strong inter-survey agreement between OSCAR and their own survey with respect to RN, LPN, and CNA full-time equivalents (FTEs) data. Another limitation of the OSCAR data is that is a facility-level data so this precludes the use of resident-level risk-adjustment methods. However, we included a facility case-mix measure to control for differences across facilities in patient acuity. Finally, the study examined only CMS certified facilities and therefore only included the 43 of the 91 state veterans' nursing homes that were certified at the time. Therefore, the results are limited to those facilities that are CMS certified.
The research reported here was supported by Department of Veterans Affairs, Veterans Health Administration, Health Services Research and Development Service (project no. IIR 02-284). The views expressed in this report are those of the authors and do not necessarily represent the views of Department of Veterans Affairs.
Abt Associates Inc.(2001).Appropriateness of minimum staffing ratios in nursing homes. Report to Congress: Phase II Final. Baltimore, MD: Centers for Medicare and Medicaid Services.
Angelelli J, Mor V, Intrator O, Feng Z, Zinn J. (2003), Oversight of nursing homes: pruning the tree or just spotting bad apples? Gerontologist. 43(special issue II):67-75.
Aronheim, J. C., M. Mulvihill, C. Sieger, P. Park, and B. E. Fries. (2001), State practice variations in the use of tube feeding for nursing home residents with severe cognitive impairment. Journal of the American Geriatrics Society 49 (2):148-152
Berlowitz DR, Rosen AK, Wang F, Tsilimingras D, Tariot PN, Engelhardt J,Kader B, Mukamel DB. (2005) Purchasing or providing nursing home care: can quality of care data provide guidance. J Am Geriatr Soc. Apr;53(4):603-8.
Binns, G. S. (1991) The relationship among quality, cost, and market share in hospitals. Topics in Healthcare Finance 18 (2):21-32
Bliesmer MM, Smayling M, Kane RL, Shannon I. (1998) The relationship between nursing staffing levels and nursing home outcomes. J Aging Health Aug;10(3):351-71.
Cowles, C. (2002). Nursing Home Statistical Yearbook. Montgomery Village, MD: Cowles Research Group
Centers for Medicare and Medicaid Services (CMS) 1998 http://www.cms.hhs.gov/apps/media/press/factshe et.asp?Counter=340&intNumPerPage=10&check Date=&checkKey=&srchType=&numDays=3500 &srchOpt=0&srchData=&keywordType=All&ch kNewsType=6&intPage=&showAll=&pYear=&y ear=&desc=true&cboOrder=title, Accessed Sept 2006
Davis, M. A. (1991). On nursing home quality: a review and analysis. Medical Care Review 48 (2):129-166.v 20 (2):16-33
Donabedian, A. (1988). The quality of care: How can it be assessed? Journal of the American Medical Association, 260(12), 1743-1748.
Feng Z, Katz PR, Intrator O, Karuza J, Mor V. (2005) Physician and nurse staffing in nursing homes: the role and limitations of the Online Survey Certification and Reporting (OSCAR) system. J Am Med Dir Assoc. Jan-Feb;6(1):27-33.
General Accounting Office (GAO). 2001. VA long Term Care. Oversight of Community Nursing homes needs strengthening. GAO-01-768. Washington, DC: GAO
General Accounting Office (GAO). 2004. VA Long-Term care: Changes in Delivery Service Raise Important Questions. GAO-04-25T. Washington, DC: GAO
General Accounting Office (GAO). 2006. VA long Term Care. Trends and Planning Challenges in Providing Nursing Home Care to Veterans. GAO-06-333T. Washington, DC: GAO.
Harrington C, Zimmerman D, Karon SL, Robinson J, Beutel P. (2000) Nursing home staffing and its relationship to deficiencies. J Gerontol B Psychol Sci Soc Sci. Sep;55(5):S278-87.
Harrington C, Woolhandler S, Mullan J, Carrillo H, Himmelstein DU.(2001), Does investor ownership of nursing homes compromise the quality of care? Am J Public Health. Sep;91(9):1452-5.
Hillner MP, Wodchis WP, Gill SS, Anderson GM, Rochon PA.(2005), Nursing home profit status and quality of care: is there any evidence of an association? Med Care Res Rev. Apr;62(2):139-66.
Johnson, C., Weech-Maldonado, R., Jia, H., Reker, D., Buchanan, R., and Laberge, A. (2006). Characteristics of Community Nursing Homes Serving Per Diem Veterans 1999-2002. Working Paper. Gainesville, FL: VA Rehabilitation Outcomes Research Center.
Johnson TM, Ouslander JG, Uman GC, Schnelle JF. (2001) Urinary incontinence treatment preferences in long-term care. J Am Geriatr Soc. Jun;49(6):710-8.
Karlsson, S., G. Bucht, S. Erikssonn, and P. O. Sandmann. (2001). Factors relating to the use of physical restraints in geriatric care settings. Journal of the American Geriatrics Society 49 (12):1722-1728
Kane, R. L. (1998). Assuring Quality in Nursing Home Care. Journal of the American Geriatric Society 46:232-237.
Munroe DJ. (1990) The influence of registered nurse staffing on the quality of nursing home care. Res urs Health. Aug;13(4):263-70.
Smolker, J. (2002). State veterans' homes digging into SNF occupancy rates. Provider 28 (9):22-4, 27-8, 33-4.
Shefer, A, L. McKibben, B. Bardenheier, D. Bratzler, and H. Roberts. (2005). Characteristics of long-term care facilities associated with standing order programs to deliver influenza and pneumococcal vaccinations to residents in 13 states. Journal of the American Medical Directors Association 6 (2):97-104
Spector, William D, Thomas M Selden, and Joel W Cohen. (1998). Impact of ownership type on nursing home outcomes. Health Economics 7:639-653
STATA. 2001. Stata Reference Manual Release 7. College Station, TX: Stata Press
Tsai AG, Kinosian B.(2003) The association between profit levels and quality of care in California nursing homes. Med Care. Dec;41(12):1315-7.
White, H. (1980). A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity. Econometrica 48:817-830
U.S Senate Committee Hearings May 12, 2005. http://veterans.senate.gov/index.cfm?FuseAction= Hearings.CurrentHearings&rID=54&hID=13,
US Census Bureau, September 15, 2006 http://www.census.gov/prod/2001pubs/c2kbr01-10.pdf
VA Fact Sheet, January 2005, Retrieved December 5, 2006 http://www1.va.gov/opa/fact/docs/ltcare.doc
University of Florida
CHRISTOPHER E. JOHNSON
Texas A&M, School of Rural Public Health
Department of Veterans Affairs Rehabilitation Outcomes
Center of Excellence
Department of Veteran Affairs Rehabilitation Outcomes
Center of Excellence
Table (1) Veteran Nursing Home Resident Demographics Race State Veterans' Community Nursing Nursing Homes Homes Black 7.1% 15.1% Hispanic 1.9% 2.3% Asian 0.5% 0.4% American Indian 0.4% 0.7% White 90.1% 81.4% Gender Female 12.8% 9.5% Male 87.2% 90.5% Age Less than 65 years 6.8% 18.4% 65 to 74 years 18.5% 20.9% 75 to 84 years 47.2% 39.8% 85 to 94 years 16.7% 11.9% greater than 94 years 1.5% 0.9% Education Less than 8th grade 13.0% 15.9% Some high school 10.7% 12.6% High school graduate 27.0% 33.7% Some college 13.2% 16.5% College graduate or graduate school 8.1% 7.3% Source: CMS Minimum Data (1999-2001) Table (2) Veteran 1ursing Home Diagnoses 15 Top Diagnoses State Veteran Homes Hypertension 34.9% Depression 25.6% Diabetes Mellitus 22.2% Cerebrovascular Accident (stroke) 21.4% Dementia other than Alzheimer's Disease 21.1% Emphysema/COPD 18.9% Allergies 17.5% Congestive Heart Failure 13.8% Arthritis 13.4% Anemia 13.2% Arteriosclerotic Heart Disease (ASHD) 12.3% Peripheral Vascular Disease 12.2% Alzheimer's Disease 11.5% Cancer 10.8% Cardiac Dysrhythmias 10.8% Per-Diem Veterans in Community Nursing Homes Hypertension 43.6% Diabetes Mellitus 26.4% Depression 24.3% Dementia other than Alzheimer's Disease 23.4% Emphysema/COPD 22.4% Cerebrovascular Accident (stroke) 21.0% Other Cardiovascular Disease 17.3% Allergies 16.9% Anemia 15.8% Congestive Heart Failure 14.9% Cancer 13.5% Arthritis 12.5% Cardiac Dysrhythmias 10.2% Arteriosclerotic Heart Disease (ASHD) 9.7% Peripheral Vascular Disease 9.4% Source: CMS Minimum Data Set (1999-2001) Table (3) Dependent Variables Definitions Structure Variables Variable Definitions Met CNA Standard = 1 if nursing home met the CMS recommended staffing level for CNA that year. Met RN / LPN Standard = 1 if nursing home met the CMS recommended staffing level for licensed staff (RN/LPN combined) that year. Met RN standard = 1 if nursing home me the CMS recommended staffing level for RN that year. Process Variables Tube feeds = ratio of tube feeding prevalence per resident. Catheters = ratio of catheterizations prevalence per resident. Mobility restraints = ratio of mobility restraints prevalence in the facility per resident. Outcome Variables Quality of care = total quality of care deficiencies per deficiencies resident on the licensing survey for that year. Quality of life = total quality of life deficiencies per deficiencies resident on the licensing survey for that year. Actual harm citation = 1 if the facility received a citation with a scope and severity level of G or higher. Pressure sores = ratio of pressure sore prevalence per resident. Table (4) Independent Variable Definitions Explanatory Variables Variable Definitions Government Community = 1 if the nursing home was a government nursing home community nursing home. Not For Profit Community = 1 if the nursing home was a not for nursing home profit community nursing home. For Profit Community = 1 if the nursing home was a for profit Nursing Home community nursing home State Veterans' Nursing = 1 if the nursing home was a state Home veterans' nursing home Control Variables Size = total number of beds in the facility. Acuity Index = Ratio of acuity of residents. (made up of the ADL index and the special treatment index) Member of a system = 1 if facility is a part of a system % Medicare % of residents that are Medicare % Medicaid % of residents that are Medicaid VISN1, ..., VISN23 = 1 for nursing home located in VISN in question. VISN8 was omitted for comparison purposes. Year1999-2002 = 1 for each year in question with 2002 being the reference. State = 1 for each state Table (5) Means (Standard Errors) for Dependent and Independent Variables State Govern- Not-For- For Veterans' ment Profit Profit N=192 N=4,122 N= 19,066 N= 43,755 STRUCTURE CNA 36.42% 36.96% 24.76% 13.15% LPN/RN 36.99% 30.13% 29.31% 22.04% RN 24.86% 13.09% 13.25% 8.08% PROCESS Catheter 12.29 9.47 6.72 6.83 (17.90) (10.55) (6.60) (5.72) Restraint 14.68 18.23 11.14 10.78 (19.19) (33.05) (14.91) (12.58) Tube feed 7.86 10.88 7.53 8.17 (10.66) (15.66) (11.54) (8.71) OUTCOME Pressure sores 8.88 9.21 7.96 7.62 (8.13) (10.92) (8.92) (6.54) Quality of Care 1.34 1.57 1.70 2.08 Deficiencies (1.62) (1.70) (1.87) (2.14) Quality of life 0.64 0.56 0.62 0.95 Deficiencies (0.90) (0.87) (0.97) (1.24) Actual Harm 0.21 0.23 0.26 0.28 CONTROL VARIABLES Total Residents 145.65 149.81 118.92 103.84 (103.13) (112.41 (84.37) (52.94) System Member 28.9% 6.4% 46.6% 70.5% Acuity index 9.56 10.23 10.08 10.16 (1.57) (1.49) (1.39) (1.54) Percent Medicare 6.80 5.59 9.20 10.03 (15.57) (10.53) (12.71) (11.03) Percent Medicaid 44.55 73.68 60.38 69.28 (29.18) (17.67) (22.10) (19.15) Year 1999 0.25% 6.29% 28.33% 65.13% Year 2000 0.27% 6.19% 28.31% 65.23% Year 2001 0.31% 6.12% 28.54% 65.03% Year 2002 0.32% 5.95% 28.42% 65.31% Table (6) Regression Results for Structure Analysis CNA Staff LPN/RN Staff Odds Ratio (SE) Odds Ratio (SE) For Profit 0.370 *** 0.522 ** (0.108) (0.141) Government 1.95 ** 1.143 (0.608) (0.337) Not For Profit 0.915 0.780 (0.268) (0.213) Acuity index 1.126 *** 1.167 *** (0.022) (0.023) System membership 0.531 *** 0.864 *** (0.029) (0.042) Percent Medicare 1.017 *** 1.031 *** (0.002) (0.024) Percent Medicaid 0.989 *** 0.985 *** (0.001) (0.001) Total Residents 0.997 *** 0.998 *** (0.001) (0.000) Yr1999 0.675 *** 1.237 *** (0.033) (0.051) Yr2000 0.698 *** 1.153 *** (0.033) (0.047) Yr2001 0.743 ** 1.047 (0.032) (0.039) RN Staff Odds Ratio (SE) For Profit 0.381 *** (0.122) Government 0.872 (0.313) Not For Profit 0.594 * (0.193) Acuity index 1.094 ** (0.032) System membership 0.887 * (0.065) Percent Medicare 1.030 *** (0.002) Percent Medicaid 0.979 *** (0.002) Total Residents 0.997 ** (0.000) Yr1999 1.999 *** (0.130) Yr2000 1.841 *** (0.117) Yr2001 1.322 *** 0.0798) Note: significance * p<.10, **p<.05, ***p<.01. Analysis used logistic regression with odds ratio. State and VISN control variables have been omitted from the table. Table (7) Regression Results for Process Analysis Catheter Tube Feed Restraints Prevalence Prevalence Prevalence Beta (S.E.) Beta (S.E. Beta (S.E.) For Profit -0.025 *** -0.012 *** -0.005 (.009) (0.004) (0.011) Government -0.027 *** -0.019 *** 0.008 (0.010) (0.005) (0.012) Not For Profit -0.030 *** -0.017 *** -0.003 (0.009) (0.004) (0.001) Acuity index 0.011 *** 0.018 *** 0.014 *** (0.000) (0.001) (0.001) System membership 0.000 -0.0013 -0.013 *** (0.001) (0.001) 0.002 Percent Medicare 0.001 *** 0.001 *** -0.001 *** (0.000) (0.000) (0.000) Percent Medicaid 0.000 * 0.001 *** -0.000 (0.000) (0.000) (0.000) Total Residents 0.000 0.000 *** -0.000 (0.000) (0.000) (0.000) Yr 1999 0.002 ** 0.007 *** 0.019 *** (0.001) (0.001) (0.002) Yr 2000 0.000 0.006 *** 0.004 *** (0.001) (0.001) (0.002) Yr 2001 -0.001 * 0.004 *** 0.004 *** (0.001) (0.000) (0.001) Note: Significance * p<.10, **p<.05, ***p<.01. Analysis used ordinary least squares regression. State and VISN control variables have been omitted from the table. Table (8) Regression Results for Outcomes Analysis Pressure Quality of Sores Care Prevalence Deficiencies Beta (S.E.) Incidence Rate Ratio (S.E.) For Profit -0.004 1.1740 (0.006) (0.129) Government -0.012 * -0.852 (0.006) (0.100) Not For Profit -0.006 1.049 (0.006) (0.117) Acuity index 0.010 *** 1.050 *** (0.000) (0.006) System membership 0.002 *** 1.099 *** (0.001) (0.019) Percent Medicare 0.001 *** 1.002 *** (0.000) (0.001) Percent Medicaid -0.000 * 1.005 *** (0.000) (0.000) Total Residents 0.000 * 1.002 ** (0.000) (0.000) Yr1999 -0.002 * 1.006 (0.001) (0.017) Yr2000 -0.001 1.030 * (0.001) (0.017) Yr2001 0.001 1.066 *** (0.001) (0.017) Quality of Actual Harm Life Odds Ratio Deficiencies (S.E.) Incidence Rate Ratio (S.E.) For Profit 0.988 1.124 (0.135) (0.231) Government 0.629 *** 0.730 (0.093) (0.165) Not For Profit 0.818 1.010 (0.114) (0.209) Acuity index 1.026 *** 1.072 *** (0.007) (0.013) System membership 1.113 *** 1.096 ** (0.025) (0.042) Percent Medicare 1.001 0.999 (0.001) (0.002) Percent Medicaid 1.008 *** 1.006 *** (0.001) (0.002) Total Residents 1.002 *** 1.003 *** (0.000) (0.000) Yr1999 1.050 ** 2.095 *** (0.024) (0.090) Yr2000 1.027 1.772 *** (0.021) (0.075) Yr2001 1.053 ** 1.302 *** (0.020) (0.055) Note: significance *p<.10, **p<.05, ***p<.01. Pressure sores analysis used ordinary least squares regression. Deficiencies analysis used negative binomial regression with incidence rate ratio. Actual harm citations analysis used logistic regression with odds ratio State and VISN control variables have been omitted from the table.
|Printer friendly Cite/link Email Feedback|
|Author:||Laberge, Alexandre; Weech-Maldonado, Robert; Johnson, Christopher E.; Jia, Huanguang; Dewald, Lloyd|
|Publication:||Journal of Health and Human Services Administration|
|Date:||Mar 22, 2008|
|Previous Article:||Weight counseling for elderly patients in primary care: how often and how much time.|
|Next Article:||Health implications of physical activity in individuals with spinal cord injury: a literature review.|