The Gerontologist
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Google Scholar
Right arrow Articles by Harrington, C.
Right arrow Articles by Carrillo, H.
PubMed
Right arrow PubMed Citation
Right arrow Articles by Harrington, C.
Right arrow Articles by Carrillo, H.
The Gerontologist 48:679-691 (2008)
© 2008 The Gerontological Society of America

Variation in the Use of Federal and State Civil Money Penalties for Nursing Homes

Charlene Harrington, PhD1, Theodore Tsoukalas, PhD1, Cynthia Rudder, PhD2, Richard J. Mollot, JD2 and Helen Carrillo, MS1

Correspondence: Address correspondence to Charlene Harrington, PhD, Department of Social & Behavioral Sciences, University of California, 3333 California Street, Suite 455, San Francisco, CA 94118. E-mail: Charlene.harrington{at}ucsf.edu


   Abstract
TOP
 Abstract
Background on Regulation of...
 Conceptual Model
 Data Collection
 Analysis of Data
 Findings
 Discussion and Conclusions
 References
 
Purpose: The study examined factors associated with state variations in the use of federal and state civil money penalties (CMPs) for nursing homes. Design and Methods: We collected federal and state CMP data from state survey and certification agencies for 2004. We also used federal CMP data from the federal enforcement action database for 2000–2004. Logistic regressions examined factors related to whether states issued CMPs, and ordinary least squares regressions examined the number and amount of federal CMPs (2000–2004) and the total federal and state CMPs (2004). Results: In 2004, 3,159 federal and state CMPs were collected, for a total of $21.6 million, but CMPs were given for only 2% of deficiencies issued. The number of federal CMPs collected was positively related to average facility occupancy rates, the percentage of facilities with deficiencies for harm or jeopardy, and state survey and certification budgets but was negatively related to the number of facility complaints per nursing home bed. Total federal and state CMPs were positively related to state senators' liberal voting records, having a democratic governor, and the percentage of Medicaid nursing home residents and were negatively related to the population aged 65 and older, complaints per nursing home bed, percentage of hospital-based facilities, and home- and community-based expenditures. Implications: The Centers for Medicare & Medicaid Services should address the state variations in CMPs by providing states and federal regional offices with guidelines on the use of federal CMPs. It should also improve accuracy and completeness by including federal and state CMPs in its enforcement database.

Key Words: Enforcement actions • Licensing and certification • State survey agencies • Deficiencies

Because poor nursing home quality is a chronic problem in the United States, the Nursing Home Reform Act (Omnibus Budget Reconciliation Act, 1987) adopted provisions designed to strengthen the federal and state nursing home survey process and to improve enforcement systems. In addition to the threat of decertifying poorly performing facilities from the federal Medicare and Medicaid programs, the new law specifically provided for the use of intermediate sanctions including civil money penalties (CMPs), denial of payment for new or current admissions, and temporary managers (U.S. General Accounting Office [GAO], 1987). Federal CMPs were implemented in 1995 with the adoption of regulations by the Health Care Financing Administration (1995). Many states also have a process by which they collect state CMPs (sometimes called fines by states but hereafter referred to as CMPs) for violations of state nursing home regulations (Edelman, 1998; Hawes, 2002). Although the Centers for Medicare & Medicaid Services (CMS) considers federal CMPs to be an important tool for encouraging a high quality of care and sanctioning nursing homes that are out of compliance with regulations, a recent study reported the infrequent use of CMPs as an enforcement tool by states and wide variations across states (Tsoukalas et al., 2006). Little is known about what accounts for the variation in the use of CMPs.

Given the lack of knowledge in this area, the present study examined states' use of federal and state CMPs as an intermediate sanction. The first aim was to describe state variation in the collection of federal and state CMPs. The second aim was to examine factors related to variation in (a) whether states collected federal and/or state CMPs, (b) the number of federal and/or state CMPs collected, and (c) the dollar amounts of federal and/or state CMPs collected. We conducted regression analyses on two data sets: (a) longitudinal cross-sectional federal CMP data for 2000–2004 and (b) survey data from state licensing and certification officials on the total federal and state CMPs collected in 2004. The variations in state enforcement practices are a concern because they can create inequities in quality of care across states and potentially serious problems for residents where quality is poor (Leichter, 1997). We expect the findings from this study to be useful to state and federal policy makers and nursing home stakeholder groups in understanding the factors associated with issuing and collecting federal and state CMPs. The study also raises questions about CMP data collection policies and practices. The study was not designed to determine the effectiveness of collecting CMPs or how CMP funds are used by states.


   Background on Regulation of Nursing Homes
TOP
 Abstract
 Background on Regulation of...
Conceptual Model
 Data Collection
 Analysis of Data
 Findings
 Discussion and Conclusions
 References
 
All states have established state licensing requirements and have contracts with the federal CMS to certify that nursing homes meet the federal standards for participation in Medicare and Medicaid. The federal government establishes federal regulations for nursing homes; sets state survey and certification procedures; funds state survey and certification programs; and provides central and regional oversight over state survey, certification, and enforcement programs (Edelman, 1998). States are largely responsible for carrying out survey, certification, and enforcement activities and issuing a range of intermediate sanctions such as CMPs depending upon the scope and severity of violations (CMS, 2004c). Some states have state nursing home regulations similar to the federal regulations, whereas other states have additional requirements (with the federal requirements serving as the minimum baseline). State nursing home surveyors may use discretion regarding whether to issue deficiencies and sanctions and whether to issue deficiencies and CMPs under federal or state regulations, although the same action cannot be cited simultaneously under both federal and state regulations.

For federal CMPs, states recommend federal CMPs and other sanctions to CMS for Medicare-only certified facilities and Medicare and Medicaid dually certified facilities, but CMS makes the final determination of action, issues CMPs and other sanctions, and collects the funds. For Medicaid-only certified nursing homes, states may issue and collect CMPs directly and use the funds for a variety of state activities under federal guidelines (e.g., paying for the costs of relocation of residents to other facilities and nursing home closures). The collection and use of state-only CMPs is entirely at the discretion of states. For a discussion of how states use CMP funds, see Tsoukalas and colleagues (2006).

There has been a general decline in the average number of deficiencies and in the scope and severity of deficiencies issued by states since 1999, which in turn could reduce the number of CMPs issued (GAO, 2000, 2003; Harrington, Carrillo, & Mercado-Scott, 2005; Office of the Inspector General [OIG], 1999). These studies have also shown wide variations and poor performance of state surveyors in issuing deficiencies and sanctions such as CMPs. In 2000–2001, $81.7 million in federal CMPs were imposed, but less than half were collected in 2002 (OIG, 2005). These reports recommended that CMS provide written guidelines to CMS staff and states regarding procedures for issuing and collecting federal CMPs and streamline the processing of federal CMPs (GAO, 2003; OIG, 2005). This paper extends the previous research on federal CMPs by using data that include federal and state CMPs and examines the factors associated with the use of CMPs.


   Conceptual Model
TOP
 Abstract
 Background on Regulation of...
 Conceptual Model
Data Collection
 Analysis of Data
 Findings
 Discussion and Conclusions
 References
 
Regulatory policies are of growing interest in health care because of the impact these policies have on shaping the delivery of health care. Brown (1992) described four features of health care regulation as centralized or decentralized and budgetary or behavioral. Nursing home quality regulations would be classified as decentralized behavioral regulations, because they are designed to ensure quality of care and quality of life rather than to address budgetary issues. Because nursing home survey and enforcement procedures are decentralized to states, there is a potential for wide variations in survey procedures and enforcement actions, which previous studies have documented (GAO, 2003; Harrington, Mullan, & Carrillo, 2004; OIG, 2005; Walshe & Harrington, 2002). E. A. Miller (2005), in a review of studies of state health policies, suggested that state policy adoption is related to states' external environments, where federal assistance, laws, and regulation, as well as what other states adopt, are important. State environmental factors are also important, and these include sociodemographic, economic, and supply factors and political system characteristics including administrative, participatory, structural, and ideological factors (E. A. Miller, 2005).

In this study, we examined the factors associated with whether states issued CMPs, the number of CMPs collected, and the dollar amounts of CMPs collected. Based on the literature, we identified six primary factors that may explain variation in the use of CMPs by states: (a) quality differences; (b) sociodemographic, economic, and political differences; (c) policy differences; (d) facility characteristics; (e) provider supply or competition; and (f) regulatory resources. We describe these factors below.

As for the first factor, the quality of facilities may vary across states and could impact the use of CMPs, but nursing home quality is very difficult to measure. Stevenson (2005) and Harrington and colleagues (2004) concluded that consumer complaints are one way to differentiate nursing home quality variations across states. Higher complaints are related to more deficiencies and more serious deficiencies issued (and possibly more CMPs; Stevenson, 2005), but more CMPs are not expected to increase the number of complaints (i.e., are not endogenous). We used the percentage of facilities with pressure ulcers as another quality measure. This measure has been a common quality outcome measure in a number of studies (e.g., Grabowski & Angelelli, 2004). Increased numbers of pressure ulcers should be asso-ciated with increased deficiencies and possibly CMPs.

The second set of factors includes sociodemographic and political variables. Harrington and colleagues (2004) found that states having a higher percentage of the population aged 65 and older and states having a democratic governor were two factors that predicted strong state nursing home enforcement. The aged population may represent a political constituency measure favoring more stringent enforcement (Kronebusch, 1997). States with democratic governors (representing public official ideology) and/or senators (representing electorate ideology) with liberal voting records are proxies for liberal ideologies that may be more likely to favor regulatory and enforcement approaches for nursing homes than conservative perspectives (Deason-Howell & Blevins, 2003; Harrington et al., 2004). Studies have found that income per capita and the percentage of non-Whites in the population are related state variations in participants and expenditures (N. A. Miller, Ramsland, Goldstein, & Harrington, 2001), so these factors could have an impact on both quality of care and regulatory activities (Davidson, 1997). States with higher per capita incomes may be more willing to allocate funds for regulatory activities, or they may have populations that are less tolerant of poor-quality care and may advocate for more enforcement. States with high minority populations may have more quality-of-care problems (Mor, Zinn, Angelelli, Teno, & Miller, 2004), or this measure may be a proxy for southern states, which take more enforcement actions than northeastern and midwestern states (Harrington et al., 2004). Thus, the percentage of the population aged 65 and older, income per capita, democratic governor, liberal voting records, and the percentage of the non-White population should increase enforcement activities and, consequently, increase the use of CMPs.

Third, state policies should have some impact on the use of CMPs. Studies have shown that states with higher Medicaid reimbursement rates tend to have higher quality nursing home care (Grabowski, 2001). Higher reimbursement rates should encourage improvements in staffing and quality, which should reduce poor quality and the need for issuing CMPs. In addition, the supplemental security income (federal SSI) and the state supplemental income program (SSP) participation rates are qualifying requirements for Medicaid eligibility. States with more generous Medicaid eligibility (allowing greater access to nursing homes) may have more active enforcement programs as a quality control mechanism. Thus, the Medicaid nursing home rate per day and SSI/SSP participation rates may be positive predictors of enforcement.

The fourth set of factors includes nursing home characteristics. For-profit facilities have more deficiencies and higher levels of deficiencies (Harrington, Woolhandler, Mullan, Carrillo, & Himmelstein, 2001; O'Neill, Harrington, Kitchener, & Saliba, 2003). The percentage of facilities in chains has sometimes been related to lower staffing and quality (Harrington et al., 2001; Walshe & Harrington, 2002) and should predict stronger state nursing home enforcement (Harrington et al., 2004). Hospital-based facilities generally have higher staffing and fewer deficiencies, so states with more hospital-based facilities are expected to have fewer quality problems and fewer CMPs issued (Harrington et al., 2004). Facilities with a greater proportion of Medicaid funding often have lower reported quality and more deficiencies (Mor et al., 2004). Medicaid rates are generally lower than private-pay and Medicare rates, so facilities with more Medicaid residents often have financial and quality problems. Facilities with high occupancy rates may have more deficiencies because these facilities may not have to compete with other facilities on the basis of quality. Facilities given deficiencies by state agencies for harm or jeopardy to residents should be more likely to be issued CMPs by those agencies. Overall, facilities that are for profit, are in a chain, are non-hospital-based, have higher percentages of Medicaid residents, have higher occupancy rates, and receive harm or jeopardy deficiencies are expected to have more CMP sanctions.

Fifth, supply and competition can also be factors influencing enforcement practices. States with more nursing home beds per population (greater competition) should have better quality of care and may have less deficiencies and CMPs (Nyman, 1989). However, surveyors may be likely to issue federal CMPs where there are adequate numbers of nursing home beds in an area because there are more available beds for residents to select than in areas with shortages of beds. Another factor is the availability of alternatives to nursing homes. The number of home health agencies per population and the home- and community-based service (HCBS) expenditures per capita would serve as alternatives (greater competition) to nursing homes and should reduce nursing home quality problems and result in fewer CMPs.

Sixth, resources available to states for nursing home regulation should be a positive factor in the use of CMPs. State resources are a measure of state administrative capacity, and this should influence bureaucratic decision making and activities (E. A. Miller, 2006). E. A. Miller (2006), in a study of state reimbursement policies, found that states with greater administrative capacity were better able to overcome obstacles to implementing state policies. Walshe and Harrington (2002) found that federal and state expenditures for the regulation of nursing facilities per bed was a positive factor in predicting which states issued more deficiencies and higher levels of deficiencies (in terms of scope and severity). Edelman (1998), in a case study of six states, also found that limited state survey and certification budgets were problematic for state enforcement activities.


   Data Collection
TOP
 Abstract
 Background on Regulation of...
 Conceptual Model
 Data Collection
Analysis of Data
 Findings
 Discussion and Conclusions
 References
 
We used a survey of state licensing and certification officials to collect information about whether states issued federal and state CMPs, the number of CMPs issued and collected, and the amount of CMPs collected. The survey used a 1-page written questionnaire. We used state Freedom of Information Act requests to obtain additional state data. The survey was necessary because CMS does not collect data on state deficiencies and state sanctions, including CMPs. Of the 41 states that provided data on federal and state CMPs, only 21 collected state CMPs in 2004. We supplemented the survey data with federal CMP data from CMS's enforcement action database (CMS, 2004b) for those states that did not respond to the survey in 2004. In addition, we used 5 years of CMS enforcement action data on federal CMPs for a longitudinal cross-sectional analysis (2000–2004).

Dependent Variables
For the dependent variables in the regression models, we used two sets of data on CMPs. The first set was longitudinal cross-sectional data for the 5-year period 2000–2004 of enforcement actions obtained from CMS. This data set included provider numbers, names, and addresses and the number and type of federal CMPs. We used the enforcement action database to examine the following: (a) whether the state collected federal CMPs, (b) the number of federal CMPs collected standardized per 100 nursing home beds in the state, and (c) the amount of federal CMP dollars collected per nursing home bed for those facilities that were issued CMPs. Complete data were available for all states and the District of Columbia (N = 51) for the 5-year period (N = 255), although the accuracy and completeness of the data were unknown.

The second data set derived from a survey of state licensing and certification officials, and we used it to examine the combination of federal and state CMPs in 2004. We used three variables: (a) whether states collected federal and/or state CMPs in 2004, (b) the number of federal and/or state CMPs collected per 100 nursing home beds in 2004, and (c) the dollar amounts of federal and/or state CMPs collected per nursing home bed in 2004. Although the total federal and state CMPs were only available for 1 year, they represented a more complete picture of CMPs than the federal enforcement action data. The accuracy and completeness of the data were a concern, but the states are the only sources for such information.

Independent Variables
Table 1 shows a listing of all of the independent variables used in this study, including the data sources and hypotheses. CMS data from the Online Survey, Certification and Reporting (OSCAR) complaint files were used for the number of complaints and were only available for 1999 and 2003–2006 (we estimated interim years by using a linear best fit trend line between the two periods). We standardized these based on the number of nursing home beds surveyed during that year from the OSCAR data. We obtained sociodemographic and political variables from the U.S. Census Bureau, Americans for Democratic Action, and the Republican Governors Associations. We obtained state Medicaid nursing home reimbursement rates from Swan (2003). We obtained the number of SSI/SSP recipients from the U.S. Social Security Administration (1999–2003). Facility characteristics were available from OSCAR provider file data (Harrington, Carrillo, et al., 2005).


View this table:
[in this window]
[in a new window]

  		Table 1. Variables, Sources, and Hypotheses.

 
The number of licensed nursing home beds per 1,000 population was available from a survey of states (Harrington, Chapman, Miller, Miller, & Newcomer, 2005). We obtained the number of certified home health agencies from unpublished data from CMS (CMS, 2004a) and state survey and certification budget data from unpublished CMS data (2005). States reported HCBS expenditures on CMS Form 64 (Burwell, Sredl, & Eiken, 2004). We standardized the total state survey and certification budgets for all facilities, including nursing homes, collected from CMS by the total number of nursing home beds in a state. Most independent variables were lagged 1 year to reduce potential problems with endogeneity, a standard procedure used in studies of states (N. A. Miller et al., 2001). Facility characteristics were not lagged, however, because we did not consider these variables to be potentially endogenous.


   Analysis of Data
TOP
 Abstract
 Background on Regulation of...
 Conceptual Model
 Data Collection
 Analysis of Data
Findings
 Discussion and Conclusions
 References
 
We compiled and completed descriptive analyses of the data by state. We present frequencies for the following variables: the number of federal CMPs and state CMPs collected and the dollar amounts in 2004. We used OSCAR data to calculate the percentage of federal CMPs per deficiency issued for 2004 (Harrington, Carrillo, et al., 2005).

With the state as the unit of analysis, this study estimated six models of CMPs by using the two data sets described above. Because the number and dollar amounts for CMP data were skewed, we used the log for each. All states and DC were included. Table 2 presents the descriptive statistics. We examined the independent variables for possible multicollinearity by completing a correlation matrix. None of the variables were highly correlated (above a.65 correlation; full results available upon request). Tolerance tests in the regression analyses did not show multicollinearity to be a problem.


View this table:
[in this window]
[in a new window]

  		Table 2. Number of Federal CMPs and Dollar Amount, 2000–2004; and Number and Dollar Amount of Total Federal and State CMPs, 2004.

 
We used the STATA program (Release 8) to conduct the analyses using two-tailed significance tests. In the first set of analyses, we used a logit panel regression model to examine factors related to whether a state collected federal CMPs (employed as a dichotomous variable) and ordinary least squares (OLS) panel regressions models to examine the log number of federal CMPs per 100 nursing home beds and the log federal CMP dollars collected per nursing home bed. These are reflected in the following equations:


Formula

where i is the ith of 50 states, t is the tth year between 2000 and 2004, and E is a random error term.

Thus, in the first analysis of the longitudinal cross-sectional federal CMP data for 2000–2004, the number of observations was 255 for the logit and 160 for the OLS panel regressions for states that collected federal CMPs. For the OLS regressions, we used a cross-sectional panel model to test our hypotheses. We used fixed effects models because this approach includes a dummy variable for states to account for differences across states that may be fixed over time. We used the fixed effects models because they do not rely on assumptions about the relationship between the fixed effects and other explanatory variables in the model (Greene, 1997). We also examined random effects models and found that the results were similar, but we present only the fixed effects model. We controlled for autocorrelation of the residuals by using the STATA autoregressive (xtregar) procedures (a linear model with an AR [1] disturbance). We also used robust standard errors to account for clustering of observations for all of the analyses.

In the second set of analyses, we examined state survey agency data for total federal and state CMPs in 2004. There were 51 observations for the logit and 44 observations for the OLS regressions for those states that collected CMPs. Because we only had 1 year of data for the 2004 analysis, we had to limit the number of variables in the model. We used a forward stepwise regression model to select five to seven variables that best explained each dependent variable. As expected, the stepwise regressions selected different variables for the three dependent variables examined. We also used robust standard errors to account for clustering of observations.


   Findings
TOP
 Abstract
 Background on Regulation of...
 Conceptual Model
 Data Collection
 Analysis of Data
 Findings
Discussion and Conclusions
 References
 
In 2004, the states surveyed 15,138 nursing homes and issued about 140,000 federal deficiencies for violations of federal quality regulations. Table 2 shows that a total of 1,401 federal CMPs resulted in the collection of a total of $2.3 million in 2004. The state survey reported about the same number of federal deficiencies issued (1,451) but substantially larger dollar amounts of federal CMPs collected (a total of $17 million was collected in 2004; no table shown). In addition, 22 states including the District of Columbia collected a total of $4.48 million from 1,708 state CMPs for violations of state regulations in 2004 (no table shown). On the state survey, 8 states reported no federal CMPs collected (although of these, 4 collected state CMPs), whereas 14 states reported no federal CMPs collected on the federal database in 2004. Table 3 shows that the mean number of federal CMPs was 57 in states in 2004 and the mean for total federal and state CMPs was 86 in 2004.


View this table:
[in this window]
[in a new window]

  		Table 3. Means and Standard Deviations.

 
The number of total federal and state CMPs collected was about 2% of the total number of federal deficiencies issued even though 15.5% of deficiencies were rated by states as having the potential for causing or actually causing harm or jeopardy in 2004 (no table shown). We found wide variations in the ratio of collected CMPs to deficiencies issued, ranging from 19% in Wisconsin to no CMPs issued in eight states (no table shown).

Table 4 shows the first panel regression models of the federal CMP data for the 2000–2004 period. The table shows the logistic panel regressions predicting whether states collected federal CMPs (yes = 1; no = 0). Controlling for the number of complaints per 100 nursing home beds, we found that states with higher percentages of nursing home residents with pressure sores were less likely to collect CMPs. The percentage of non-Whites, the SSI/SSP participation rates per 1,000 population, the percentages of for-profit and chain-owned facilities, and nursing home beds per 1,000 population were positive predictors of state collection of federal CMPs.


View this table:
[in this window]
[in a new window]

  		Table 4. Fixed Effects Regressions on Whether Federal CMPs Were Collected, Number of Federal CMPs Collected, and CMP Dollar Amount Collected, OSCAR Data 2000–2004.

 
The OLS panel regression for the number of federal CMPs collected per 100 nursing home beds showed a positive relationship with facility occupancy rates, the percentage of facilities with deficiencies that cause harm or jeopardy, and the state survey and certification budget per nursing home bed in a state (see Table 4). The number of complaints per 100 nursing home beds was negatively related to the number of federal CMPs collected per 100 nursing homes beds.

The OLS panel regression model for the dollar amount of federal CMPs collected per nursing home bed showed that the positive predictors were income per capita, percent hospital-based facilities, and facility occupancy rate (see Table 4).

Table 5 shows the regressions for the total federal and state CMPs collected in 2004. The forward stepwise logistic regression identified five factors related to state collection of federal and state CMPs in 2004. The percentage of non-Whites in the population and nursing home beds per 1,000 population were positive predictors of whether states collected federal and state CMPs. Negative factors were personal income per capita, democratic governors, and the number of certified home health agencies per 1,000 population.


View this table:
[in this window]
[in a new window]

  		Table 5. Regressions on Whether Federal and State CMPs/Fines Were Collected, Number of Federal and State CMPs/Fines Collected, and Dollar Amount Collected in 2004.

 
For the OLS regression on the number of federal and state CMPs collected per 100 nursing home beds, three factors were positive: state senators with a liberal voting record, democratic governors, and higher percentages of nursing home Medicaid residents in a facility. Negative factors were the number of complaints per 100 nursing home beds, the percentage of people aged 65 and older in the population, the percentage of hospital-based facilities, and HCBS expenditures per 1,000 population (see Table 5). For the dollar amount of federal and state CMPs collected per nursing home bed, facility occupancy rate was a positive factor, whereas the percentage of the population aged 65 and older, the state Medicaid nursing home reimbursement rate, the percentage of hospital-based facilities, and the amount of HCBS expenditures per 1,000 population were negative factors (see Table 5).


   Discussion and Conclusions
TOP
 Abstract
 Background on Regulation of...
 Conceptual Model
 Data Collection
 Analysis of Data
 Findings
 Discussion and Conclusions
References
 
The findings show that few federal deficiencies (about 2% of deficiencies) resulted in CMPs that were collected and that states vary widely in the number of deficiencies and CMPs collected. All states except four collected either federal and/or state CMPs in 2004. Although state survey agencies are required to use intermediate sanctions for facilities with serious violations, between 16% and 27% of states were not using federal CMPs in 2004. These findings are consistent with previous studies that have shown wide variations in state survey processes and enforcement activities that have been attributed to weak federal oversight by CMS over the past decade (GAO, 1987, 2000, 2003; Harrington & Carrillo, 1999; OIG, 1999).

Although states reported collecting about $21.6 million in total federal and state CMPs in 2004, the number of state CMPs collected was greater than the number of federal CMPs, although the average amount of state CMPs was smaller. In previous studies, states have reported that the federal CMP process is cumbersome and ineffective (Harrington & Carrillo, 1999; Harrington et al., 2004; Tsoukalas et al., 2006), which may explain why some states prefer to use state CMPs. In addition, the funds collected from state CMPs and Medicaid-only facilities are not shared with the federal government, whereas federal CMPs from dually certified facilities are shared between the federal and state governments, giving states an incentive to use state CMPs. Because state CMPs play an important role in state enforcement activities, information about state enforcement, including the levying of state CMPs, as well as other enforcement actions, needs to be collected and made available to the public and to policy makers.

There are several concerns about the accuracy and completeness of the CMP data. Some states were unable to provide federal data because they did not maintain detailed reports of enforcement actions and did not have access to federal enforcement action data. As noted above, the CMS federal enforcement action database does not collect data on state deficiencies and CMPs. States reported about the same number of federal CMPs collected but higher amounts of federal CMPs than reported in the CMS federal enforcement action database. There may be many reasons for the discrepancies between state and federal reports of federal CMP data, including lag times in reporting or entering data into the federal database, lost reports, changes or reductions in the amounts of CMPs by CMS, and inaccuracies in state and/or federal records. These data problems support the need for more accurate and complete federal and state data reporting systems on enforcement actions, as recommended in previous studies (GAO, 2003; OIG, 2005). CMS should issue guidelines to states on the recording of federal and state enforcement actions and maintain a publicly accessible database of both state and federal enforcement actions.

Contrary to expectations, higher numbers of complaints were not associated with the state collection of federal CMPs or federal and state CMPs. Moreover, the number of complaints was negatively associated with the number of federal CMPs and the total number of federal and state CMPs collected per 100 nursing home beds. The percentage of residents with pressure sores was also negatively associated with the collection of federal CMPs. There may be a number of reasons for this. When controlling for the severity of the deficiencies issued, perhaps states with fewer complaints and pressure sores were less tolerant of poor quality and therefore more apt to issue CMPs than those states with more quality problems overall. Alternatively, the number of complaints may not be a good indicator of poor care; when numbers are low, deficiencies may be underreported because of fear of reprisal and cultural values that can affect the likelihood of making a complaint. In addition, the estimated complaint data for 3 of the 5 years may not have been accurate.

A number of factors were associated with whether states issued federal CMPs. As expected, the percentage of non-Whites was a positive predictor of state collection of federal CMPs (2000–2004) as well as the collection of federal and state CMPs in 2004. States with higher non-White populations may have poorer quality nursing home care that requires more enforcement actions such as CMPs, or higher percentages of non-Whites in the population may be more indicative of southern states, which have traditionally had stronger nursing home enforcement than those in the midwest and the northeast (Harrington et al., 2004).

Personal income per capita was positively associated with the amount of federal CMPs collected but, contrary to expectations, negatively associated with whether states collected federal and state CMPs in 2004. It seems logical that states with higher incomes and costs of living would collect higher amounts of CMPs, but the negative relationship was difficult to explain. Perhaps high-income individuals with access to care at home or residential care facilities have little interest in nursing home enforcement. Having a democratic governor was a negative predictor of whether federal and state CMPs were collected in 2004 but a positive predictor of the number of federal CMPs collected. Perhaps decisions about using state CMPs were made after the Omnibus Budget Reconciliation Act 1987 was passed, and states continued the policies over the years so that the party of the current governor had no effect. State senators' liberal voting records were positively associated with the number of federal and state CMPs collected in 2004. This partially supports the hypothesis that liberal ideology is associated with greater CMP enforcement.

As expected, the percentages of for-profit and chain-owned facilities were positive predictors of state collection of federal CMPs although not predictors of the number and amount of federal and state CMPs collected. Previous studies have documented lower staffing and higher deficiencies in for-profit and chain-owned facilities, so states with more of these types of facilities may be more enforcement oriented (Harrington et al., 2001, 2004; O'Neill et al., 2003). The percentage of hospital-based facilities was positively associated with the amount of federal CMPs collected but negatively associated with the number and amount of federal and state CMPs collected, as expected. Previous studies have shown that these facilities have higher staffing and receive fewer deficiencies and CMPs (Harrington et al., 2004), so they would receive fewer CMPs. Because hospital-based nursing homes may appear to have more resources than free-standing nursing homes, surveyors may give higher CMPs when deficiencies are issued.

The positive relationship of bed supply to the collection of federal CMPs may have occurred because surveyors may be less concerned about the potential negative consequences of facilities receiving CMPs in areas with greater supply. Presumably, patients have more options to choose a facility where the supply is greater. At the same time, higher facility occupancy rates were positively associated with the number and the amount of federal CMPs collected and the amount of combined federal and state CMPs collected in 2004. High occupancy rates should improve the facility's financial status, so surveyors may be less concerned about the negative financial consequences of issuing CMPs, or perhaps these facilities have poorer care because they do not need to compete on the basis of quality.

The percentage of facilities with deficiencies that cause harm or jeopardy was positively associated with the number of federal CMPs collected, as expected, but not with the amount of federal CMPs collected. Federal guidelines urge more sanctions for serious quality problems, but the lack of association with higher amounts of CMPs was puzzling. States only propose the imposition of federal CMPs to regional offices of CMS, and CMS must approve both the actual imposition of the CMP and the amount. Facilities routinely ask the regional offices for a reduction in fines based upon financial hardship. CMS offices are often reported to grant reductions in CMPs (OIG, 2005). Perhaps the lack of higher CMPs for more serious deficiencies is related to later CMS reductions in the amount of CMPs collected.

Having more Medicaid residents in nursing homes did not correlate with the collection of federal CMPs, the number of federal CMPs, or the amount of federal CMPs collected in 2000–2004, but it was a positive predictor of the number of total federal and state CMPs collected in 2004. States with more Medicaid nursing homes may have more quality-of-care problems and may be more likely to give state, rather than federal, CMPs to address these problems. This is consistent with previous findings showing poorer quality of care in nursing homes with more Medicaid residents (Mor et al., 2004).

The number of certified home health agencies was negatively related to whether a state collected federal and state CMPs, and expenditures for HCBS were negatively related to the number and the amount of federal and state CMPs collected in 2004. These services may serve as alternatives to nursing home care, which may encourage nursing homes to offer higher quality. Higher survey and certification budgets per nursing home bed was a positive predictor of the number of federal CMPs collected, consistent with the findings in a previous study by Walshe and Harrington (2002). This suggests that providing higher and/or more uniform federal funding to states may improve state agency enforcement responsiveness.

Overall, the models give support for the independent variables selected for the analysis even though the findings varied by model. The predictions might be improved considerably if state CMP data were combined with the federal CMP data over the 5-year period. With only 1 year of total state and federal CMPs in 2004, the power for the models was limited. In addition, other potentially important factors were not included in the models, such as the viewpoints and policies of state attorneys general, the political power of the nursing home industry, and the influence of consumer advocacy organizations.

Congress adopted intermediate sanctions in the Nursing Home Reform Act (Omnibus Budget Reconciliation Act, 1987) in order to encourage compliance with federal regulations and to forestall or prevent having to use the sanction of decertification (removal from the Medicare and Medicaid program). Overall, the findings show that federal and state CMPs are a seldom used intermediate sanction with wide variations in state policies and practices. The fact that some states are not using federal CMPs suggests the need to reexamine the federal CMP system. CMS should provide more guidance to states to help them understand when they can and should propose federal CMPs and the amount of CMPs that should be recommended. CMS should also give better guidance to regional CMS offices in terms of approving CMPs, deciding appropriate amounts of CMPs, and responding to claims of financial hardship by facilities. There may also be ways to improve and streamline the federal CMP enforcement system to optimize its effectiveness as both a remedy and a source of information for stakeholders and the public. For example, some have proposed that CMPs be collected immediately and put in a holding account until the appeal process is complete and the funds returned to the facility if the facility wins the appeal (Tsoukalas et al., 2006).

Perhaps there are better enforcement remedies than CMPs for serious violations, such as holds on the admission of residents for poor-performing facilities. An evaluation of the effectiveness of different enforcement approaches would be worthwhile. Such an evaluation could examine ways to make the use of CMPs more effective and to improve the uniformity in enforcement actions across states. CMS must pay greater attention to studying the use and collection of CMPs, developing standard policies and practices for the number and amount of federal CMPs issued by states, and collecting complete data on federal and state CMPs and other enforcement actions.


   Footnotes
 
This research was funded by Commonwealth Fund Grant 20050012. The article is our own responsibility and does not reflect the opinions of the Commonwealth Fund. We would like to acknowledge the assistance we received from Gail MacInnes and Alice Hedt at the National Citizens Coalition for Nursing Home Reform. Back

1 Department of Social & Behavioral Sciences, University of California, San Francisco. Back

2 Long Term Care Community Coalition, New York, NY. Back

Decision Editor: William J. McAuley, PhD

Received for publication April 21, 2007. Accepted for publication January 8, 2008.


   References
TOP
 Abstract
 Background on Regulation of...
 Conceptual Model
 Data Collection
 Analysis of Data
 Findings
 Discussion and Conclusions
 References
 





This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Google Scholar
Right arrow Articles by Harrington, C.
Right arrow Articles by Carrillo, H.
PubMed
Right arrow PubMed Citation
Right arrow Articles by Harrington, C.
Right arrow Articles by Carrillo, H.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
All GSA journals Journals of Gerontology Series A: Biological Sciences and Medical Sciences Journals of Gerontology Series B: Psychological Sciences and Social Sciences