Limitations of the facility-specific standardized mortality ratio for profiling health care quality in dialysis

Can we trust the standardized mortality ratio? A formal analysis and evaluation based on axiomatic requirements

BACKGROUND

The standardized mortality ratio (SMR) is often used to assess and compare hospital performance. While it has been recognized that hospitals may differ in their SMRs due to differences in patient composition, there is a lack of rigorous analysis of this and other-largely unrecognized-properties of the SMR.

METHODS

This paper proposes five axiomatic requirements for adequate standardized mortality measures: strict monotonicity (monotone relation to actual mortality rates), case-mix insensitivity (independence of patient composition), scale insensitivity (independence of hospital size), equivalence principle (equal rating of hospitals with equal actual mortality rates in all patient groups), and dominance principle (better rating of unambiguously better performing hospitals). Given these axiomatic requirements, effects of variations in patient composition, hospital size, and actual and expected mortality rates on the SMR were examined using basic algebra and calculus. In this regard, we distinguished between standardization using expected mortality rates derived from a different dataset (external standardization) and standardization based on a dataset including the considered hospitals (internal standardization). The results were illustrated by hypothetical examples.

RESULTS

Under external standardization, the SMR fulfills the axiomatic requirements of strict monotonicity and scale insensitivity but violates the requirement of case-mix insensitivity, the equivalence principle, and the dominance principle. All axiomatic requirements not fulfilled under external standardization are also not fulfilled under internal standardization. In addition, the SMR under internal standardization is scale sensitive and violates the axiomatic requirement of strict monotonicity.

CONCLUSIONS

The SMR fulfills only two (none) out of the five proposed axiomatic requirements under external (internal) standardization. Generally, the SMRs of hospitals are differently affected by variations in case mix and actual and expected mortality rates unless the hospitals are identical in these characteristics. These properties hamper valid assessment and comparison of hospital performance based on the SMR.

Characterizing Comorbidity in Dialysis Patients: Principles of Measurement and Applications in Risk Adjustment and Patient Care

Comorbid conditions are highly prevalent in dialysis patients and are significant predictors of mortality and other adverse outcomes. Accordingly, it is important to account for differences in comorbid illness burden among groups of dialysis patients being compared. At present, there is no consensus on what conditions matter, how each should be defined, and what weights each carries when defining an individual's risk or case-mix severity. A number of comorbidity instruments, generic or disease specific, have been employed in dialysis populations. They differ by the representation and definition of conditions as well as instrument scoring. No instrument has been found to be superior to another in terms of predictive accuracy for mortality, and accuracy across the board is low. Further studies are needed to determine whether improvements would be found with the use of more specifically defined items and through assignment of item weights based on relationships for outcomes specifically in a dialysis population. The roles of other factors in risk prediction, such as markers of nutritional status, inflammation, or other physiological parameters, relative to comorbid conditions also need to be defined. Outcomes other than mortality are likely to identify different factors and/or different relationships than those noted for mortality, which also require study.

Comorbidity is important for risk adjusting comparative analyses in nonrandomized trials and quality of care assessments and may, in future, influence payment for dialysis services. Efforts to improve the management of comorbid illnesses are needed. Comorbid conditions must be documented accurately and uniformly in all dialysis patients to enable these applications.

Center-Specific Risk-Adjusted Standardized Mortality Rates on Continuous Ambulatory Peritoneal Dialysis in China

Background

The aim of this study was to determine if there were centers in China with unusually high levels of risk-adjusted mortality in continuous ambulatory peritoneal dialysis (CAPD) patients.

Methods

We analyzed an inception cohort commencing CAPD between 1 January 2005 and 13 August 2015, followed until death, dropout defined as discontinuation of Baxter products, loss to follow-up, or 13 November 2015, whichever occurred first. We calculated standardized mortality ratios (SMRs) from Cox proportional hazards models, adjusting for age, gender, employment status, insurance status, primary renal disease, size of peritoneal dialysis (PD) program, and year of dialysis inception. We calculated 2 SMRs, 1 from models including a fixed effect for center of treatment, and 1 from stratified models.

Results

In this study, there was a 9.9% annual mortality rate in China, with decreasing mortality risk over time. There was significant variation of outcomes between Chinese centers, with up to 20% of facilities having SMRs indicating a higher risk-adjusted mortality rate than average. In particular, larger centers had better than expected mortality than smaller ones. There was significant misclassification of SMRs calculated using stratification versus fixed-effects models, although both showed directionally similar results.

Conclusion

Despite overall satisfactory and improving outcomes, our study showed a significant proportion of PD centers with higher than expected mortality. This is a signal for further assessment of these centers in China, after which there might be a range of actions taken depending on the results of the assessment and context, bearing in mind that the variation seen may be driven by factors unrelated to quality of care or beyond the control of hospital.

Risk Adjustment and the Assessment of Disparities in Dialysis Mortality Outcomes

Standardized mortality ratios (SMRs) reported by Medicare compare mortality at individual dialysis facilities with the national average, and are currently adjusted for race. However, whether the adjustment for race obscures or clarifies disparities in quality of care for minority groups is unknown. Cox model-based SMRs were computed with and without adjustment for patient race for 5920 facilities in the United States during 2010. The study population included virtually all patients treated with dialysis during this period. Without race adjustment, facilities with higher proportions of black patients had better survival outcomes; facilities with the highest percentage of black patients (top 10%) had overall mortality rates approximately 7% lower than expected. After adjusting for within-facility racial differences, facilities with higher proportions of black patients had poorer survival outcomes among black and non-black patients; facilities with the highest percentage of black patients (top 10%) had mortality rates approximately 6% worse than expected. In conclusion, accounting for within-facility racial differences in the computation of SMR helps to clarify disparities in quality of health care among patients with ESRD. The adjustment that accommodates within-facility comparisons is key, because it could also clarify relationships between patient characteristics and health care provider outcomes in other settings.

Semiparametric methods for center effect measures based on the ratio of survival functions

The survival function is often of chief interest in epidemiologic studies of time to an event. We develop methods for evaluating center-specific survival outcomes through a ratio of survival functions. The proposed method assumes a center-stratified additive hazards model, which provides a convenient framework for our purposes. Under the proposed methods, the center effects measure is cast as the ratio of subject-specific survival functions under two scenarios: the scenario in which the subject is treated at center [Formula: see text]; and that wherein the subject is treated at a hypothetical center with survival function equal to the population average. The proposed measure reduces to the ratio of baseline survival functions, but is invariant to the choice of baseline covariate level. We derive the asymptotic properties of the proposed estimators, and assess finite-sample characteristics through simulation. The proposed methods are applied to national kidney transplant data.

Hemodialysis facility-based quality-of-care indicators and facility-specific patient outcomes

BACKGROUND

We evaluated whether incremental achievement of up to 8 facility quality goals was associated with improvement in facility-specific mortality and hospitalization rates.

STUDY DESIGN

Prospective observational study.

SETTING & PARTICIPANTS

1,085 Fresenius Medical Care, North America facilities providing hemodialysis (HD) for 25 or more patients during January 2006.

MEASUREMENTS

The facility average for the period up to December 31, 2006, was used to determine achievement of each goal for equilibrated Kt/V, missed HD treatments, hemoglobin level, bicarbonate level, albumin level, phosphorus level, fistulae, and HD catheters. Linear regression models were used to relate facility-wide achievement of goals with facility-specific hospital days and standardized mortality ratios.

RESULTS

Most facilities (64%) achieved 2 to 4 of 8 goals, with only 8% meeting more than 5 quality goals. Achieving more than 5 goals averaged 3.5 fewer hospital days/patient-year and 20% lower standardized mortality ratios (all P < 0.001). The incremental number of goals met also was associated with improvement in facility mortality (P < 0.001) and hospital days (P < 0.001). Catheter and albumin level goals were achieved least (6% and 9% of facilities, respectively), but they had the best outcomes. Facilities achieving more than 5 goals had older patients (64.0 versus 61.5 years; P < 0.001), fewer African American patients (16% versus 38%; P < 0.001), and fewer women (44% versus 46%; P = 0.003) compared with the average.

LIMITATIONS

Observational design with residual confounding from unmeasured patient-, facility-, and treatment-related factors.

CONCLUSIONS

Achieving more facility quality goals was significantly associated with better facility-based measurements of patient outcomes. Although these results do not establish a causal relationship, findings agree with the present practice of monitoring facility performance for continuous quality improvement.

Associates of mortality and hospitalization in hemodialysis: potentially actionable laboratory variables and vascular access

BACKGROUND

To determine the most significant potentially actionable clinical variables associated with mortality and hospitalization risk in hemodialysis (HD) patients.

STUDY DESIGN

Cohort study.

SETTING & PARTICIPANTS

Adult maintenance HD patients in the Fresenius Medical Care, North America database as of January 1, 2004, with baseline information from October 1, 2003, to December 31, 2003, comprising approximately 26% of the US HD population.

PREDICTORS

Case-mix (age, sex, race, diabetes, vintage, and body surface area), vascular access, and laboratory (albumin, equilibrated Kt/V, hemoglobin, calcium, phosphorus, creatinine, bicarbonate, biointact parathyroid hormone, transferrin saturation, and white blood cell count) variables.

OUTCOMES

1-year mortality and hospitalization risk from January 1 to December 31, 2004.

MEASUREMENTS

Cox proportional hazards models for death and hospitalization.

RESULTS

The cohort (N = 78,420) had a mean age of 61.4 +/- 15.0 years, 47% were women, 49% were white, 41% were black race (10% defined as "other"), and 52% had diabetes. The top 5 actionable variables were the same for mortality and hospitalization. Final case-mix plus laboratory-adjusted hazard ratios for these top 5 actionable variables indicate 177% increased risk of death and 67% increased risk of hospitalization per 1-g/dL decrease in albumin level, 39% and 45% greater risk with catheters compared with fistulas, 18% and 9% greater risk per 1-mg/dL greater phosphorus level, 11% and 9% lower risk per 1-g/dL greater hemoglobin level, and 5% and 2% greater risk per 0.1-unit decrease in equilibrated Kt/V, respectively (all P < 0.0001).

LIMITATIONS

Observational cross-sectional study with limited comorbidity adjustment (for diabetes).

CONCLUSION

The same variables are associated with both mortality and hospitalization in HD patients. The top 5 potentially actionable variables are readily identifiable, with albumin level and catheter use the most prominent, and all 5 are appropriate targets for improvement.

Ranking USRDS provider specific SMRs from 1998-2001

Provider profiling (ranking/percentiling) is prevalent in health services research. Bayesian models coupled with optimizing a loss function provide an effective framework for computing non-standard inferences such as ranks. Inferences depend on the posterior distribution and should be guided by inferential goals. However, even optimal methods might not lead to definitive results and ranks should be accompanied by valid uncertainty assessments. We outline the Bayesian approach and use estimated Standardized Mortality Ratios (SMRs) in 1998-2001 from the United States Renal Data System (USRDS) as a platform to identify issues and demonstrate approaches. Our analyses extend Liu et al. (2004) by computing estimates developed by Lin et al. (2006) that minimize errors in classifying providers above or below a percentile cut-point, by combining evidence over multiple years via a first-order, autoregressive model on log(SMR), and by use of a nonparametric prior. Results show that ranks/percentiles based on maximum likelihood estimates of the SMRs and those based on testing whether an SMR = 1 substantially under-perform the optimal estimates. Combining evidence over the four years using the autoregressive model reduces uncertainty, improving performance over percentiles based on only one year. Furthermore, percentiles based on posterior probabilities of exceeding a properly chosen SMR threshold are essentially identical to those produced by minimizing classification loss. Uncertainty measures effectively calibrate performance, showing that considerable uncertainty remains even when using optimal methods. Findings highlight the importance of using loss function guided percentiles and the necessity of accompanying estimates with uncertainty assessments.

N-terminal Pro-B–Type Natriuretic Peptide (NT-proBNP) Concentrations in Hemodialysis Patients: Prognostic Value of Baseline and Follow-up Measurements

Background: Increased N-terminal pro-B–type natriuretic peptide (NT-proBNP) concentrations are associated with increased cardiovascular mortality in chronic hemodialysis patients. Previous studies focused on prevalent dialysis patients and examined single measurements of NT-proBNP in time.

Methods: We measured NT-proBNP concentrations in 2990 incident hemodialysis patients to examine the risk of 90-day and 1-year mortality associated with baseline NT-proBNP concentrations. In addition, we calculated the change in concentrations after 3 months in a subset of 585 patients to examine the association between longitudinal changes in NT-proBNP and subsequent mortality.

Results: Increasing quartiles of NT-proBNP were associated with a monotonic increase in 90-day [quartile 1, referent; from quartile 2 to quartile 4, hazard ratio (HR) 1.7–6.3, P &lt; 0.001] and 1-year (quartile 1, referent; from quartile 2 to quartile 4, HR 1.7–4.9, P &lt; 0.001) all-cause mortality. After multivariable adjustment, these associations remained robust. When examined using a multivariable fractional polynomial, increased NT-proBNP concentrations were associated with increased 90-day (HR per unit increase in log NT-proBNP 1.5, 95% CI 1.3–1.7) and 1-year (HR per unit increase in log NT-proBNP 1.4, 95% CI 1.3–1.5) all-cause mortality. In addition, patients with the greatest increase in NT-proBNP after 3 months of dialysis had a 2.4-fold higher risk of mortality than those with the greatest decrease in NT-proBNP.

Conclusions: NT-proBNP concentrations are independently associated with mortality in incident hemodialysis patients. Furthermore, the observation that longitudinal changes in NT-proBNP concentrations were associated with subsequent mortality suggests that monitoring serial NT-proBNP concentrations may represent a novel tool for assessing adequacy and guiding therapy in patients initiating hemodialysis.

Vitamin D levels and early mortality among incident hemodialysis patients

Vitamin D deficiency is associated with cardiovascular disease, the most common cause of mortality in hemodialysis patients. To investigate the relation between blood levels of 25-hydroxyvitamin D (25D) and 1,25-dihydroxyvitamin D (1,25D) with hemodialysis outcomes, we measured baseline vitamin D levels in a cross-sectional analysis of 825 consecutive patients from within a prospective cohort of incident US hemodialysis patients. Of these patients, 78% were considered vitamin D deficient with 18% considered severely deficient. Calcium, phosphorus, and parathyroid hormone levels correlated poorly with 25D and 1,25D concentrations. To test the association between baseline vitamin D levels and 90-day mortality, we selected the next 175 consecutive participants who died within 90 days and compared them to the 750 patients who survived in a nested case-control analysis. While low vitamin D levels were associated with increased mortality, significant interaction was noted between vitamin D levels, subsequent active vitamin D therapy, and survival. Compared to patients with the highest 25D or 1,25D levels who received therapy, untreated deficient patients were at significantly increased risk for early mortality. Our study shows that among incident hemodialysis patients, vitamin D deficiency is common, correlates poorly with other components of mineral metabolism and is associated with increased early mortality.

Attainment of clinical performance targets and improvement in clinical outcomes and resource use in hemodialysis care: a prospective cohort study

BACKGROUND

Clinical performance targets are intended to improve patient outcomes in chronic disease through quality improvement, but evidence of an association between multiple target attainment and patient outcomes in routine clinical practice is often lacking.

METHODS

In a national prospective cohort study (ESRD Quality, or EQUAL), we examined whether attainment of multiple targets in 668 incident hemodialysis patients from 74 U.S. not-for-profit dialysis clinics was associated with better outcomes. We measured whether the following accepted clinical performance targets were met at 6 months after study enrollment: albumin (> or =4.0 g/dl), hemoglobin (> or =11 g/dl), calcium-phosphate product (<55 mg2/dl2), dialysis dose (Kt/V> or =1.2), and vascular access type (fistula). Outcomes included mortality, hospital admissions, hospital days, and hospital costs.

RESULTS

Attainment of each of the five targets was associated individually with better outcomes; e.g., patients who attained the albumin target had decreased mortality [relative hazard (RH) = 0.55, 95% confidence interval (CI), 0.41-0.75], hospital admissions [incidence rate ratio (IRR) = 0.67, 95% CI, 0.62-0.73], hospital days (IRR = 0.61, 95% CI, 0.58-0.63), and hospital costs (average annual cost reduction = 3,282 dollars, P = 0.002), relative to those who did not. Increasing numbers of targets attained were also associated, in a graded fashion, with decreased mortality (P = 0.030), fewer hospital admissions and days (P < 0.001 for both), and lower costs (P = 0.029); these trends remained statistically significant for all outcomes after adjustment (P < 0.001), except cost, which was marginally significant (P = 0.052).

CONCLUSION

Attainment of more clinical performance targets, regardless of which targets, was strongly associated with decreased mortality, hospital admissions, and resource use in hemodialysis patients.

Predicting hospitalization and mortality in end-stage renal disease (ESRD) patients using an Index of Coexisting Disease (ICED)-based risk stratification model

We evaluated the use of an additive Index of Coexisting Diseases (ICED)-based stratification schema to determine subsequent hospitalization and mortality in a hemodialysis population. Patients from five commercial health plans were stratified into low-, medium-, and high-risk groups and followed for up to 1 year. Patients were reassessed and restratified at 90-day intervals and censored when disease management ceased. Outcome measures collected through selfreports and health plan records were captured in an active database. Survival to first hospitalization/ mortality was compared by Kaplan Meier curves, survivor function differences by the Wilcoxon test, and group comparisons by ANOVA and chi square. Population characteristics included mean age of 63.0, 57.7% male, and 58.8% diabetic. Mortality was 13.0% per patient year (standardized mortality ratio 0.43) and the hospitalization rate was 0.59 per patient year (standardized hospitalization ratio 0.24). Survival curves demonstrated differences in mortality and hospitalization between the patients in different initial risk categories (p < 0.01). Mean hospitalizations were 0.81 +/- 1.53 per patient year (high risk), 0.45 +/- 0.99 (medium risk), and 0.15 +/- 0.51 for the low-risk group (p < 0.001). Stratification was dynamic; 47.3% decreased and 4.7% increased risk level between the first and second assessment. These changes were associated with survival differences for initial low (p = 0.06) or medium patients (p < 0.01), and hospital-free survival for initial medium (p = 0.08) or high patients (p < 0.05). In conclusion, this ICED-based stratification schema predicted mortality and hospitalization for hemodialysis patients participating in our disease management program.

Loss Function Based Ranking in Two-Stage, Hierarchical Models

Performance evaluations of health services providers burgeons. Similarly, analyzing spatially related health information, ranking teachers and schools, and identification of differentially expressed genes are increasing in prevalence and importance. Goals include valid and efficient ranking of units for profiling and league tables, identification of excellent and poor performers, the most differentially expressed genes, and determining "exceedances" (how many and which unit-specific true parameters exceed a threshold). These data and inferential goals require a hierarchical, Bayesian model that accounts for nesting relations and identifies both population values and random effects for unit-specific parameters. Furthermore, the Bayesian approach coupled with optimizing a loss function provides a framework for computing non-standard inferences such as ranks and histograms.Estimated ranks that minimize Squared Error Loss (SEL) between the true and estimated ranks have been investigated. The posterior mean ranks minimize SEL and are "general purpose," relevant to a broad spectrum of ranking goals. However, other loss functions and optimizing ranks that are tuned to application-specific goals require identification and evaluation. For example, when the goal is to identify the relatively good (e.g., in the upper 10%) or relatively poor performers, a loss function that penalizes classification errors produces estimates that minimize the error rate. We construct loss functions that address this and other goals, developing a unified framework that facilitates generating candidate estimates, comparing approaches and producing data analytic performance summaries. We compare performance for a fully parametric, hierarchical model with Gaussian sampling distribution under Gaussian and a mixture of Gaussians prior distributions. We illustrate approaches via analysis of standardized mortality ratio data from the United States Renal Data System.Results show that SEL-optimal ranks perform well over a broad class of loss functions but can be improved upon when classifying units above or below a percentile cut-point. Importantly, even optimal rank estimates can perform poorly in many real-world settings; therefore, data-analytic performance summaries should always be reported.

Activated injectable vitamin D and hemodialysis survival: a historical cohort study

Patients with ESRD commonly experience secondary hyperparathyroidism, a condition primarily managed with activated injectable vitamin D. The biologic effects of vitamin D, however, are widespread, and it is possible that activated injectable vitamin D alters survival in ESRD. This hypothesis was tested in a historical cohort study of incident hemodialysis patients who lived throughout the United States between January 1996 and December 1999. The primary outcome was 2-yr survival among those who survived for at least 90 d after initiation of chronic hemodialysis. During this period, 51,037 chronic hemodialysis patients survived for at least 90 d from the initiation of hemodialysis, and in the ensuing 2 yr, 37,173 received activated injectable vitamin D and 13,864 did not. At 2 yr, mortality rates were 13.8/100 person-years in the group that received injectable vitamin D compared with 28.6/100 person-years in the group that did not (P < 0.001). Cox proportional hazards analyses adjusting for several potential confounders and examining injectable vitamin D therapy as a time-dependent exposure suggested that compared with patients who did not receive injectable vitamin D, the 2-yr survival advantage associated with the group that did receive injectable vitamin D was 20% (hazard ratio, 0.80; 95% confidence interval, 0.76 to 0.83). The incidence of cardiovascular-related mortality was 7.6/100 person-years in the injectable vitamin D group, compared with 14.6/100 person-years in the non-vitamin D group (P < 0.001). The benefit of injectable vitamin D was evident in 48 of 49 strata examined, including those with low serum levels of intact parathyroid hormone and elevated levels of serum calcium and phosphorus, situations in which injectable vitamin D is often withheld. Repeating the entire analysis using marginal structural models to adjust for time-dependent confounding by indication yielded a survival advantage of 26% (hazard ratio, 0.74; 95% confidence interval, 0.71 to 0.79) associated with the injectable vitamin D group. In this historical cohort study, chronic hemodialysis patients in the group that received injectable vitamin D had a significant survival advantage over patients who did not. Randomized clinical trials would permit definitive conclusions.

Methods for Estimating and Interpreting Provider-Specific Standardized Mortality Ratios

Standardized Mortality Ratios (SMRs) are widely used as a measurement of quality of care for profiling and otherwise comparing medical care providers. Invalid estimation or inappropriate interpretation may have serious local and national consequences. Estimating an SMR entails producing provider-specific expected deaths via a statistical model and then computing the "observed/expected" ratio. Appropriate comparison of estimated SMRs requires considering both estimated values and statistical uncertainty. With statistical uncertainty that varies over providers, hypothesis testing to identify poor performers unfairly penalizes large providers; use of direct estimates unfairly penalizes small providers. Since neither approach suffices, we report on a suite of comparisons, each addressing an important aspect of the comparison. Our approach is based on a hierarchical statistical model. Goals include estimating and ranking (percentiling) provider-specific SMRs and calculating the probability that a provider's true SMR percentile falls within a specified percentile range. We present the issues and related statistical models for comparing SMRs and apply our approaches to the 1998 United States Renal Data System (USRDS) dialysis provider data.

Lack of correlation between facility-based standardized rates of transplantation and mortality

The standardized mortality ratio (SMR) has been used to provide information about adjusted survival outcomes at dialysis facilities. There has been concern that high rates of transplantation could unjustly lead to unfavorable SMR profiles for individual dialysis units because healthier patients would be removed from dialysis therapy, leaving less healthy patients in the dialysis pool. We correlated 1999 overall adjusted SMR and 1999 standardized transplantation ratio (STR) weighted for mortality patient count and count of first transplantations of patients younger than 65 years. A total of 2,362 facilities were included in analyses. We found no correlation between rates of transplantation (by STR) and overall mortality profile (by SMR) based on Pearson's correlation coefficients (r), either unweighted, weighted by number of patients included in the 1999 mortality calculation (SMR), or weighted by number of patients included in the 1999 transplantation calculation (r = -0.016, r = -0.015, and r = -0.015, respectively; P > 0.40 for each). Sensitivity analyses using SMR and STR over 3- and 3.5-year periods (January 1997 to June 2000) also showed no correlation between SMR and STR, respectively. We conclude that reported standardized rates for transplantation do not correlate with those reported for mortality by dialysis facilities.

Cost-quality trade-offs in dialysis care: a national survey of dialysis facility administrators

Dialysis facilities face important trade-offs between cost and quality under constrained capitated reimbursement. How management at dialysis facilities makes decisions affecting cost and quality of care and views opportunities and threats is unknown. We conducted a national survey of dialysis facility administrators. We asked administrators what changes they would make in response to increases or decreases in reimbursement, their views on linking dialysis care payment to quality-of-care measures, and their views on providing patients with treatment options and outcomes information. One hundred fifty-seven of 280 dialysis facility administrators (56%) responded. If dialysis reimbursement were to increase by 20%, the five most common responses were to: improve patient education programs (62% of respondents), improve facility amenities (42%), purchase new equipment (30%), provide more money for staff salaries (28%), and increase number of nursing staff (21%). Conversely, if dialysis reimbursement were to decrease by 20%, the most common responses were to: limit staff salary (45% of respondents), decrease nursing staff (41%), not replace dialysis equipment (43%), increase dialyzer reuse (37%), and return less to investors (36%). Differences in rank order of responses were observed according to professional training of the administrator and profit status of the facility. Administrators uniformly believe that it is very acceptable to provide facility-specific outcomes data to the public, as well as information on modalities of treatment provided by facilities. However, administrators varied in their views regarding whether reimbursement should be based on quality by using a process-of-care measure, such as the average dose of dialysis, or an outcome-of-care measure, such as case-mix-adjusted mortality rates. We conclude that increases in facility reimbursement generally would be used by dialysis facility administrators for the benefit of patients, whereas decreases (or inflation erosion) in payment rates might compromise staffing. US dialysis administrators support sharing treatment options and outcomes information with patients, but appear to be ambivalent with regard to linking reimbursement to adequacy of dialysis or patient outcomes. These results have important implications regarding proposed changes in the US capitated dialysis payment rate and current efforts to empower consumers of dialysis care.

Association between prevalent care process measures and facility-specific mortality rates

BACKGROUND

Medical communities often develop practice guidelines recommending certain care processes intended to promote better clinical outcome among patients. Conformance with those guidelines by facilities is then monitored to evaluate care quality, presuming that the process is associated with and can be used reliably to predict clinical outcome. Outcome is often monitored as a facility-specific mortality rate (SMR) standardized to the mix of patients treated, also presuming that inferior outcome implies a suboptimal process. The U.S. Health Care Financing Administration monitors three practice guidelines, called Core Indicators, in dialysis facilities to assist management of its end-stage renal disease program: (1) patients' hematocrit values should exceed 30 vol%, (2) the urea reduction ratio (URR) during dialysis should equal or exceed 65%, and (3) patients' serum albumin concentrations should equal or exceed 3.5 g/dL.

METHODS

The associations of a facility-specific SMR were evaluated with the fractions of hemodialysis patients not conforming to (that is, at variance with) the Core Indicators during three successive years (1993 to 1995) in large numbers of facilities (394, 450, and 498) using one-variable and multivariable statistical models. Three related strategies were used. First, the association of the SMR with the fraction of patients not meeting the guideline was evaluated. Second, each facility was classified by whether its SMR exceeded the 80% confidence interval above 1.0 (worse than 1.0, Group 3), was less than the interval below 1.0 (better than 1.0, Group 1), or was within the interval (Group 2). The fraction of those patients who did not meet the Indicator guidelines was then evaluated in each group. Third, the ability of variance from Indicator guidelines to predict into which of the three SMR groups a facility would be categorized was evaluated.

RESULTS

SMR was directly correlated with variance from the Indicator guidelines, but the strengths of the associations were weak particularly for the hematocrit (R(2) = 2.2%, 5.6, and 2.2 for each of the 3 years) and URR Indicators (R(2) = 2.6, 0.6, 3.3). It was stronger for the albumin Indicator (R(2) = 11.6, 20.4, 21.8). The fractions of patients falling outside of the Indicator guidelines tended to be higher in the highest SMR group. The groups were not well separated, however, particularly for the hematocrit and URR Indicators, and there was substantial overlap between them. Finally, although the likelihood that a facility would be a member of the high or low SMR group was associated with fractional variance from Core Indicator guidelines, the strengths of association were weak, and the probability that a facility would be a member of the high or low group could not be easily distinguished from the probability that it would be a member of the middle group.

CONCLUSIONS

While there were statistical associations between SMR and the fraction of patients in facilities who were at variance with these guidelines, they were weak and variances from the guidelines could not be used reliably to predict high or low SMR. Such findings do not imply that measures reflecting anemia, dialysis dose, or medical processes that influence serum albumin concentration are irrelevant to the quality of care. They do suggest, however, that more attention needs be paid to these and other associates and causes of mortality among dialysis patients when developing care process indicator guidelines.