Similar risk of complete revision for infection with single-dose versus multiple-dose antibiotic prophylaxis in primary arthroplasty of the hip and knee: results of an observational cohort study in the Dutch Arthroplasty Register in 242,179 patients

Background and purpose - The optimal type and duration of antibiotic prophylaxis for primary arthroplasty of the hip and knee are subject to debate. We compared the risk of complete revision (obtained by a 1- or 2-stage procedure) for periprosthetic joint infection (PJI) after primary total hip or knee arthroplasty between patients receiving a single dose of prophylactic antibiotics and patients receiving multiple doses of antibiotics for prevention of PJI. Patients and methods - A cohort of 130,712 primary total hip and 111,467 knee arthroplasties performed between 2011 and 2015 in the Netherlands was analyzed. We linked data from the Dutch arthroplasty register to a survey collected across all Dutch institutions on hospital-level antibiotic prophylaxis policy. We used restricted cubic spline Poisson models adjusted for hospital clustering to compare the risk of revision for infection according to type and duration of antibiotic prophylaxis received. Results - For total hip arthroplasties, the rates of revision for infection were 31/10,000 person-years (95% CI 28-35), 39 (25-59), and 23 (15-34) in the groups that received multiple doses of cefazolin, multiple doses of cefuroxime, and a single dose of cefazolin, respectively. The rates for knee arthroplasties were 27/10,000 person-years (95% CI 24-31), 40 (24-62), and 24 (16-36). Similar risk of complete revision for infection among antibiotic prophylaxis regimens was found when adjusting for confounders. Interpretation - In a large observational cohort we found no apparent association between the type or duration of antibiotic prophylaxis and the risk of complete revision for infection. This does question whether there is any advantage to the use of prolonged antibiotic prophylaxis beyond a single dose.

Variability in Smoking Status for Lobectomy Among Society of Thoracic Surgeons Database Participants

BACKGROUND

Current smokers undergoing lobectomy are at greater risk of complications than are former smokers. The Society of Thoracic Surgeons (STS) composite score for rating program performance for lobectomy adjusts for smoking status, a modifiable risk factor. This study examined variability in the proportion of current smokers undergoing lobectomy among STS database participants. Additionally, the study determined whether each participant's rating changed if smoking was excluded from the risk adjustment model.

METHODS

This is a retrospective analysis of the STS cohort used to develop the composite score for rating program performance for lobectomy. The study summarized the variability among STS database participants for performing lobectomy on current smokers and compared star ratings developed from models with and without smoking status.

RESULTS

There were 24,912 patients with smoking status data: 23% current smokers, 62% former smokers, and 15% never smokers. There was significant variability among participants in the proportion of current smokers undergoing lobectomy (3% to 48.6%; P < .001). Major morbidity or mortality (composite) was greater in current smokers (12.1%) than in former smokers (8.6%) and never smokers (4.2%) (P < .001). Using the current risk adjustment model, participant star ratings were as follows: 1 star, n = 6 (3.2%); 2 stars, n = 170 (91.4%); and 3 stars, n = 10 (5.4%). When smoking status was excluded from the model, 1 participant shifted from a 2-star to a 3-star program.

CONCLUSIONS

There is substantial variability among STS database participants with regard to the proportion of current smokers undergoing lobectomy. However, exclusion of smoking status from the model did not significantly affect participant star rating.

Risk Adjustment In Medicare ACO Program Deters Coding Increases But May Lead ACOs To Drop High-Risk Beneficiaries

The Medicare Shared Savings Program (MSSP) adjusts savings benchmarks by beneficiaries' baseline risk scores. To discourage increased coding intensity, the benchmark is not adjusted upward if beneficiaries' risk scores rise while in the MSSP. As a result, accountable care organizations (ACOs) have an incentive to avoid increasingly sick or expensive beneficiaries. We examined whether beneficiaries' exposure to the MSSP was associated with within-beneficiary changes in risk scores and whether risk scores were associated with entry to or exit from the MSSP. We found that the MSSP was not associated with consistent changes in within-beneficiary risk scores. Conversely, beneficiaries at the ninety-fifth percentile of risk score had a 21.6 percent chance of exiting the MSSP, compared to a 16.0 percent chance among beneficiaries at the fiftieth percentile. The decision not to upwardly adjust risk scores in the MSSP has successfully deterred coding increases but might discourage ACOs to care for high-risk beneficiaries in the MSSP .

Assessing the Impact of Body Mass Index Information on the Performance of Risk Adjustment Models in Predicting Health Care Costs and Utilization

BACKGROUND

Using electronic health records (EHRs) for population risk stratification has gained attention in recent years. Compared with insurance claims, EHRs offer novel data types (eg, vital signs) that can potentially improve population-based predictive models of cost and utilization.

OBJECTIVE

To evaluate whether EHR-extracted body mass index (BMI) improves the performance of diagnosis-based models to predict concurrent and prospective health care costs and utilization.

METHODS

We used claims and EHR data over a 2-year period from a cohort of continuously insured patients (aged 20-64 y) within an integrated health system. We examined the addition of BMI to 3 diagnosis-based models of increasing comprehensiveness (ie, demographics, Charlson, and Dx-PM model of the Adjusted Clinical Group system) to predict concurrent and prospective costs and utilization, and compared the performance of models with and without BMI.

RESULTS

The study population included 59,849 patients, 57% female, with BMI class I, II, and III comprising 19%, 9%, and 6% of the population. Among demographic models, R improvement from adding BMI ranged from 61% (ie, R increased from 0.56 to 0.90) for prospective pharmacy cost to 29% (1.24-1.60) for concurrent medical cost. Adding BMI to demographic models improved the prediction of all binary service-linked outcomes (ie, hospitalization, emergency department admission, and being in top 5% total costs) with area under the curve increasing from 2% (0.602-0.617) to 7% (0.516-0.554). Adding BMI to Charlson models only improved total and medical cost predictions prospectively (13% and 15%; 4.23-4.79 and 3.30-3.79), and also improved predicting all prospective outcomes with area under the curve increasing from 3% (0.649-0.668) to 4% (0.639-0.665; and, 0.556-0.576). No improvements in prediction were seen in the most comprehensive model (ie, Dx-PM).

DISCUSSION

EHR-extracted BMI levels can be used to enhance predictive models of utilization especially if comprehensive diagnostic data are missing.

Risk-adjusted monitoring of surgical performance

We propose a nonparametric risk-adjusted cumulative sum chart to monitor surgical outcomes for patients with different risks of post-operative mortality due to risk factors that exist before the surgery. Using varying-coefficient logistic regression models, we accomplish the risk adjustment. Unknown coefficient functions are estimated by global polynomial spline approximation based on the maximum likelihood principle. We suggest a bisection minimization approach and a bootstrap method to determine the chart testing limit value. Compared with the previous (parametric) risk-adjusted cumulative sum chart, a major advantage of our method is that the morality rate can be modeled more flexibly by related covariates, which significantly enhances the monitoring efficiency. Simulations demonstrate nice performance of our proposed procedure. An application to a UK cardiac surgery dataset illustrates the use of our methodology.

Medicare risk-adjusted outcomes in elective major vascular surgery

BACKGROUND

Risk-adjusted outcomes of elective major vascular surgery that is inclusive of inpatient and 90-day post-discharge adverse outcomes together have not been well studied.

METHODS

We studied 2012-2014 Medicare inpatients who received open aortic procedures, open peripheral vascular procedures, endovascular aortic procedures, and percutaneous angioplasty procedures of the lower extremity for risk-adjusted adverse outcomes of inpatient deaths, 3-sigma prolonged length-of-stay outliers, 90-day post-discharge deaths without readmission, and 90-day post-discharge associated readmissions after excluding unrelated events. Observed and predicted total adverse outcomes for hospitals meeting minimum risk-volume criteria were assessed and hospital-specific z-scores and risk-adjusted adverse outcomes were calculated to compare performance.

RESULTS

The total adverse-outcome rate was 27.8% for open aortic procedures, 31.5% for open peripheral vascular procedures, 19.6% for endovascular aortic procedures, and 36.4% for percutaneous angioplasty procedures. The difference in risk-adjusted adverse-outcome rates between the best- and the poorest-performing deciles were 32.2% for open aortic procedures, 29.5% for open peripheral vascular procedures, 21.5% for endovascular aortic procedures, and 37.1% for percutaneous angioplasty procedures. The 90-day post-discharge deaths and readmissions were the major driver of overall adverse-outcome rates.

CONCLUSION

The variability in risk-adjusted outcomes among best- and poorest-performing hospitals is over 20% in all major vascular procedures and indicates that a large opportunity exists for improvement in results.

Outlier classification performance of risk adjustment methods when profiling multiple providers

BACKGROUND

When profiling multiple health care providers, adjustment for case-mix is essential to accurately classify the quality of providers. Unfortunately, misclassification of provider performance is not uncommon and can have grave implications. Propensity score (PS) methods have been proposed as viable alternatives to conventional multivariable regression. The objective was to assess the outlier classification performance of risk adjustment methods when profiling multiple providers.

METHODS

In a simulation study based on empirical data, the classification performance of logistic regression (fixed and random effects), PS adjustment, and three PS weighting methods was evaluated when varying parameters such as the number of providers, the average incidence of the outcome, and the percentage of outliers. Traditional classification accuracy measures were considered, including sensitivity and specificity.

RESULTS

Fixed effects logistic regression consistently had the highest sensitivity and negative predictive value, yet a low specificity and positive predictive value. Of the random effects methods, PS adjustment and random effects logistic regression performed equally well or better than all the remaining PS methods for all classification accuracy measures across the studied scenarios.

CONCLUSIONS

Of the evaluated PS methods, only PS adjustment can be considered a viable alternative to random effects logistic regression when profiling multiple providers in different scenarios.

A risk-model for hospital mortality among patients with severe sepsis or septic shock based on German national administrative claims data

BACKGROUND

Sepsis is a major cause of preventable deaths in hospitals. Feasible and valid methods for comparing quality of sepsis care between hospitals are needed. The aim of this study was to develop a risk-adjustment model suitable for comparing sepsis-related mortality between German hospitals.

METHODS

We developed a risk-model using national German claims data. Since these data are available with a time-lag of 1.5 years only, the stability of the model across time was investigated. The model was derived from inpatient cases with severe sepsis or septic shock treated in 2013 using logistic regression with backward selection and generalized estimating equations to correct for clustering. It was validated among cases treated in 2015. Finally, the model development was repeated in 2015. To investigate secular changes, the risk-adjusted trajectory of mortality across the years 2010-2015 was analyzed.

RESULTS

The 2013 deviation sample consisted of 113,750 cases; the 2015 validation sample consisted of 134,851 cases. The model developed in 2013 showed good validity regarding discrimination (AUC = 0.74), calibration (observed mortality in 1st and 10th risk-decile: 11%-78%), and fit (R2 = 0.16). Validity remained stable when the model was applied to 2015 (AUC = 0.74, 1st and 10th risk-decile: 10%-77%, R2 = 0.17). There was no indication of overfitting of the model. The final model developed in year 2015 contained 40 risk-factors. Between 2010 and 2015 hospital mortality in sepsis decreased from 48% to 42%. Adjusted for risk-factors the trajectory of decrease was still significant.

CONCLUSIONS

The risk-model shows good predictive validity and stability across time. The model is suitable to be used as an external algorithm for comparing risk-adjusted sepsis mortality among German hospitals or regions based on administrative claims data, but secular changes need to be taken into account when interpreting risk-adjusted mortality.

Imperfect risk adjustment, risk preferences, and sorting in competitive health insurance markets

I develop a model of insurer price-setting and consumer welfare under risk-adjustment, a policy commonly used to combat inefficient sorting due to adverse selection in health insurance markets. I use the model to illustrate graphically that risk-adjustment causes health plan prices to be based on costs not predicted by the risk-adjustment model ("residual costs") rather than total costs, either weakening or exacerbating selection problems depending on the correlation between demand and costs predicted by the risk-adjustment model. I then use a structural model to estimate the welfare consequences of risk-adjustment, finding a welfare gain of over $600 per person-year.

Adjusted Analyses in Studies Addressing Therapy and Harm: Users' Guides to the Medical Literature

Observational studies almost always have bias because prognostic factors are unequally distributed between patients exposed or not exposed to an intervention. The standard approach to dealing with this problem is adjusted or stratified analysis. Its principle is to use measurement of risk factors to create prognostically homogeneous groups and to combine effect estimates across groups.The purpose of this Users' Guide is to introduce readers to fundamental concepts underlying adjustment as a way of dealing with prognostic imbalance and to the basic principles and relative trustworthiness of various adjustment strategies.One alternative to the standard approach is propensity analysis, in which groups are matched according to the likelihood of membership in exposed or unexposed groups. Propensity methods can deal with multiple prognostic factors, even if there are relatively few patients having outcome events. However, propensity methods do not address other limitations of traditional adjustment: investigators may not have measured all relevant prognostic factors (or not accurately), and unknown factors may bias the results.A second approach, instrumental variable analysis, relies on identifying a variable associated with the likelihood of receiving the intervention but not associated with any prognostic factor or with the outcome (other than through the intervention); this could mimic randomization. However, as with assumptions of other adjustment approaches, it is never certain if an instrumental variable analysis eliminates bias.Although all these approaches can reduce the risk of bias in observational studies, none replace the balance of both known and unknown prognostic factors offered by randomization.

Comparison of the Properties of Regression and Categorical Risk-Adjustment Models

Clinical risk-adjustment, the ability to standardize the comparison of individuals with different health needs, is based upon 2 main alternative approaches: regression models and clinical categorical models. In this article, we examine the impact of the differences in the way these models are constructed on end user applications.

High-risk centers and the benefits for lower-risk transplants

OBJECTIVES

Allogeneic hematopoietic cell transplantation (HCT) is the transplantation of stem cells from a donor and an effective treatment for many hematologic malignancies. We sought to compare allogeneic HCT survival outcomes and hazard of death among US centers that treat higher-risk patients versus those in centers that do not perform lower-risk HCT procedures.

STUDY DESIGN

We utilized 2008 to 2010 Center for International Blood and Marrow Transplant Research data. We categorized patients into 4 risk categories that align with factors shown in the literature to be associated with HCT survival. We stratified centers into those that do and do not conduct high-risk pre-transplant HCT.

METHODS

To further evaluate the association between pre-transplant mortality risk and HCT survival by transplant center, we examined the association between risk category score and hazard of death using Cox proportional hazard modeling.

RESULTS

There were 12,436 HCT recipients at 147 transplant centers. Of the 147 centers, 74 performed HCT for patients ranging from the lowest risk category to the highest category, and 73 centers performed only lower-risk HCT. Adjusting for all other factors, lower-risk patients that underwent transplants in lower- or higher-risk centers had a similar relative hazard of death (P ≤ .05).

CONCLUSIONS

Low-risk patients had similar survival outcomes irrespective of whether they underwent transplant at higher- or lower-risk centers. Patient and payer policy implications could include initiatives that reduce travel for low-risk patients. Similarly, HCT center administrators and providers that manage higher-risk patients need not expect commensurate benefits in survival for lower-risk patients.

The interaction of direct and indirect risk selection

This paper analyzes the interaction of direct and indirect risk selection in health insurance markets. It is shown that direct risk selection - using measures unrelated to the benefit package like selective advertising or 'losing' applications of high risk individuals - nevertheless has an influence on the distortions of the benefit package caused by indirect risk selection. Direct risk selection (DRS) may either increase or decrease these distortions, depending on the type of equilibrium (pooling or separating), the type of DRS (positive or negative) and the type of cost for DRS (individual-specific or not). Regulators who succeed in reducing DRS by, e.g., banning excessive advertising or implementing fines for 'losing' applications, may therefore (unintendedly) mitigate or exacerbate the distortions of the benefit package caused by indirect risk selection. It is shown that the interaction of direct and indirect risk selection also alters the formula for optimal risk adjustment.

Hospital diagnostic aggregation and risk-adjusted quality

OBJECTIVE

To demonstrate the importance of diagnostic aggregation when assessing hospitals.

DATA SOURCES

Patient data from the Victorian Admitted Episodes Database (VAED), 1999/2000 to 2004/2005. Financial statements from public hospitals, 2002/2003 to 2004/2005.

STUDY DESIGN

Risk-adjusted quality computed for each hospital using two aggregation levels. Each is then used to estimate the relationship between hospital efficiency and quality using two-stage DEA/Tobit model by Wilson and Simar (2006).

DATA COLLECTION

Selected variables from the VAED were obtained from the Department of Health in Victoria, then linked anonymously with financial statements.

PRINCIPAL FINDINGS

Hospital quality and, in some cases, its relationship with efficiency differs depending on aggregations.

CONCLUSIONS

Patient risk adjustment should be conducted using more than one aggregation level whenever possible.

A population health approach to reducing observational intensity bias in health risk adjustment: cross sectional analysis of insurance claims

OBJECTIVE

To compare the performance of two new approaches to risk adjustment that are free of the influence of observational intensity with methods that depend on diagnoses listed in administrative databases.

SETTING

Administrative data from the US Medicare program for services provided in 2007 among 306 US hospital referral regions.

DESIGN

Cross sectional analysis.

PARTICIPANTS

20% sample of fee for service Medicare beneficiaries residing in one of 306 hospital referral regions in the United States in 2007 (n = 5,153,877).

MAIN OUTCOME MEASURES

The effect of health risk adjustment on age, sex, and race adjusted mortality and spending rates among hospital referral regions using four indices: the standard Centers for Medicare and Medicaid Services--Hierarchical Condition Categories (HCC) index used by the US Medicare program (calculated from diagnoses listed in Medicare's administrative database); a visit corrected HCC index (to reduce the effects of observational intensity on frequency of diagnoses); a poverty index (based on US census); and a population health index (calculated using data on incidence of hip fractures and strokes, and responses from a population based annual survey of health from the Centers for Disease Control and Prevention).

RESULTS

Estimated variation in age, sex, and race adjusted mortality rates across hospital referral regions was reduced using the indices based on population health, poverty, and visit corrected HCC, but increased using the standard HCC index. Most of the residual variation in age, sex, and race adjusted mortality was explained (in terms of weighted R2) by the population health index: R2=0.65. The other indices explained less: R2=0.20 for the visit corrected HCC index; 0.19 for the poverty index, and 0.02 for the standard HCC index. The residual variation in age, sex, race, and price adjusted spending per capita across the 306 hospital referral regions explained by the indices (in terms of weighted R2) were 0.50 for the standard HCC index, 0.21 for the population health index, 0.12 for the poverty index, and 0.07 for the visit corrected HCC index, implying that only a modest amount of the variation in spending can be explained by factors most closely related to mortality. Further, once the HCC index is visit corrected it accounts for almost none of the residual variation in age, sex, and race adjusted spending.

CONCLUSION

Health risk adjustment using either the poverty index or the population health index performed substantially better in terms of explaining actual mortality than the indices that relied on diagnoses from administrative databases; the population health index explained the majority of residual variation in age, sex, and race adjusted mortality. Owing to the influence of observational intensity on diagnoses from administrative databases, the standard HCC index over-adjusts for regional differences in spending. Research to improve health risk adjustment methods should focus on developing measures of risk that do not depend on observation influenced diagnoses recorded in administrative databases.

Do Medicare Advantage plans select enrollees in higher margin clinical categories?

The CMS-HCC risk adjustment system for Medicare Advantage (MA) plans calculates weights, which are effectively relative prices, for beneficiaries with different observable characteristics. To do so it uses the relative amounts spent per beneficiary with those characteristics in Traditional Medicare (TM). For multiple reasons one might expect relative amounts in MA to differ from TM, thereby making some beneficiaries more profitable to treat than others. Much of the difference comes from differences in how TM and MA treat different diseases or diagnoses. Using data on actual medical spending from two MA-HMO plans, we show that the weights calculated from MA costs do indeed differ from those calculated using TM spending. One of the two plans (Plan 1) is more typical of MA-HMO plans in that it contracts with independent community providers, while the other (Plan 2) is vertically integrated with care delivery. We calculate margins, or average revenue/average cost, for Medicare beneficiaries in the two plans who have one of 48 different combinations of medical conditions. The two plans' margins for these 48 conditions are correlated (r=0.39, p<0.01). Both plans have margins that are more positive for persons with conditions that are managed by primary care physicians and where medical management can be effective. Conversely they have lower margins for persons with conditions that tend to be treated by specialists with greater market power than primary care physicians and for acute conditions where little medical management is possible. The two plan's margins among beneficiaries with different observable characteristics vary over a range of 160 and 98 percentage points, respectively, and thus would appear to offer substantial incentive for selection by HCC. Nonetheless, we find no evidence of overrepresentation of beneficiaries in high margin HCC's in either plan. Nor, using the margins from Plan 1, the more typical plan, do we find evidence of overrepresentation of high margin HCC's in Medicare more generally. These results do not permit a conclusion on overall social efficiency, but we note that selection according to margin could be socially efficient. In addition, our findings suggest there are omitted interaction terms in the risk adjustment model that Medicare currently uses.

Integrating risk adjustment and enrollee premiums in health plan payment

In two important health policy contexts - private plans in Medicare and the new state-run "Exchanges" created as part of the Affordable Care Act (ACA) - plan payments come from two sources: risk-adjusted payments from a Regulator and premiums charged to individual enrollees. This paper derives principles for integrating risk-adjusted payments and premium policy in individual health insurance markets based on fitting total plan payments to health plan costs per person as closely as possible. A least squares regression including both health status and variables used in premiums reveals the weights a Regulator should put on risk adjusters when markets determine premiums. We apply the methods to an Exchange-eligible population drawn from the Medical Expenditure Panel Survey (MEPS).

Truly inefficient or providing better quality of care? Analysing the relationship between risk-adjusted hospital costs and patients' health outcomes

Observed variation in hospital costs may be attributable to differences in patients' health outcomes. Previous studies have resorted to inherently incomplete outcome measures such as mortality or re-admission rates to assess this claim. This study makes use of a novel dataset of routinely collected patient-reported outcome measures (PROMs) linked to inpatient records to (i) access the degree to which cost variation is associated with variation in patients' health gain and (ii) explore how far judgement about hospital cost performance changes when health outcomes are accounted for. We use multilevel modelling to address the clustering of patients in providers and isolate unexplained cost variation. We find some evidence of a U-shaped relationship between risk-adjusted costs and outcomes for hip replacement surgery. For three other procedures (knee replacement, varicose vein and groin hernia surgery), the estimated relationship is sensitive to the choice of PROM instrument. We do not observe substantial changes in cost performance estimates when outcomes are explicitly accounted for.

Adjusting National Health and Nutrition Examination Survey sample weights for women of childbearing age

BACKGROUND

Maternal risk factors have been tabulated for women of childbearing age using defined age ranges. However, statistics for factors strongly related to age may be overly influenced by values for the youngest and oldest women in a range, because pregnancies are most likely for ages 20-35.

OBJECTIVE

This report evaluates adjustment methods, based on the probability of pregnancy, for calculating estimates of risk factors for women of childbearing age.

METHODS

Adjusted and unadjusted estimates for environmental and nutritional variables were calculated from the 1999-2008 National Health and Nutrition Examination Survey (NHANES) for women aged 16-49. U.S. births were used to determine the probability of pregnancy.

RESULTS

Adjusted and unadjusted estimates differed for some, but not all, examined variables. More marked differences were observed for the environmental variables compared with the nutritional variables. Adjusted estimates were within about 5% of the unadjusted estimates for the nutritional variables. Adjusted geometric means for lead and mercury were about 7%-10% lower, and for polychlorinated biphenyl (or PCB) about 25% lower, than their respective unadjusted geometric means. With few exceptions, different adjustment methods led to similar estimates.

CONCLUSION

When calculating statistics for women of childbearing age, the decision to adjust for age or not to adjust appears to be more important than the choice of adjustment method. Although the results suggest only small differences among adjustment methods, approaches based on the NHANES design and sample weighting methodology may be the most robust for other applications.

Quantifying morbidity burdens and medical utilization of children with intellectual disabilities in Taiwan: a nationwide study using the ACG case-mix adjustment system

The purpose of this study was to quantify morbidity burdens of children with intellectual disability (ID) and to examine its association with total medical utilization and expenditure on a national basis in Taiwan. People under 18 years of age that had been continuously enrolled in the National Health Insurance (NHI) between year 2008 and 2010 were selected from one million randomly-sampled NHI beneficiaries. The Johns Hopkins Adjusted Clinical Group (ACG) System was applied to evaluate an individual's morbidity burden using 2008-2010 claims data, including age, sex, diagnosis, pharmacy, ambulatory, and inpatient utilization and expenditure (in New Taiwan Dollars, NTDs). The ID prevalence rate was 0.69% for people aged under 18. People with ID could be assigned to 20 mutually exclusive ACGs and to five simplified morbidity categories: healthy (0.1%), low (1.5%), moderate (31.9%), high (44.0%), and very high (22.4%). People with ID had more per capita visits (108.4 vs. 51.5, p<0.001), hospital admission (27.7% vs. 13.1%, p<0.001), pharmacy (NTD 21,069 vs. 4983, p<0.001) and total expenditure (NTD 144,962 vs. 29,764, p<0.001) than those without ID over 3 years. Those who assigned to the high-morbid categories cost more in ambulatory and inpatient services than those with low to moderate morbidities. In conclusion, the morbidity burdens of people with ID can be quantified by the ACG System based on readily available data. Regularly evaluating morbidity burdens and medical utilization has particular relevance for planning high-quality and efficient care. People's disabilities and comorbid illnesses shall be treated by integrated multidisciplinary teams.

How should risk adjustment data be collected?

Risk adjustment has broad general application and is a key part of the Patient Protection and Affordable Care Act (ACA). Yet, little has been written on how data required to support risk adjustment should be collected. This paper offers analytical support for a distributed approach, in which insurers retain possession of claims but pass on summary statistics to the risk adjustment authority as needed. It shows that distributed approaches function as well as or better than centralized ones-where insurers submit raw claims data to the risk adjustment authority-in terms of the goals of risk adjustment. In particular, it shows how distributed data analysis can be used to calibrate risk adjustment models and calculate payments, both in theory and in practice--drawing on the experience of distributed models in other contexts. In addition, it explains how distributed methods support other goals of the ACA, and can support projects requiring data aggregation more generally. It concludes that states should seriously consider distributed methods to implement their risk adjustment programs.

Risk adjustment for inter-hospital comparison of caesarean delivery rates in low-risk deliveries

BACKGROUND

Caesarean delivery (CD) rates have been frequently used as quality measures for maternity service comparisons. More recently, primary CD rates (CD in women without previous CD) or CD rates within selected categories such as nulliparous, term, cephalic singleton deliveries (NTCS) have been used. The objective of this study is to determine the extent to which risk adjustment for clinical and socio-demographic variables is needed for inter-hospital comparisons of CD rates in women without previous CD and in NTCS deliveries.

METHODS

Hospital discharge records of women who delivered in Emilia-Romagna Region (Italy) from January, 2007 to June 2009 and in Tuscany Region for year 2009 were linked with birth certificates. Adjusted RRs of CD in women without a previous Caesarean and NTCS were estimated using Poisson regression. Percentage differences in RR before and after adjustment were calculated and hospital rankings, based on crude and adjusted RRs, were examined.

RESULTS

Adjusted RR differed substantially from crude RR in women without a previous Caesarean and only marginally in NTCS group. Hospital ranking was markedly affected by adjustment in women without a previous CD, but less in NTCS.

CONCLUSION

Risk adjustment is warranted for inter-hospital comparisons of primary CD rates but not for NTCS CD rates. Crude NTCS CD rates are a reliable estimate of adjusted NTCS CD.

[Risk-adjusted assessment of neonatology wards by the new quality indicator "transfer rate of mature newborns"]

OBJECTIVE

1. The transfer rate of mature newborns will be presented as a new quality indicator. 2. Another objective of this study was to adjust the transfer rate of mature newborns of different hospitals according to their "risk" profile of patients by multivariate analysis.

METHOD

The perinatal database of 118,416 newborns of the Saxonian quality surveillance from 2001 to 2004 was analysed. Based on 17 clinical and 3 structural factors, a logistic regression model was used to develop a specific "risk" predictor for the quality indicator "transfer rate".

RESULTS

For care level III (basic care) a "risk" predictor for the transfer rate was developed, which consists of 15 factors. The AUC(ROC)-value of this quality indicator was 78.6%, which is sufficient. The hospital ranking based on the adjusted risk assessment was different from the hospital ranking prior to this adjustment. The average correction of ranking position was 10.4 for 43 clinics.

CONCLUSION

1. The new quality indicator "transfer rate of mature newborns" can be recommended. 2. The application of the risk adjustment method proposed here allows for a more objective comparison of the quality indicator "transfer rate" among different hospitals.

Patient volume affects outcome in critically ill patients

CONTEXT

A positive relationship between patient volume and outcome has been demonstrated for a variety of clinical conditions and procedures, but the evidence is sparse for critically ill patients.

OBJECTIVE

To evaluate the relationship between patient volume and outcome in a large cohort of critically ill patients.

DESIGN

Prospective multicenter cohort study, January 1998 through December 2005.

SETTING

40 intensive care units in Austria.

PATIENTS

A total of 83,259 consecutively admitted patients.

MAIN OUTCOME MEASURES

Structural quality of participating ICUs was evaluated using a questionnaire and merged with the prospectively collected data. Volume related indices were then calculated, representing patient turnover, occupancy rate, nursing workload and diagnostic variability.

RESULTS

Univariate analysis revealed that several volume variables were associated with outcome: more patients treated per year per bed in the intensive care unit and more patients treated in the same diagnostic category reduced the risk of dying in the hospital (odds ratios, 0.967 and 0.991 for each additional 10 patients treated, respectively). In contrast, an increase in the patient-to-nurse ratio and an increase in the number of diagnostic categories were associated with increased mortality rates. Multivariate analysis confirmed these results. The relationship between the number of patients treated in the same diagnostic category and their outcomes showed not a linear but a U shape, with increasing mortality rates below and above a certain patient volume.

CONCLUSIONS

Our results provide evidence for a relationship between patient volume and outcome in critically ill patients. Besides the total number of patients, diagnostic variability plays an important role. The relationship between volume and outcome seems, however, to be complex and to be influenced by other variables, such as workload of nursing staff.

Evidence of methodological bias in hospital standardised mortality ratios: retrospective database study of English hospitals

OBJECTIVE

To assess the validity of case mix adjustment methods used to derive standardised mortality ratios for hospitals, by examining the consistency of relations between risk factors and mortality across hospitals.

DESIGN

Retrospective analysis of routinely collected hospital data comparing observed deaths with deaths predicted by the Dr Foster Unit case mix method.

SETTING

Four acute National Health Service hospitals in the West Midlands (England) with case mix adjusted standardised mortality ratios ranging from 88 to 140.

PARTICIPANTS

96 948 (April 2005 to March 2006), 126 695 (April 2006 to March 2007), and 62 639 (April to October 2007) admissions to the four hospitals.

MAIN OUTCOME MEASURES

Presence of large interaction effects between case mix variable and hospital in a logistic regression model indicating non-constant risk relations, and plausible mechanisms that could give rise to these effects.

RESULTS

Large significant (P

CONCLUSIONS

The Dr Foster Unit hospital standardised mortality ratio is derived from an internationally adopted/adapted method, which uses at least two variables (the Charlson comorbidity index and emergency admission) that are unsafe for case mix adjustment because their inclusion may actually increase the very bias that case mix adjustment is intended to reduce. Claims that variations in hospital standardised mortality ratios from Dr Foster Unit reflect differences in quality of care are less than credible.

Collapse

Key Words Collapse

MESH Headings

• Bias
• Emergencies/epidemiology
• England/epidemiology
• Hospital Mortality
• Length of Stay/statistics & numerical data
• Patient Admission/statistics & numerical data
• Regression Analysis
• Retrospective Studies
• Risk Adjustment/statistics & numerical data
• Risk Factors

Collapse

Grants

• G0800808 Medical Research Council

Collapse

Affiliation(s)

Mohammed A Mohammed Unit of Public Health, Epidemiology and Biostatistics, University of Birmingham, Birmingham B15 2TT.
Jonathan J Deeks
Alan Girling
Gavin Rudge
Martin Carmalt
Andrew J Stevens
Richard J Lilford

Collapse

How can we know whether short term trends in a hospital's HSMR are significant?

The Hospital Standardized Mortality Ratio (HSMR) has been chosen by CIHI as its primary mortality measure. The indirect standardization used in the calculation of HSMR does not allow for valid comparison between hospitals but it does invite the assessment of quarterly trends in hospital mortality. However, statistical methods for assessing HSMR trends are not well-developed. In 2007 one large hospital in our health authority had four consecutive quarters of apparently increasing HSMR. As a result, we needed to assess the significance of this trend which, if it were to continue into the next quarter, would lead to an HSMR that significantly exceeded 100. We explored four methods to assess statistical significance of time trends in HSMR data: the WINPEPI "Describe" module, the CUSUM representation of Observed-Expected differences, the Variable Life Adjusted Display (VLAD) plots with CUSUM overlays, and the Change Point Analysis using Monte Carlo simulation.

Review of the application of risk-adjusted charts to analyse mortality outcomes in critical care

This review describes the methods for displaying riskadjusted mortality data for critical care units. Two applications are considered. The comparison within a cohort of risk-adjusted mortality performance uses standardised mortality ratios (SMRs), league tables, caterpillar plots and funnel plots. Monitoring of riskadjusted performance over time is considered using SMRs, risk-adjusted p (RAP), observed minus expected outcome (VLAD), risk-adjusted cumulative sum (RACUSUM), riskadjusted sequential probability ratio test (RASPRT), and riskadjusted exponentially weighted moving average (RAEWMA) charts. Examples of the charts are provided, and calculation of the statistics and design of the charts are described in the Appendix. This overview is an introduction to the use of riskadjustment methods to track mortality rates. The importance of model performance and relevance of the risk-adjustment models is emphasised. The relative merits of different methods are discussed. Risk-adjusted monitoring plays a role in the context of a holistic quality development strategy. The importance of a planned approach to response and intervention is stressed.

Use of a diagnosis-based risk adjustment model to estimate costs of indigent care in a community at Medicaid reimbursement rates

OBJECTIVES

This study used a diagnosis-based risk adjustment model to estimate the annual costs of uninsured patients in Austin, Texas, and describe the prevalence and costs of their chronic conditions. The data were supplied by the Indigent Care Collaboration, a partnership of local safety-net hospitals and clinics.

METHODS

This study used the Diagnostic Cost Groups prospective Medicaid All-Encounters model, which uses diagnoses, age and gender to assign relative risk scores to patients. The relative risk scores were multiplied by the per capita Texas Medicaid expenditure to obtain estimated annual costs. Chronic diseases were described in terms of prevalence and total estimated annual cost.

RESULTS

A total of 471,194 encounters were recorded for 163,729 patients meeting the study inclusion criteria between the 1st March 2004 and the 28th February 2005. The mean estimated patient yearly cost was US $1,307, and the total estimated yearly population cost was $228,909,529. The most common chronic conditions included hypertension, diabetes, depression, substance abuse, pregnancy, asthma, chronic obstructive pulmonary disease and congestive heart failure.

CONCLUSIONS

This study demonstrates how the unknown costs associated with caring for indigent uninsured patients in a community can be estimated at Medicaid reimbursement rates using the Diagnostic Cost Group model on aggregated patient encounter data.

Adjustments for unmeasured confounders in pharmacoepidemiologic database studies using external information

BACKGROUND

Nonexperimental studies of drug effects in large automated databases can provide timely assessment of real-life drug use, but are prone to confounding by variables that are not contained in these databases and thus cannot be controlled.

OBJECTIVES

To describe how information on additional confounders from validation studies can help address the problem of unmeasured confounding in the main study.

RESEARCH DESIGN

Review types of validation studies that allow adjustment for unmeasured confounding and illustrate these with an example.

SUBJECTS

Main study: New Jersey residents age 65 years or older hospitalized between 1995 and 1997, who filled prescriptions within Medicaid or a pharmaceutical assistance program. Validation study: representative sample of Medicare beneficiaries.

MEASURES

Association between nonsteroidal antiinflammatory drugs (NSAIDs) and mortality.

RESULTS

Validation studies are categorized as internal (ie, additional information is collected on participants of the main study) or external. Availability of information on disease outcome will affect choice of analytic strategies. Using an external validation study without data on disease outcome to adjust for unmeasured confounding, propensity score calibration (PSC) leads to a plausible estimate of the association between NSAIDs and mortality in the elderly, if the biases caused by measured and unmeasured confounders go in the same direction.

CONCLUSIONS

Estimates of drug effects can be adjusted for confounders that are not available in the main, but can be measured in a validation study. PSC uses validation data without information on disease outcome under a strong assumption. The collection and integration of validation data in pharmacoepidemiology should be encouraged.

Heterogeneity and the Interpretation of Treatment Effect Estimates From Risk Adjustment and Instrumental Variable Methods

OBJECTIVES

To contrast the interpretations of treatment effect estimates using risk adjustment and instrumental variable (IV) estimation methods using observational data when the effects of treatment are heterogeneous across patients. We demonstrate these contrasts by examining the effect of breast conserving surgery plus irradiation (BCSI) relative to mastectomy on early stage breast cancer (ESBC) survival.

METHODS

We estimated discrete time survival models for 6185 ESBC patients in the 1989-1994 Iowa Cancer Registry via IV estimation using 2 distinct instruments (distance of the patient's residence from the nearest radiation center, and local area BCSI rate) and controlling for cancer stage, grade, and location; age; comorbidity; hospital access; payer; diagnosis year; and area poverty level. We then estimated comparable risk adjustment survival models using linear probability methods with robust standard errors.

RESULTS

Risk adjustment models yielded average survival estimates similar to trial results. With favorable BCSI selection, these estimates represent an upper bound of the true effect for patients receiving BCSI. IV estimates showed a BCSI survival risk for patients whose surgery choices were affected by the instruments and these estimates varied with the instrument specification.

CONCLUSIONS

When treatment benefits are heterogeneous across patients, treatment effect estimates from observational data can still be useful to policymakers, but they must be interpreted correctly. Risk adjustment methods yield estimates that can assess whether the patients who received treatment benefited from the treatment, but the direction of bias must be considered. In contrast, IV estimates can assess the effect of treatment rate changes, but characteristics of patients whose choices were affected by the instruments must be considered when making such inferences.

The choice between different statistical approaches to risk-adjustment influenced the identification of outliers

OBJECTIVE

Many statistical approaches have been applied to compare health care providers' performance, but few studies have examined how the outlier status of providers depends on the choice between risk-adjustment techniques.

STUDY DESIGN AND SETTING

We analyzed the recourse to breast-conserving surgery (BCS) for breast carcinoma across 31 hospitals of the Veneto Region (Italy). The following methods were compared: the ratio of observed to expected events (O/E), regression models with provider effects introduced as dummy variables obtained by standard or weighted effect coding, and multilevel analysis.

RESULTS

The O/E method classified seven hospitals (one with high and six with low BCS rates) as outliers. The regression model with the weighted parameterization gave similar results, whereas through standard effect coding all odds ratios shifted and different outliers were identified. Multilevel analysis was quite conservative in identifying small hospitals with BCS rates lower than the regional mean.

CONCLUSION

Whenever feasible, results obtained through different statistical methodologies should be compared. If providers are modeled as dummy variables obtained by effect coding, departures of the model intercept from the regional mean should be checked. The increasing use of multilevel models could entail a lower sensitivity in identifying low-quality outliers if a volume-outcome relationship exists.

Comparing self-reported health status and diagnosis-based risk adjustment to predict 1- and 2 to 5-year mortality

OBJECTIVES

To compare the ability of two diagnosis-based risk adjustment systems and health self-report to predict short- and long-term mortality.

DATA SOURCES/STUDY SETTING

Data were obtained from the Department of Veterans Affairs (VA) administrative databases. The study population was 78,164 VA beneficiaries at eight medical centers during fiscal year (FY) 1998, 35,337 of whom completed an 36-Item Short Form Health Survey for veterans (SF-36V) survey.

STUDY DESIGN

We tested the ability of Diagnostic Cost Groups (DCGs), Adjusted Clinical Groups (ACGs), SF-36V Physical Component score (PCS) and Mental Component Score (MCS), and eight SF-36V scales to predict 1- and 2-5 year all-cause mortality. The additional predictive value of adding PCS and MCS to ACGs and DCGs was also evaluated. Logistic regression models were compared using Akaike's information criterion, the c-statistic, and the Hosmer-Lemeshow test.

PRINCIPAL FINDINGS

The c-statistics for the eight scales combined with age and gender were 0.766 for 1-year mortality and 0.771 for 2-5-year mortality. For DCGs with age and gender the c-statistics for 1- and 2-5-year mortality were 0.778 and 0.771, respectively. Adding PCS and MCS to the DCG model increased the c-statistics to 0.798 for 1-year and 0.784 for 2-5-year mortality.

CONCLUSIONS

The DCG model showed slightly better performance than the eight-scale model in predicting 1-year mortality, but the two models showed similar performance for 2-5-year mortality. Health self-report may add health risk information in addition to age, gender, and diagnosis for predicting longer-term mortality.

Risk-adjusting outcomes of mental health and substance-related care: a review of the literature

Risk adjustment is increasingly recognized as crucial to refining health care reimbursement and to comparing provider performance in terms of quality and outcomes of care. Risk adjustment for mental and substance use conditions has lagged behind other areas of medicine, but model development specific to these conditions has accelerated in recent years. After describing outcomes of mental health and substance-related care and associated risk factors, we review research studies on risk adjustment meeting the following criteria: (1) publication in a peer-reviewed journal between 1980 and 2002, (2) evaluation of one or more multivariate models used to risk-adjust comparisons of utilization, cost, or clinical outcomes of mental or substance use conditions across providers, and (3) quantitative assessment of the proportion of variance explained by patient characteristics in the model (e.g., R(2) or c-statistic). We identified 36 articles that included 72 models addressing utilization, 74 models of expenditures, and 15 models of clinical outcomes. Models based on diagnostic and sociodemographic information available from administrative data sets explained an average 6.7% of variance, whereas models using more detailed sources of data explained a more robust 22.8%. Results are appraised in the context of the mental health care system's needs for risk adjustment; we assess what has been accomplished, where gaps remain, and directions for future development.

The effect of complementary and alternative medicine claims on risk adjustment

OBJECTIVE

We sought to assess how the inclusion of claims from complementary and alternative medicine (CAM) providers affects measures of morbidity burden and expectations of health care resource use for insured patients.

METHODS

Claims data from Washington State were used to create 2 versions of a case-mix index. One version included claims from all provider types; the second version omitted claims from CAM providers who are covered under commercial insurance. Expected resource use was also calculated. The distribution of expected and actual resource use was then compared for the 2 indices.

RESULTS

Inclusion of claims from CAM providers shifted 19,650 (32%) CAM users into higher morbidity categories. When morbidity categories were defined using claims from all providers, CAM users in the highest morbidity category had average (+/-SD) annual expenditures of $6661 (+/-$13,863). This was less than those in the highest morbidity category when CAM provider claims were not included in the index ($8562 +/- $16,354), and was also lower than the highest morbidity patients who did not use any CAM services ($8419 +/- $18,885).

CONCLUSIONS

Inclusion of services from CAM providers under third-party payment increases risk scores for their patients but expectations of costs for this group are lower than expected had costs been estimated based only on services from traditional providers. Risk adjustment indices may need recalibration when adding services from provider groups not included in the development of the index.

Scale of interest versus scale of estimation: comparing alternative estimators for the incremental costs of a comorbidity

We investigate how the scale of estimation in risk-adjustment models for health-care costs affects the covariate effect, where the scale of interest for the covariate effect may be different from the scale of estimation. As an illustrative example, we use claims data to estimate the incremental costs associated with heart failure within one year subsequent to myocardial infarction. Here, the scale of interest for the effect of heart failure on costs is additive. However, traditional methods for modeling costs use predetermined scale of estimation - for example, ordinary least squares (OLS) regression assumes an additive scale while log-transformed OLS and generalized linear models with log-link assume a multiplicative scale of estimation. We compare these models with a new flexible model that lets the data determine the appropriate scale of estimation. We use a variety of goodness-of-fit measures along with a modified Copas test to assess robustness, lack of fit, and over-fitting properties of the alternative estimators. Biases up to 19% in the scale of interest are observed due to the misrepresentation of the scale of estimation. The new flexible model is found to appropriately represent the scale of estimation and less susceptible to over-fitting despite estimating additional parameters in the link and the variance functions.

[Options for a future risc structure compensation in Germany]

AIM OF THE ARTICLE: The risc structure compensation scheme within the German compulsory health insurance system is intended to enforce the principle of solidarity all over the statutory health insurance and not only within the different sickness funds. Differences in the contribution rates should not reflect different risc profiles, but the differences of the efficiency in social care. The criticism against the current adjustment system in Germany is multifarious and points e. g. on the missing orientation to morbidity. This article follows the question, whether this criticism is valid.

METHODS

The variables and methods, which are currently used to calculate the risc structure adjustment are discussed and compared to an alternative proposal for the future form of the risc structure adjustment, which includes both a higher orientation to riscs and incentives for social health insurance funds to decline the costs for the social care system on long-term.

RESULTS

Currently, for the calculation of the risc structure adjustment the following variables are used: age, sex, income, number of family members who are exempted from contributions and persons who get occupational disability pension, and number of insured persons who are registered to an accredited Disease-Management-Program (DMP). Especially the last variable includes a high control effort, because the higher co-payments of the adjustment system are aligned to the voluntariness of participation and active collaboration of the patients in DMP. The argument, a further development to a morbidity-oriented risc structure adjustment leads to less cost management of the sickness funds is not totally correct, because not actual, but standardised costs are the basis for compensation. On the other hand the morbidity determined cost components should not totally be adjusted, as a proper distribution of savings to the risc structure adjustment and the single funds would still be an incentive for cost management and prevention.

CONCLUSION

An ongoing refining of the risc structure adjustment might cause new incentive problems. Instead a morbidity orientated risc structure compensation scheme should leave a part of the savings due to better social care structures in the sickness funds and should include outpatient care parameters. The change to a new honorarium system could create a better data basis for this improved form of risc structure adjustment in the future.

Development and Validation of a Psychiatric Case-Mix System

BACKGROUND

Although difficulties in applying risk-adjustment measures to mental health populations are increasingly evident, a model designed specifically for patients with psychiatric disorders has never been developed.

OBJECTIVE

Our objective was to develop and validate a case-mix classification system, the "PsyCMS," for predicting concurrent and future mental health (MH) and substance abuse (SA) healthcare costs and utilization.

SUBJECTS

Subjects included 914,225 veterans who used Veterans Administration (VA) healthcare services during fiscal year 1999 (FY99) with any MH/SA diagnosis (International Classification of Diseases, 9th Revision, Clinical Modification [ICD-9-CM] codes 290.00-312.99, 316.00-316.99).

METHODS

We derived diagnostic categories from ICD-CM codes using Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition definitions, clinical input, and empiric analyses. Weighted least-squares regression models were developed for concurrent (FY99) and prospective (FY00) MH/SA costs and utilization. We compared the predictive ability of the PsyCMS with several case-mix systems, including adjusted clinical groups, diagnostic cost groups, and the chronic illness and disability payment system. Model performance was evaluated using R-squares and mean absolute prediction errors (MAPEs).

RESULTS

Patients with MH/SA diagnoses comprised 29.6% of individuals seen in the VA during FY99. The PsyCMS accounted for a distinct proportion of the variance in concurrent and prospective MH/SA costs (R=0.11 and 0.06, respectively), outpatient MH/SA utilization (R=0.25 and 0.07), and inpatient MH/SA utilization (R=0.13 and 0.05). The PsyCMS performed better than other case-mix systems examined with slightly higher R-squares and lower MAPEs.

CONCLUSIONS

The PsyCMS has clinically meaningful categories, demonstrates good predictive ability for modeling concurrent and prospective MH/SA costs and utilization, and thus represents a useful method for predicting mental health costs and utilization.

The potential of risk adjustment for the Military Health System TRICARE Program

Using industry leading methods and software products, this article explores the potential for health risk adjustment in the Department of Defense (DoD) Military Health System. From an epidemiologic perspective, the study assesses whether DoD populations enrolled to and served by different provider groups exhibit sufficient variation in risk to justify changes in resource allocation, such as staffing, services, and facility modification. In addition, the study investigates whether systematic differences in risk exist between TRICARE enrollees and commercially insured enrollees. Given the interest of the DoD in managed care, risk prediction tools are shown to project future risk for individuals with sufficient accuracy to justify development of disease management programs for patients with chronic conditions. Analyses demonstrate that Military Health System data support industry requirements to perform risk grouping among a random sample of 200,000 enrollees selected from 495,941 eligible enrollees.

The role of psychiatric diagnosis in satisfaction with primary care: data from the department of veterans affairs

BACKGROUND

Although patient satisfaction is widely used as a quality indicator, most such measures do not account for patient subgroups such as those with psychiatric illness. There is also very little data on satisfaction of psychiatric patients with their medical care.

OBJECTIVE

The objective of this study was to assess the role of psychiatric illness in satisfaction with outpatient primary care services in the Department of Veterans Affairs (VA).

METHOD

Data from the VA Customer Feedback Survey (n = 50,532) were merged with administrative data to determine diagnoses and other characteristics. Satisfaction ratings were compared across psychiatric diagnoses and across various aspects of satisfaction with care.

RESULTS

After controlling for patient characteristics (eg, gender, age, disability, acute vs. routine visit) and subjective health, patients with schizophrenia, posttraumatic stress disorder, drug abuse, depression, and other psychiatric disorders reported significantly lower satisfaction with their outpatient primary care. Dissatisfaction was particularly reported for access to care and overall coordination of care.

CONCLUSIONS

Despite VA characteristics that might be thought to improve satisfaction (eg, easier access to specialty mental health services as a result of the integrated VA system), patients with psychiatric disorders are significantly less satisfied than patients without such disorders. Possible explanations include both lower technical quality of care and poorer interpersonal communication between providers and patients with mental illness, including the negative effects of stigma. These findings highlight the need for satisfaction ratings to be case-mix-adjusted, including the incorporation of health and mental health diagnoses, and the need for further research that elucidates the reasons behind lower satisfaction ratings.

A matter of classes: stratifying health care populations to produce better estimates of inpatient costs

OBJECTIVE

To stratify traditional risk-adjustment models by health severity classes in a way that is empirically based, is accessible to policy makers, and improves predictions of inpatient costs.

DATA SOURCES

Secondary data created from the administrative claims from all 829,356 children aged 21 years and under enrolled in Georgia Medicaid in 1999.

STUDY DESIGN

A finite mixture model was used to assign child Medicaid patients to health severity classes. These class assignments were then used to stratify both portions of a traditional two-part risk-adjustment model predicting inpatient Medicaid expenditures. Traditional model results were compared with the stratified model using actuarial statistics.

PRINCIPAL FINDINGS

The finite mixture model identified four classes of children: a majority healthy class and three illness classes with increasing levels of severity. Stratifying the traditional two-part risk-adjustment model by health severity classes improved its R(2) from 0.17 to 0.25. The majority of additional predictive power resulted from stratifying the second part of the two-part model. Further, the preference for the stratified model was unaffected by months of patient enrollment time.

CONCLUSIONS

Stratifying health care populations based on measures of health severity is a powerful method to achieve more accurate cost predictions. Insurers who ignore the predictive advances of sample stratification in setting risk-adjusted premiums may create strong financial incentives for adverse selection. Finite mixture models provide an empirically based, replicable methodology for stratification that should be accessible to most health care financial managers.

Benzodiazepines and injury: a risk adjusted model

BACKGROUND

Benzodiazepines (BZD) are one class of medications that are generally acknowledged to be a risk factor for injuries.

OBJECTIVE

Our objective was to link outpatient prescription data with clinical data in order to develop a risk adjusted binary model that associates BZD usage with the risk for a healthcare encounter for an injury.

METHODS

In total, 3 years of outpatient BZD prescription data, totaling 133 872 outpatient BZD prescriptions for 13 745 patients for a VA medical center, were combined with data from inpatient and outpatient administrative databases. The model incorporated Elixhauser comorbidity measures with 1-year look back period, along with hospital discharges, marital status, age, mean arterial pressure and body mass index. The model also included the dose of the drug, converted to valium equivalents and its duration. The model was analyzed using generalized estimation equations (GEE).

RESULTS

Dose, duration, discharges and various comorbidities were associated with an increased risk for injury, while being married reduced the risk. Increased body mass was associated with increased injury risk. Increased mean arterial pressure was associated with decreased risk.

CONCLUSIONS

These findings offer guidance on how specific combinations of risk factors and potential protective effects may impact accidental injury risk. Clinicians prescribing or adjusting BZDs can use these results to more accurately tailor medication regimens for a patient. Our findings suggest that clinicians should also consider the nature of the social support system available to the patient in assessing total injury risk.

Kassenwechsel und Risikostrukturausgleich in der gesetzlichen Krankenversicherung - empirische Befunde der Kooperativen Gesundheitsforschung in der Region Augsburg (KORA)

Since 1996, all citizens of the Federal Republic of Germany who are insured in the statutory health insurance system are entitled to switch their sickness fund. The rationale of this regulation was to strengthen elements of competition in this system in order to stimulate the sickness funds to improve the efficiency of health care and to respond to consumers' preferences. Simultaneously, to avoid the implicit incentives for sickness funds to engage in risk selection, a risk compensation mechanism was introduced, including as morbidity-related risk adjusters age, sex and incapacity to work. Based on the KORA survey S4 (1999/2001) we take the case of switching behaviour in the region of Augsburg, and analyse whether this risk adjustment scheme was working effectively. The results show that persons changing their sickness fund were characterised by a comparatively smaller burden of chronic diseases and by a less frequent utilization of inpatient health care. Under these conditions, differences in the contribution rates do not accurately reflect differences in the performance and efficiency of sickness funds. Moreover, the migration of good risk to sickness funds with favourable contribution rates threatens the principle of financial solidarity. Therefore, the system of risk equalisation has to be developed towards measuring the risk volume borne by the sickness funds more precisely than hitherto.