Diagnostic Cost Groups Research Articles

A Novel Machine Learning Algorithm for Creating Risk-Adjusted Payment Formulas

Models predicting health care spending and other outcomes from administrative records are widely used to manage and pay for health care, despite well-documented deficiencies. New methods are needed that can incorporate more than 70 000 diagnoses without creating undesirable coding incentives. To develop a machine learning (ML) algorithm, building on Diagnostic Item (DXI) categories and Diagnostic Cost Group (DCG) methods, that automates development of clinically credible and transparent predictive models for policymakers and clinicians. DXIs were organized into disease hierarchies and assigned an Appropriateness to Include (ATI) score to reflect vagueness and gameability concerns. A novel automated DCG algorithm iteratively assigned DXIs in 1 or more disease hierarchies to DCGs, identifying sets of DXIs with the largest regression coefficient as dominant; presence of a previously identified dominating DXI removed lower-ranked ones before the next iteration. The Merative MarketScan Commercial Claims and Encounters Database for commercial health insurance enrollees 64 years and younger was used. Data from January 2016 through December 2018 were randomly split 90% to 10% for model development and validation, respectively. Deidentified claims and enrollment data were delivered by Merative the following November in each calendar year and analyzed from November 2020 to January 2024. Concurrent top-coded total health care cost. Model performance was assessed using validation sample weighted least-squares regression, mean absolute errors, and mean errors for rare and common diagnoses. This study included 35 245 586 commercial health insurance enrollees 64 years and younger (65 901 460 person-years) and relied on 19 clinicians who provided reviews in the base model. The algorithm implemented 218 clinician-specified hierarchies compared with the US Department of Health and Human Services (HHS) hierarchical condition category (HCC) model's 64 hierarchies. The base model that dropped vague and gameable DXIs reduced the number of parameters by 80% (1624 of 3150), achieved an R2 of 0.535, and kept mean predicted spending within 12% ($3843 of $31 313) of actual spending for the 3% of people with rare diseases. In contrast, the HHS HCC model had an R2 of 0.428 and underpaid this group by 33% ($10 354 of $31 313). In this study, by automating DXI clustering within clinically specified hierarchies, this algorithm built clinically interpretable risk models in large datasets while addressing diagnostic vagueness and gameability concerns.

Weight of Risk Factors for Adjusting Capitation in Primary Health Care: A Systematic Review.

Background: Capitation payment is the best-known strategy for paying providers in primary health care. Since health care needs and personal characteristics play an essential role in health care utilization and resource spending, there is a growing tendency on risk adjustment models among health researchers. The objective of this systematic review was to examine the weights used for risk adjustment in primary health care capitation payment. Methods: We systematically searched Scopus, ProQuest, Web of Science, and PubMed in March 2018. Two authors independently apprised the included articles and they also evaluated, identified, and categorized different factors on capitation payments mentioned in the included studies. Results: A total of 742 studies were identified and 12 were included in the systematic review after the screening process. Risk factors for capitation adjustment included age, gender, and income with the weighted average being 1.76 and 1.03, respectively. Moreover, the weighted average disease incidence adjusted clinical groups (ACGs), diagnostic cost groups (DCGs), principal in patient diagnostic cost groups (PIP-DCGs), and hierarchical coexisting conditions (HCCs) were reported as 1.31, 24.7-.99, 10.4-.65, and 11.7-1.01, respectively. Conclusion: In low-income countries, the most effective factors used in capitation adjustment are age and sex. Moreover, the most applied factor in high-income countries is adjusted clinical groups, and income factors can have a better impact on the reduction of costs in low-income countries. Each country can select its most efficient factors based on the weight of the factor, income level, and geographical condition.

Abstract P154: Short-Term Health Care Expenditures Increase After Workplace Wellness Program

Introduction: A major goal of workplace wellness programs is to improve employee health and lower health care costs. Research on wellness program effectiveness is often limited by either a lack of a control comparison groups or confounding when the intervention and control groups differ on baseline health. Our study objective was to assess the effect on health expenditures of an employee wellness program (BeFit), already shown to improve diet, physical activity, cholesterol, and BP. We compared participants to a matched non-participant control group. Hypothesis: We hypothesized that health care costs would be lower for participants than for non-participants. Methods: BeFit is a 10 week nutrition/exercise program that includes measurement of cardiovascular risk factors, weekly meetings with a nutritionist and personal trainer, gym membership, and team competition to achieve weight and exercise goals. For this study, participants were matched by age, sex, period of employment, and pre-participation Diagnostic Cost Group (DCG) risk score (a predictor of future health care use based on past-year use), to two non-participant control employees. Health expenditures were gathered from claims data obtained by the employer’s human resource department. Subjects in the analysis had at least 12 months pre- and post-program claims data. Within-subject quarterly health care cost trends were compared after subjects’ BeFit start dates using multivariable generalized least squares regression controlling for age, sex, and baseline health. Control subjects were assigned the start date of the subject to whom they were matched. Expenditures were capped at the 99 th percentile ($42,393) so results were not unduly influenced by extreme events unlikely to be affected by a wellness program. We analyzed overall, domain-specific (outpatient, inpatient, pharmaceutical), and diagnosis-specific costs. Results: From 2010-2014, 412 BeFit participants and 824 controls met the inclusion criteria (77% female, mean age 41 [range 23,78], mean baseline DCG risk score of 92.3 [range 3, 2315]). After participation in BeFit, the regression-adjusted health expenditures decreased by $15/quarter among BeFit participants compared to $85/quarter among control subjects (95% CI for difference $27, $113). Trends were significantly more negative in the control group for outpatient, inpatient, and pharmaceutical costs, as well. There were no differences in trends for cardiovascular or diabetes-related costs. Conclusions: Contrary to our hypothesis, trends in health expenditures decreased more for the matched control group than for participants in the employee wellness program. This result may be due to increased attention to health and higher health care use spurred by participation in the program. Further work is necessary to understand the long-term impact of wellness programs on health and health expenditures.

Factors Associated With Having a Physician, Nurse Practitioner, or Physician Assistant as Primary Care Provider for Veterans With Diabetes Mellitus.

Expanded use of nurse practitioners (NPs) and physician assistants (PAs) is a potential solution to workforce issues, but little is known about how NPs and PAs can best be used. Our study examines whether medical and social complexity of patients is associated with whether their primary care provider (PCP) type is a physician, NP, or PA. In this national retrospective cohort study, we use 2012-2013 national Veterans Administration (VA) electronic health record data from 374 223 veterans to examine whether PCP type is associated with patient, clinic, and state-level factors representing medical and social complexity, adjusting for all variables simultaneously using a generalized logit model. Results indicate that patients with physician PCPs are modestly more medically complex than those with NP or PA PCPs. For the group having a Diagnostic Cost Group (DCG) score >2.0 compared with the group having DCG <0.5, odds of having an NP or a PA were lower than for having a physician PCP (NP odds ratio [OR] = 0.83, 95% confidence interval [CI]: 0.79-0.88; PA OR = 0.85, CI: 0.80-0.89). Social complexity is not consistently associated with PCP type. Overall, we found minor differences in provider type assignment. This study improves on previous work by using a large national dataset that accurately ascribes the work of NPs and PAs, analyzing at the patient level, analyzing NPs and PAs separately, and addressing social as well as medical complexity. This is a requisite step toward studies that compare patient outcomes by provider type.

Association between bariatric surgery and long-term survival.

Accumulating evidence suggests that bariatric surgery improves survival among patients with severe obesity, but research among veterans has shown no evidence of benefit. To examine long-term survival in a large multisite cohort of patients who underwent bariatric surgery compared with matched control patients. In a retrospective cohort study, we identified 2500 patients (74% men) who underwent bariatric surgery in Veterans Affairs (VA) bariatric centers from 2000-2011 and matched them to 7462 control patients using sequential stratification and an algorithm that included age, sex, geographic region, body mass index, diabetes, and Diagnostic Cost Group. Survival was compared across patients who underwent bariatric surgery and matched controls using Kaplan-Meier estimators and stratified, adjusted Cox regression analyses. Bariatric procedures, which included 74% gastric bypass, 15% sleeve gastrectomy, 10% adjustable gastric banding, and 1% other. All-cause mortality through December 2013. Surgical patients (n = 2500) had a mean age of 52 years and a mean BMI of 47. Matched control patients (n = 7462) had a mean age of 53 years and a mean BMI of 46. At the end of the 14-year study period, there were a total of 263 deaths in the surgical group (mean follow-up, 6.9 years) and 1277 deaths in the matched control group (mean follow-up, 6.6 years). Kaplan-Meier estimated mortality rates were 2.4% at 1 year, 6.4% at 5 years, and 13.8% at 10 years for surgical patients; for matched control patients, 1.7% at 1 year, 10.4% at 5 years, and 23.9% at 10 years. Adjusted analysis showed no significant association between bariatric surgery and all-cause mortality in the first year of follow-up (adjusted hazard ratio [HR], 1.28 [95% CI, 0.98-1.68]), but significantly lower mortality after 1 to 5 years (HR, 0.45 [95% CI, 0.36-0.56]) and 5 to 14 years (HR, 0.47 [95% CI, 0.39-0.58]). The midterm (>1-5 years) and long-term (>5 years) relationships between surgery and survival were not significantly different across subgroups defined by diabetes diagnosis, sex, and period of surgery. Among obese patients receiving care in the VA health system, those who underwent bariatric surgery compared with matched control patients who did not have surgery had lower all-cause mortality at 5 years and up to 10 years following the procedure. These results provide further evidence for the beneficial relationship between surgery and survival that has been demonstrated in younger, predominantly female populations.

Mispricing in the medicare advantage risk adjustment model.

The Centers for Medicare and Medicaid Services (CMS) implemented hierarchical condition category (HCC) models in 2004 to adjust payments to Medicare Advantage (MA) plans to reflect enrollees’ expected health care costs. We use Verisk Health’s diagnostic cost group (DxCG) Medicare models, refined “descendants” of the same HCC framework with 189 comprehensive clinical categories available to CMS in 2004, to reveal 2 mispricing errors resulting from CMS’ implementation. One comes from ignoring all diagnostic information for “new enrollees” (those with less than 12 months of prior claims). Another comes from continuing to use the simplified models that were originally adopted in response to assertions from some capitated health plans that submitting the claims-like data that facilitate richer models was too burdensome. Even the main CMS model being used in 2014 recognizes only 79 condition categories, excluding many diagnoses and merging conditions with somewhat heterogeneous costs. Omitted conditions are typically lower cost or “vague” and not easily audited from simplified data submissions. In contrast, DxCG Medicare models use a comprehensive, 394-HCC classification system. Applying both models to Medicare’s 2010-2011 fee-for-service 5% sample, we find mispricing and lower predictive accuracy for the CMS implementation. For example, in 2010, 13% of beneficiaries had at least 1 higher cost DxCG-recognized condition but no CMS-recognized condition; their 2011 actual costs averaged US$6628, almost one-third more than the CMS model prediction. As MA plans must now supply encounter data, CMS should consider using more refined and comprehensive (DxCG-like) models.

Comparative effectiveness of first-line chemotherapies in advanced non-small cell lung cancer.

293 Background: There is a variety of chemotherapy regimens used as first-line treatment in advanced non-small cell lung cancer (NSCLC). Few studies have used real-world data to compare the clinical outcomes associated with different chemotherapy combinations. The study objective was to compare overall survival between chemotherapy regimens used for patients with advanced NSCLC in an integrated healthcare system. Methods: This retrospective cohort analysis included patients age 18 or older from Northern and Southern California Kaiser Permanente with a diagnosis of Stage IIIB/IV NSCLC identified in the cancer registry. First chemotherapy date occurred from January 2008 through September 2012 with at least 6 months of follow up to March 31, 2013. Overall survival was measured as the number of months from initial chemotherapy until death, end of enrollment or March 31, 2013. Chemotherapy regimens were categorized into five mutually exclusive groups: pemetrexed-based, pemetrexed + bevacizumab, bevacizumab-based, platinum doublets and singlets. Survival was compared between the five groups using Cox proportional hazard models, adjusting for age, gender, race, region, stage, histology, days from diagnosis to initial chemotherapy and Diagnostic Cost Group (DxCG) risk score. The analysis was performed for all age groups and by age over and under 65. Results: Among 1,989 adults, those that received singlet chemotherapy were at significantly increased risk of mortality compared to doublets (HR 1.23; 95% CI 1.02-1.49). Patients younger than 65 who had received pemetrexed + bevacizumab regimens were at significantly reduced risk of mortality (HR 0.58; 95% CI 0.35-0.95) when compared to platinum doublets. Among patients over 65, there were no significant differences in survival between the chemotherapy regimens. Conclusions: Adult patients with advanced NSCLC who are younger than 65 years old have improved survival after receiving pemetrexed + bevacizumab chemotherapy regimens compared to platinum doublet regimens. No survival differences were found between regimens in patients over the age of 65.

Association of age and comorbidity on 2009 influenza A pandemic H1N1-related intensive care unit stay in Massachusetts.

We compared comorbidity measures by age group and risk factors for influenza-like illness (ILI)-related intensive care unit (ICU) stay during the 2009 seasonal influenza and influenza A (pH1N1) pandemic. We identified all patients discharged from Massachusetts hospitals with ILI-related diagnoses between October 1, 2008, and April 25, 2009, and pH1N1-related diagnoses between April 26 and September 30, 2009. We calculated the Diagnostic Cost Group (DxCG) risk score as a measure of comorbidity. We used logistic regression predictive models to compare ICU stay predictors. Mean DxCG scores were similar for pH1N1 and seasonal influenza time periods (0.69 and 0.70). Compared with those aged 45 to 64 years, patients younger than 5, 5 to 12, and 13 to 18 years had an increased risk of pH1N1-related ICU stay. Within the pH1N1 cohort, an asthma diagnosis was highly predictive of ICU admission among those younger than 5, 5 to 12, and 13 to 18 years, and pregnancy among those aged 26 to 44 years. High-risk groups, including children with asthma or pregnant women, would benefit from improved surveillance and resource allocation during influenza outbreaks to prevent serious ILI-related complications.

Risk Adjustment for Health Care Financing in Chronic Disease

Adjustment for differing risks among patients is usually incorporated into newer payment approaches, and current risk models rely on age, sex, and diagnosis codes. It is unknown the extent to which controlling additionally for disease severity improves cost prediction. Failure to adjust for within-disease variation may create incentives to avoid sicker patients. We address this issue among patients with chronic obstructive pulmonary disease (COPD). Cost and clinical data were collected prospectively from 1202 COPD patients at Kaiser Permanente. Baseline analysis included age, sex, and diagnosis codes (using the Diagnostic Cost Group Relative Risk Score) in a general linear model predicting total medical costs in the following year. We determined whether adding COPD severity measures-forced expiratory volume in 1 second, 6-Minute Walk Test, dyspnea score, body mass index, and BODE Index (composite of the other 4 measures)-improved predictions. Separately, we examined household income as a cost predictor. Mean costs were $12,334/y. Controlling for Relative Risk Score, each ½ SD worsening in COPD severity factor was associated with $629 to $1135 in increased annual costs (all P<0.01). The lowest stratum of forced expiratory volume in 1 second (<30% normal) predicted $4098 (95% confidence interval, $576-$8773) additional costs. Household income predicted excess costs when added to the baseline model (P=0.038), but this became nonsignificant when also incorporating the BODE Index. Disease severity measures explain significant cost variations beyond current risk models, and adding them to such models appears important to fairly compensate organizations that accept responsibility for sicker COPD patients. Appropriately controlling for disease severity also accounts for costs otherwise associated with lower socioeconomic status.

Effect of Risk Adjustment Method on Comparisons of Health Care Utilization Between Complementary and Alternative Medicine Users and Nonusers

Complementary and alternative medicine (CAM) providers are becoming more integrated into the United States health care system. Because patients self-select CAM use, risk adjustment is needed to make the groups more comparable when analyzing utilization. This study examined how the choice of risk adjustment method affects assessment of CAM use on overall health care utilization. Insurance claims data for 2000-2003 from Washington State, which mandates coverage of CAM providers, were analyzed. Three (3) risk adjustment methods were compared in patients with musculoskeletal conditions: Adjusted Clinical Groups (ACG), Diagnostic Cost Groups (DCG), and the Charlson Index. Relative Value Units (RVUs) were used as a proxy for expenditures. Two (2) sets of median regression models were created: prospective, which used risk adjustments from the previous year to predict RVU in the subsequent year, and concurrent, which used risk adjustment measures to predict RVU in the same year. The sample included 92,474 claimants. Prospective models showed little difference in the effect of CAM use on RVU among the three risk adjustment methods, and all models had low predictive power (R(2) ≤0.05). In the concurrent models, coefficients were similar in direction and magnitude for all risk adjustment methods, but in some models the predicted effect of CAM use on RVU differed by as much as double between methods. Results of DCG and ACG models were similar and were stronger than Charlson models. Choice of risk adjustment method may have a modest effect on the outcome of interest.

Determining population based mortality risk in the Department of Veterans Affairs

We develop a patient level hierarchical regression model using administrative claims data to assess mortality outcomes for a national VA population. This model, which complements more traditional process driven performance measures, includes demographic variables and disease specific measures of risk classified by Diagnostic Cost Groups (DCGs). Results indicate some ability to discriminate survivors and non-survivors with an area under the Receiver Operating Characteristic Curve (C-statistic) of .86. Observed to expected mortality ranges from .86 to 1.12 across predicted mortality deciles while Risk Standardized Mortality Rates (RSMRs) range from .76 to 1.29 across 145 VA hospitals. Further research is necessary to understand mortality variation which persists even after adjusting for case mix differences. Future work is also necessary to examine the role of personal behaviors on patient outcomes and the potential impact on population survival rates from changes in treatment policy and infrastructure investment.

Does Medication Adherence Following a Copayment Increase Differ by Disease Burden?

To compare changes in medication adherence between patients with high- or low-comorbidity burden after a copayment increase. We conducted a retrospective observational study at four Veterans Affairs (VA) medical centers by comparing veterans with hypertension or diabetes required to pay copayments with propensity score-matched veterans exempt from copayments. Disease cohorts were stratified by Diagnostic Cost Group risk score: low- (<1) and high-comorbidity (>1) burden. Medication adherence from February 2001 to December 2003, constructed from VA pharmacy claims data based on the ReComp algorithm, were assessed using generalized estimating equations. Veterans with lower comorbidity were more responsive to a U.S.$5 copayment increase than higher comorbidity veterans. In the lower comorbidity groups, veterans with diabetes had a greater reduction in adherence than veterans with hypertension. Adherence trends were similar for copayment-exempt and nonexempt veterans with higher comorbidity. Medication copayment increases are associated with different impacts for low- and high-risk patients. High-risk patients incur greater out-of-pocket costs from continued adherence, while low-risk patients put themselves at increased risk for adverse health events due to greater nonadherence.

Racial Disparities in Health Status

OBJECTIVEAmerican Indians and Alaska Natives are 2.3 times more likely to have diabetes than are individuals in the U.S. general population. The objective of this study was to compare morbidity among American Indian and U.S. adults with diabetes.RESEARCH DESIGN AND METHODSWe extracted demographic and health service utilization data for an adult American Indian population aged 18–64 years (n = 30,121) served by the Phoenix Service Unit from the Indian Health Service clinical reporting system. Similar data for a U.S. population (n = 1,500,002) with commercial health insurance, matched by age and sex to the American Indian population, were drawn from the MartketScan Research Database. We used Diagnostic Cost Groups to identify medical conditions for which each individual was treated and to assign a risk score to quantify his or her morbidity burden. We compared the prevalence of comorbidities and morbidity burden of American Indian and U.S. adults with diabetes.RESULTSAmerican Indians with diabetes had significantly higher rates of hypertension, cerebrovascular disease, renal failure, lower-extremity amputations, and liver disease than commercially insured U.S. adults with diabetes (P < 0.05). The American Indian prevalence rates were 61.2, 6.9, 3.9, 1.8, and 7.1%, respectively. The morbidity burden among the American Indian with diabetes exceeded that of the insured U.S. adults with diabetes by 50%.CONCLUSIONSThe morbidity burden associated with diabetes among American Indians seen at the Phoenix Service Unit far exceeded that of commercially insured U.S. adults. These findings point to the urgency of enhancing diabetes prevention and treatment services for American Indians/Alaska Natives to reduce diabetes-related disparities.

Using Information on Clinical Conditions to Predict High‐Cost Patients

To compare the ability of different models to predict prospectively whether someone will incur high medical expenditures. Using nationally representative data from the Medical Expenditure Panel Survey (MEPS), prediction models were developed using cohorts initiated in 1996-1999 (N=52,918), and validated using cohorts initiated in 2000-2003 (N=61,155). We estimated logistic regression models to predict being in the upper expenditure decile in Year 2 of a cohort, based on data from Year 1. We compared a summary risk score based on diagnostic cost group (DCG) prospective risk scores to a count of chronic conditions and indicators for 10 specific high-prevalence chronic conditions. We examined whether self-rated health and functional limitations enhanced prediction, controlling for clinical conditions. Models were evaluated using the Bayesian information criterion and the c-statistic. Medical condition information substantially improved prediction of high expenditures beyond gender and age, with the DCG risk score providing the greatest improvement in prediction. The count of chronic conditions, self-reported health status, and functional limitations were significantly associated with future high expenditures, controlling for DCG score. A model including these variables had good discrimination (c=0.836). The number of chronic conditions merits consideration in future efforts to develop expenditure prediction models. While significant, self-rated health and indicators of functioning improved prediction only slightly.

Performance of Comorbidity, Risk Adjustment, and Functional Status Measures in Expenditure Prediction for Patients With Diabetes

OBJECTIVE—To compare the ability of generic comorbidity and risk adjustment measures, a diabetes-specific measure, and a self-reported functional status measure to explain variation in health care expenditures for individuals with diabetes.RESEARCH DESIGN AND METHODS—This study included a retrospective cohort of 3,092 diabetic veterans participating in a multisite trial. Two comorbidity measures, four risk adjusters, a functional status measure, a diabetes complication count, and baseline expenditures were constructed from administrative and survey data. Outpatient, inpatient, and total expenditure models were estimated using ordinary least squares regression. Adjusted R2 statistics and predictive ratios were compared across measures to assess overall explanatory power and explanatory power of low- and high-cost subgroups.RESULTS—Administrative data–based risk adjusters performed better than the comorbidity, functional status, and diabetes-specific measures in all expenditure models. The diagnostic cost groups (DCGs) measure had the greatest predictive power overall and for the low- and high-cost subgroups, while the diabetes-specific measure had the lowest predictive power. A model with DCGs and the diabetes-specific measure modestly improved predictive power.CONCLUSIONS—Existing generic measures can be useful for diabetes-specific research and policy applications, but more predictive diabetes-specific measures are needed.

Implications of Comorbidity on Costs for Patients With Alzheimer Disease

No prior studies have used a comprehensive clinical classification system to examine the effect of differences in overall illness burden and the presence of other diseases on costs for patients with Alzheimer disease (AD) when compared with demographically matched nondemented controls. Of a total of 627,775 enrollees who were eligible for medical and pharmacy benefits for 2003 and 2004 in the MarketScan Medicare Supplemental and Coordination of Benefits Database, we found 25,109 AD patients. For each case, 3 demographically matched nondemented controls were selected using propensity scores. Applying the diagnostic cost groups (DCGs) model to all enrollees, 2003 diagnoses were used to estimate prospective relative risk scores (RRSs) that predict 2004 costs from all illness other than AD. RRSs were then used to control for illness burden to estimate AD's independent effect on costs. Compared with the control group, the AD cohort has more comorbid conditions (8.1 vs. 6.5) and higher illness burden (1.23 vs. 1.04). Individuals with AD are more likely to have mental health conditions, neurologic conditions, cognitive disorders, cerebrovascular disease, diabetes with acute complications, and injuries. Annual costs for AD patients are $3567 (34%) higher than for controls. Excess costs attributable to AD, after controlling for non-AD illness burden, are estimated at $2307 per year with outpatient pharmacy being the key driver ($1711 in excess costs). AD patients are sicker and more expensive than demographically matched controls. Even after adjusting for differences in illness burden, costs remain higher for AD patients.

Healthcare costs and utilization for Medicare beneficiaries with Alzheimer's

BackgroundAlzheimer's disease (AD) is a neurodegenerative disorder incurring significant social and economic costs. This study uses a US administrative claims database to evaluate the effect of AD on direct healthcare costs and utilization, and to identify the most common reasons for AD patients' emergency room (ER) visits and inpatient admissions.MethodsDemographically matched cohorts age 65 and over with comprehensive medical and pharmacy claims from the 2003–2004 MEDSTAT MarketScan® Medicare Supplemental and Coordination of Benefits (COB) Database were examined: 1) 25,109 individuals with an AD diagnosis or a filled prescription for an exclusively AD treatment; and 2) 75,327 matched controls. Illness burden for each person was measured using Diagnostic Cost Groups (DCGs), a comprehensive morbidity assessment system. Cost distributions and reasons for ER visits and inpatient admissions in 2004 were compared for both cohorts. Regression was used to quantify the marginal contribution of AD to health care costs and utilization, and the most common reasons for ER and inpatient admissions, using DCGs to control for overall illness burden.ResultsCompared with controls, the AD cohort had more co-morbid medical conditions, higher overall illness burden, and higher but less variable costs ($13,936 s. $10,369; Coefficient of variation = 181 vs. 324). Significant excess utilization was attributed to AD for inpatient services, pharmacy, ER visits, and home health care (all p < 0.05). In particular, AD patients were far more likely to be hospitalized for infections, pneumonia and falls (hip fracture, syncope, collapse).ConclusionPatients with AD have significantly more co-morbid medical conditions and higher healthcare costs and utilization than demographically-matched Medicare beneficiaries. Even after adjusting for differences in co-morbidity, AD patients incur excess ER visits and inpatient admissions.

Risk-Adjusting Hospital Inpatient Mortality Using Automated Inpatient, Outpatient, and Laboratory Databases

To develop a risk-adjustment methodology that maximizes the use of automated physiology and diagnosis data from the time period preceding hospitalization. : Retrospective cohort study using split-validation and logistic regression. Seventeen hospitals in a large integrated health care delivery system. Patients (n = 259,699) hospitalized between January 2002 and June 2005. Inpatient and 30-day mortality. Inpatient mortality was 3.50%; 30-day mortality was 4.06%. We tested logistic regression models in a randomly chosen derivation dataset consisting of 50% of the records and applied their coefficients to the validation dataset. The final model included sex, age, admission type, admission diagnosis, a Laboratory-based Acute Physiology Score (LAPS), and a COmorbidity Point Score (COPS). The LAPS integrates information from 14 laboratory tests obtained in the 24 hours preceding hospitalization into a single continuous variable. Using Diagnostic Cost Groups software, we categorized patients as having up to 40 different comorbidities based on outpatient and inpatient data from the 12 months preceding hospitalization. The COPS integrates information regarding these 41 comorbidities into a single continuous variable. Our best model for inpatient mortality had a c statistic of 0.88 in the validation dataset, whereas the c statistic for 30-day mortality was 0.86; both models had excellent calibration. Physiologic data accounted for a substantial proportion of the model's predictive ability. Efforts to support improvement of hospital outcomes can take advantage of risk-adjustment methods based on automated physiology and diagnosis data that are not confounded by information obtained after hospital admission.

Analysis of predictive value of four risk models in medicaid recipients with chronic obstructive pulmonary disease in Texas

Objective: The aim of this study was to assess the predictive value of 4 risk models when applied to Medicaid recipients with chronic obstructive pulmonary disease (COPD) in Texas. Methods: Four risk models were used to predict future medical costs: Diagnostic Cost Group (DCG) Inpatient + RxGroups Commercial, DCG All-Encounter + RxGroups Commercial, DCG All-Encounter + RxCost Medicaid, and prior cost. Diagnostic codes, National Drug Code categories, and 2001 total health care costs were input into the risk models. The study included Texas Medicaid recipients aged 40 to <65 years who had an inpatient diagnosis of COPD in 2000 and were continuously enrolled in Medicaid in 2000–2002. The risk models were used to predict total health care costs for 2002. Predicted costs were compared with actual 2002 costs using R 2, and predictions of high-cost cases were analyzed using the positive predictive value (PPV) of the top 5% of high-cost patients. Results: Information from 7967 patients was used in this analysis. The R 2 values were 0.11, 0.08, 0.05, and 0.04 with the DCG Inpatient + RxGroups Commercial, DCG All-Encounter + RxGroups Commercial, DCG All-Encounter + RxCost Medicaid, and prior cost models, respectively. In the top 5% of high-cost patients, the corresponding PPVs were 26% (105/399), 23% (90/399), 18% (72/399), and 28% (112/399). Conclusions: The DCG Inpatient + RxGroups Commercial model was the best predictor of future costs as measured using R 2. However, diagnosis-based risk-adjustment models may not be as useful as prior cost data when attempting to identify future high-cost cases among Texas Medicaid recipients with COPD. A combination of diagnosis-based risk adjustment models and prior cost data may prove to be the most effective way to use claims data to identify COPD candidates for case management.

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.