[Predicting individual risk of high healthcare cost to identify complex chronic patients]. / Predicción del riesgo individual de alto coste sanitario para la identificación de pacientes crónicos complejos.

OBJECTIVE: To develop a predictive model for the risk of high consumption of healthcare resources, and assess the ability of the model to identify complex chronic patients. METHODS: A cross-sectional study was performed within a healthcare management organization by using individual data from 2 consecutive years (88,795 people). The dependent variable consisted of healthcare costs above the 95th percentile (P95), including all services provided by the organization and pharmaceutical consumption outside of the institution. The predictive variables were age, sex, morbidity-based on clinical risk groups (CRG)-and selected data from previous utilization (use of hospitalization, use of high-cost drugs in ambulatory care, pharmaceutical expenditure). A univariate descriptive analysis was performed. We constructed a logistic regression model with a 95% confidence level and analyzed sensitivity, specificity, positive predictive values (PPV), and the area under the ROC curve (AUC). RESULTS: Individuals incurring costs >P95 accumulated 44% of total healthcare costs and were concentrated in ACRG3 (aggregated CRG level 3) categories related to multiple chronic diseases. All variables were statistically significant except for sex. The model had a sensitivity of 48.4% (CI: 46.9%-49.8%), specificity of 97.2% (CI: 97.0%-97.3%), PPV of 46.5% (CI: 45.0%-47.9%), and an AUC of 0.897 (CI: 0.892 to 0.902). CONCLUSIONS: High consumption of healthcare resources is associated with complex chronic morbidity. A model based on age, morbidity, and prior utilization is able to predict high-cost risk and identify a target population requiring proactive care.

The effect of hepatitis C treatment response on medical costs: a longitudinal analysis in an integrated care setting.

BACKGROUND: Studies suggest that chronic hepatitis C patients who achieve sustained virologic response (SVR) have lower risks of liver-related morbidity and mortality. Given the substantial costs and complexity of hepatitis C virus (HCV) antiviral treatment, post-treatment benefits are important to understand.   OBJECTIVE: To determine whether health care costs and utilization for up to 5 years after treatment differed between patients who achieved SVR and those who did not.  METHODS: Kaiser Permanente Medical Care Program patients receiving HCV treatment with pegylated interferon and ribavirin (Peg-IFN/RBV) from 2002 to 2007 were retrospectively analyzed, excluding those with human immunodeficiency virus (HIV) or chronic hepatitis B. Health care utilization and costs for up to 5 years after treatment completion were derived from electronic records. We compared mean annual cost and overall post-treatment costs (standardized to year-2007 dollars), and yearly utilization counts between the SVR and non-SVR groups, adjusting for pretreatment costs, age, sex, baseline cirrhosis, and race using gamma and Poisson regression models.  RESULTS: The 1,924 patients eligible for inclusion were a mean age of 50 years; 63% male; 58% white, non-Hispanic; 62% with genotype 1; and 48% who had achieved SVR. The mean duration of post-treatment time was 3 years, and patients without SVR incurred significantly higher health care costs than patients with SVR. For each post-treatment year, total adjusted costs were significantly higher in the non-SVR group than in the SVR group, with rate ratios (RRs) and 95% CIs ranging from 1.26 (95% CI, 1.13-1.40) to 1.64 (95% CI, 1.38-1.96), driven mostly by hospital and outpatient pharmacy costs. When all post-treatment years were considered collectively, the non-SVR group had significantly higher costs overall (RR=1.41; 95% CI, 1.17-1.69) and in each category of costs. The adjusted difference in yearly total mean costs was $2,648 (95% CI, 737-4,560). In post-treatment years 2-5, adjusted liver-specific laboratory test rates were 1.8 to 2.3 times higher in the non-SVR group than in the SVR group (each year, P less than 0.001). During post-treatment years 1-5, adjusted yearly liver-related hospitalization rates were up to 2.45 times higher (95% CI, 1.56-3.85), and medicine/GI clinic visit rates were up to 1.39 times higher (95% CI, 1.23-1.54) in the non-SVR group compared with the SVR group.  CONCLUSION: Health care utilization and costs after HCV antiviral therapy with Peg-IFN/RBV, particularly for liver-related tests, outpatient drugs, and hospitalizations, were significantly lower for patients who achieved SVR than for those without SVR. Our observations are consistent with the potentially lower risk of severe liver disease among patients with SVR. 

Paying for quality not quantity: a wisconsin health maintenance organization proposes an incentive model for reimbursement of chiropractic services.

OBJECTIVE: The purpose of this study is to describe a reimbursement model that was developed by one Health Maintenance Organization (HMO) to transition from fee-for-service to add a combination of pay for performance and reporting model of reimbursement for chiropractic care. METHODS: The previous incentive program used by the HMO provided best-practice education and additional reimbursement incentives for achieving the National Committee for Quality Assurance Back Pain Recognition Program (NCQA-BPRP) recognition status. However, this model had not leveled costs between doctors of chiropractic (DCs). Therefore, the HMO management aimed to develop a reimbursement model to incentivize providers to embrace existing best-practice models and report existing quality metrics. The development goals included the following: it should (1) be as financially predictable as the previous system, (2) cost no more on a per-member basis, (3) meet the coverage needs of its members, and (4) be able to be operationalized. The model should also reward DCs who embraced best practices with compensation, not simply tied to providing more procedures, the new program needed to (1) cause little or no disruption in current billing, (2) be grounded achievable and defined expectations for improvement in quality, and (3) be voluntary, without being unduly punitive, should the DC choose not to participate in the program. RESULTS: The generated model was named the Comprehensive Chiropractic Quality Reimbursement Methodology (CCQRM; pronounced "Quorum"). In this hybrid model, additional reimbursement, beyond pay-for-procedures will be based on unique payment interpretations reporting selected, existing Physician Quality Reporting System (PQRS) codes, meaningful use of electronic health records, and achieving NCQA-BPRP recognition. This model aims to compensate providers using pay-for-performance, pay-for-quality reporting, pay-for-procedure methods. CONCLUSION: The CCQRM reimbursement model was developed to address the current needs of one HMO that aims to transition from fee-for-service to a pay-for-performance and quality reporting for reimbursement for chiropractic care. This model is theoretically based on the combination of a fee-for-service payment, pay for participation (NCQA Back Pain Recognition Program payment), meaningful use of electronic health record payment, and pay for reporting (PQRS-BPMG payment). Evaluation of this model needs to be implemented to determine if it will achieve its intended goals.

Barriers to physical and mental condition integrated service delivery.

OBJECTIVE: To assess pragmatic challenges faced when implementing, delivering, and sustaining models of integrated mental health intervention in primary care settings. Thirty percent of primary care patients with chronic medical conditions and up to 80% of those with health complexity have mental health comorbidity, yet primary care clinics rarely include onsite mental health professionals and only one in eight patients receive evidence-based mental health treatment. Integrating specialty mental health into primary care improves outcomes for patients with common disorders, such as depression. METHODS: We used key informant interviews documenting barriers to implementation and components that inhibited or enhanced operational success at 11 nationally established integrated physical and mental condition primary care programs. RESULTS: All but one key informant indicated that the greatest barrier to the creation and sustainability of integrated mental condition care in primary care settings was financial challenges introduced by segregated physical and mental health reimbursement practices. For integrated physical and mental health program initiation and outcome changing care to be successful, key components included a clinical and administrative champion-led culture shift, which valued an outcome orientation; cross-disciplinary training and accountability; use of care managers; consolidated clinical record systems; a multidisease, total population focus; and active, respectful coordination of colocated interdisciplinary clinical services. CONCLUSIONS: Correction of disparate physical and mental health reimbursement practices is an important activity in the development of sustainable integrated physical and mental condition care in primary care settings, such as a medical home. Multiple clinical, administrative, and economic factors contribute to operational success.

The direct medical cost of cardiovascular diseases, hypertension, diabetes, cancer, pregnancy and female infertility in a large HMO in Israel.

OBJECTIVES: The aim of this study was to assess the direct medical cost of treating major chronic illnesses in Maccabi Healthcare Services, a 1.8 million member health maintenance organization in Israel. METHODS: Direct medical costs were calculated for each member in 2006. We used multiple linear regression models to evaluate the overall costs of chronic conditions (cardiovascular diseases, diabetes mellitus, hypertension, female infertility treatments, and cancer), pregnancy and treatments for female infertility. RESULTS: According to the study model, hypertension was associated with the largest direct medical costs in both sexes. Cardiovascular diseases accounted for 9.5% of the total direct medical costs in men, but only 5.9% in women. Diabetes mellitus accounted for 3.5% of the total medical costs both in men and women and is comparable to the total pregnancy-related costs in women. CONCLUSIONS: The findings indicate that hypertension, diabetes mellitus and female infertility treatments impose a considerable economic burden on public healthcare services in Israel which is comparable with the costs of cancer and cardiovascular diseases.

Integrated delivery systems: the cure for fragmentation.

Our healthcare system is fragmented, with a misalignment of incentives, or lack of coordination, that spawns inefficient allocation of resources. Fragmentation adversely impacts quality, cost, and outcomes. Eliminating waste from unnecessary, unsafe care is crucial for improving quality and reducing costs--and making the system financially sustainable. Many believe this can be achieved through greater integration of healthcare delivery, more specifically via integrated delivery systems (IDSs). An IDS is an organized, coordinated, and collaborative network that links various healthcare providers to provide a coordinated, vertical continuum of services to a particular patient population or community. It is also accountable, both clinically and fiscally, for the clinical outcomes and health status of the population or community served, and has systems in place to manage and improve them. The marketplace already contains numerous styles and degrees of integration, ranging from Kaiser Permanente-style full integration, to more loosely organized individual practice associations, to public-private partnerships. Evidence suggests that IDSs can improve healthcare quality, improve outcomes, and reduce costs--especially for patients with complex needs--if properly implemented and coordinated. No single approach or public policy will fix the fragmented healthcare system, but IDSs represent an important step in the right direction.

Cancer screening before and after switching to a high-deductible health plan.

BACKGROUND: Health plans with high deductibles could lead patients to avoid preventive care, such as cancer screening. OBJECTIVE: To determine the effect of membership in a high-deductible health plan on cervical, breast, and colorectal cancer screening. DESIGN: Before-after comparison between groups. SETTING: A high-deductible health plan and an HMO in Massachusetts. The high-deductible health plan fully covered mammography, Papanicolaou tests, and fecal occult blood testing (FOBT) but not colonoscopy, flexible sigmoidoscopy, or double-contrast barium enema (DCBE). PARTICIPANTS: 3169 high-deductible health plan members and 27,022 HMO members (who served as controls). MEASUREMENTS: Change in the proportions of patients undergoing breast, cervical, and colorectal cancer screening. RESULTS: Cancer screening in the high-deductible health plan group was unchanged from baseline to follow-up (adjusted ratios of change, 1.04 [95% CI, 0.91 to 1.19] for breast cancer, 1.04 [CI, 0.92 to 1.17] for cervical cancer, and 1.02 [CI, 0.89 to 1.16] for colorectal cancer). High-deductible health plan members had colonoscopy, flexible sigmoidoscopy, and DCBE less often (ratio of change, 0.73 [CI, 0.56 to 0.95]) and FOBT more often (ratio of change, 1.16 [CI, 1.01 to 1.33]) than HMO members. LIMITATIONS: Population screening frequency was probably underestimated because the study could not assess screening before the baseline year. The study may have included people ineligible for screening because of previous colectomy, mastectomy, or hysterectomy. The findings are limited to a population with relatively high socioeconomic status, which is typical of employed, commercially insured populations. CONCLUSION: Members of a high-deductible health plan did not seem to change their use of breast, cervical, and colorectal cancer screening when tests were fully covered. However, members may have substituted a fully covered screening test (FOBT) for tests subject to the deductible (colonoscopy, flexible sigmoidoscopy, and DCBE).

Utilization and drug cost outcomes of a step-therapy edit for generic antidepressants in an HMO in an integrated health system.

OBJECTIVE: Antidepressants do not differ significantly in their ability to treat depression. Excluding the tricyclic antidepressants (TCAs), these drugs also do not differ significantly in their incidence of adverse events. Therefore, the initial choice of antidepressant medication should be based, in part, on cost. The objective of this study was to evaluate the impact on utilization and costs of a generic steptherapy edit for antidepressant drugs excluding TCAs in a health maintenance organization (HMO) in an integrated health system (IHS). METHODS: The pharmacy department of the 440,000-member HMO in an IHS collaborated with the Behavioral Health Clinical Program to design an intervention that required generic antidepressants as first-line pharmacotherapy. Under the GenericStart! Program, a brand-name antidepressant was covered only after trial with a generic antidepressant, excluding TCAs. A step-therapy edit was added to the pharmacy claims processing system on January 1, 2005. All new starts, defined as members with no claims history of antidepressant treatment within the preceding 6 months, were required to use a generic antidepressant. The member copayment was waived for the first prescription. All generic antidepressants were in tier 1 of the drug formulary, with an average copayment of $5 to $10. All brand-name antidepressants were in either tier 2 (preferred brand), with an average copayment of $20 to $25 or 25% coinsurance, or tier 3 (nonformulary brand), with an average copayment of $40 to $45 or 50% coinsurance. Pharmacy claims data from a national pharmacy benefit manager (PBM) without interventions for antidepressants in 2004 or 2005 were used for the comparison group. RESULTS: The generic antidepressant dispensing rate increased by 20 points (32.5% to 52.5%) in the intervention group but only 7.4 points (24.9% to 32.3%) in the comparison group in 2005 compared with 2004. The principal measure of antidepressant drug cost per day of therapy in the intervention group decreased by 11.7% (from $2.40 to $2.12) in 2005 compared with 2004 versus a 2.7% decrease (from $2.60 to $2.53) in the comparison group (P <0.001). Days of antidepressant drug therapy per member per month (PMPM) dropped by 1.5% (from 1.74 to 1.71) in the intervention group versus a decrease of 5.0% (from 1.37 to 1.30) in the comparison group in 2005 compared with 2004. The combination of change in drug cost and utilization resulted in a 13.0% decrease in antidepressant drug cost, from $4.16 PMPM in 2004 to $3.62 in 2005, compared with a 7.6% decrease (from $3.57 to $3.30 PMPM) in the comparison group. The 9.0% difference in drug cost per day represents drug cost savings of approximately $0.36 PMPM or $1,880,562 in 2005 dollars for this HMO of approximately 440,000 members. CONCLUSION: A step-therapy edit requiring HMO members to use a generic antidepressant, excluding tricyclics, prior to use of a brand-name antidepressant resulted in drug cost savings of 9.0% for the entire class of antidepressants, equal to $1,880,562 ($0.36 PMPM) in 2005 dollars in the first year of the intervention. A small (-1.5%) decrease in use of antidepressants occurred in the intervention group, which was less than the 5.0% decrease in utilization of antidepressants in the comparison group.

[Annual cost of ischemic heart disease in Brazil. Public and private perspective]. / Custo anual do manejo da cardiopatia isquêmica crônica no Brasil. Perspectiva pública e privada.

OBJECTIVE: To estimate the annual cost of coronary artery disease (CAD) management in Public Health Care System (SUS) and HMOs values in Brazil. METHODS: Cohort study, including ambulatory patients with proven CAD. Clinic visits, exams, procedures, hospitalizations and medications were considered to estimate direct costs. Values of appointments and exams were obtained from the SUS and the Medical Procedure List (LPM 1999) reimbursement tables. Costs of cardiovascular events were obtained from admissions in public and private hospitals with similar diagnoses-related group classifications in 2002. The price of medications used was the lowest found in the market. RESULTS: The 147 patients (65 +/- 12 years old, 63% men, 69% hypertensive, 35% diabetic and 59% with previous AMI) had an average follow-up of 24 +/- 8 months. The average estimated annual cost per patient was R$ 2,733.00, for the public sector, and R$ 6,788.00, for private and fee-for-service plans. Expenses with medications (R$ 1,154.00) represented 80% and 55% of outpatient costs, and 41% and 17% of total expenses, in public and non-public sectors, respectively. The occurrence of cardiovascular event had a great impact (R$ 4,626.00 vs. R$ 1,312.00, in SUS, and R$ 13,453.00 vs. R$ 1,789.00, for HMOs, p<0.01) on the results. CONCLUSION: The average annual cost of CAD management was high, being the pharmacological treatment the main determinant of public costs. Such estimates may subsidize economical analyses in this area, and foster related healthcare policies.

Custo anual do manejo da cardiopatia isquêmica crônica no Brasil: perspectiva pública e privada / Annual cost of ischemic heart disease in Brazil: public and private perspective

OBJETIVO: Estimar o custo anual do manejo da doenca arterial coronária (DAC) em valores do SUS e convênios. MÉTODOS: Estudo de coorte, incluindo pacientes ambulatoriais com DAC comprovada. Considerou-se para estimar custos diretos: consultas, exames, procedimentos, internacões e medicamentos. Valores de consultas e exames foram obtidos da tabela SUS e da Lista de Procedimentos Médicos (LPM). Valores de eventos cardiovasculares foram obtidos de internacões em hospital público e privado com estas classificacões diagnósticas em 2002. O preco dos fármacos utilizado foi o de menor custo no mercado. RESULTADOS: Os 147 pacientes (65n12 anos, 63 por cento homens, 69 por cento hipertensos, 35 por cento diabéticos e 59 por cento com IAM prévio) tiveram acompanhamento médio de 24n8 meses. O custo anual médio estimado por paciente foi de R$ 2.733,00, pelo SUS, e R$ 6.788,00, para convênios. O gasto com medicamentos ($ 1.154,00) representou 80 por cento e 55 por cento dos custos ambulatoriais, e 41 por cento e 17 por cento dos gastos totais, pelo SUS e para convênios, respectivamente. A ocorrência de evento cardiovascular teve grande impacto (R$ 4.626,00 vs. R$ 1.312,00, pelo SUS, e R$ 13.453,00 vs. R$ 1.789,00, para convênios, p<0,01). CONCLUSAO: O custo médio anual do manejo da DAC foi elevado, sendo o tratamento farmacológico o principal determinante dos custos públicos. Essas estimativas podem subsidiar análises econômicas nesta área, sendo úteis para nortear políticas de saúde pública.

Economic evaluation of four treatments for low-back pain: results from a randomized controlled trial.

OBJECTIVE: We sought to compare total outpatient costs of 4 common treatments for low-back pain (LBP) at 18-months follow-up. METHODS: Our work reports on findings from a randomized controlled trial within a large medical group practice treating HMO patients. Patients (n = 681) were assigned to 1 of 4 treatment groups, ie, medical care only (MD), medical care with physical therapy (MDPt), chiropractic care only (DC), or chiropractic care with physical modalities (DCPm). Total outpatient costs, excluding pharmaceuticals, were measured at 18 months. We did not perform a cost-effectiveness analysis because previously published findings showed no clinically meaningful difference in outcomes among the 4 treatment groups. Thirty-seven participants were lost to follow-up at 18 months, leaving a final sample size of n = 654. RESULTS: Adjusting for covariates, DC was 51.9% more expensive than MD (P < 0.001), DCPm 3.2% more expensive than DC (P = 0.76), and MDPt 105.8% more expensive than MD (P < 0.001). The adjusted mean outpatient costs per treatment group were 369 US dollars for MD, 560 US dollars for DC, 579 US dollars for DCPm, and 760 US dollars for MDPt. CONCLUSIONS: This study is the first randomized trial to show higher costs for chiropractic care without producing better clinical outcomes, but our findings are likely to understate the costs of medical care with or without physical therapy because of the absence of pharmaceutical data. Physical therapy provided in combination with medical care and physical modalities provided in combination with chiropractic care do not appear to be cost-effective strategies for treatment of LBP; they produce higher costs without clinically significant improvements in outcome.

Cost-effectiveness with multiple outcomes.

In a large number of situations, activities in health care have to be measured in terms of outcome and cost. However, the cases where outcome is fully captured by a single measure are rather few, so that one uses some index for outcome, computed by weighing together several outcome measures using subjective and somewhat arbitrary weights. In the paper we propose an approach to cost-effectiveness analysis where such artificial aggregation is avoided. This is achieved by assigning to each activity the weights which are the most favourable in a comparison with the other options available, so that activities which have a poor score in this method are guaranteed to be inferior. The method corresponds to applying Data envelopment analysis, known from the theory of productivity, to the context of health economic evaluations. The method is applied to an analysis of the cost-effectiveness of alternative health plans using data from the Medical Outcome Study (JAMA 1996; 276: 1039-1047), where outcome is measured as improvement in mental and physical health.

Credibility and creativity: a conversation with Kaiser Permanente's George C. Halvorson.

In 2002 Kaiser Permanente's board surprised the industry by reaching outside its organization and selecting a nonphysician leader, George C. Halvorson, then CEO of HealthPartners of Minneapolis. In this interview Halvorson talks about returning to Kaiser's strengths--its sixty-year-plus history of integrated health care organization and its power base along the Pacific Coast--and about how he and his physician colleagues intend to leverage clinical information technology to improve their subscribers' health. Halvorson also discusses the new Medicare prescription drug legislation, the sources of the current run-up in health costs, and how he intends to position Kaiser for future growth.

International care models for chronic kidney disease: methods and economics--United States.

In the United States, there is a major chronic kidney disease (CKD) problem with over 8 million adults having stage 3 or 4 CKD. There is good medical evidence that many of these patients can benefit from focused interventions. And while there are strong theoretical reasons to believe these interventions are cost-effective, there are little published data to back up this assertion. However, despite the lack of financial data proving cost-effectiveness and against the background of a disorganized health care system in the US, some models of CKD care are being employed. At the present time, the most comprehensive models of care in the US are emerging in vertically integrated health care programs. Other models of care are developing in the setting of managed care health plans that employ CKD disease management programs, either developed internally or in partnership with renal disease management companies.

Hospital tiers in health insurance: balancing consumer choice with financial incentives.

Variations in efficiency and market power are generating wide variations in the prices charged by hospitals to health insurance plans. Insurers are developing new network structures that expose the consumer to some of the cost differences, to encourage but not mandate differential use of the more economical facilities. The three leading designs include hospital "tiers" within a single broad network, multiple-network products, and the replacement of copayments by coinsurance in HMO as well as PPO products. This paper describes the new network designs and evaluates the challenges they face in influencing consumers' behavior, incorporating information on clinical quality, and supporting medical education and uncompensated care.

Impact of subsidizing effective anti-osteoporosis drugs on compliance with management guidelines in patients following low-impact fractures.

Early in 2000, proven-effective antiresorptive drugs (alendronate and raloxifene) were included in the national "health basket" in Israel. We carried out the present study to evaluate the effect of subsidizing antiosteoporosis drugs on the use of antiosteoporosis drugs in patients following low-impact fractures. The rates of dispensation of antiosteoporosis drugs, in the hospital and in the community, before and after an incident of a newly diagnosed low-impact fracture, respectively, were evaluated during January and February 1998 and 1999 ("pre-basket") and the corresponding months of 2000 and 2001 ("post-basket"). The study was carried out in a 950-bed teaching hospital, the only one serving the area, and the largest health maintenance organization in the area. Hospital charts of women and men age 50 years and older with new fractures following low- or moderate-impact trauma treated in the emergency room, or admitted to the orthopedic surgery and rehabilitation departments, were reviewed. A centralized pharmacy computerized database was used to follow antiosteoporosis drug dispensation in the community. A significant, approximately two-fold, increase in the baseline (before fracture) rate of osteoporosis drug dispensation was observed between the pre- and post-basket periods. The rate of patients treated after a fracture incident also increased significantly, 1.6 fold, in the post-basket period; however, even in the post-basket period, two-thirds of the patients remained untreated following a fracture incident, and most of those treated received only calcium and vitamin D; only 17% received potent antiosteoporosis drugs. In a multivariate analysis, female gender, hospitalization, having the incident of fracture in the post-basket period, and above all being treated for osteoporosis before the fracture incident, had the greatest effect on the likelihood of being treated following a low-impact fracture incident. The increase in the pooled use of antiosteoporosis drugs and/or calcium/vitamin D supplements was continuous, and subsidizing created no step-up effect, besides a transient increase in the use of potent antiosteoporosis drugs in the first year following the health-basket amendment. We conclude that while subsidizing may have a significant, positive effect on antiosteoporosis drug utilization, other factors may be even more important. There is an ongoing need to find ways to encourage the use of effective pharmacological interventions for primary and secondary prevention of osteoporotic fractures.

Utilization and cost impact of integrating substance abuse treatment and primary care.

OBJECTIVE: To examine the impact of integrating medical and substance abuse treatment on health care utilization and cost. RESEARCH DESIGN: Randomized clinical trial assigning patients to one of two treatment modalities: an Integrated Care model where primary health care is provided along with substance abuse treatment within the unit and an Independent Care model where medical care is provided in the HMO's primary care clinics independently from substance abuse treatment. SUBJECTS: Adult patients entering treatment at the outpatient Chemical Dependency Recovery Program in Kaiser Sacramento. MEASURES: Medical utilization and cost for 12 months pretreatment and 12 months after treatment entry. RESULTS: For the full, randomized cohort, there were no statistically significant differences between the two treatment groups over time. However, among the subset of patients with substance abuse related medical conditions (SAMC), Integrated Care patients had significant decreases in hospitalization rates (P = 0.04), inpatient days (P = 0.05) and ER use (P = 0.02). Total medical costs per member-month declined from 431.12 US dollars to 200.03 US dollars (P = 0.02). Among SAMC Independent Care patients, there was a downward trend in inpatient days (P = 0.08) and ER costs (P = 0.05) but no statistically significant decrease in total medical cost. CONCLUSIONS: (Non)findings for the full sample suggest that integrating substance abuse treatment with primary care, may not be necessary or appropriate for all patients. However, it may be beneficial to refer patients with substance abuse related medical conditions to a provider also trained in addiction medicine. There appear to be large cost impacts of providing integrated care for such patients.