Multi-payer Claims Database Research Articles

Race/ethnicity and socioeconomic position in emergency department utilization in patients with hepatocellular carcinoma.

Aim: Evaluate the association of race/ethnicity and socioeconomic position (SEP) on emergency department (ED) visits for patients with hepatocellular carcinoma (HCC), which may reflect access to and quality of cancer care.Materials & methods: Patients with HCC identified from a commercial multi-payer claims database between 2015 and2018 were matched to near-neighborhood social determinants of health (SDOH) and stratified by race/ethnicity and SEP (proxied by annual household income). Analyses evaluated the effect of race/ethnicity and SEP on ED utilization, adjusting for SDOH, demographic and clinical characteristics using multivariable regression methods.Results: A total of22,247 patients were included. Black and Hispanic patients had 43and 18% higher ED utilization than White patients at higher-income levels (p<0.01); these differences were nonsignificant at lower-income. Regardless of income level, Asian patients had lower ED utilization.Conclusion: Further research on the intersectionality between race/ethnicity, SEP and other SDOH may guide structural-level interventions to address health inequities.

Dietary Counseling Documentation Among Patients Recently Hospitalized for Cardiovascular Disease

BackgroundDiet intervention forms the cornerstone for cardiovascular disease (CVD) management. ObjectiveThe objective was to measure the frequency of dietary counseling documentation for patients recently hospitalized with CVD. DesignThis was an observational study. Participants and settingPatients were included from the Michigan Value Collaborative Multipayer Claims Registry from October 2015 to February 2020. Main outcome measureThe study measured the frequency of medical claims that document dietary counseling ≤90 days after hospitalization (ie, an episode of care) for CVD events (coronary artery bypass grafting, acute myocardial infarction, congestive heart failure, and percutaneous coronary intervention). Dietary counseling documentation was defined as having an encounter-level International Classification of Diseases 10th Revision code for dietary counseling or current procedural terminology code for medical nutrition therapy or cardiac rehabilitation. Statistical analyses performedMultivariable logistic regression was used to measure variation in documentation across gender, age, comorbidities, hospital geography, CVD event, and insurer. ResultsThere were 175,631 episodes of care (congesitve heart failure 47.1%, acute myocardial infarction 28.7%, percutaneous coronary intervention 17.0%, and coronary artery bypass grafting 7.3%) among 146,185 individuals. Most episodes occurred among men (55.8%) and those older than age 65 years (71.9%). Dietary counseling was documented for 22.8% of episodes and was more common as cardiac rehabilitation (18.6%) than other encounter types (5.1%). In multivariable analysis, there was lower odds for dietary counseling documentation among those older than age 65 years (odds ratio [OR] 0.77; P < .001), women (OR 0.83; P < .001), with chronic kidney disease (OR 0.74; P < .001), or diabetes (OR 0.95; P < .001), but greater odds for those with obesity (OR 1.28; P < .001) and nonmetropolitan hospitals (OR 1.31; P < .001). Compared with coronary artery bypass grafting, acute myocardial infarction (OR 0.29; P < .001), confestive heart failure (OR 0.12; P < .001), and percutaneous coronary intervention (OR 0.36; P < .001) episodes had lower odds to have dietary counseling coded. Compared with Traditional Medicare, Medicaid and Medicare Advantage health maintenance organization plans had lower odds, whereas Commercial or Medicare Advantage preferred provider organization and Commercial health maintenance organization plans had higher odds to have dietary counseling documented. Results were mostly similar when evaluated by race. ConclusionsDietary counseling was infrequently documented after hospitalization for CVD episodes in medical claims in a Michigan-based multipayer claims database with large variation by reason for hospitalization and patient factors.

Abstract WP108: Predictive Accuracy of Multivariable Models for 30-day and 1-year All Cause Readmission After Stroke Using Different Combinations of Registry, Hospital and Claims Based Data Sources

Introduction: Hospital readmissions are often used as an indicator of quality of care. However, identifying patients at risk of readmission after stroke is challenging and predictive models have historically not performed well, in part because they often rely on single data sources. Data linkage might offer a solution. Methods: We probabilistically linked data from the Michigan’s Get With The Guidelines Stroke registry and Michigan Value Collaborative multipayer claims database from Medicare and Blue Cross Blue Shield beneficiaries discharged alive following acute stroke (ICD-10 I61-I63) between 2016-2020. The registry dataset included 64 variables covering demographics, stroke presentation, medical history, procedures, and complications. The hospital dataset included 20 variables from the American Hospital Association’s database. The claims dataset included payer and 79 HCC comorbidity codes. Using combinations of the 3 data sources, we examined the performance of multivariable LASSO logistic regression models to predict all cause readmission at 30-days and 1-year post discharge. We generated hospital-specific testing and training models and reported the mean model discrimination (AUC) of all combinations of the 3 datasets. Results: Of 19,382 linked stroke discharges, 2,724 (14.1%) and 8,169 (42.2%) were readmitted within 30-days and 1-year, respectively. For 30-day readmission, the model based on only registry data produced the best performance (M1, Table). However, for prediction of 1-year readmission, the combination of registry and claims data produced the best performing model (M13, Table). Hospital level characteristics did not have any significant impact on prediction accuracy. Conclusions: Clinical registry data was the best data source for predicting 30-day readmission. However, claims based data were additive when predicting readmission within 1-year, probably because HCC codes add information about total comorbidity burden.

Abstract 10521: Frequency and Impact of Virtual Follow-Up Visits After Heart Failure Hospitalizations: Insights from Michigan

Introduction: Outpatient follow-up after heart failure hospitalization has been shown to reduce readmission and emergency department (ED) visits. Hypothesis: We sought to evaluate the use of virtual follow up-care heart failure hospitalizations in a commercially insured population. Methods: We queried the Michigan Value Collaborative multipayer claims database for heart failure hospitalizations in patients discharged to home or home health between 7/20-11/20. We identified follow-up visits as outpatient office visits within 30 days of discharge, and categorized each patient as having an in-person only, virtual-only (GT or 95 modifier code), both virtual and in-person, and no follow-up visit. Multivariable logistic and linear regression was used to evaluate the association between follow-up visit type and 30-day readmissions, ED visits, and price-standardized episode spending. Results: A total of 2415 patients were discharged alive after a heart failure hospitalization: 94 (3.9%) patients had a virtual-only follow-up visit, 258 (10.7%) had both in-person and virtual visits, 1551 (64.2%) had in-person only follow-up visits, and 512 (21.2%) had no follow-up within 30 days of discharge. Adjusted analyses revealed that patients receiving both virtual and in-person visits had similar rates of 30-day readmissions (aOR=0.73, 95% CI: 0.42, 1.27), ED visits (aOR=0.52, 0.26, 1.07), and episode spending (adjusted difference= -$1508, 95% CI: -$3835, $821; Table ). However, patients receiving only virtual visits or no follow-up had significantly higher 30-day readmissions (aOR=2.31, 95% CI: 1.21, 4.42) and episode spending (adjusted difference= $1974, 95% CI: $206, $3742) compared to in-person visits. Conclusions: Virtual follow-up visits occurred in nearly 15% of heart failure patients, but had mixed results on patient outcomes. More evidence is needed to support the role of virtual-only follow-up visits after heart failure hospitalizations.

Antipsychotic Medication Adherence and Healthcare Services Utilization in Two Cohorts of Patients with Serious Mental Illness.

ObjectiveTo evaluate differences in patient characteristics and real-world outcomes in two distinct high-risk cohorts of patients with serious mental illness (SMI).MethodsRetrospective cross-sectional analysis using a national multi-payer claims database. Two SMI cohorts identified by a technical expert panel—patients recently discharged (RD) from an SMI-related hospitalization and early episode (EE) patients—were evaluated for antipsychotic medication adherence, healthcare utilization, and spending patterns.ResultsThe analysis included 51,705 patients with bipolar disorder, major depressive disorder, and schizophrenia. More than half were over age 46 and >60% were female. Adherence to psychiatric medications was low (52.5% RD and 16.1% EE). More than half of RD and 100% of EE patients switched medications at least once annually, but 19% of RD patients switched ≥2 times compared to 14% of EE. The RD cohort (generally older and sicker) had higher psychiatric related utilization and higher annual costs (US$21,171 versus US$15,398). In both cohorts, women were more likely to have an emergency department (ED) and primary care provider (PCP) visit, but less likely to be hospitalized. Patients age <46 were less likely to have a PCP visit and more likely to have an ED visit, but younger patients age 18–24 were less likely to be hospitalized.ConclusionEfforts to manage SMI are confounded by heterogeneity and low adherence to treatment. By better understanding which patients are at higher risk for specific adverse outcomes, clinicians can target interventions more appropriately to reduce the significant burden of SMI.

Tamsulosin dispensation patterns in the United States: a real-world, longitudinal, population claims database analysis.

Tamsulosin remains the single most popular uroselective alpha adrenoceptor antagonist approved for the treatment of lower urinary tract symptoms (LUTS) attributable to benign prostatic hyperplasia (BPH). Over the last 3 decades, the utilization of tamsulosin has extended to conditions beyond its original indication. To identify potential changes to prescribing patterns and the extent of tamsulosin use for conditions beyond its original indication, we evaluated tamsulosin dispensing patterns in the United States using a large, multi-payer claims database. We conducted a retrospective analysis using IMS PharMetrics Plus™. Patients with a tamsulosin dispensation/BPH diagnosis code (index dates), identified during a 12-month selection period (October 2012-September 2013), were included if continuously enrolled in a health plan during the 18-month analysis period (12 months pre-index-6 months post-index). Patient and provider characteristics were evaluated using descriptive statistics and were contrasted with previously reported data from the literature. Of 133,977 patients dispensed tamsulosin during the analysis period, 72,583 (54.2%) were new users [59,197 (81.6%) men; 13,386 (18.4%) women]. Tamsulosin was newly initiated in men and women mostly by primary care physicians (PCPs; 31.6%) and emergency medicine physicians (21.6%). During the analysis period, 35,071 (59.2%) male new tamsulosin users did not receive a BPH diagnosis code during the analysis period. Of 199,468 men with a BPH diagnosis code, 143,444 (71.9%) were newly diagnosed, mostly [70,412 (49.1%)] by urologists. Few men received hypotension diagnosis: 252 (0.4%) new tamsulosin users within 1 month of starting tamsulosin and 640 (0.4%) within 1 month of a new BPH diagnosis. Tamsulosin was prescribed in patients without a recorded diagnosis of BPH and in women. Physicians were comfortable prescribing tamsulosin in the presence of comorbidity and polypharmacy, and PCPs and emergency medicine physicians were the primary prescribers. These results have important implications for future retrospective research for tamsulosin.

Lessons From Reporting National Performance Measures in a Regional Setting: Washington State Community Cancer Care Report.

Regional public reporting of performance measures in oncology can facilitate local decision making across stakeholders, but small numbers of patients and clinics pose a challenge to creating statistically robust measures. In this article, we describe our development of the Community Cancer Care in Washington State: Quality and Cost Report, the first publicly available report showing clinic-level quality and cost measures at the regional level. We learned key lessons in how to adapt national performance reporting to our regional setting using a registry-linked multipayer claims database. In short, limited numbers of eligible patients for some nationally recognized metrics led us to group metrics and use a 3-year performance window. After completing clinic attribution and other requirements of metric construction, the final metrics included between 62.9% and 88.4% of the eligible patients. To link total costs to some quality measures, we had to define a treatment and surveillance episode of care. Risk adjustment was challenged by the ability to include a limited number of risk adjustors and their potential concentration in a few clinics. We used a different quality score than national performance reporting to account for variation in the range of risk-standardized rates. Current methodology does not permit us to determine whether clinically meaningful differences in quality or costs exist, which inhibits value comparisons. Stakeholder engagement was critical for providing methodologic feedback. In conclusion, we found that refining national metrics was necessary to facilitate public reporting in a regional setting. Further methodologic development can strengthen public reporting and future applications.

Utilization of fetal fibronectin testing and pregnancy outcomes among women with symptoms of preterm labor

ObjectivesTo identify pregnant health plan members triaged through the emergency department (ED), including labor and delivery (ELD) units, with symptoms of preterm labor (PTL), and evaluate the use of fetal fibronectin (fFN) testing; and to calculate the rate of hospitalization and timing of delivery in relation to the ED visit.MethodsRetrospective cohort study using Medical Outcomes Research for Effectiveness and Economics Registry®, a national multipayer claims database. A cohort of pregnant women evaluated in an ELD with a diagnosis of PTL from June 2012 through November 2015 was identified. The proportion of women with PTL who received fFN testing was calculated.ResultsA total of 23,062 patients met the criteria for inclusion in the study. The rate of fFN testing prior to delivery was 12.0%. Of the 23,062 patients included in the analysis, 75.9% were discharged home. Of those who were discharged from the emergency room, one in five went on to deliver within 3 days and almost 96% of this group was not screened for the presence of fFN. Of the remaining 24.1% of patients admitted to the hospital, 91.3% delivered during their stay. In a sensitivity analysis, the percentage of women who delivered within 3 days of the ELD encounter was lower for women who received fFN testing only (6.6%) versus those who had a history of transvaginal ultrasound (TVUS) only (21.6%). Furthermore, the rate of delivery within 3 days was lowest among patients who had both fFN testing and TVUS (4.7%).ConclusionThe utilization of fFN testing is 12%. The majority of pregnant patients triaged through the ELD with symptomatic PTL do not receive an fFN test. As part of PTL evaluation, fFN testing may identify women at increased risk for preterm delivery and help determine appropriate patient management.

High Mortality Rate Found in Youth Newly Diagnosed With Psychosis

Back to table of contents Previous article Next article Clinical and Research NewsFull AccessHigh Mortality Rate Found in Youth Newly Diagnosed With PsychosisJoAnn BlakeJoAnn BlakeSearch for more papers by this authorPublished Online:31 May 2017https://doi.org/10.1176/appi.pn.2017.5a11AbstractPsychiatrists know that young patients with psychosis entering adulthood are particularly vulnerable. A new study shows just how serious a risk they face.Young people experiencing first-episode psychosis (FEP) are 24 times more likely to die within the year of diagnosis than their age-matched peers, reported a study published April 6 in Schizophrenia Bulletin. The findings highlight the importance of intensive clinical intervention at the early stages of psychotic illness and the need for widely available coordinated specialty care programs. NIMH’s Michael Schoenbaum, Ph.D., hopes this study’s findings will alert mental health professionals to improve the follow-up care of young patients with psychosis.“We expected to see the elevated mortality from previous research, but none of that prepared us for the excess mortality on the scale we observed,” said Michael Schoenbaum, Ph.D., lead author and senior advisor for mental health services, epidemiology, and economics at the National Institute of Mental Health (NIMH).The unusually high 12-month death rate in youth with FEP—1,968 per 100,000—should be a “wake-up call,” Schoenbaum told Psychiatric News. In the general population, only people over 70 years old have 12-month mortality approaching the rates observed among patients with psychosis in this study.“Right now, we’re a long way from providing meaningful and effective care. Here is one particular catastrophic result of what happens when patients fall through the cracks,” he said. “We need to invest heavily in establishing and maintaining a close therapeutic relationship with the psychosis patient.”Furthermore, the health care system is failing these patients, said William T. Carpenter, Jr., M.D., a professor of psychiatry and pharmacology at the University of Maryland School of Medicine and editor-in-chief of Schizophrenia Bulletin. “These data substantiate the human cost of a health care system that doesn’t implement evidence-based care for people with psychotic disorders.” Carpenter was not directly involved with the study.Schoenbaum and colleagues used health care insurance claims data from the Multi-Payer Claims Database (MPCD) to assess the 12-month health outcomes in a group of 5,488 insured young adults aged 16 to 30 years who received a diagnosis of FEP. MPCD links to information from the Social Security Administration’s full Death Master File, allowing documentation associated with all-cause mortality associated with FEP.“It’s rare to be able to look at mortality in population health data. We happen to have access to a population-based data set that contains health information linked to information on mortality,” Schoenbaum said. The data, however, did not tell the researchers how these young people died. Given the high death rates, the team expected that patients with FEP would have received more care. However, that’s not what they found. A total of 61 percent of the patients studied did not fill any antipsychotic prescriptions, and 41 percent did not receive psychotherapy in the year following FEP diagnosis. “On the one hand, it’s encouraging that many did receive some psychotherapy, which has been declining in the United States, but it’s surprising that so few received antipsychotic medication,” Schoenbaum said. He added that detailed clinical information would be needed to determine whether the kind of treatment these patients received was appropriate or effective. Overall, 69 percent of the youth with FEP had at least one visit with a mental health specialty provider for medication management, psychotherapy, or both over the 12-month period. These individuals had about one outpatient contact per month. Among those using hospital services, 23 percent were hospitalized in the year after the initial diagnosis. More than half of these individuals had at least one visit to the emergency department in the year following the diagnosis. Although several countries have implemented comprehensive treatment programs for young FEP patients, including lower-dose antipsychotic medication, psychotherapy, family education, and support services, the United States has been slow to adopt early intervention, according to Schoenbaum. “People in these countries that have national health systems are less likely to fall through the cracks because they are treated by the same system throughout their lives,” Schoenbaum said. “Other studies have shown that early coordinated treatment for psychosis produces the best results. However, we know that the typical duration of untreated psychosis in the United States is around 17 months,” Robert Heinssen, Ph.D., director of the Division of Intervention Services at NIMH and co-author on the paper, said in a press release. “This study reinforces federal and state support for funding evidence-based psychosis treatment programs across the country, and the need for communities to invest in more treatment programs.”The study was funded by NIMH. ■An abstract of “Twelve-Month Health Care Use and Mortality in Commercially Insured Young People With Incident Psychosis in the United States” can be accessed here. ISSUES NewArchived

Twelve-Month Health Care Use and Mortality in Commercially Insured Young People With Incident Psychosis in the United States.

To assess 12-month mortality and patterns of outpatient and inpatient treatment among young people experiencing an incident episode of psychosis in the United States. Prospective observational analysis of a population-based cohort of commercially insured individuals aged 16-30 receiving a first observed (index) diagnosis of psychosis in 2008-2009. Data come from the US Department of Health and Human Services' Multi-Payer Claims Database Pilot. Outcomes are all-cause mortality identified via the Social Security Administration's full Death Master File; and inpatient, outpatient, and psychopharmacologic treatment based on health insurance claims data. Outcomes are assessed for the year after the index diagnosis. Twelve-month mortality after the index psychosis diagnosis was 1968 per 100000 under our most conservative assumptions, some 24 times greater than in the general US population aged 16-30; and up to 7372 per 100000, some 89 times the corresponding general population rate. In the year after index, 61% of the cohort filled no antipsychotic prescriptions and 41% received no individual psychotherapy. Nearly two-thirds (62%) of the cohort had at least one hospitalization and/or one emergency department visit during the initial year of care. The hugely elevated mortality observed here underscores that young people experiencing psychosis warrant intensive clinical attention-yet we found low rates of pharmacotherapy and limited use of psychosocial treatment. These patterns reinforce the importance of providing coordinated, proactive treatment for young people with psychosis in US community settings.

Inappropriate Utilization in Fee-for-Service Medicare and Medicare Advantage Plans.

This study uses a national multi-payer claims database to test for differences in potentially inappropriate emergency department (ED) visits and ambulatory care sensitive (ACS) admissions in fee-for-service (FFS) Medicare and Medicare Advantage (MA) plans. Rates of ACS admissions for MA enrollees were approximately one third those of FFS beneficiaries, controlling for covariates, which included the beneficiary's health status as represented by their risk score. This study then compared FFS and MA beneficiaries when they moved from one type of health plan to another. Again, controlling for covariates, potentially inappropriate ED visits and ACS admissions remained at their low baseline values for FFS beneficiaries who switched from FFS Medicare to MA plans, but rose for MA enrollees switching to FFS Medicare.

Allopurinol use and the risk of acute cardiovascular events in patients with gout and diabetes

BackgroundFew studies, if any, have examined cardiovascular outcomes in patients with diabetes and gout. Both diabetes and gout are risk factors for cardiovascular disease. The objective of this study was to examine the effect of allopurinol on the risk of incident acute cardiovascular events in patients with gout and diabetes.MethodsWe used the 2007–2010 Multi-Payer Claims Database (MPCD) that linked health plan data from national commercial and governmental insurances, representing beneficiaries with United Healthcare, Medicare, or Medicaid coverage. In patients with gout and diabetes, we assessed the current allopurinol use, defined as a new filled prescription for allopurinol, as the main predictor of interest. Our outcome of interest was the occurrence of the first Incident hospitalized myocardial infarction (MI) or stroke (composite acute cardiovascular event), after which observations were censored. We employed multivariable-adjusted Cox proportional hazards models that simultaneously adjusted for patient demographics, cardiovascular risk factors and other medical comorbidities. We calculated hazard ratios [HR] (95% confidence intervals [CI]) for incident composite (MI or stroke) acute cardiovascular events. We performed sensitivity analyses that additionally adjusted for the presence of immune diseases and colchicine use, as potential confounders.ResultsThere were 2,053,185 person days (5621.3 person years) of current allopurinol use and 1,671,583 person days (4576.5 person years) of prior allopurinol use. There were 158 incident MIs or strokes in current and 151 in prior allopurinol users, respectively. Compared to previous allopurinol users, current allopurinol users had significantly lower adjusted hazard of incident acute cardiovascular events (incident stroke or MI), with an HR of 0.67 (95% CI, 0.53, 0.84). Sensitivity analyses, additionally adjusted for immune diseases or colchicine use, confirmed this association.ConclusionsCurrent allopurinol use protected against the occurrence of acute cardiovascular events in patients with gout and diabetes. The underlying mechanisms for this potential cardio-protective effect of allopurinol need further exploration.

FRI0193 Herpes Zoster Infection across Auto-Immune and Inflammatory Diseases: Implications for Vaccination

BackgroundMore than one million herpes zoster (HZ) cases occur in the United States every year. Currently, the live zoster vaccine is recommended for healthy older patients age ≥60 years. Whether...

Direct Medical Costs for Complications Among Children and Adults with Diabetes in the US Commercial Payer Setting

Complications associated with diabetes are a major contributor to the burden of the disease. To better inform decision modelling, there is a need for cost estimates of specific diabetes-related complications, stratified by diabetes type and patient age group. To obtain direct medical costs of managing and treating diabetes-related complications over a 2-year period, for adults and children with type1 diabetes mellitus (T1DM) or type2 diabetes mellitus (T2DM), using data from a large commercially insured US subscriber database. We examined records from a large US multi-payer claims database to identify patients with any diabetes-related complications included in nine pre-specified categories, filed between January 2009 and September 2010, and with pre-index evidence of T1DM or T2DM. Patients were required to have continuous health plan enrolment 12months before and 24months after each index complication. Patients were classified into cohorts based on their diabetes type and age status at the time of the complication. The direct medical cost associated with each complication was calculated for the 12- and 24-month follow-up periods. Mean paid and allowed total costs were calculated and inflation-adjusted to the year 2011. Of the 119,715 patients who met the inclusion criteria, 211 (0.2%) were categorized as children with T1DM, 55 (0.05%) as children with T2DM, 6,227 (5.2%) as adults with T1DM and 113,222 (94.6%) as adults with T2DM. The respective mean cohort ages were 13.5, 14.9, 48.5 and 58years. Proteinuria/albuminuria was the most common complication for T1DM and T2DM child cohorts, with this complication occurring in almost one third of these children. Among the child cohorts, renal disease accounted for the highest mean paid cost for T1DM patients (US$6,053) whereas for T2DM patients, the complication associated with the highest mean paid cost was lactic acidosis (US$25,053). For the adult T1DM cohort, the complications with the highest occurrence and highest mean total paid cost were non-proliferative retinopathy (40.3%) and renal disease (US$28,076), respectively. Similarly, for the adult T2DM cohort, these complications were neuropathy (26.8%) and peritoneal dialysis (US$32,826). With the continuing and increasing interest in child and adult T1DM and T2DM, stakeholders will need relevant and timely information to guide treatment decision making. This cost research may directly inform the economic models that are often developed to better identify, understand and manage key economic considerations that drive the costs of this chronic disease.

Do decision-analytic models identify cost-effective treatments? A retrospective look at helicobacter pylori eradication.

Pharmacoeconomic models of Helicobacter (H) pylori eradication have been frequently cited but never validated. Examine retrospectively whether H pylori pharmacoeconomic models direct decision makers to cost-effective therapeutic choices. We first replicated and then validated 2 models, replacing model assumptions with empirical data from a multipayer claims database. Database subjects were 435 commercially insured U.S. patients treated with bismuthmetronidazole- tetracycline (BMT), proton pump inhibitor (PPI)-clarithromycin, or PPI-amoxicillin. Patients met >1 clinical requirement (ulcer disease, gastritis/duodenitis, stomach function disorder, abdominal pain, H pylori infection, endoscopy, or H pylori assay). Sensitivity analyses included only patients with ulcer diagnosis or gastrointestinal specialist care. Outcome measures were: (1) rates of eradication retreatment; (2) use of office visits, hospitalizations, endoscopies, and antisecretory medication; and (3) cost per effectively treated (nonretreated) patient. Model results overstated the cost-effectiveness of PPI-clarithromycin and underestimated the cost-effectiveness of BMT. Prior to empirical adjustment, costs per effectively treated patient were 1,001 US dollars, 980 US dollars, and 1,730 US dollars for BMT, PPIclarithromycin, and PPI-amoxicillin, respectively. Estimates after adjustment were US dollars for BMT, 1,118 US dollars for PPI-clarithromycin, and 1,131 US dollars for PPI-amoxicillin. Key model assumptions that proved retrospectively incorrect were largely unsupported by either empirical evidence or systematic assessment of expert opinion. Organizations with access to medical and pharmacy claims databases should test key assumptions of influential models to determine their validity. Journal peer-review processes should pay particular attention to the basis of model assumptions.