Procedure for Organizing a Post-FDA-approval Evaluation of Antidepressants

Purpose: The study reports the construction of a cohort used to study the effectiveness of antidepressants.

Methods: The cohort includes experiences of 3,678,082 patients with depression in the United States on antidepressants between January 1, 2001, and December 31, 2018. A total of 10,221,145 antidepressant treatment episodes were analyzed. Patients who had no utilization of health services for at least two years, or who had died, were excluded from the analysis. Follow-up was passive, automatic, and collated from fragmented clinical services of diverse providers.

Results: The average follow-up was 2.93 years, resulting in 15,096,055 person-years of data. The mean age of the cohort was 46.54 years (standard deviation of 17.48) at first prescription of antidepressant, which was also the enrollment event (16.92% were over 65 years), and most were female (69.36%). In 10,221,145 episodes, within the first 100 days of start of the episode, 4,729,372 (46.3%) continued their treatment, 1,306,338 (12.8%) switched to another medication, 3,586,156 (35.1%) discontinued their medication, and 599,279 (5.9%) augmented their treatment.

Conclusions: We present a procedure for constructing a cohort using claims data. A surrogate measure for self-reported symptom remission based on the patterns of use of antidepressants has been proposed to address the absence of outcomes in claims. Future studies can use the procedures described here to organize studies of the comparative effectiveness of antidepressants.

Effectiveness of common antidepressants: a post market release study

BACKGROUND

This study summarizes the experiences of patients, who have multiple comorbidities, with 15 mono-treated antidepressants.

METHODS

This is a retrospective, observational, matched case control study. The cohort was organized using claims data available through OptumLabs for depressed patients treated with antidepressants between January 1, 2001 and December 31, 2018. The cohort included patients from all states within United States of America. The analysis focused on 3,678,082 patients with major depression who had 10,221,145 antidepressant treatments. Using the robust, and large predictors of remission, and propensity to prescribe an antidepressant, the study created 16,770 subgroups of patients. The study reports the remission rate for the antidepressants within the subgroups. The overall impact of antidepressant on remission was calculated as the common odds ratio across the strata.

FINDINGS

The study accurately modelled clinicians' prescription patterns (cross-validated Area under the Receiver Operating Curve, AROC, of 82.0%, varied from 77% to 90%) and patients' remission (cross-validated AROC of 72.0%, varied from 69.5% to 78%). In different strata, contrary to published randomized studies, remission rates differed significantly and antidepressants were not equally effective. For example, in age and gender subgroups, the best antidepressant had an average remission rate of 50.78%, 1.5 times higher than the average antidepressant (30.30% remission rate) and 20 times higher than the worst antidepressant. The Breslow-Day chi-square test for homogeneity showed that across strata a homogenous common odds-ratio did not exist (alpha<0.0001). Therefore, the choice of the optimal antidepressant depended on the strata defined by the patient's medical history.

INTERPRETATION

Study findings may not be appropriate for specific patients. To help clinicians assess the transferability of study findings to specific patient, the web site http://hi.gmu.edu/ad assesses the patient's medical history, finds similar cases in our data, and recommends an antidepressant based on the experience of remission in our data. Patients can share this site's recommendations with their clinicians, who can then assess the appropriateness of the recommendations.

FUNDING

This project was funded by the Robert Wood Johnson foundation grant #76786. The development of related web site was supported by grant 247-02-20 from Virginia's Commonwealth Health Research Board.

Expenditures among young adults with acute lymphoblastic leukemia by site of care

BACKGROUND

Individuals diagnosed with acute lymphoblastic leukemia (ALL) between the ages of 22 and 39 years experience worse outcomes than those diagnosed when they are 21 years old or younger. Treatment at National Cancer Institute-designated Comprehensive Cancer Centers (CCC) mitigates these disparities but may be associated with higher expenditures.

METHODS

Using deidentified administrative claims data (OptumLabs Data Warehouse), the cancer-related expenditures were examined among patients with ALL diagnosed between 2001 and 2014. Multivariable generalized linear model with log-link modeled average monthly health-plan-paid (HPP) expenditures and amount owed by the patient (out-of-pocket [OOP]). Cost ratios were used to calculate excess expenditures (CCC vs non-CCC). Incidence rate ratios (IRRs) compared CCC and non-CCC monthly visit rates. Models adjusted for sociodemographics, comorbidities, adverse events, and months enrolled.

RESULTS

Clinical and sociodemographic characteristics were comparable between CCC (n = 160) and non-CCC (n = 139) patients. Higher monthly outpatient expenditures in CCC patients ($15,792 vs $6404; P < .001) were driven by outpatient hospital HPP expenditures. Monthly visit rates and per visit expenditures for nonchemotherapy visits (IRR = 1.6; P = .001; CCC = $8247, non-CCC = $1191) drove higher outpatient hospital expenditures among CCCs. Monthly OOP expenditures were higher at CCCs for outpatient care (P = .02). Inpatient HPP expenditures were significantly higher at CCCs ($25,918 vs $13,881; ꞵ = 0.9; P < .001) before accounting for adverse events but were no longer significant after adjusting for adverse events (ꞵ = 0.4; P = .1). Hospitalizations and length of stay were comparable.

CONCLUSIONS

Young adults with ALL at CCCs have higher expenditures, likely reflecting differences in facility structure, billing practices, and comprehensive patient care. It would be reasonable to consider CCCs comparable to the oncology care model and incentivize the framework to achieve superior outcomes and long-term cost savings.

LAY SUMMARY

Health care expenditures in young adults (aged 22-39 years) with acute lymphoblastic leukemia (ALL) are higher among patients at National Cancer Institute-designated Comprehensive Cancer Centers (CCC) than those at non-CCCs. The CCC/non-CCC differences are significant among outpatient expenditures, which are driven by higher rates of outpatient hospital visits and outpatient hospital expenditures per visit at CCCs. Higher expenditures and visit rates of outpatient hospital visits among CCCs may also reflect how facility structure and billing patterns influence spending or comprehensive care. Young adults at CCCs face higher inpatient HPP expenditures; these are driven by serious adverse events.

Adverse events among chronic myelogenous leukemia patients treated with tyrosine kinase inhibitors: a real-world analysis of health plan enrollees

With tyrosine kinase inhibitor (TKI) therapy, chronic myelogenous leukemia (CML) is now a chronic disease. CML patients treated with TKIs (n = 1200) were identified from the OptumLabs^® Data Warehouse (de-identified claims and electronic health records) between 2000 and 2016 and compared with a non-cancer cohort (n = 7635). The 5-year cumulative incidence of all organ system outcomes was significantly greater for the TKI versus non-cancer group. In the first year, compared with imatinib, later generation TKIs were associated with primary infections (hazard ratios [HR] 1.43, 95% CI 1.02-2.00), circulatory events (HR 1.15, 95% CI 1.01-1.31), and skin issues (HR 1.43, 95% CI 1.13-1.80); musculoskeletal and nervous system/sensory issues were less common (HRs 0.83-0.84, p < 0.05). Increased risk of infections, cardiopulmonary and skin issues associated with later generation TKIs persisted in subsequent years. In this real-world population, TKI therapy was associated with a high burden of adverse events. Later generation TKIs may have greater toxicity than imatinib.

Health Care Cost Associated With Contemporary Chronic Myelogenous Leukemia Therapy Compared With That of Other Hematologic Malignancies

PURPOSE

Given the widespread introduction of tyrosine kinase inhibitors (TKIs), we evaluated the cost associated with chronic myelogenous leukemia (CML) care compared with the cost of care for patients with hematologic malignancies (HEM) and for patients without cancer (GEN), to aid with resource allocation and clinical decision making.

METHODS

A retrospective cohort was constructed from the OptumLabs Data Warehouse using claims from 2000 to 2016. Eligible patients had ≥ 2 CML claims and were enrolled continuously for ≥ 6 months before diagnosis and ≥ 1 year afterward (n = 1,909). Patients with CML were frequency matched 4:1 with HEM and GEN cohorts and were observed through October 2017. We used generalized linear models to assess the variation in total mean annualized health care costs in the 3 cohorts and to examine the influence of factors associated with costs.

RESULTS

Mean annualized costs for CML were $82,054 (ie, $25,471 [95% CI, $20,808 to $30,133] more than those for HEM and $74,993 [95% CI, $70,818 to $79,167] more than those for GEN); these differences were driven by pharmacy costs in the CML group. The cost of CML care exceeded that for HEM and GEN for all index years in this study and increased over each diagnostic interval until 2015, peaking at $91,990. The mean annual cost of all TKIs increased. Imatinib's mean annualized cost was $41,546 in the period 2000-2004 but increased to $105,069 in the period 2015-2017. In multivariable analysis, percent days on TKIs had the greatest influence on cost: ≥ 75% of the time versus none showed a difference in cost of $108,716 (95% CI, $99,193 to $118,239).

CONCLUSION

Contemporary CML costs exceeded the cost of treatment of other hematologic malignancies. Cost was primarily driven by TKIs, whose cost continued to increase over time.

A cross-sectional survey of work and income loss consideration among patients with herpes zoster when completing a quality of life questionnaire

BACKGROUND

Prior research suggests that many patients do not spontaneously include work/income loss when responding to utility assessments, although this remains unconfirmed in the US due to almost no published US-based studies to date, and has not been previously studied among patients with herpes zoster (HZ). The objective of this study was to examine whether patients with HZ consider work and income loss when completing a quality of life survey.

METHODS

A cross-sectional survey was administered to 2000 US adult commercial health plan enrollees aged 50-64 years with ≥ 1 HZ medical claim during 2014. The survey collected information related to health status (EQ-5D), work productivity, and HZ severity and clinical features.

RESULTS

Mean respondent age was 58.4 years [standard deviation (SD) 4.1] and 62.0% were female. About 3 in 4 (76.8%) patients (N = 772) were employed either full (69.9%) or part time (6.9%). Less than half (45%) spontaneously considered work/income loss when responding to EQ-5D, and mean EQ-5D scores for patients who considered work/income loss were lower than for patients who did not [0.56 (SD = 0.28) vs. 0.69 (SD = 0.24); p < 0.001]. Overall, 43% of patients reported at least one full day missed (mean = 9 full days) and 29% reported at least one partial day missed (mean = 6 partial days) during the most recent shingles episode. Patients who considered work loss were more likely to have missed full (76.4% vs 26.0%, p < 0.001) or partial (70.9% vs. 35.2%, p < 0.001) days. Patients with absenteeism were more likely to consider work/income loss when completing EQ-5D [odds ratio (OR) = 7.91, 95% confidence interval (CI) 5.01-12.31]. Odds of absenteeism/presenteeism increased significantly with increasing levels of HZ severity, and higher odds were associated with pain located on the face/scalp/neck/eye/ear (OR 1.90, 95% CI 1.06-3.40) and with pain lasting 12+ months (OR = 2.91, 95% CI 1.14-7.42).

CONCLUSIONS

HZ has considerable impact on the work and productivity of adults aged 50-64 years old. However, many patients with HZ do not spontaneously consider work/income loss when completing a standardized quality of life questionnaire. Studies that use health state utilities in HZ based on EQ-5D may not fully reflect the societal costs of work loss.

Swimming exercise demonstrates advantages over running exercise in reducing proteinuria and glomerulosclerosis in spontaneously hypertensive rats

Experimental studies in animal models have described the benefits of physical exercise (PE) to kidney diseases associated with hypertension. Land- and water-based exercises induce different responses in renal function. Our aim was to evaluate the renal alterations induced by different environments of PE in spontaneously hypertensive rats (SHRs). The SHRs were divided into sedentary (S), swimming exercise (SE), and running exercise (RE) groups, and were trained for 8 weeks under similar intensities (60 min/day). Arterial pressure (AP) and heart rate (HR) were recorded. The renal function was evaluated through urinary volume at each week of training; sodium and potassium excretions, plasma and urinary osmolarities, glomerular filtration rate (GFR), levels of proteinuria, and renal damage were determined. SE and RE rats presented reduced mean AP, systolic blood pressure, and HR in comparison with S group. SE and RE rats showed higher urine osmolarity compared with S. SE rats showed higher free water clearance (P < 0.01), lower urinary density (P < 0.0001), and increased weekly urine volume (P < 0.05) in comparison with RE and S groups. GFR was increased in both SE and RE rats. The proteinuria of SE (7.0 ± 0.8 mg/24 h) rats was decreased at the 8th week of the PE in comparison with RE (9.6 ± 0.8 mg/24 h) and S (9.8 ± 0.5 mg/24 h) groups. The glomerulosclerosis was reduced in SE rats (P < 0.02). SE produced different response in renal function in comparison with RE, in which only swimming-trained rats had better profile for proteinuria and glomerulosclerosis.

Factors associated with second-line triplet therapy in routine care in relapsed/refractory multiple myeloma

WHAT IS KNOWN AND OBJECTIVE

Second-line therapy (SLT) trials in relapsed/refractory multiple myeloma (RRMM) report superior outcomes with triplet combinations. We sought to determine factors associated with triplet SLT in routine practice.

METHODS

A retrospective cohort with claims for MM between 01/01/2008 and 03/31/2015 was grouped by 1-2 ("doublet") or 3+ ("triplet") agent therapy. Charlson comorbidity index (CCI) and disability status; CRAB symptoms (hypercalcaemia, renal/bone disease, anaemia); and relapse risk were determined.

RESULTS

Among 623 patients, the triplet group (n=146 [23%]) was younger (65.2 vs 69.8 years) and more likely to have high-risk relapse (67% vs 50%), CRAB symptoms (94.5% vs 81.1%), triplet first-line treatment (75% vs 51%) and frontline stem cell transplant (38% vs 20%) (P<0.001 for all). In multivariate analyses, CRAB symptoms (OR: 3.22, 95% CI: 1.47, 7.10), high-risk relapse (OR: 1.71, 95% CI: 1.12, 2.62) and prior triplet therapy (OR: 2.16, 95% CI: 1.38, 3.40), but neither CCI nor disability, were associated with triplet SLT. A trend towards triplets among younger patients (<65 vs >75 years, OR: 1.73, 95% CI: 0.99, 3.04) was observed.

WHAT IS NEW AND CONCLUSION

The majority of patients did not receive triplet regimens. Treatment selection with triplet therapy for RRMM should carefully consider comorbidities and patient-specific characteristics.

Outcomes of tumor necrosis factor inhibitor cycling versus switching to a disease-modifying anti-rheumatic drug with a new mechanism of action among patients with rheumatoid arthritis

OBJECTIVES

To examine treatment patterns, treatment effectiveness, and treatment costs for 1 year after patients with rheumatoid arthritis switched from a tumor necrosis factor inhibitor (TNFi) (adalimumab, certolizumab pegol, etanercept, golimumab, or infliximab), either cycling to another TNFi ("TNFi cyclers") or switching to a new mechanism of action (abatacept, tocilizumab, or tofacitinib) ("new MOA switchers").

METHODS

This retrospective cohort study used administrative claims data for a national insurer. Treatment persistence (without switching again, restarting, or discontinuing), treatment effectiveness (defined below), and costs were assessed for the 12-month post-switch period. Patients were "effectively treated" if they satisfied all six criteria for a treatment effectiveness algorithm (high adherence, no dose increase, no new conventional synthetic disease-modifying anti-rheumatic drug, no subsequent switch in therapy, no new/increased oral glucocorticoids, and <2 glucocorticoid injections). Multivariable logistic models were used to adjust for baseline factors.

RESULTS

The database included 581 new MOA switchers and 935 TNFi cyclers. New MOA switchers were 39% more likely than TNFi cyclers to persist after the switch (odds ratio [OR] = 1.39; 95% confidence interval [CI] = 1.12-1.74; p = .003) and 36% less likely to switch therapy again (OR = 0.64; 95% CI = 0.51-0.81; p < .001). New MOA switchers were 43% more likely than TNFi cyclers to be effectively treated (OR = 1.43; 95% CI = 1.11-1.85; p = .006). New MOA switchers had 16% lower drug costs than TNFi cyclers (cost ratio = 0.84; 95% CI = 0.79-0.88; p < .001) and 11% lower total costs of rheumatoid arthritis-related medical care (cost ratio = 0.89; 95% CI = 0.84-0.94; p < .001).

LIMITATIONS

Claims payments may not reflect rebates or other cost offsets. Medical and pharmacy claims do not include clinical end-points or reasons that lead to new MOA switching vs TNFi cycling.

CONCLUSIONS

These results support switching to a new MOA after a patient fails treatment with a TNFi, which is consistent with recent guidelines for the pharmacologic management of established rheumatoid arthritis.

Abstract P5-08-15: Real-world treatment patterns and survival among triple negative breast cancer patients versus patients with other breast cancer subtypes in early stage breast cancer

Background: It is generally accepted that outcomes of breast cancer (BC) patients (pts) differ by tumor characteristics, but relatively few studies have examined treatment patterns and outcomes in early stage disease by hormone receptor (HR) and HER2 status.

Methods: This retrospective study used physician-reported clinical information (date and stage at diagnosis, HER2 and HR status) for commercially insured BC pts from the Optum Oncology Management Registry linked with the Optum Research Database (medical and pharmacy claims from a national US health plan) and death data from Social Security Administration files from 01/2008 to 6/2014. Pts diagnosed with early stage breast cancer (ESBC) were examined. The index date was diagnosis of BC. Adult pts (≥18 yrs old) had to have known HR and HER2 status, continuous enrollment in the health plan from initial diagnosis for ≥6 months, surgery and systemic cancer therapy. Neo-adjuvant and adjuvant therapy (initial phase of care) based on timing of claims for NCCN-recommended therapy were examined, as well as survival and development of metastasis (based on claims with ICD-9 codes 196.0x, 196.2x, 196.5x-199.0x) after initial phase of care during the study period. Differences across the HR/HER2 subtypes were examined using t-test for continuous variables and chi-square test for categorical variables.

Results: Among 6,881 ESBC pts identified, 15% (n=1038) were TNBC pts, 65% HR+/HER2-, 14% HR+/HER2+ and 6% HR-/HER2+. Mean (median) age was 52 (53) years. Treatment patterns and outcomes vary by tumor subtype (Table).

 HR-/HER2-HR+/HER2-HR-/HER2+HR+/HER2+ N=1038N=4452N=401N=990Length of follow-up, mths, mean(SD)*, median27 (16), 2328 (16), 2529 (16), 2527 (15), 24Length of initial phase of care – mths, mean(SD)*, media10 (7), 922 (15), 1915 (9), 1423 (14), 19Pts with systemic neo-adjuvant therapy N (%)*303 (29)531 (12)140 (35)216 (22)Most common neo-adjuvant regimens (%)Paclitaxel, cyclophosphamide, doxorubicin (45%)Paclitaxel, cyclophosphamide, doxorubicin (31%)Docetaxel, carboplatin, trastuzumab (44%)Docetaxel, carboplatin, trastuzumab (31%)Pts with systemic adjuvant therapy N (%)*740 (71)4065 (91)366 (91)935 (94)Most common adjuvant regimens (%)Paclitaxel, cyclophosphamide, doxorubicin (30%)Tamoxifen (19%)Docetaxel, carboplatin, trastuzumab (30%)Docetaxel, carboplatin, trastuzumab, tamoxifen (9%)Evidence of distant metastasis N (%)*78 (7.5)76 (1.7)17 (4.2)15 (1.5)Mortality N (%)*23(2.2)23 (0.5)1 (0.3)3 (0.3)*p&lt;0.05

Conclusions: The most common neo-adjuvant and adjuvant chemotherapy agents received were similar across HER2 subtypes and regimens for HR+ and HER2+ patients also included hormonal or HER2 targeted agents. TNBC pts had the shortest length of initial phase of care (mean 10 mths), though this cohort had the second highest percentage of patients receiving systemic neo-adjuvant therapy. Development of distant metastasis and mortality during the study period were highest for the HR-/HER2- patients, at 7.5% and 2.2% respectively. Development of novel targets and therapies for TNBC may significantly benefit patients with this BC subtype, in both neo-adjuvant and adjuvant setting to improve outcomes.

Citation Format: DaCosta Byfield S, Abushamaa AM, Becker LK, Shepherd SP, Ricker JL, Bonnet P. Real-world treatment patterns and survival among triple negative breast cancer patients versus patients with other breast cancer subtypes in early stage breast cancer [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P5-08-15.

Costs and mortality of recurrent versus de novo hormone receptor-positive/HER2(-) metastatic breast cancer

AIM

To examine cost and mortality differences in postmenopausal women with HR(+)/HER2(-) advanced breast cancer.

METHODS

Using claims data (2007-2013), women with newly diagnosed (de novo) stage IV, or early- or late-recurring metastatic breast cancer were identified.

RESULTS

Compared with de novo (n = 121) and late-recurrent (n = 106), early-recurrent (n = 172) patients had significantly higher costs in total and for anticancer systemic agents. Adjusted per patient per month costs for early-recurrent patients were US$13,404, versus US$9955 (de novo) and US$9721 (late-recurrent; p = 0.02). Early-recurrent patients' risk of death was twice that of de novo patients (p = 0.02).

CONCLUSION

Compared with new diagnosis or late recurrence, early recurrence of HR+/HER2- metastatic breast cancer was associated with higher mortality and healthcare costs.

Application of the new GOLD COPD staging system to a US primary care cohort, with comparison to physician and patient impressions of severity

Background

In 2011, the traditional Global Initiative for Chronic Obstructive Lung Disease (GOLD) COPD spirometry-based severity classification system was revised to also include exacerbation history and COPD Assessment Test (CAT) and modified Medical Research Council Dyspnea Scale (mMRC) scores. This study examined how COPD patients treated in primary care are reclassified by the new GOLD system compared to the traditional system, and each system’s level of agreement with patient’s or physician’s severity assessments.

Methods

In this US multicenter cross-sectional study, COPD patients were recruited by 83 primary care practitioners (PCPs) to complete spirometry testing and a survey. Patients were classified by the traditional spirometry-based system (stages 1–4) and under the new system (grades A, B, C, D) using spirometry, exacerbation history, mMRC, and/or CAT results. Concordance between physician and patient-reported severity, spirometry stage, and ABCD grade based on either mMRC or CAT scores was examined.

Results

Data from 445 patients with spirometry-confirmed COPD were used. As compared to the traditional system, the GOLD mMRC system reclassifies 47% of patients, and GOLD CAT system reclassifies 41%, but the distributions are very different. The GOLD mMRC system resulted in relatively equal distributions by ABCD grade (33%, 22%, 19%, 26%, respectively), but the GOLD CAT system put most into either B or D groups (9%, 45%, 4%, and 42%). The addition of exacerbation history reclassified only 19 additional patients. Agreement between PCPs’ severity rating or their patients’ self-assessment and the new ABCD grade was very poor (κ=0.17 or less).

Conclusion

As compared to the traditional system, the GOLD 2011 multidimensional system reclassified nearly half of patients, but how they were reclassified varied greatly by whether the mMRC or CAT questionnaire was chosen. Either way, the new system had little correlation with the PCPs or their patients’ impressions about the COPD severity.

Clostridium difficile-Associated Disease in Patients in a Small Rural Hospital

Objective.

To determine the risk factors for Clostridium difficile–associated disease (CDAD) in a 25-bed rural hospital and to compare antimicrobial use ratios at the study hospital with those at a large academic medical center.

Design.

Case-control study.

Setting.

A 25-bed rural hospital in Iowa during the period from August 2002 through January 2005.

Patients.

A total of 17 case patients with CDAD and 34 control patients matched for age (ie, within 10 years of the case patient's age), sex, and admission date (ie, within 2 weeks of the case patient's admission date).

Methods.

Retrospective medical record review was performed to obtain data on antimicrobial exposures during the 6 weeks before hospital admission for both case and control patients. Exact conditional logistic regression was used for univariable and multivariable analyses. Antimicrobial use ratios were calculated to compare the rates of antimicrobial use for case and control patients at the study hospital with the rates for patients evaluated in a study of CDAD at a nearly 700-bed teaching hospital.

Results.

Case patients had a larger cumulative number of days of antimicrobial use (P = .004), and they received a larger total number of antimicrobial agents during hospitalization (P = .001). Antimicrobial use ratios were higher for both case and control patients at the smaller hospital, compared with the larger hospital.

Conclusions.

CDAD at a small rural hospital was not associated with exposure to the antimicrobial classes that are typically associated with CDAD, but was instead related to the total number of antimicrobials used to treat patients. The rate of antimicrobial use for case and control patients was about 40% higher at the small rural hospital, compared with the corresponding rates at a large academic medical center.

Comparative effectiveness of colony-stimulating factors in febrile neutropenia prophylaxis: how results are affected by research design

Aims: To examine the impact of research design on results in two published comparative effectiveness studies. Methods: Guidelines for comparative effectiveness research have recommended incorporating disease process in study design. Based on the recommendations, we develop a checklist of considerations and apply the checklist in review of two published studies on comparative effectiveness of colony-stimulating factors. Both studies used similar administrative claims data, but different methods, which resulted in directionally different estimates. Results: Major design differences between the two studies include: whether the timing of intervention in disease process was identified and whether study cohort and outcome assessment period were defined based on this temporal relationship. Conclusion: Disease process and timing of intervention should be incorporated into the design of comparative effectiveness studies.

Service setting impact on costs for bevacizumab-treated oncology patients

OBJECTIVES

To investigate treatment patterns and healthcare costs of patients with metastatic colorectal cancer (mCRC) or lung cancer (LC) who were treated with bevacizumab in a physician office (OFF) setting versus a hospital outpatient (HOP) setting.

STUDY DESIGN

Retrospective analysis of claims from a national US health plan.

METHODS

mCRC and LC patients initiating treatment with bevacizumab (index date) between January 1, 2006, and July 31, 2012, were identified. Patients were aged ≥18 years with ≥6-month pre- (baseline) and ≥6-month post index (follow-up) data, retaining patients who died with <6 months of follow-up. Differences by site of service were analyzed by χ2 and t test (bevacizumab administrations, dose) and general linear model adjusted for demographic and clinical characteristics (all-cause healthcare costs).

RESULTS

A total of 1687 mCRC (OFF: 1292; HOP: 395) and 1232 LC patients (OFF: 983; HOP: 249) were identified. Mean age was 61.3 years, 56.3% were male, and 78% were treated in OFF. Treatment in OFF declined from 2006 (84% of patients) to 2012 (61%). For OFF versus HOP, mean length of treatment (208.3 vs 191.0 days; P=.007), number of bevacizumab administrations per month (1.4 vs 1.1; P<.001), and mean weekly dose (eg, for 2012, 4.34 vs 3.11 mg/kg, P<.05) were higher in OFF. Adjusted monthly HOP costs (vs OFF) were higher by 37.8% for mCRC patients (cost ratio=1.378; 95% CI, 1.282-1.482) and 31.1% for LC patients (cost ratio=1.311; 95% CI, 1.204-1.427) CONCLUSIONS: Despite fewer administrations and lower weekly dose of bevacizumab in HOP, adjusted total costs were 31% to 38% higher for mCRC and LC patients treated in the HOP setting.

Abstract 305: Post-discharge Costs in Heart Failure: Kidney Disease Severity as a Key Contributor

Background: This study examined post-discharge costs among individuals hospitalized with heart failure (HF) by stage of chronic kidney disease (CKD).

Methods: A retrospective claims-based study of medical and pharmacy data linked to laboratory results from a large US health plan (Jan2007-Sep2012) was conducted. Earliest inpatient stay with HF diagnosis was defined as index hospitalization. Post-index costs were calculated per-member-per-month (PMPM) for the subset of individuals with a post-discharge observation up to 6 months. Individuals were assigned to CKD stages 1-5 using the CKD-Epidemiology Collaboration Group equation. One-way ANOVA was used to examine differences in follow-up health care costs across CKD stages.

Results: A total of 24,023 individuals hospitalized with HF (mean age 70 y, 51% female, 71% white, 65% Medicare Advantage) were included. Subjects with at least moderately reduced renal function (eGFR <60 mL/min, stages 3-5) when compared to individuals with normal/mild reduction (eGFR ≥60 mL/min, stages 1-2) experienced higher follow-up ambulatory ($1,508 vs. $1,236) and inpatient ($3,725 vs. $3,152) all-cause costs. HF-related medical costs accounted for approximately 40% of all-cause medical cost for CKD stages 2-4 (table).

Conclusion: Findings from this study suggest that medical costs following HF hospitalization increase by severity of CKD stage, with HF-related costs comprising a large portion of inpatient and outpatient costs.

Relationship between physician specialty and allopurinol prescribing patterns: a study of patients with gout in managed care settings

OBJECTIVE

Allopurinol is used to lower serum uric acid (sUA) levels in gout patients. The objective of this study was to investigate the influence of physician specialty on allopurinol treatment patterns and sUA control.

DESIGN AND METHODS

This was a retrospective study using claims from a managed care database of US health plan enrollees. Gout patients at least 18 years of age who received allopurinol were identified from the database between January 1, 2002 and April 30, 2007. The index date was defined as the date of the earliest allopurinol claim, and patients were required to have health plan enrollment for at least 365 days prior to and following the index date for inclusion. Physician specialty was determined using the index allopurinol claim. Dosage of allopurinol prescription(s) and number of gout flares were determined from claims data. sUA measurements were used to assess goal attainment over a period of at least one year following the index allopurinol prescription.

RESULTS

There were 3363 patients with gout of whom 69.9% received an index allopurinol prescription from a generalist/internist, 5.7% from a rheumatologist, 2.6% from a nephrologist, and 21.8% from a physician with other specialty. Of patients receiving their index prescription from a nephrologist, 38.7% reached the sUA goal of <6 mg/dL (357 μmol/L), as compared to patients prescribed by a rheumatologist, generalist/internist, or other physician (35.4%, 31.4%, and 39.4%, respectively; P = 0.015). When controlling for patient characteristics, multivariate analysis did not reveal statistically significant different odds of sUA goal attainment based on prescribing physician specialty, though separate analyses indicated that patients prescribed by a nephrologist had fewer gout flares. Change in allopurinol dosage from initial to final dose was more frequent among patients prescribed by rheumatologists and nephrologists.

CONCLUSION

There is significant heterogeneity in the specialists' management of sUA levels in patients with gout, possibly reflecting differences in case mix and treatment approaches. Limitations related to the use of claims data, such as inability to observe medications filled over-the-counter, should be considered when interpreting study results.