Zachary Burningham

The Epidemiology of Sarcoma

Sarcomas account for over 20% of all pediatric solid malignant cancers and less than 1% of all adult solid malignant cancers. The vast majority of diagnosed sarcomas will be soft tissue sarcomas, while malignant bone tumors make up just over 10% of sarcomas. The risks for sarcoma are not well-understood. We evaluated the existing literature on the epidemiology and etiology of sarcoma. Risks for sarcoma development can be divided into environmental exposures, genetic susceptibility, and an interaction between the two. HIV-positive individuals are at an increased risk for Kaposi’s sarcoma, even though HHV8 is the causative virus. Radiation exposure from radiotherapy has been strongly associated with secondary sarcoma development in certain cancer patients. In fact, the risk of malignant bone tumors increases as the cumulative dose of radiation to the bone increases (p for trend <0.001). A recent meta-analysis reported that children with a history of hernias have a greater risk of developing Ewing’s sarcoma (adjusted OR 3.2, 95% CI 1.9, 5.7). Bone development during pubertal growth spurts has been associated with osteosarcoma development. Occupational factors such as job type, industry, and exposures to chemicals such as herbicides and chlorophenols have been suggested as risk factors for sarcomas. A case-control study found a significant increase in soft tissue sarcoma risk among gardeners (adjusted OR 4.1, 95% CI 1.00, 14.00), but not among those strictly involved in farming. A European-based study reported an increased risk in bone tumors among blacksmiths, toolmakers, or machine-tool operators (adjusted OR 2.14, 95% CI 1.08, 4.26). Maternal and paternal characteristics such as occupation, age, smoking status, and health conditions experienced during pregnancy also have been suggested as sarcoma risk factors and would be important to assess in future studies. The limited studies we identified demonstrate significant relationships with sarcoma risk, but many of these results now require further validation on larger populations. Furthermore, little is known about the biologic mechanisms behind each epidemiologic association assessed in the literature. Future molecular epidemiology studies may increase our understanding of the genetic versus environmental contributions to tumorigenesis in this often deadly cancer in children and adults.

Evaluation of the Effectiveness of a Patient Centered Educational Mailer Designed to Improve Statin Adherence: A Pragmatic Trial

Background: Patients with high total cholesterol have increased risk of cardiovascular disease. National Cholesterol Education Program (NCEP) and American Heart Association (AHA) guidelines recommend cholesterol lowering with statin medications; however, statin adherence remains poor. We hypothesized that patient-centered education on the 10-year risk for each of the major constituents of cardiovascular disease would increase statin adherence and achievement of the low-density lipoprotein cholesterol (LDL-C) goal. Methods: Veterans within the Salt Lake City Veterans Affairs (VA) Medical Center initiating statin therapy from October 2008 to December 2011 were randomized in a pragmatic design to receive either an educational mailer or usual care. The mailer outlined their 10-year global cardiovascular risk, separated into coronary heart disease, stroke, and congestive heart failure. The study was unblinded and followed an intention-to-treat analysis where outcome measures were obtained during normal care process. The primary outcome measure was the achievement of the LDL-C goal during the 12-month follow-up. Results: Two hundred and seven patients were randomly assigned to either the intervention arm (95) or the control arm (112). No differences in the proportion of patients meeting the LDL-C goal were detected during 12-months [Relative Risk (RR): 0.95 (95 percent confidence interval (CI): 0.77–1.17)] or 18-months [RR: 1.03 (95 percent CI: 0.84, 1.25)]. Patients in the intervention arm had higher adherence on average, e.g., intervention patients were more likely to have 70 percent or more days of statin therapy compared to patients who received standard care—though this did not reach statistical significance—RR: 1.33 (95 percent CI: 1.00, 1.78). There were no statistical differences in cardiovascular outcomes or mortality. Conclusion: Patient education mailers sent to patients starting statin treatment did not have a clear impact on LDL-C goal achievement or adherence to statin therapy.

Use of Disease-modifying Antirheumatic Drugs for Inflammatory Arthritis in US Veterans: Effect of Specialty Care and Geographic Distance

Objective. To evaluate the effect of access to and distance from rheumatology care on the use of disease-modifying antirheumatic drugs (DMARD) in US veterans with inflammatory arthritis (IA). Methods. Provider encounters and DMARD dispensations for IA (rheumatoid arthritis, psoriatic arthritis, and ankylosing spondylitis) were evaluated in national Veterans Affairs (VA) datasets between January 1, 2015, and December 31, 2015. Results. Among 12,589 veterans with IA, 23.5% saw a rheumatology provider. In the general IA population, 25.3% and 13.6% of veterans were exposed to a synthetic DMARD (sDMARD) and biologic DMARD (bDMARD), respectively. DMARD exposure was 2.6- to 3.4-fold higher in the subpopulation using rheumatology providers, compared to the general IA population. The distance between veterans’ homes and the closest VA rheumatology site was < 40 miles (Near) for 55.9%, 40–99 miles (Intermediate) for 31.7%, and ≥ 100 miles (Far) for 12.4%. Veterans in the Intermediate and Far groups were less likely to see a rheumatology provider than veterans in the Near group (RR = 0.72 and RR = 0.49, respectively). Exposure to bDMARD was 34% less frequent in the Far group than the Near group. In the subpopulation who used rheumatology care, the bDMARD exposure discrepancy did not persist between distance groups. Conclusion. Use of rheumatology care and DMARD was low for veterans with IA. DMARD exposure was strongly associated with rheumatology care use. Veterans in the general IA population living far from rheumatology sites accessed rheumatology care and bDMARD less frequently than veterans living close to rheumatology sites.

Evaluation of the Case–Crossover (CCO) Study Design for Adverse Drug Event Detection

IntroductionThe case–crossover (CCO) design was originally intended to study exposures characterized as intermittent with acute effects. The performance of the CCO design is not well characterized under alternative exposure and outcome relationships.ObjectiveThe purpose of this study was to evaluate the ability of the CCO to identify simulated treatment effects under different drug exposures and outcomes relationships while varying the duration of the 1:1 matched risk and control windows.MethodsThe simulated data were obtained from the Observational Medical Dataset Simulator, version 2 (OSIM2). The area under the receiver operator characteristic curve (AUC) was calculated to compare CCO performance across outcome types, simulated relative risk (RR), and duration of risk and control windows.ResultsThe AUC for acute outcomes was higher for shorter risk and control windows and improved with higher simulated RR. For example, the AUC for the simulated RR of 4 was 0.95 for a 30-day window length and 0.78 for a 360-day window length. The AUC for the accumulative outcomes increased with longer risk and control windows and stronger simulated RR. For example, the AUC for the simulated RR of 4 was 0.85 for a 360-day window length and 0.23 for a 30-day window length. Risk and control window lengths did not appear to sufficiently alter the AUC for insidious onset outcomes.ConclusionsThe CCO performed best for acute-onset outcomes, but may be useful for exploring adverse outcomes with accumulative effects. Careful consideration must be given to the hypothesized drug exposure and outcome distribution because specification of risk and control window duration affects CCO performance.

Predicting Psychiatric Hospitalizations among Elderly Veterans with a History of Mental Health Disease

Introduction: Patient Aligned Care Team (PACT) care managers are tasked with identifying aging Veterans with psychiatric disease in attempt to prevent psychiatric crises. However, few resources exist that use real-time information on patient risk to prioritize coordinating appropriate care amongst a complex aging population. Objective: To develop and validate a model to predict psychiatric hospital admission, during a 90-day risk window, in Veterans ages 65 or older with a history of mental health disease. Methods: This study applied a cohort design to historical data available in the Veterans Affairs (VA) Corporate Data Warehouse (CDW). The Least Absolute Shrinkage and Selection Operator (LASSO) regularization regression technique was used for model development and variable selection. Individual predicted probabilities were estimated using logistic regression. A split-sample approach was used in performing external validation of the fitted model. The concordance statistic (C-statistic) was calculated to assess model performance. Results: Prior to modeling, 61 potential candidate predictors were identified and 27 variables remained after applying the LASSO method. The final model’s predictive accuracy is represented by a C-statistic of 0.903. The model’s predictive accuracy during external validation is represented by a C-statistic of 0.935. Having a previous psychiatric hospitalization, psychosis, bipolar disorder, and the number of mental-health related social work encounters were strong predictors of a geriatric psychiatric hospitalization. Conclusion: This predictive model is capable of quantifying the risk of a geriatric psychiatric hospitalization with acceptable performance and allows for the development of interventions that could potentially reduce such risk.

Identifying Axial Spondyloarthritis in Electronic Medical Records of US Veterans.

Large database research in axial spondyloarthritis (SpA) is limited by a lack of methods for identifying most types of axial SpA. Our objective was to develop methods for identifying axial SpA concepts in the free text of documents from electronic medical records.

Comparison of the effectiveness and side effects of dofetilide and dronedarone in the treatment of atrial fibrillation during an indicated period in time with perceived equipoise [version 1; referees: 1 approved, 1 approved with reservations]

Dronedarone is an anti-arrhythmic drug (AAD) originally approved for the treatment of atrial arrhythmias. The effectiveness and side effects of dronedarone have not been adequately compared to other commonly used AADs using observational data. We compared rates of recurrent atrial arrhythmias, incidence of side effects, and discontinuation rates of dronedarone to another class III AAD, dofetilide. We included patients from a single academic medical center between 2003 and 2010. Chart review was utilized to collect historical data of baseline clinical characteristics, side effects, arrhythmia recurrence, and drug discontinuation. Propensity score matching was used to balance baseline covariates. Cox-proportional hazard models were used to compare rates of recurrence between dronedarone and dofetilide. Patients were excluded if they failed to acutely achieve sinus rhythm, developed side effects leading to immediate discontinuation, or did not have sufficient follow-up. The final analysis included 127 dofetilide patients and 57 dronedarone patients. Fifty-nine patients (46.5%) experienced recurrence in the dofetilide group within the first year of treatment compared to 42 dronedarone patients (71.2%) (p<0.01). The adjusted hazard rate of recurrence was 2.42 times greater for dronedarone compared to dofetilide (95% CI: 1.44, 4.07; p-value<0.01). Side effects leading to drug discontinuation, including significant QT prolongation, developed more frequently with dofetilide (24.1% vs. 9.9%; p<0.01). Dronedarone is less effective than dofetilide in arrhythmia suppression. Our findings suggest dofetilide is associated with more serious side effects and a higher rate of discontinuation.

Errata to NLP study of infusion notes to identify outpatient infusions in the VA.

In the recent article titled, “The use of natural language processing of infusion notes to identify outpatient infusions,” we reported that Healthcare Common Procedure Coding System (HCPCS) codes were not capable of adequately capturing infliximab infusions administered in the Veteran Health Administration (VHA) health care system.1 Furthermore, we reported that the use of Natural Language Processing (NLP) software on clinical notes was superior in capturing infliximab infused dose and administered date. As a result, we recommended a combined approach to capturing infliximab infusions in the Corporate Data Warehouse (CDW) using both HCPCS codes and NLP on clinical notes. Thus, we concluded that NLP would be required to accurately identify outpatient infliximab infusions and accompanying details. However, we have recently discovered additional approaches capable of identifying infliximab infusions that warranted further examination. Other data-marts that serve distinct purposes in the VA exist beyond the CDW. For example, the Managerial Cost Accounting (MCA) data system serves as the only means the VA has of complying with public laws that mandate the use of a data system capable of assigning costs for health care encounters.2 In addition, the Pharmacy Benefits Management (PBM) data-mart was developed to assist the national VA drug formulary process.3 It was our impression that these additional data-marts were being phased out of the VHA. However, we have since learned that these particular sources of VHA data were simply undergoing structural changes in order to improve their accuracy and utility for VHA research investigators. We felt that it was appropriate to include these additional data-marts in order to improve the completeness of our original investigation. Furthermore, an additional data source found within the CDW-raw domain, referred to as the CDW Inpatient Medications Intravenous (IV) database, is also known to capture drugs administered in an outpatient setting and has been recently made accessible to us for review. With the addition of these data-marts and the CDW-raw data domain, we were able to successfully compare eight different approaches potentially suitable for accurately identifying outpatient infliximab infusion administered dose and date. They are as follows:

Abstract A84: Risk factors for lung cancer in Nepal

Over 672,000 lung cancer patients, half of all cases in the world, are diagnosed in low-and middle-income countries (LMC) each year, yet the majority of epidemiologic studies on cancer have been conducted in high-income countries. Cancer is already a major burden in LMCs, and the burden is expected to increase in the next decades. Nepal is one of the poorest countries in the world, with a gross domestic product per capita of

Socioeconomic status and lung cancer risk in Nepal.

470 in 2009. Lung cancer is the most common cancer among men and the third most common cancer among women in Nepal, where cancer patients remain among the most severely medically underserved. Primary exposures of concern in Nepal with regard to lung cancer are indoor air pollution from solid fuels and smoking tobacco including local tobacco products such as choor and kankat. We have initiated a case-control study of lung cancer in Nepal, with the aims to study the lung cancer risk factor profile in a LMC and estimate attributable fractions and to identify genetic susceptibility factors for lung cancer in Nepal. The lung cancer cases are being recruited from the main cancer hospital in Nepal, the B.P. Koirala Memorial Cancer Hospital (BPKMCH). We aim to recruit 600 lung cancer cases and 600 controls, who will undergo a face-to-face interview on tobacco consumption, occupational history, residential history, type of fuel used for cooking, reproductive factors (for women only), dietary and other lifestyle habits. Biosample collection will include blood, toenail clippings, hair, urine, blood spots and tumor tissue for cases. Thus far we have recruited 103 lung cancer cases and 156 controls. Approximately 46.6% (n=48) of lung cancer patients were females and the largest proportion of cases were in the 60-69 year age group (35.0%). The proportion of never-smokers among controls was more (39%) compared to lung cancer cases (14%), as expected. Among men, 10.9% of cases and 28.5% of controls were never smokers. Among women, 16.7% of cases and 70% of controls were never smokers. The most common type of tobacco smoked was cigarettes without filters (61.2% of cases and 34.0% of controls). The odds ratio adjusted for age, sex, education and ethnicity were 3.96 (95%CI=1.71-9.22) for subjects who currently smoked cigarettes without filters and 1.94 (95%CI=0.96-3.92) for subjects who formerly smoked cigarettes without filters. The majority of subjects lived in rural areas for all of their lives (79.9%) and used wood for cooking and heating in the home (78.2%). Approximately 65.4% of subjects had been exposed to involuntary smoking in the home and 48.4% had ever consumed alcohol. As more subjects are recruited, we will analyze the data further for risk factors. In the future, we would like to evaluate clinical management, quality of life and survival among lung cancer patients in Nepal, and to conduct cost utility analysis for primary and tertiary prevention efforts against lung cancer in Nepal. Our study will generate crucial data needed to formulate more effective cancer care and control policies in Nepal, to improve the situation for this severely medically underserved population. Citation Information: Cancer Epidemiol Biomarkers Prev 2010;19(10 Suppl):A84.