Romain Neugebauer

HIV Infection, Immunodeficiency, Viral Replication, and the Risk of Cancer

Background: Few studies have compared cancer risk between HIV-infected individuals and a demographically similar HIV-uninfected internal comparison group, adjusting for cancer risk factors. Methods: We followed 20,775 HIV-infected and 215,158 HIV-uninfected individuals enrolled in Kaiser Permanente (KP) California for incident cancer from 1996 to 2008. Rate ratios (RR) were obtained from Poisson models comparing HIV-infected (overall and stratified by recent CD4 count and HIV RNA) with HIV-uninfected individuals, adjusted for age, sex, race/ethnicity, calendar period, KP region, smoking, alcohol/drug abuse, and overweight/obesity. Results: We observed elevated RRs for Kaposi sarcoma (KS; RR = 199; P < 0.001), non-Hodgkin lymphoma (NHL; RR = 15; P < 0.001), anal cancer (RR = 55; P < 0.001), Hodgkin lymphoma (HL; RR = 19; P < 0.001), melanoma (RR = 1.8; P = 0.001), and liver cancer (RR = 1.8; P = 0.013), a reduced RR for prostate cancer (RR = 0.8; P = 0.012), and no increased risk for oral cavity/pharynx (RR = 1.4; P = 0.14), lung (RR = 1.2; P = 0.15), or colorectal (RR = 0.9; P = 0.34) cancers. Lung and oral cavity/pharynx cancers were elevated for HIV-infected subjects in models adjusted only for demographics. KS, NHL, anal cancer, HL, and colorectal cancer had significant (P < 0.05) trends for increasing RRs with decreasing recent CD4. The RRs for lung and oral cavity/pharynx cancer were significantly elevated with CD4 < 200 cells/μL and for melanoma and liver cancer with CD4 < 500 cells/μL. Only KS and NHL were associated with HIV RNA. Conclusion: Immunodeficiency was positively associated with all cancers examined except prostate cancer among HIV-infected compared with HIV-uninfected individuals, after adjustment for several cancer risk factors. Impact: Earlier antiretroviral therapy initiation to maintain high CD4 levels might reduce the burden of cancer in this population. Cancer Epidemiol Biomarkers Prev; 20(12); 2551–9. ©2011 AACR.

Construction of a Multisite DataLink Using Electronic Health Records for the Identification, Surveillance, Prevention, and Management of Diabetes Mellitus: The SUPREME-DM Project

Introduction Electronic health record (EHR) data enhance opportunities for conducting surveillance of diabetes. The objective of this study was to identify the number of people with diabetes from a diabetes DataLink developed as part of the SUPREME-DM (SUrveillance, PREvention, and ManagEment of Diabetes Mellitus) project, a consortium of 11 integrated health systems that use comprehensive EHR data for research. Methods We identified all members of 11 health care systems who had any enrollment from January 2005 through December 2009. For these members, we searched inpatient and outpatient diagnosis codes, laboratory test results, and pharmaceutical dispensings from January 2000 through December 2009 to create indicator variables that could potentially identify a person with diabetes. Using this information, we estimated the number of people with diabetes and among them, the number of incident cases, defined as indication of diabetes after at least 2 years of continuous health system enrollment. Results The 11 health systems contributed 15,765,529 unique members, of whom 1,085,947 (6.9%) met 1 or more study criteria for diabetes. The nonstandardized proportion meeting study criteria for diabetes ranged from 4.2% to 12.4% across sites. Most members with diabetes (88%) met multiple criteria. Of the members with diabetes, 428,349 (39.4%) were incident cases. Conclusion The SUPREME-DM DataLink is a unique resource that provides an opportunity to conduct comparative effectiveness research, epidemiologic surveillance including longitudinal analyses, and population-based care management studies of people with diabetes. It also provides a useful data source for pragmatic clinical trials of prevention or treatment interventions.

Causal inference in epidemiological studies with strong confounding.

One of the identifiability assumptions of causal effects defined by marginal structural model (MSM) parameters is the experimental treatment assignment (ETA) assumption. Practical violations of this assumption frequently occur in data analysis when certain exposures are rarely observed within some strata of the population. The inverse probability of treatment weighted (IPTW) estimator is particularly sensitive to violations of this assumption; however, we demonstrate that this is a problem for all estimators of causal effects. This is due to the fact that the ETA assumption is about information (or lack thereof) in the data. A new class of causal models, causal models for realistic individualized exposure rules (CMRIER), is based on dynamic interventions. CMRIER generalize MSM, and their parameters remain fully identifiable from the observed data, even when the ETA assumption is violated, if the dynamic interventions are set to be realistic. Examples of such realistic interventions are provided. We argue that causal effects defined by CMRIER may be more appropriate in many situations, particularly those with policy considerations. Through simulation studies, we examine the performance of the IPTW estimator of the CMRIER parameters in contrast to that of the MSM parameters. We also apply the methodology to a real data analysis in air pollution epidemiology to illustrate the interpretation of the causal effects defined by CMRIER.

Occult Uterine Sarcoma and Leiomyosarcoma: Incidence of and Survival Associated With Morcellation.

OBJECTIVE: To estimate the incidence of occult uterine sarcoma and leiomyosarcoma in hysterectomies for leiomyomas and the risk associated with their morcellation. METHODS: We conducted a population-based cohort study. All uterine sarcomas from 2006–2013 in an integrated health care system were identified. Age- and race-specific incidences of occult uterine sarcoma were calculated. Kaplan-Meier survival analysis was performed. Crude and adjusted risk ratios of recurrence and death associated with morcellation at 1, 2, and 3 years were estimated using Poisson regression with inverse probability weighting. RESULTS: There were 125 hysterectomies with occult uterine sarcomas identified among 34,728 hysterectomies performed for leiomyomas. The incidence of occult uterine sarcoma and leiomyosarcoma was 1 of 278 or 3.60 (95% confidence interval [CI] 2.97–4.23) and 1 of 429 or 2.33 (95% CI 1.83–2.84) per 1,000 hysterectomies. For stage I leiomyosarcoma (n=111), eight (7.2%) were power and 27 (24.3%) nonpower-morcellated. The unadjusted 3-year probability of disease-free survival for no morcellation, power and nonpower morcellation was 0.54, 0.19, and 0.51, respectively (P=.15); overall survival was 0.64, 0.75, and 0.68, respectively (P=.97). None of the adjusted risk ratios for recurrence or death were significant except for death at 1 year for power and nonpower morcellation groups combined (6/33) compared with no morcellation (4/76) (5.12, 95% CI 1.33–19.76, P=.02). We had inadequate power to infer differences for all other comparisons including 3-year survival and power morcellation. CONCLUSION: Morcellation is associated with decreased early survival of women with occult leiomyosarcomas. We could not accurately assess associations between power morcellation and 3-year survival as a result of small numbers.

Use of Abacavir and Risk of Cardiovascular Disease Among HIV-Infected Individuals.

Background:Evidence is conflicting about the association of abacavir use and cardiovascular disease (CVD) among HIV-infected individuals. Previous studies may have been biased by the preferential initiation or continuation of abacavir in patients with renal dysfunction. Methods:We conducted a cohort study in Kaiser Permanente California during 1998–2011, following HIV-infected adults initiating antiretroviral therapy until the earliest of CVD (ie, coronary heart disease or ischemic stroke), health plan disenrollment, death, or end of study. We used inverse-probability weighting to fit marginal structural models to estimate hazard ratios (HRs) for CVD comparing regimens with and without abacavir. Propensity score models included demographics, HIV-specific factors, and CVD risk factors, including alcohol/drug use, smoking, overweight/obesity, diabetes, lipid-lowering and hypertension therapy, and renal dysfunction (ie, estimated glomerular filtration rate <60 mL·min−1·1.73 m−2). Results:Among 8154 subjects, 178 had ≥1 CVD event, with 24/704 (3.4%) in the abacavir group and 154/7450 (2.1%) in the group initiating regimens without abacavir. Abacavir users had more renal dysfunction at antiretroviral therapy initiation (7.0% vs. 3.3%, P < 0.001). Compared with patients initiating regimens without abacavir, abacavir users had a 2.2-fold higher risk of CVD in intention-to-treat analysis [HR 2.2, 95% confidence interval (CI): 1.4 to 3.5], a 2.7-fold higher risk when remaining on their initial regimens for ≥1 year (HR 2.7, 95% CI: 1.5 to 5.0), and a 2.1-fold higher risk in per-protocol analysis (HR 2.1, 95% CI: 0.9 to 5.0). Conclusions:Abacavir was associated with an over 2-fold increased risk of CVD, which was not explained by renal dysfunction or other CVD risk factors.

Analysis of Body Mass Index and Mortality in Patients With Colorectal Cancer Using Causal Diagrams.

IMPORTANCE Physicians and investigators have sought to determine the relationship between body mass index (BMI [calculated as weight in kilograms divided by height in meters squared]) and colorectal cancer (CRC) outcomes, but methodologic limitations including sampling selection bias, reverse causality, and collider bias have prevented the ability to draw definitive conclusions. OBJECTIVE To evaluate the association of BMI at the time of, and following, colorectal cancer (CRC) diagnosis with mortality in a complete population using causal diagrams. DESIGN, SETTING, AND PARTICIPANTS This retrospective observational study with prospectively collected data included a cohort of 3408 men and women, ages 18 to 80 years, from the Kaiser Permanente Northern California population, who were diagnosed with stage I to III CRC between 2006 and 2011 and who also had surgery. EXPOSURES Body mass index at diagnosis and 15 months following diagnosis. MAIN OUTCOMES AND MEASURES Hazard ratios (HRs) for all-cause mortality and CRC-specific mortality compared with normal-weight patients, adjusted for sociodemographics, disease severity, treatment, and prediagnosis BMI. RESULTS This study investigated a cohort of 3408 men and women ages 18 to 80 years diagnosed with stage I to III CRC between 2006 and 2011 who also had surgery. At-diagnosis BMI was associated with all-cause mortality in a nonlinear fashion, with patients who were underweight (BMI <18.5; HR, 2.65; 95% CI, 1.63-4.31) and patients who were class II or III obese (BMI ≥35; HR, 1.33; 95% CI, 0.89-1.98) exhibiting elevated mortality risks, compared with patients who were low-normal weight (BMI 18.5 to <23). In contrast, patients who were high-normal weight (BMI 23 to <25; HR, 0.77; 95% CI, 0.56-1.06), low-overweight (BMI 25 to <28; HR, 0.75; 95% CI, 0.55-1.04), and high-overweight (BMI 28 to <30; HR, 0.52; 95% CI, 0.35-0.77) had lower mortality risks, and patients who were class I obese (BMI 30 to <35) showed no difference in risk. Spline analysis confirmed a U-shaped relationship in participants with lowest mortality at a BMI of 28. Associations with CRC-specific mortality were similar. Associations of postdiagnosis BMI and mortality were also similar, but patients who were class I obese had significantly lower all-cause and cancer-specific mortality risks. CONCLUSIONS AND RELEVANCE In this study, body mass index at the time of diagnosis and following diagnosis of CRC was associated with mortality risk. Though evidence shows that exercise in patients with cancer should be encouraged, findings suggest that recommendations for weight loss in the immediate postdiagnosis period among patients with CRC who are overweight may be unwarranted.

Health-plan and employer-based wellness programs to reduce diabetes risk: The Kaiser Permanente Northern California NEXT-D Study.

Primary prevention of diabetes is increasingly recognized by both health plans and employers as an important strategy to improve the health of insured populations. As a part of the Natural Experiments in Translation for Diabetes (NEXT-D) network, the Kaiser Permanente Northern California (KPNC) Division of Research is assessing the effectiveness of 2 health plan-initiated programs to prevent the onset of diabetes in patients at high risk. The first study evaluates a telephonic health-coaching program that provides counseling on healthful eating, active living, and weight loss to KPNC members. The second evaluation examines a postpartum glucose screening and educational diabetes prevention program for women with gestational diabetes mellitus that KPNC implemented in 2006. Identifying effective approaches to preventing diabetes will be of value to health care systems, policy makers, and public health officials seeking to understand the roles systems and employers can play in preventing chronic illness.

The association of race/ethnicity and risk of atypical femur fracture among older women receiving oral bisphosphonate therapy

PURPOSE Several epidemiologic studies suggest that compared to white women, Asians have a greater propensity to suffer an atypical femur fracture (AFF) while taking bisphosphonate therapy. This study examines the relative risk of AFF following bisphosphonate initiation for Asian compared to white women. METHODS Using data from a large integrated northern California healthcare delivery system, we examined diaphyseal femur fracture outcomes among women age≥50years old who initiated oral bisphosphonate therapy during 2002-2007. An AFF was defined by the 2013 American Society of Bone and Mineral Research Task Force criteria. The risk of radiographically-confirmed AFF was examined for Asian compared to white women, adjusting for differences in bisphosphonate exposure and other potential risk factors. RESULTS Among 48,390 women (65.3% white, 17.1% Asian) who newly initiated bisphosphonate therapy and were followed for a median of 7.7years, 68 women experienced an AFF. The rate of AFF was 18.7 per 100,000 person-years overall and eight-fold higher among Asian compared to white women (64.2 versus 7.6 per 100,000 person-years). Asians were also more likely to have longer bisphosphonate treatment duration compared to whites (median 3.8 versus 2.7years). The age-adjusted relative hazard for AFF was 8.5 (95% confidence interval 4.9-14.9) comparing Asian to white women, and was only modestly reduced to 6.6 (3.7-11.5) after adjusting for bisphosphonate duration and current use. CONCLUSIONS Our study confirms marked racial disparity in AFF risk that should be further investigated, particularly the mechanisms accounting for this difference. These findings also underscore the need to further examine the association of bisphosphonate duration and AFF in women of Asian race, as well as differential risk across Asian subgroups. In the interim, counseling of Asian women about osteoporosis drug continuation should include consideration of their potentially higher AFF risk.

Targeted learning in real-world comparative effectiveness research with time-varying interventions

In comparative effectiveness research (CER), often the aim is to contrast survival outcomes between exposure groups defined by time-varying interventions. With observational data, standard regression analyses (e.g., Cox modeling) cannot account for time-dependent confounders on causal pathways between exposures and outcome nor for time-dependent selection bias that may arise from informative right censoring. Inverse probability weighting (IPW) estimation to fit marginal structural models (MSMs) has commonly been applied to properly adjust for these expected sources of bias in real-world observational studies. We describe the application and performance of an alternate estimation approach in such a study. The approach is based on the recently proposed targeted learning methodology and consists in targeted minimum loss-based estimation (TMLE) with super learning (SL) within a nonparametric MSM. The evaluation is based on the analysis of electronic health record data with both IPW estimation and TMLE to contrast cumulative risks under four more or less aggressive strategies for treatment intensification in adults with type 2 diabetes already on 2+ oral agents or basal insulin. Results from randomized experiments provide a surrogate gold standard to validate confounding and selection bias adjustment. Bootstrapping is used to validate analytic estimation of standard errors. This application does the following: (1) establishes the feasibility of TMLE in real-world CER based on large healthcare databases; (2) provides evidence of proper confounding and selection bias adjustment with TMLE and SL; and (3) motivates their application for improving estimation efficiency. Claims are reinforced with a simulation study that also illustrates the double-robustness property of TMLE.

Super learning to hedge against incorrect inference from arbitrary parametric assumptions in marginal structural modeling

OBJECTIVE Clinical trials are unlikely to ever be launched for many comparative effectiveness research (CER) questions. Inferences from hypothetical randomized trials may however be emulated with marginal structural modeling (MSM) using observational data, but success in adjusting for time-dependent confounding and selection bias typically relies on parametric modeling assumptions. If these assumptions are violated, inferences from MSM may be inaccurate. In this article, we motivate the application of a data-adaptive estimation approach called super learning (SL) to avoid reliance on arbitrary parametric assumptions in CER. STUDY DESIGN AND SETTING Using the electronic health records data from adults with new-onset type 2 diabetes, we implemented MSM with inverse probability weighting (IPW) estimation to evaluate the effect of three oral antidiabetic therapies on the worsening of glomerular filtration rate. RESULTS Inferences from IPW estimation were noticeably sensitive to the parametric assumptions about the associations between both the exposure and censoring processes and the main suspected source of confounding, that is, time-dependent measurements of hemoglobin A1c. SL was successfully implemented to harness flexible confounding and selection bias adjustment from existing machine learning algorithms. CONCLUSION Erroneous IPW inference about clinical effectiveness because of arbitrary and incorrect modeling decisions may be avoided with SL.