Ronald B. Geskus

Competing risks in epidemiology: possibilities and pitfalls

BACKGROUND In studies of all-cause mortality, the fundamental epidemiological concepts of rate and risk are connected through a well-defined one-to-one relation. An important consequence of this relation is that regression models such as the proportional hazards model that are defined through the hazard (the rate) immediately dictate how the covariates relate to the survival function (the risk). METHODS This introductory paper reviews the concepts of rate and risk and their one-to-one relation in all-cause mortality studies and introduces the analogous concepts of rate and risk in the context of competing risks, the cause-specific hazard and the cause-specific cumulative incidence function. RESULTS The key feature of competing risks is that the one-to-one correspondence between cause-specific hazard and cumulative incidence, between rate and risk, is lost. This fact has two important implications. First, the naïve Kaplan-Meier that takes the competing events as censored observations, is biased. Secondly, the way in which covariates are associated with the cause-specific hazards may not coincide with the way these covariates are associated with the cumulative incidence. An example with relapse and non-relapse mortality as competing risks in a stem cell transplantation study is used for illustration. CONCLUSION The two implications of the loss of one-to-one correspondence between cause-specific hazard and cumulative incidence should be kept in mind when deciding on how to make inference in a competing risks situation.

Endoscopic Transgastric vs Surgical Necrosectomy for Infected Necrotizing Pancreatitis: A Randomized Trial

CONTEXT Most patients with infected necrotizing pancreatitis require necrosectomy. Surgical necrosectomy induces a proinflammatory response and is associated with a high complication rate. Endoscopic transgastric necrosectomy, a form of natural orifice transluminal endoscopic surgery, may reduce the proinflammatory response and reduce complications. OBJECTIVE To compare the proinflammatory response and clinical outcome of endoscopic transgastric and surgical necrosectomy. DESIGN, SETTING, AND PATIENTS Randomized controlled assessor-blinded clinical trial in 3 academic hospitals and 1 regional teaching hospital in The Netherlands between August 20, 2008, and March 3, 2010. Patients had signs of infected necrotizing pancreatitis and an indication for intervention. INTERVENTIONS Random allocation to endoscopic transgastric or surgical necrosectomy. Endoscopic necrosectomy consisted of transgastric puncture, balloon dilatation, retroperitoneal drainage, and necrosectomy. Surgical necrosectomy consisted of video-assisted retroperitoneal debridement or, if not feasible, laparotomy. MAIN OUTCOME MEASURES The primary end point was the postprocedural proinflammatory response as measured by serum interleukin 6 (IL-6) levels. Secondary clinical end points included a predefined composite end point of major complications (new-onset multiple organ failure, intra-abdominal bleeding, enterocutaneous fistula, or pancreatic fistula) or death. RESULTS We randomized 22 patients, 2 of whom did not undergo necrosectomy following percutaneous catheter drainage and could not be analyzed for the primary end point. Endoscopic transgastric necrosectomy reduced the postprocedural IL-6 levels compared with surgical necrosectomy (P = .004). The composite clinical end point occurred less often after endoscopic necrosectomy (20% vs 80%; risk difference [RD], 0.60; 95% CI, 0.16-0.80; P = .03). Endoscopic necrosectomy did not cause new-onset multiple organ failure (0% vs 50%, RD, 0.50; 95% CI, 0.12-0.76; P = .03) and reduced the number of pancreatic fistulas (10% vs 70%; RD, 0.60; 95% CI, 0.17-0.81; P = .02). CONCLUSION In patients with infected necrotizing pancreatitis, endoscopic necrosectomy reduced the proinflammatory response as well as the composite clinical end point compared with surgical necrosectomy. TRIAL REGISTRATION isrctn.org Identifier: ISRCTN07091918.

Effective therapy has altered the spectrum of cause-specific mortality following HIV seroconversion.

Introduction:Although HAART has led to a reduction in overall mortality among HIV-infected individuals, its impact on death from specific causes is unknown. Methods:Twenty-two cohorts of HIV-infected individuals with known dates of seroconversion are pooled in the CASCADE collaboration. Causes of death (COD) were categorized into three AIDS-related and seven non-AIDS-related causes. The unknown causes were assigned a separate category. The cumulative incidence for each COD was calculated in the presence of the other competing COD, for the pre-HAART and HAART eras. A multivariate regression analyses for the cumulative rate of progression to the different COD was performed. Results:A total of 1938 of 7680 HIV-seroconverters died. Pre-HAART, AIDS opportunistic infections (OI) was the most common COD, followed by unknown and HIV/AIDS-unspecified. In the HAART era, the cumulative incidence for all AIDS-related COD decreased, OI remaining the most important. Large reductions in death due to other infections and organ failure were seen. Cumulative death risk decreased in the HAART era for most causes. The effect of HAART was not the same for all risk groups. The cumulative risk of death from AIDS-related malignancies, OI and non-AIDS-related malignancies decreased significantly among homosexual men (MSM), whereas the risk of dying from (un)-intentional death increased significantly among injecting drug users (IDU). A non-significant increase in hepatitis/liver-related death was seen in MSM, IDU and haemophiliacs. Conclusion:Overall and cause specific mortality decreased following the introduction of HAART. OI remain the most common COD in the HAART era, suggesting that AIDS-related events will continue to be important in the future. Future trends in COD should be monitored using standardized guidelines.

High Risk of Symptomatic Cardiac Events in Childhood Cancer Survivors

PURPOSE To evaluate the long-term risk for validated symptomatic cardiac events (CEs) and associated risk factors in childhood cancer survivors (CCSs). PATIENTS AND METHODS We determined CEs grade 3 or higher: congestive heart failure (CHF), cardiac ischemia, valvular disease, arrhythmia and/or pericarditis (according to Common Terminology Criteria for Adverse Events [CTCAE], version 3.0) in a hospital-based cohort of 1,362 5-year CCSs diagnosed between 1966 and 1996. We calculated both marginal and cause-specific cumulative incidence of CEs and cause-specific cumulative incidence of separate events. We analyzed different risk factors in multivariable Cox regression models. RESULTS Overall, 50 CEs, including 27 cases of CHF, were observed in 42 survivors (at a median attained age of 27.1 years). The 30-year cause-specific cumulative incidence of CEs was significantly increased after treatment with both anthracyclines and cardiac irradiation (12.6%; 95% CI, 4.3% to 20.3%), after anthracyclines (7.3%; 95% CI, 3.8% to 10.7%), and after cardiac irradiation (4.0%; 95% CI, 0.5% to 7.4%) compared with other treatments. In the proportional hazards analyses, anthracycline (dose), cardiac irradiation (dose), combination of these treatments, and congenital heart disease were significantly associated with developing a CE. We demonstrated an exponential relationship between the cumulative anthracycline dose, cardiac irradiation dose, and risk of CE. CONCLUSION CCSs have a high risk of developing symptomatic CEs at an early age. The most common CE was CHF. Survivors treated with both anthracyclines and radiotherapy have the highest risk; after 30 years, one in eight will develop severe heart disease. The use of potentially cardiotoxic treatments should be reconsidered for high-risk groups, and frequent follow-up for high-risk survivors is needed.

A resurgent HIV-1 epidemic among men who have sex with men in the era of potent antiretroviral therapy

Objective:Reducing viral load, highly active antiretroviral therapy has the potential to limit onwards transmission of HIV-1 and thus help contain epidemic spread. However, increases in risk behaviour and resurgent epidemics have been widely reported post-highly active antiretroviral therapy. The aim of this study was to quantify the impact that highly active antiretroviral therapy had on the epidemic. Design:We focus on the HIV-1 epidemic among men who have sex with men in the Netherlands, which has been well documented over the past 20 years within several long-standing national surveillance programs. Methods:We used a mathematical model including highly active antiretroviral therapy use and estimated the changes in risk behaviour and diagnosis rate needed to explain annual data on HIV and AIDS diagnoses. Results:We show that the reproduction number R(t), a measure of the state of the epidemic, declined early on from initial values above two and was maintained below one from 1985 to 2000. Since 1996, when highly active antiretroviral therapy became widely used, the risk behaviour rate has increased 66%, resulting in an increase of R(t) to 1.04 in the latest period 2000–2004 (95% confidence interval 0.98–1.09) near or just above the threshold for a self-sustaining epidemic. Hypothetical scenario analysis shows that the epidemiological benefits of highly active antiretroviral therapy and earlier diagnosis on incidence have been entirely offset by increases in the risk behaviour rate. Conclusion:We provide the first detailed quantitative analysis of the HIV epidemic in a well defined population and find a resurgent epidemic in the era of highly active antiretroviral therapy, most likely predominantly caused by increasing sexual risk behaviour.

HIV incidence on the increase among homosexual men attending an Amsterdam sexually transmitted disease clinic: using a novel approach for detecting recent infections.

ObjectiveDramatic increases have occurred in sexually transmitted diseases (STD) and in sexual risk behaviour among homosexual men in Amsterdam and internationally. We investigated whether these trends indicate a resurgence of the HIV epidemic. MethodsHIV incidence was determined among homosexual attendees of an STD clinic in Amsterdam, who had participated in semi-annual anonymous unlinked cross-sectional HIV prevalence studies from 1991 to 2001. Stored HIV-seropositive samples were tested with a less-sensitive HIV assay and, if non-reactive, were further tested for the presence of antiretroviral drugs, indicative of the use of highly active antiretroviral therapy. Seropositive men who tested non-reactive on the less-sensitive assay and had not used antiretroviral drugs were classified as recently infected (< 170 days). Annual HIV incidence and its changes were examined. ResultsAmong 3090 homosexual participants (median age 34 years), 454 were HIV infected, of whom 37 were recently infectioned. From 1991 to 2001 the overall incidence was 3.0 infections/100 person-years. Incidence increased over time (P = 0.02) and, strikingly, the increase was evident in older (⩾ 34 years) men (P < 0.01), but not in the young. Of men recently infected, 84% (n = 31) were unaware of their infection and 70.3% (n = 26) had a concurrent STD. These 26 men reportedly had sex with a total of 315 men in the preceding 6 months. ConclusionHIV incidence is increasing among homosexual attendees of an STD clinic. It is imperative to trace recently infected individuals, because they are highly infectious, and can thus play a key role in the spread of HIV.

The hepatitis C epidemic among HIV-positive MSM: incidence estimates from 1990 to 2007

Background:Outbreaks of acute hepatitis C virus (HCV) infection among HIV-infected MSM have been described since 2000. However, phylogenetic analysis suggests that the spread of HCV started around 1996. We estimated the incidence of HCV in HIV-infected MSM with well estimated dates of HIV seroconversion from 1990 to 2007. Methods:Data from 12 cohorts within the Concerted Action on SeroConversion to AIDS and Death in Europe (CASCADE) Collaboration were used. HCV incidence was estimated using standard incidence methods and methods for interval-censored data. We accounted for the fact that routine HCV data collection in each cohort started in different calendar years. Results:Of 4724 MSM, 3014 had an HCV test result and were included. Of these, 124 (4%) had only positive HCV test results, 2798 (93%) had only negative results and 92 (3%) had both. In 1990, HCV incidence ranged from 0.9 to 2.2 per 1000 person-years, depending on the analysis strategy used. HCV incidence increased up to 1995 when it was estimated to range between 5.5 and 8.1 per 1000 person-years. From 2002 onwards, it increased substantially to values between 16.8 and 30.0 per 1000 person-years in 2005 and between 23.4 and 51.1 per 1000 person-years in 2007. Conclusion:Our data support phylodynamic findings that HCV incidence had already increased among HIV-infected MSM from the mid-1990s. However, the main expansion of the HCV epidemic started after 2002. Incidence estimates obtained from cohort studies may help identify changes in the spread of important infections earlier and should guide routine testing policies to minimize further disease burden.

Time From Human Immunodeficiency Virus Seroconversion to Reaching CD4+ Cell Count Thresholds <200, <350, and <500 Cells/mm3: Assessment of Need Following Changes in Treatment Guidelines

BACKGROUND Recent updates of human immunodeficiency virus (HIV) treatment guidelines have raised the CD4+ cell count thresholds for antiretroviral therapy initiation from 350 to 500 cells/mm(3) in the United States and from 200 to 350 cells/mm³ in mid- and low-income countries. Robust data of time from HIV seroconversion to CD4+ cell counts of 200, 350, and 500 cells/mm³ are lacking but are needed to inform health care planners of the likely impact and cost effectiveness of these and possible future changes in CD4+ cell count initiation threshold. METHODS Using Concerted Action on Seroconversion to AIDS and Death in Europe data from individuals with well-estimated dates of HIV seroconversion, we fitted mixed models on the square root of CD4+ cell counts measured before combined antiretroviral therapy (cART) initiation. Restricting analyses to adults (age >16 years), we predicted time between seroconversion and CD4+ cell count <200, <350, and <500 cells/mm³ as well as CD4+ cell count distribution and proportions reaching these thresholds at 1, 2, and 5 years after seroconversion. RESULTS Median (interquartile range [IQR]) follow-up for the 18495 eligible individuals from seroconversion while cART-free was 3.7 years (1.5, 7). Most of the subjects were male (78%), had a median age at seroconversion of 30 years (IQR, 25-37 years), and were infected through sex between men (55%). Estimated median times (95% confidence interval [CI]) from seroconversion to CD4+ cell count <500, <350, and <200 cells/mm(3) were 1.19 (95% CI, 1.12-1.26), 4.19 (95% CI, 4.09-4.28), and 7.93 (95% CI, 7.76-8.09) years, respectively. Almost half of infected individuals would require treatment within 1 year of seroconversion for guidelines recommending its initiation at 500 cells/mm³, compared with 26% and 9% for guidelines recommending initiation at 350 and 200 cells/mm³, respectively. CONCLUSIONS These data suggest substantial increases in the number of individuals who require treatment and call for early HIV testing.

Major decline of hepatitis C virus incidence rate over two decades in a cohort of drug users

Injecting drug users (DU) are at high risk for hepatitis C virus (HCV) and HIV infections. To examine the prevalence and incidence of these infections over a 20-year period (1985–005), the authors evaluated 1276 DU from the Amsterdam Cohort Studies who had been tested prospectively for HIV infection and retrospectively for HCV infection. To compare HCV and HIV incidences, a smooth trend was assumed for both curves over calendar time. Risk factors for HCV seroconversion were determined using Poisson regression. Among ever-injecting DU, the prevalence of HCV antibodies was 84.5% at study entry, and 30.9% were co-infected with HIV. Their yearly HCV incidence dropped from 27.5/100 person years (PY) in the 1980s to 2/100 PY in recent years. In multivariate analyses, ever-injecting DU who currently injected and borrowed needles were at increased risk of HCV seroconversion (incidence rate ratio 29.9, 95% CI 12.6, 70.9) compared to ever-injecting DU who did not currently inject. The risk of HCV seroconversion decreased over calendar time. The HCV incidence in ever-injecting DU was on average 4.4 times the HIV incidence, a pattern seen over the entire study period. The simultaneous decline of both HCV and HIV incidence probably results from reduced risk behavior at the population level.

Cause-Specific Cumulative Incidence Estimation and the Fine and Gray Model Under Both Left Truncation and Right Censoring

Summary The standard estimator for the cause-specific cumulative incidence function in a competing risks setting with left truncated and/or right censored data can be written in two alternative forms. One is a weighted empirical cumulative distribution function and the other a product-limit estimator. This equivalence suggests an alternative view of the analysis of time-to-event data with left truncation and right censoring: individuals who are still at risk or experienced an earlier competing event receive weights from the censoring and truncation mechanisms. As a consequence, inference on the cumulative scale can be performed using weighted versions of standard procedures. This holds for estimation of the cause-specific cumulative incidence function as well as for estimation of the regression parameters in the Fine and Gray proportional subdistribution hazards model. We show that, with the appropriate filtration, a martingale property holds that allows deriving asymptotic results for the proportional subdistribution hazards model in the same way as for the standard Cox proportional hazards model. Estimation of the cause-specific cumulative incidence function and regression on the subdistribution hazard can be performed using standard software for survival analysis if the software allows for inclusion of time-dependent weights. We show the implementation in the R statistical package. The proportional subdistribution hazards model is used to investigate the effect of calendar period as a deterministic external time varying covariate, which can be seen as a special case of left truncation, on AIDS related and non-AIDS related cumulative mortality.