Flexible Estimation of Differences in Treatment-Specific Recurrent Event Means in the Presence of a Terminating Event

Summary .  In this article, we consider the setting where the event of interest can occur repeatedly for the same subject (i.e., a recurrent event; e.g., hospitalization) and may be stopped permanently by a terminating event (e.g., death). Among the different ways to model recurrent/terminal event data, the marginal mean (i.e., averaging over the survival distribution) is of primary interest from a public health or health economics perspective. Often, the difference between treatment-specific recurrent event means will not be constant over time, particularly when treatment-specific differences in survival exist. In such cases, it makes more sense to quantify treatment effect based on the cumulative difference in the recurrent event means, as opposed to the instantaneous difference in the rates. We propose a method that compares treatments by separately estimating the survival probabilities and recurrent event rates given survival, then integrating to get the mean number of events. The proposed method combines an additive model for the conditional recurrent event rate and a proportional hazards model for the terminating event hazard. The treatment effects on survival and on recurrent event rate among survivors are estimated in constructing our measure and explain the mechanism generating the difference under study. The example that motivates this research is the repeated occurrence of hospitalization among kidney transplant recipients, where the effect of expanded criteria donor (ECD) compared to non-ECD kidney transplantation on the mean number of hospitalizations is of interest.     

Penalized survival models for the analysis of alternating recurrent event data

Recurrent event data are widely encountered in clinical and observational studies. Most methods for recurrent events treat the outcome as a point process and, as such, neglect any associated event duration. This generally leads to a less informative and potentially biased analysis. We propose a joint model for the recurrent event rate (of incidence) and duration. The two processes are linked through a bivariate normal frailty. For example, when the event is hospitalization, we can treat the time to admission and length-of-stay as two alternating recurrent events. In our method, the regression parameters are estimated through a penalized partial likelihood, and the variance-covariance matrix of the frailty is estimated through a recursive estimating formula. Moreover, we develop a likelihood ratio test to assess the dependence between the incidence and duration processes. Simulation results demonstrate that our method provides accurate parameter estimation, with a relatively fast computation time. We illustrate the methods through an analysis of hospitalizations among end-stage renal disease patients.     

Outcomes of Renal Transplantation in HIV-1 Associated Nephropathy

Introduction

Several studies have demonstrated that renal transplantation in HIV positive patients is both safe and effective. However, none of these studies have specifically examined outcomes in patients with HIV-associated nephropathy (HIVAN).

Methods

Medical records of all HIV-infected patients who underwent kidney transplantation at Johns Hopkins Hospital between September 2006 and January 2014 were reviewed. Data was collected to examine baseline characteristics and outcomes of transplant recipients with HIVAN defined pathologically as collapsing focal segmental glomerulosclerosis (FSGS) with tubulo-interstitial disease.

Results and Discussion

During the study period, a total of 16 patients with HIV infection underwent renal transplantation. Of those, 11 patients were identified to have biopsy-proven HIVAN as the primary cause of their end stage renal disease (ESRD) and were included in this study. They were predominantly African American males with a mean age of 47.6 years. Seven (64%) patients developed delayed graft function (DGF), and 6 (54%) patients required post-operative dialysis within one week of transplant. Graft survival rates at 1 and 3 years were 100% and 81%, respectively. Acute rejection rates at 1 and 3 years were 18% and 27%, respectively. During a mean follow up of 3.4 years, one patient died.

Conclusions

Acute rejection rates in HIVAN patients in this study are higher than reported in the general ESRD population, which is similar to findings from prior studies of patients with HIV infection and ESRD of various causes. The high rejection rates appear to have no impact on short or intermediate term graft survival.     

The separation of timescales in Bayesian survival modeling of the time-varying effect of a time-dependent exposure

In this paper, we apply flexible Bayesian survival analysis methods to investigate the risk of lymphoma associated with kidney transplantation among patients with end-stage renal disease. Of key interest is the potentially time-varying effect of a time-dependent exposure: transplant status. Bayesian modeling of the baseline hazard and the effect of transplant requires consideration of 2 timescales: time since study start and time since transplantation, respectively. Previous related work has not dealt with the separation of multiple timescales. Using a hierarchical model for the hazard function, both timescales are incorporated via conditionally independent stochastic processes; smoothing of each process is specified via intrinsic conditional Gaussian autoregressions. Features of the corresponding posterior distribution are evaluated from draws obtained via a Metropolis-Hastings-Green algorithm.     

Clinical Value of Cadaveric Renal Homotransplantation

Cadaveric renal transplantation was found to be potentially applicable to 80% of a uremic population as observed over a six-year period. Death prior to presentation of a donor kidney occurred in 25%, 40% have received transplants and 15% are awaiting transplantation. Transplantation resulted in restoration of near normal renal function and homograft survival rates of 45% at one year, 40% at two years and 17% at four years. Failure of the therapeutic procedure resulted more often from death of the patient than from failure of the donor organ. Patient death was most frequently ascribed to complications of the immunosuppressive therapy, but cardiovascular accidents were also a significant cause. Early renal failure was due to hyperacute or acute rejection, while the cause of late renal failure remains unproved. Re-transplantation was effective treatment for late failure and, as a result, the four-year patient survival is nearly 40% compared to four-year initial kidney survival of 17%.     

Kidney transplantation in Canada, 1981-84.

The Canadian Renal Failure Register was established in 1980. Data have been collected annually for all Canadian patients in whom irreversible kidney failure developed and who required dialysis or transplantation. The authors present actuarial patient and graft survival rates for 1981-84. In 1984, patients with a functioning renal graft accounted for 43.9% of the patients with end-stage renal disease. The number of transplants performed increased from 482 in 1981 to 662 in 1984; however, 1,022 patients undergoing dialysis (25.2%) were on an active waiting list for a transplant at the end of 1984. Greater effort is needed to increase the transplantation rate.     

肾移植术后人微小病毒B19 感染导致纯红细胞增生障碍性贫血

目的:探讨肾移植术后人微小病毒B19感染所致纯红细胞再生障碍性贫血的诊断及其治疗。方法:回顾性分析4例肾移植术后纯红细胞再生障碍性贫血患者的临床特点,诊断方法,治疗过程及预后。结果:两周内在本中心接受肾移植手术的6例患者中,有4例在术后60天内均出现发热、血红蛋白进行性下降等相似症状;综合骨髓穿刺、ELISA方法检测血清特异性IgG、IgM等方法诊断为人微小病毒B19感染;经静脉注射免疫球蛋白、调整免疫抑制方案等综合治疗后,4例患者病情均明显缓解。结论:(1)贫血是肾移植术后患者感染人微小病毒B19的典型临床症状;(2)PCR检测和/或ELISA方法,结合骨髓穿刺及其他实验室指标可诊断人微小病毒感染;(3)静脉注射免疫球蛋白是肾移植术后人微小病毒B19感染导致PRCA的首选治疗方法,病情反复时,再次应用仍然有效。同时予以调整免疫抑制剂方案等综合治疗,可获得理想疗效。     

An approach to model clustered survival data with dependent censoring

In this study we introduce a likelihood-based method, via the Weibull and piecewise exponential distributions, capable of accommodating the dependence between failure and censoring times. The methodology is developed for the analysis of clustered survival data and it assumes that failure and censoring times are mutually independent conditional on a latent frailty. The dependent censoring mechanism is accounted through the frailty effect and this is accomplished by means of a key parameter accommodating the correlation between failure and censored observations. The full specification of the likelihood in our work simplifies the inference procedures with respect to Huang and Wolfe since it reduces the computation burden of working with the profile likelihood. In addition, the assumptions made for the baseline distributions lead to models with continuous survival functions. In order to carry out inferences, we devise a Monte Carlo EM algorithm. The performance of the proposed models is investigated through a simulation study. Finally, we explore a real application involving patients from the Dialysis Outcomes and Practice Patterns Study observed between 1996 and 2015.     

Semiparametric Analysis for Recurrent Event Data with Time-Dependent Covariates and Informative Censoring

Summary .  Recurrent event data analyses are usually conducted under the assumption that the censoring time is independent of the recurrent event process. In many applications the censoring time can be informative about the underlying recurrent event process, especially in situations where a correlated failure event could potentially terminate the observation of recurrent events. In this article, we consider a semiparametric model of recurrent event data that allows correlations between censoring times and recurrent event process via frailty. This flexible framework incorporates both time-dependent and time-independent covariates in the formulation, while leaving the distributions of frailty and censoring times unspecified. We propose a novel semiparametric inference procedure that depends on neither the frailty nor the censoring time distribution. Large sample properties of the regression parameter estimates and the estimated baseline cumulative intensity functions are studied. Numerical studies demonstrate that the proposed methodology performs well for realistic sample sizes. An analysis of hospitalization data for patients in an AIDS cohort study is presented to illustrate the proposed method.     

Pre-operative risk factors predict post-operative respiratory failure after liver transplantation

Objective

Post-operative pulmonary complications significantly affect patient survival rates, but there is still no conclusive evidence regarding the effect of post-operative respiratory failure after liver transplantation on patient prognosis. This study aimed to predict the risk factors for post-operative respiratory failure (PRF) after liver transplantation and the impact on short-term survival rates.

Design

The retrospective observational cohort study was conducted in a twelve-bed adult surgical intensive care unit in northern Taiwan. The medical records of 147 liver transplant patients were reviewed from September 2002 to July 2007. Sixty-two experienced post-operative respiratory failure while the remaining 85 patients did not.

Measurements and Main Results

Gender, age, etiology, disease history, pre-operative ventilator use, molecular adsorbent re-circulating system (MARS) use, source of organ transplantation, model for end-stage liver disease score (MELD) and Child-Turcotte-Pugh score calculated immediately before surgery were assessed for the two groups. The length of the intensive care unit stay, admission duration, and mortality within 30 days, 3 months, and 1 year were also evaluated. Using a logistic regression model, post-operative respiratory failure correlated with diabetes mellitus prior to liver transplantation, pre-operative impaired renal function, pre-operative ventilator use, pre-operative MARS use and deceased donor source of organ transplantation (p<0.05). Once liver transplant patients developed PRF, their length of ICU stay and admission duration were prolonged, significantly increasing their mortality and morbidity (p<0.001).

Conclusions

The predictive pre-operative risk factors significantly influenced the occurrence of post-operative respiratory failure after liver transplantation.     

Integrated dialysis and renal transplantation: small is beautiful.

Many patients in Britain with chronic renal failure suitable for renal replacement treatment die because not enough treatment facilities are available. Moreover, the number of renal transplants performed is insufficient to meet even present needs, so the number of patients on dialysis is rising. The integrated dialysis and transplant unit in Aberdeen, which has a population base much smaller than the average British unit, meets community needs for dialysis and transplantation. The problem of harvesting cadaver kidneys has been solved; the present supply has not only enabled the number of patients on dialysis to remain stable but has resulted in a net export of kidneys. The Aberdeen unit shows how estimated needs for chronic dialysis and renal transplantation may be met.     

Dialysis and transplantation among Aboriginal children with kidney failure

Background:

Relatively little is known about the management and outcomes of Aboriginal children with renal failure in Canada. We evaluated differences in dialysis modality, time spent on dialysis, rates of kidney transplantation, and patient and allograft survival between Aboriginal children and non-Aboriginal children.

Methods:

For this population-based cohort study, we used data from a national pediatric end-stage renal disease database. Patients less than 18 years old who started renal replacement treatment (dialysis or kidney transplantation) in nine Canadian provinces (Quebec data were not available) and all three territories between 1992 and 2007 were followed until death, loss to follow-up or end of the study period. We compared initial modality of dialysis and time to first kidney transplant between Aboriginal children, white children and children of other ethnicity. We examined the association between ethnicity and likelihood of kidney transplantation using adjusted Cox proportional hazard models for Aboriginal and white children (data for the children of other ethnicity did not meet the assumptions of proportional hazards).

Results:

Among 843 pediatric patients included in the study, 104 (12.3%) were Aboriginal, 521 (61.8%) were white, and 218 (25.9%) were from other ethnic minorities. Hemodialysis was the initial modality of dialysis for 48.0% of the Aboriginal patients, 42.7% of the white patients and 62.6% of those of other ethnicity (p < 0.001). The time from start of dialysis to first kidney transplant was longer among the Aboriginal children (median 1.75 years, interquartile range 0.69–2.81) than among the children in the other two groups (p < 0.001). After adjustment for confounders, Aboriginal children were less likely than white children to receive a transplant from a living donor (hazard ratio [HR] 0.36, 95% confidence interval [CI] 0.21–0.61) or a transplant from any donor (HR 0.54, 95% CI 0.40–0.74) during the study period.

Interpretation:

The time from start of dialysis to first kidney transplant was longer among Aboriginal children than among white children. Further evaluation is needed to determine barriers to transplantation among Aboriginal children.Compared with non-Aboriginal people, Aboriginal adults with end-stage renal disease in Canada have lower rates of kidney transplantation, the optimal treatment for renal failure.^1–4 Most studies to date that have examined health outcomes among Canadian Aboriginal people with kidney disease have focused on adults.^1–8 Relatively little is known about the outcomes among Aboriginal children with renal failure. A single-centre cohort study from the province of British Columbia reported that Aboriginal children who received a kidney transplant had similar short-term, but poorer long-term allograft survival than white children.⁹ No further studies have examined differences in modality of renal replacement treatment or the likelihood of kidney transplantation among Aboriginal children with renal failure.We performed an observational cohort study of children beginning renal replacement treatment in Canada. We compared differences in dialysis modality, time spent on dialysis, rates of kidney transplantation, and graft and patient survival between Aboriginal children, white children and children of other ethnicities.     

Cardiac transplantation in severely ill patients requiring intensive support in hospital

Sixty four patients were referred for cardiac transplantation from a single cardiac team at this hospital between October 1984 and December 1986. Of these patients, 33 were referred for urgent transplantation, all of whom required intensive treatment in hospital with intravenous infusions of cardiac drugs, intra-aortic balloon counterpulsation, peritoneal dialysis, ventilation, or any combination of these to sustain life. Of these 33 patients, six died while awaiting transplantation, one was removed from the waiting list for a transplant, and 26 received cardiac transplants. There were five deaths within 24 hours of operation and one death 10 days after the operation. Twenty of those who had surgery had a successful outcome of transplantation, but there was one late death 10 weeks postoperatively and a further death 31 months after surgery. Eighteen patients were alive and well 10 to 33 months (mean 19·4 months) after transplantation, with an overall survival rate after surgery of 69%.Provided that surgery can be performed before renal failure has progressed such that renal transplantation is necessary, the results are excellent (surgical survival 85·5%) and, we believe, justify the expenditure and staffing requirements necessary to treat these terminally ill patients.     

Proportional Hazards Regression for the Analysis of Clustered Survival Data from Case–Cohort Studies

Summary Case–cohort sampling is a commonly used and efficient method for studying large cohorts. Most existing methods of analysis for case–cohort data have concerned the analysis of univariate failure time data. However, clustered failure time data are commonly encountered in public health studies. For example, patients treated at the same center are unlikely to be independent. In this article, we consider methods based on estimating equations for case–cohort designs for clustered failure time data. We assume a marginal hazards model, with a common baseline hazard and common regression coefficient across clusters. The proposed estimators of the regression parameter and cumulative baseline hazard are shown to be consistent and asymptotically normal, and consistent estimators of the asymptotic covariance matrices are derived. The regression parameter estimator is easily computed using any standard Cox regression software that allows for offset terms. The proposed estimators are investigated in simulation studies, and demonstrated empirically to have increased efficiency relative to some existing methods. The proposed methods are applied to a study of mortality among Canadian dialysis patients.     

Semiparametric analysis of recurrent events data in the presence of dependent censoring

Dependent censoring occurs in longitudinal studies of recurrent events when the censoring time depends on the potentially unobserved recurrent event times. To perform regression analysis in this setting, we propose a semiparametric joint model that formulates the marginal distributions of the recurrent event process and dependent censoring time through scale-change models, while leaving the distributional form and dependence structure unspecified. We derive consistent and asymptotically normal estimators for the regression parameters. We also develop graphical and numerical methods for assessing the adequacy of the proposed model. The finite-sample behavior of the new inference procedures is evaluated through simulation studies. An application to recurrent hospitalization data taken from a study of intravenous drug users is provided.     

Equity of access to renal transplant waiting list and renal transplantation in Scotland: cohort study

Objective To examine the access to the renal transplant waiting list and renal transplantation in Scotland.Design Cohort study.Setting Renal and transplant units in Scotland.Participants 4523 adults starting renal replacement therapy in Scotland between 1 January 1989 and 31 December 1999.Main outcome measures Impact of age, sex, social deprivation, primary renal disease, renal or transplant unit, and geography on access to the waiting list and renal transplantation.Results 1736 of 4523 (38.4%) patients were placed on the waiting list for renal transplantation and 1095 (24.2%) underwent transplantation up to 31 December 2000, the end of the study period. Patients were less likely to be placed on the list if they were female, older, had diabetes, were in a high deprivation category, and were treated in a renal unit in a hospital with no transplant unit. Patients living furthest away from the transplant centre were listed more quickly. The only factors governing access to transplantation once on the list were age, primary renal disease, and year of listing. A significant centre effect was found in access to the waiting list and renal transplantation.Conclusions A major disparity exists in access to the renal transplant waiting list and renal transplantation in Scotland. Comorbidity may be an important factor.     

Renal Transplant Aspiration Cytology. Role For Simple Morphological Criteria

Fine-needle aspiration (FNA) is a well-recognized technique for sampling solid organs. It is used in renal transplantation to clarify the cause of a poorly functioning graft. Differential scoring techniques with respect to peripheral blood cell populations, and immunocytochemistry have been employed in this context. We describe the use of simple morphological criteria alone in renal transplant FNA. We compare these with needle biopsy and clinical parameters and show their value in the detection of active cellular rejection. Their limitations are discussed within the framework of other patterns of transplant pathology.     

Renal replacement treatment for diabetic patients in Newcastle upon Tyne and the Northern region, 1964-88.

OBJECTIVES--To review the experience of renal replacement treatment in diabetic patients treated in Newcastle upon Tyne and the Northern region from 1964 to 1988, and to compare the morbidity and mortality of diabetic patients treated with dialysis or transplantation with those of matched controls of non-diabetic patients. DESIGN--Retrospective study of clinical case notes. SETTING--Renal units of the Northern region, particularly that in Newcastle upon Tyne. PATIENTS--All 65 diabetic patients treated by renal replacement treatment in Newcastle upon Tyne from 1964 to 1987; 42 diabetic patients were matched with 42 non-diabetic patients according to age, sex, year of starting treatment, and type of treatment (dialysis or transplantation). MAIN OUTCOME MEASURES--Sex, age, renal biopsy findings, blood pressure, history of diabetic treatment, and plasma creatinine concentration at the start of renal replacement treatment. History of renal replacement treatments, suitability for transplantation, history of transplantation, cumulative survival, and cause of death during follow up. Survival of technique, cumulative survival of the first peritoneal catheter and history of peritonitis in patients treated with continuous ambulatory peritoneal dialysis; source of graft, histocompatibility antigens, duration of associated stay in hospital, and graft survival in patients receiving renal or pancreatic transplant. RESULTS--1259 Patients with chronic renal failure were accepted for renal replacement treatment in Newcastle upon Tyne, of whom 65 (5%) had diabetes. The first was accepted in 1974, and between 1974 and 1980 another 15 were treated (mean age 42 years; 4% of new patients). From 1981 to 1987, 49 diabetic patients (mean age 44; 9% of new patients) were treated. Fifty patients (77%) had insulin dependent diabetes and the remaining 15 (23%) non-insulin dependent diabetes. On average, the patients were aged 25 (range 5-57) when diabetes was first diagnosed and 44 (range 24-70) at the start of renal replacement treatment. The mean age at the start of treatment was 40 for patients with non-insulin dependent diabetes and 58 for patients with non-insulin dependent diabetes. Transplantation was performed in 33 of the diabetic patients, whose mean age was lower than that of those who did not receive a transplant (41 v 48 respectively, p less than 0.05). Comparison between the 42 diabetic patients and matched controls showed that the overall survival at five years was 46% and 77% respectively. The three year survival of the diabetic patients who did not receive a transplant was poor (41% v 79% respectively). Of patients transplanted, survival at five years was 73% in the diabetic patients and 90% in the controls. However, there was no significant difference in the five year graft survival (64% v 46% respectively). CONCLUSIONS--Diabetes adversely affects morbidity and mortality in patients having renal replacement treatment, but renal transplantation seems to be the best option for treating diabetic patients with end stage renal failure.     

Shared frailty models for recurrent events and a terminal event

There has been an increasing interest in the analysis of recurrent event data (Cook and Lawless, 2002, Statistical Methods in Medical Research 11, 141-166). In many situations, a terminating event such as death can happen during the follow-up period to preclude further occurrence of the recurrent events. Furthermore, the death time may be dependent on the recurrent event history. In this article we consider frailty proportional hazards models for the recurrent and terminal event processes. The dependence is modeled by conditioning on a shared frailty that is included in both hazard functions. Covariate effects can be taken into account in the model as well. Maximum likelihood estimation and inference are carried out through a Monte Carlo EM algorithm with Metropolis-Hastings sampler in the E-step. An analysis of hospitalization and death data for waitlisted dialysis patients is presented to illustrate the proposed methods. Methods to check the validity of the proposed model are also demonstrated. This model avoids the difficulties encountered in alternative approaches which attempt to specify a dependent joint distribution with marginal proportional hazards and yields an estimate of the degree of dependence.     

Results of 250 consecutive cadaver kidney transplants.

The results of 250 cadaver kidney transplants performed at this hospital between May 1968 and December 1974 were analysed. A functioning transplant was defined strictly as one that maintained the recipient in good health with a serum creatinine below 442 mumol/1 (5 mg/100 ml) without any need for dialysis. The proportions of kidneys functioning after one, two, and three years were 40-4%, 33-9%, and 31-1% respectively, the corresponding survival rates of patients being 62-6%, 57-4%, and 43-8%. The primary failure rate overall was 21-6%, while the failure rates for first, second, and third transplants were 18-1%, 39-9%, and 66-7% respectively. Half of the primary failures were attributed to the use of cadaver kidneys with abnormal vasculature or long ischaemic times or originating from non-ventilated donors. Of the initially successful transplants 49% were subsequently lost due to rejection (53%) or death of the patient with a functioning transplant (46%), and of the secondary losses 58% occurred within three months of transplantation. HLA matching of donor and recipient for two or more antigens was associated with a significant increase in transplant survival--46% at three years as opposed to 9-5% at three years for kidneys with poorer matches.