Estimating Sensitivity and Sojourn Time in Screening for Colorectal Cancer A Comparison of Statistical Approaches
TL;DR: Various analytic strategies for fitting exponential models to data from a screening program for colorectal cancer conducted in Calvados, France, between 1991 and 1994 are considered, yielding estimates of mean sojourn time and sensitivity.
Abstract: The effectiveness of cancer screening depends crucially on two elements: the sojourn time (that is, the duration of the preclinical screen-detectable period) and the sensitivity of the screening test. Previous literature on methods of estimating mean sojourn time and sensitivity has largely concentrated on breast cancer screening. Screening for colorectal cancer has been shown to be effective in randomized trials, but there is little literature on the estimation of sojourn time and sensitivity. It would be interesting to demonstrate whether methods commonly used in breast cancer screening could be used in colorectal cancer screening. In this paper, the authors consider various analytic strategies for fitting exponential models to data from a screening program for colorectal cancer conducted in Calvados, France, between 1991 and 1994. The models yielded estimates of mean sojourn time of approximately 2 years for 45- to 54-year-olds, 3 years for 55- to 64-year-olds, and 6 years for 65- to 74-year-olds. Estimates of sensitivity were approximately 75%, 50%, and 40% for persons aged 45-54, 55-64, and 65-74 years, respectively. There is room for improvement in all models in terms of goodness of fit, particularly for the first year after screening, but results from randomized trials indicate that the sensitivity estimates are roughly correct. Am J Epidemiol 1998;148:609-19.
Citations
More filters
TL;DR: An overview of Markov models in cancer screening evaluation and focuses on two specific models, incorporating lymph node involvement as a prognostic factor and a serious limitation to the five-state model are pointed out.
Abstract: This work presents a brief overview of Markov models in cancer screening evaluation and focuses on two specific models. A three-state model was first proposed to estimate jointly the sensitivity of the screening procedure and the average duration in the preclinical phase, i.e. the period when the cancer is asymptomatic but detectable by screening. A five-state model, incorporating lymph node involvement as a prognostic factor, was later proposed combined with a survival analysis to predict the mortality reduction associated with screening. The strengths and limitations of these two models are illustrated using data from French breast cancer service screening programmes. The three-state model is a useful frame but parameter estimates should be interpreted with caution. They are highly correlated and depend heavily on the parametric assumptions of the model. Our results pointed out a serious limitation to the five-state model, due to implicit assumptions which are not always verified. Although it may still be useful, there is a need for more flexible models. Over-diagnosis is an important issue for both models and induces bias in parameter estimates. It can be addressed by adding a non-progressive state, but this may provide an uncertain estimation of over-diagnosis. When the primary goal is to avoid bias, rather than to estimate over-diagnosis, it may be more appropriate to correct for over-diagnosis assuming different levels in a sensitivity analysis. This would be particularly relevant in a perspective of mortality reduction estimation.
50 citations
TL;DR: In most cancer cases detected by a symptom-based programme, the symptoms are caused by cancer, but in a significant minority of cases cancer detection is serendipitous and experiences the benefits of a standard screening programme, a substantial mean lead time and a higher probability of early-stage diagnosis.
Abstract: Symptom lead times in lung and colorectal cancers: what are the benefits of symptom-based approaches to early diagnosis?
48 citations
Cites background from "Estimating Sensitivity and Sojourn ..."
...Mean lead times for these cases can be inferred from mean sojourn time estimates based on screening trials with multiple rounds (Day and Walter, 1984; Duffy et al, 1995; Prevost et al, 1998)....
[...]
...For colorectal cancer a range of estimates between 2 and 6 years have been published (Prevost et al, 1998; Brenner et al, 2011; Zheng and Rutter, 2012) with an average of B4 years....
[...]
TL;DR: The new estimates indicate that screening detectable phase is longer than that found in previous mammography trials/programmes, but also that the sensitivity of the screening test is lower, which indicates that the NBCSP detects more cancer cases than most previous trials/ programs.
Abstract: Objective: To assess if new screening techniques, increased use of hormone replacement therapy, or the transition from breast cancer screening trials to large scale screening programmes may influence the average time in preclinical screening detectable phase (mean sojourn time [MST]) or screening test sensitivity (STS).Setting: Screening and interval data for 395,188 women participating in the Norwegian Breast Cancer Screening Programme (NBCSP).Methods: Weighted non-linear least-square regression estimates using a tree step Markov chain model, and a sensitivity analysis of the possible impact by opportunistic screening between ordinary breast cancer screening rounds.Results: MST was estimated to 6.1 (95% confidence interval [CI] 5.1–7.0) years for women aged 50–59 years, and 7.9 (95% CI 6.0–7.9) years for those aged 60–69 years. Correspondingly, STS was estimated to 58% (95% CI 52–64 %) and 73 % (67–78 %), respectively. Simulations revealed that opportunistic screening may give a moderate estimation bias ...
42 citations
TL;DR: Symptom-based investigation would lengthen lead times and result in earlier stage at diagnosis in a small proportion of cases, but would be far less effective than standard screening targeted at smokers.
Abstract: Background: Before their diagnosis, patients with cancer present in primary care more frequently than do matched controls. This has raised hopes that earlier investigation in primary care could lead to earlier stage at diagnosis. Methods: We re-analysed primary care symptom data collected from 247 lung cancer cases and 1235 matched controls in Devon, UK. We identified the most sensitive and specific definition of symptoms, and estimated its incidence in cases and controls prior to diagnosis. We estimated the symptom lead time (SLT) distribution (the time between symptoms attributable to cancer and diagnosis), taking account of the investigations already carried out in primary care. The impact of route of diagnosis on stage at diagnosis was also examined. Results: Symptom incidence in cases was higher than in controls 2 years before diagnosis, accelerating markedly in the last 6 months. The median SLT was under 3 months, with mean 5.3 months [95% credible interval (CrI) 4.5‐6.1] and did not differ by stage at diagnosis. An earlier stage at diagnosis was observed in patients identified through chest X-ray originated in primary care. Conclusions: Most symptoms preceded clinical diagnosis by only a few months. Symptom-based investigation would lengthen lead times and result in earlier stage at diagnosis in a small proportion of cases, but would be far less effective than standard screening targeted at smokers.
41 citations
Cites background from "Estimating Sensitivity and Sojourn ..."
...Negatively correlated estimates of mean sojourn time and test sensitivity can be jointly derived from screening trials with two or more screening rounds.(13) Estimates of mean sojourn time in lung cancer have ranged from 1....
[...]
01 Sep 2020
TL;DR: Heterogeneity in breast cancer defies assumptions necessary for screening mammography in its current form to be maximally effective; strategies beyond routine screenings mammography are needed to prevent, detect, and avert deaths from the more lethal interval breast cancers.
Abstract: Importance Breast cancer comprises a highly heterogeneous group of diseases. Many breast cancers, particularly the more lethal ones, may not satisfy the assumptions about biology and natural history of breast cancer necessary for screening mammography to be effective. Objectives To compare tumor characteristics of breast cancers diagnosed within 2 years of a normal screening mammogram (interval breast cancer [IBC]) with those of screen-detected breast cancers (SBC) and to compare breast cancer-specific mortality of IBC with SBC. Design, Setting, and Participants In this registry-based cohort study, we collected data about relevant tumor- and patient-related variables on women diagnosed with breast cancer between January 2004 and June 2010 who participated in the population-based screening program in Manitoba, Canada, and those diagnosed with breast cancer outside the screening program in the province. We performed multinomial logistic regression analysis to assess tumor and patient characteristics associated with a diagnosis of IBC compared with SBC. Competing risk analysis was performed to examine risk of death by cancer detection method. Exposures Breast cancer diagnosis. Main Outcomes and Measures Differences in tumor characteristics and breast cancer-specific mortality. Results A total of 69 025 women aged 50 to 64 years had 212 screening mammograms during the study period. There were 1687 breast cancer diagnoses (705 SBC, 206 IBC, 275 were noncompliant, and 501 were detected outside the screening program), and 225 deaths (170 breast cancer-specific deaths). Interval cancers were more likely than SBC to be of high grade and estrogen receptor negative (odds ratio [OR], 6.33; 95% CI, 3.73-10.75; P < .001; and OR, 2.88; 95% CI, 2.01-4.13; P < .001, respectively). After a median follow-up of 7 years, breast cancer-specific mortality was significantly higher for IBC compared with SBC cancers (hazard ratio [HR] 3.55; 95% CI, 2.01-6.28; P < .001), for a sojorn time of 2 years. Non-breast cancer mortality was similar between IBC and SBC (HR, 1.33; 95% CI, 0.43-4.15). Conclusions and Relevance In this cohort study, interval cancers were highly prevalent in women participating in population screening, represented a worse biology, and had a hazard for breast cancer death more than 3-fold that for SBC. Strategies beyond current mammographic screening practices are needed to reduce incidence, improve detection, and reduce deaths from these potentially lethal breast cancers.
39 citations
References
More filters
TL;DR: The focus is on applied inference for Bayesian posterior distributions in real problems, which often tend toward normal- ity after transformations and marginalization, and the results are derived as normal-theory approximations to exact Bayesian inference, conditional on the observed simulations.
Abstract: The Gibbs sampler, the algorithm of Metropolis and similar iterative simulation methods are potentially very helpful for summarizing multivariate distributions. Used naively, however, iterative simulation can give misleading answers. Our methods are simple and generally applicable to the output of any iterative simulation; they are designed for researchers primarily interested in the science underlying the data and models they are analyzing, rather than for researchers interested in the probability theory underlying the iterative simulations themselves. Our recommended strategy is to use several independent sequences, with starting points sampled from an overdispersed distribution. At each step of the iterative simulation, we obtain, for each univariate estimand of interest, a distributional estimate and an estimate of how much sharper the distributional estimate might become if the simulations were continued indefinitely. Because our focus is on applied inference for Bayesian posterior distributions in real problems, which often tend toward normality after transformations and marginalization, we derive our results as normal-theory approximations to exact Bayesian inference, conditional on the observed simulations. The methods are illustrated on a random-effects mixture model applied to experimental measurements of reaction times of normal and schizophrenic patients.
13,884 citations
TL;DR: In this paper, three sampling-based approaches, namely stochastic substitution, the Gibbs sampler, and the sampling-importance-resampling algorithm, are compared and contrasted in relation to various joint probability structures frequently encountered in applications.
Abstract: Stochastic substitution, the Gibbs sampler, and the sampling-importance-resampling algorithm can be viewed as three alternative sampling- (or Monte Carlo-) based approaches to the calculation of numerical estimates of marginal probability distributions. The three approaches will be reviewed, compared, and contrasted in relation to various joint probability structures frequently encountered in applications. In particular, the relevance of the approaches to calculating Bayesian posterior densities for a variety of structured models will be discussed and illustrated.
6,294 citations
Journal Article•
TL;DR: Stochastic substitution, the Gibbs sampler, and the sampling-importance-resampling algorithm can be viewed as three alternative sampling- (or Monte Carlo-) based approaches to the calculation of numerical estimates of marginal probability distributions.
Abstract: Stochastic substitution, the Gibbs sampler, and the sampling-importance-resampling algorithm can be viewed as three alternative sampling- (or Monte Carlo-) based approaches to the calculation of numerical estimates of marginal probability distributions. The three approaches will be reviewed, compared, and contrasted in relation to various joint probability structures frequently encountered in applications. In particular, the relevance of the approaches to calculating Bayesian posterior densities for a variety of structured models will be discussed and illustrated.
6,223 citations
Book•
[...]
01 Jan 1996
TL;DR: Mathematica has defined the state of the art in technical computing for over a decade, and has become a standard in many of the world's leading companies and universities as discussed by the authors.
Abstract: From the Publisher:
Mathematica has defined the state of the art in technical computing for over a decade, and has become a standard in many of the world's leading companies and universities From simple calculator operations to large-scale programming and the preparation of interactive documents, Mathematica is the tool of choice
3,566 citations
[...]
01 Jan 1996
TL;DR: From the Publisher: Mathematica has defined the state of the art in technical computing for over a decade, and has become a standard in many of the world's leading companies and universities.
Abstract: From the Publisher:
Mathematica has defined the state of the art in technical computing for over a decade, and has become a standard in many of the world's leading companies and universities. From simple calculator operations to large-scale programming and the preparation of interactive documents, Mathematica is the tool of choice.
3,115 citations