Feasibility study of adjustment for contamination and non-compliance in a prostate cancer screening trial

Instrumental variable estimation of early treatment effect in randomized screening trials

The primary analysis of randomized screening trials for cancer typically adheres to the intention-to-screen principle, measuring cancer-specific mortality reductions between screening and control arms. These mortality reductions result from a combination of the screening regimen, screening technology and the effect of the early, screening-induced, treatment. This motivates addressing these different aspects separately. Here we are interested in the causal effect of early versus delayed treatments on cancer mortality among the screening-detectable subgroup, which under certain assumptions is estimable from conventional randomized screening trial using instrumental variable type methods. To define the causal effect of interest, we formulate a simplified structural multi-state model for screening trials, based on a hypothetical intervention trial where screening detected individuals would be randomized into early versus delayed treatments. The cancer-specific mortality reductions after screening detection are quantified by a cause-specific hazard ratio. For this, we propose two estimators, based on an estimating equation and a likelihood expression. The methods extend existing instrumental variable methods for time-to-event and competing risks outcomes to time-dependent intermediate variables. Using the multi-state model as the basis of a data generating mechanism, we investigate the performance of the new estimators through simulation studies. In addition, we illustrate the proposed method in the context of CT screening for lung cancer using the US National Lung Screening Trial data.

Contamination: How much can an individually randomized trial tolerate?

Cluster randomization results in an increase in sample size compared to individual randomization, referred to as an efficiency loss. This efficiency loss is typically presented under an assumption of no contamination in the individually randomized trial. An alternative comparator is the sample size needed under individual randomization to detect the attenuated treatment effect due to contamination. A general framework is provided for determining the extent of contamination that can be tolerated in an individually randomized trial before a cluster randomized design yields a larger sample size. Results are presented for a variety of cluster trial designs including parallel arm, stepped-wedge and cluster crossover trials. Results reinforce what is expected: individually randomized trials can tolerate a surprisingly large amount of contamination before they become less efficient than cluster designs. We determine the point at which the contamination means an individual randomized design to detect an attenuated effect requires a larger sample size than cluster randomization under a nonattenuated effect. This critical rate is a simple function of the design effect for clustering and the design effect for multiple periods as well as design effects for stratification or repeated measures under individual randomization. These findings are important for pragmatic comparisons between a novel treatment and usual care as any bias due to contamination will only attenuate the true treatment effect. This is a bias that operates in a predictable direction. Yet, cluster randomized designs with post-randomization recruitment without blinding, are at high risk of bias due to the differential recruitment across treatment arms. This sort of bias operates in an unpredictable direction. Thus, with knowledge that cluster randomized trials are generally at a greater risk of biases that can operate in a nonpredictable direction, results presented here suggest that even in situations where there is a risk of contamination, individual randomization might still be the design of choice.

Noncompliance in cancer screening trials

Background Randomized trials evaluating new cancer screening technologies may underestimate the efficacy of screening to reduce cancer mortality if study participants are noncompliant. Participants may fail to comply with the screening itself or fail to obtain appropriate diagnostic follow-up and treatment. Noncompliance with screening has drawn wide attention, but little attention has been paid to noncompliance with diagnostic follow-up and treatment.

Purpose To examine the importance of noncompliance with screening, follow-up, and treatment in cancer screening trials.

Methods The unique problems associated with noncompliance in screening trials are described and provide an example illustrating the potential impact of noncompliance in a screening trial. I discuss issues that arise with measurement of follow-up and therapeutic noncompliance, and the benefit of collecting information on health system and participant characteristics associated with noncompliance.

Results The estimate of the efficacy of a screening program on cancer mortality can be adjusted for screening, follow-up, and treatment noncompliance. Noncompliance needs to be measured in a rigorous, systematic manner across all arms of the trial. Information on health system and participant characteristics associated with compliance may also be incorporated into statistical models to estimate screening effects with full compliance, plan interventions to increase compliance, and extrapolate results of screening trials from one population to another.

Limitations Measuring compliance with follow-up and treatment can be difficult when these occur outside the trial, and when there is variation among providers in follow-up and treatment practices.

Conclusions Noncompliance may alter the estimate of a screening effect on cancer mortality in clinical trials. It is possible to adjust screening efficacy estimates for noncompliance using existing statistical techniques. It is important that data describing compliance with screening, follow-up, and treatment are collected as part of standard data collection in cancer screening trials. Clinical Trials 2007; 4: 341—349. http://ctj.sagepub.com

Adjustment for compliance behavior in trials of epidural analgesia in labor using instrumental variable meta-analysis

OBJECTIVES

Intention-to-treat (ITT) analysis of randomized controlled trials (RCTs) may cause bias when compliance is poor. Noncompliance describes failure to comply with allocation in the intervention arm, and contamination describes uptake of the intervention in the control arm. Instrumental variable (IV) analysis can be applied in addition to the primary ITT analysis to estimate the causal effect adjusted for noncompliance and contamination, assuming that noncompliers would have had the same treatment benefit as compliers. We aimed to compare ITT and IV meta-analysis of the association between epidural analgesia in labor and cesarean section.

STUDY DESIGN AND SETTING

The study was restricted to 27 trials in a Cochrane Systematic Review. The association between epidural analgesia in labor and cesarean section was calculated using ITT and IV analyses. Pooled risk ratios (RRs) were calculated using fixed-effects meta-analysis.

RESULTS

In 18 trials with compliance data, noncompliance was 23% and contamination was 27%. In 10 trials with outcome data stratified by compliance, the pooled RR for cesarean section following epidural analgesia was 1.37 [95% confidence interval (CI): 1.00, 1.89; P = 0.049] using IV compared with 1.19 (95% CI: 0.93, 1.51; P = 0.16) using ITT.

CONCLUSION

ITT meta-analysis underestimates the effect of receiving epidural analgesia in labor on cesarean section compared with IV meta-analysis.

Efficacy versus effectiveness study design within the European screening trial for prostate cancer: consequences for cancer incidence, overall mortality and cancer-specific mortality

OBJECTIVE

To assess the impact of different study designs on outcome data within the European Randomized Study of Screening for Prostate Cancer (ERSPC).

METHODS

Observed data from the Gothenburg centre (effectiveness trial with upfront randomization before informed consent) and the Rotterdam centre (efficacy trial with randomization after informed consent) were compared with expected data, which were retrieved from national cancer registries and life tables. Endpoints were 11-year cumulative prostate cancer (PC) incidence, overall mortality and PC-specific mortality.

RESULTS

In Gothenburg, the 11-year PC incidence was higher than predicted (5.8%) in both the intervention (12.4%) and control arms (7.3%). The observed overall mortality was higher than predicted (15.9%) in both the intervention (17.8%) and control arms (18.5%). The observed PC-specific mortality in the intervention arm was 0.56% versus 0.83% in the control arm, while the expected mortality was 0.83%. In Rotterdam, the observed PC incidence in the intervention arm (10.4%) was higher than expected (4.4%). The incidence in the control arm was 4.6%. The observed overall mortality was lower than expected: 13.6% in the intervention arm and 14.0% in the control arm versus an expected mortality of 16.1%. The observed PC-specific mortality was lower than expected (0.65%) in both the intervention (0.27%) and control arms (0.41%).

CONCLUSIONS

Our results suggest that an efficacy trial with informed consent prior to randomization may have introduced a 'healthy screenee bias'. Therefore, an effectiveness trial with consent after randomization may more accurately estimate the PC-specific mortality reduction if population-based screening is introduced.

Screening for prostate cancer

BACKGROUND

Any form of screening aims to reduce disease-specific and overall mortality, and to improve a person's future quality of life. Screening for prostate cancer has generated considerable debate within the medical and broader community, as demonstrated by the varying recommendations made by medical organizations and governed by national policies. To better inform individual patient decision-making and health policy decisions, we need to consider the entire body of data from randomised controlled trials (RCTs) on prostate cancer screening summarised in a systematic review. In 2006, our Cochrane review identified insufficient evidence to either support or refute the use of routine mass, selective, or opportunistic screening for prostate cancer. An update of the review in 2010 included three additional trials. Meta-analysis of the five studies included in the 2010 review concluded that screening did not significantly reduce prostate cancer-specific mortality. In the past two years, several updates to studies included in the 2010 review have been published thereby providing the rationale for this update of the 2010 systematic review.

OBJECTIVES

To determine whether screening for prostate cancer reduces prostate cancer-specific mortality or all-cause mortality and to assess its impact on quality of life and adverse events.

SEARCH METHODS

An updated search of electronic databases (PROSTATE register, the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, EMBASE, CANCERLIT, and the NHS EED) was performed, in addition to handsearching of specific journals and bibliographies, in an effort to identify both published and unpublished trials.

SELECTION CRITERIA

All RCTs of screening versus no screening for prostate cancer were eligible for inclusion in this review.

DATA COLLECTION AND ANALYSIS

The original search (2006) identified 99 potentially relevant articles that were selected for full-text review. From these citations, two RCTs were identified as meeting the inclusion criteria. The search for the 2010 version of the review identified a further 106 potentially relevant articles, from which three new RCTs were included in the review. A total of 31 articles were retrieved for full-text examination based on the updated search in 2012. Updated data on three studies were included in this review. Data from the trials were independently extracted by two authors.

MAIN RESULTS

Five RCTs with a total of 341,342 participants were included in this review. All involved prostate-specific antigen (PSA) testing, with or without digital rectal examination (DRE), though the interval and threshold for further evaluation varied across trials. The age of participants ranged from 45 to 80 years and duration of follow-up from 7 to 20 years. Our meta-analysis of the five included studies indicated no statistically significant difference in prostate cancer-specific mortality between men randomised to the screening and control groups (risk ratio (RR) 1.00, 95% confidence interval (CI) 0.86 to 1.17). The methodological quality of three of the studies was assessed as posing a high risk of bias. The European Randomized Study of Screening for Prostate Cancer (ERSPC) and the US Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial were assessed as posing a low risk of bias, but provided contradicting results. The ERSPC study reported a significant reduction in prostate cancer-specific mortality (RR 0.84, 95% CI 0.73 to 0.95), whilst the PLCO study concluded no significant benefit (RR 1.15, 95% CI 0.86 to 1.54). The ERSPC was the only study of the five included in this review that reported a significant reduction in prostate cancer-specific mortality, in a pre-specified subgroup of men aged 55 to 69 years of age. Sensitivity analysis for overall risk of bias indicated no significant difference in prostate cancer-specific mortality when referring to the meta analysis of only the ERSPC and PLCO trial data (RR 0.96, 95% CI 0.70 to 1.30). Subgroup analyses indicated that prostate cancer-specific mortality was not affected by the age at which participants were screened. Meta-analysis of four studies investigating all-cause mortality did not determine any significant differences between men randomised to screening or control (RR 1.00, 95% CI 0.96 to 1.03). A diagnosis of prostate cancer was significantly greater in men randomised to screening compared to those randomised to control (RR 1.30, 95% CI 1.02 to 1.65). Localised prostate cancer was more commonly diagnosed in men randomised to screening (RR 1.79, 95% CI 1.19 to 2.70), whilst the proportion of men diagnosed with advanced prostate cancer was significantly lower in the screening group compared to the men serving as controls (RR 0.80, 95% CI 0.73 to 0.87). Screening resulted in a range of harms that can be considered minor to major in severity and duration. Common minor harms from screening include bleeding, bruising and short-term anxiety. Common major harms include overdiagnosis and overtreatment, including infection, blood loss requiring transfusion, pneumonia, erectile dysfunction, and incontinence. Harms of screening included false-positive results for the PSA test and overdiagnosis (up to 50% in the ERSPC study). Adverse events associated with transrectal ultrasound (TRUS)-guided biopsies included infection, bleeding and pain. No deaths were attributed to any biopsy procedure. None of the studies provided detailed assessment of the effect of screening on quality of life or provided a comprehensive assessment of resource utilization associated with screening (although preliminary analyses were reported).

AUTHORS' CONCLUSIONS

Prostate cancer screening did not significantly decrease prostate cancer-specific mortality in a combined meta-analysis of five RCTs. Only one study (ERSPC) reported a 21% significant reduction of prostate cancer-specific mortality in a pre-specified subgroup of men aged 55 to 69 years. Pooled data currently demonstrates no significant reduction in prostate cancer-specific and overall mortality. Harms associated with PSA-based screening and subsequent diagnostic evaluations are frequent, and moderate in severity. Overdiagnosis and overtreatment are common and are associated with treatment-related harms. Men should be informed of this and the demonstrated adverse effects when they are deciding whether or not to undertake screening for prostate cancer. Any reduction in prostate cancer-specific mortality may take up to 10 years to accrue; therefore, men who have a life expectancy less than 10 to 15 years should be informed that screening for prostate cancer is unlikely to be beneficial. No studies examined the independent role of screening by DRE.

Prostate-specific antigen testing rates remain low in UK general practice: a cross-sectional study in six English cities

OBJECTIVE • To estimate rates of prostate-specific antigen (PSA) testing in UK general practices by age, deprivation index and geographical location. SUBJECTS AND METHODS • Practice-based, retrospective data on PSA testing patterns in 2007 were collected from a random sample of 87 general practices using EMIS LV computer systems within the passively observed non-intervention arm of a cluster-randomized controlled trial. • Information for a total of 126 716 men aged 45-89 years with no recorded diagnosis of prostate cancer prior to 1 January 2007 was collected. RESULTS • In all, 7902 (6.2%) of 126 716 men aged 45-89 without a prior diagnosis of prostate cancer underwent at least one PSA test from their general practitioner during 2007 [95% confidence interval (CI) 5.6-7.0%; practice-based inter-quartile range 3.6-8.4%]. • PSA testing rates were 1.4% (95% CI 1.1-1.6%) in men aged 45-49, rising to 11.3% (95% CI 10.0-12.9%) at age 75-79 years (P for trend <0.001). • Testing rates were lowest in the three northern centres (3.5-5.7%) vs the three more southern centres (7.1-8.9%; P < 0.001). • For every 20 points increase in the index of multiple deprivation score, the proportion of men tested fell by 1.7% (95% CI -2.5 to -0.8%; P < 0.001). • Lower proportions of men were subsequently diagnosed with prostate cancer in practices testing more men (odds ratio for a one unit increase in the natural log of testing 0.76; 95% CI 0.60-0.97; P= 0.025). CONCLUSION • Overall levels of PSA testing in UK general practice remain low, but for those tested there are important variations by age, deprivation and geographical location that do not appear to reflect clinical need or the intention of current policy. • PSA testing in general practice is currently skewed towards older men, and current policy enabling all men to make an informed choice about PSA testing is not being effectively implemented as uptake clearly varies by socioeconomic status. • This reinforces the need for robust evidence regarding the costs and benefits of using the PSA test for the detection of localized prostate cancer in the UK, a full assessment of the health economic implications and a revision of the current policy.

Assessing contamination and compliance in the prostate component of the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial

Background Recently, the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial published 7-year complete prostate cancer mortality results, which showed no benefit of screening with prostate specific antigen (PSA) and digital rectal examination (DRE). An issue of concern was the substantial level of ‘contamination’, or use of PSA and DRE in control arm men.

Purpose To provide a detailed description of contamination in PLCO.

Methods Surveys inquiring about the most recent PSA and DRE use were given to a sample of control arm men throughout the screening phase of PLCO (years 0—5). A probability model was utilized to translate survey results into actual frequency counts of tests. To assess the impact of contamination, Surveillance, Epidemiology, and End Results (SEER) incidence rates from the pre-screening era (1985—1987) as well as contemporaneous rates, were applied to PLCO person-years of observation.

Results Of 38,350 control arm men, 2427 were surveyed. Pre-trial screening and college education were statistically significantly associated with increased contamination rates. The estimated mean number of screening PSAs (DREs) in the control arm was 2.7 (1.1); this compares to 5.0 (3.5) in the screened arm. 1984 and 2538 prostate cancers were observed in the control and screened arms, respectively, during the screening phase. In the absence of screening, 960 and 949 would have been expected; with contemporaneous incidence rates, 1630 and 1611 were expected.

Limitations Due to the limitations of the surveys, in terms of both reach and scope, the exact level of PSA and DRE use in control arm men cannot be known.

Conclusions Use of prostate screening by control arm men was substantial, but also substantially less than in screened arm men. Detailed quantitative analyses of screening use across arms are critical for understanding current and future findings from the prostate component of PLCO. Clinical Trials 2010; 7: 303—311. http:// ctj.sagepub.com

Prostate cancer mortality reduction by prostate-specific antigen-based screening adjusted for nonattendance and contamination in the European Randomised Study of Screening for Prostate Cancer (ERSPC)

BACKGROUND

Prostate-specific antigen (PSA) based screening for prostate cancer (PCa) has been shown to reduce prostate specific mortality by 20% in an intention to screen (ITS) analysis in a randomised trial (European Randomised Study of Screening for Prostate Cancer [ERSPC]). This effect may be diluted by nonattendance in men randomised to the screening arm and contamination in men randomised to the control arm.

OBJECTIVE

To assess the magnitude of the PCa-specific mortality reduction after adjustment for nonattendance and contamination.

DESIGN, SETTING, AND PARTICIPANTS

We analysed the occurrence of PCa deaths during an average follow-up of 9 yr in 162,243 men 55-69 yr of age randomised in seven participating centres of the ERSPC. Centres were also grouped according to the type of randomisation (ie, before or after informed written consent).

INTERVENTION

Nonattendance was defined as nonattending the initial screening round in ERSPC. The estimate of contamination was based on PSA use in controls in ERSPC Rotterdam.

MEASUREMENTS

Relative risks (RRs) with 95% confidence intervals (CIs) were compared between an ITS analysis and analyses adjusting for nonattendance and contamination using a statistical method developed for this purpose.

RESULTS AND LIMITATIONS

In the ITS analysis, the RR of PCa death in men allocated to the intervention arm relative to the control arm was 0.80 (95% CI, 0.68-0.96). Adjustment for nonattendance resulted in a RR of 0.73 (95% CI, 0.58-0.93), and additional adjustment for contamination using two different estimates led to estimated reductions of 0.69 (95% CI, 0.51-0.92) to 0.71 (95% CI, 0.55-0.93), respectively. Contamination data were obtained through extrapolation of single-centre data. No heterogeneity was found between the groups of centres.

CONCLUSIONS

PSA screening reduces the risk of dying of PCa by up to 31% in men actually screened. This benefit should be weighed against a degree of overdiagnosis and overtreatment inherent in PCa screening.

Cancer screenings, diagnostic technology evolution, and cancer control

Screening should allow for the anticipation of cancer diagnosis at an earlier stage, when curative treatment is possible. Screening for cervical, large bowel, and breast cancer were shown to be effective in reducing mortality. The wide acceptance of the screening concept led to the wide diffusion also of screening of uncertain benefit against prostate cancer and skin melanoma. Diagnostic technologies are continuously evolving, and new tests are proposed to improve existing screenings or as screening tests for additional cancer sites (e.g., lung cancer). Cancer screening, however, is a complex and costly intervention that does not result only in benefits but also may cause harm. A major emerging problem of screening is overdiagnosis, or the detection of cases that would have not progressed to the symptomatic phase in the absence of screening. Thus, both experimental and observational evaluation studies are needed to reduce harm caused by screenings and to select effective interventions among many proposed innovations. Finally, the research of markers to assess the aggressive nature of screen-detected lesions is of great importance to improve screenings ' harm/benefit ratio.

Overall survival in the intervention arm of a randomized controlled screening trial for prostate cancer compared with a clinically diagnosed cohort

OBJECTIVES

This population-based study provides comparisons of prostate cancer characteristics at diagnosis of two cohorts of men from two well-defined geographical areas exposed to different intensities of prostate cancer screening. Overall survival in both cohorts was compared with that in the general population.

METHODS

A cohort of 822 men randomized to the intervention arm of a prostate cancer screening trial and subsequently diagnosed with prostate cancer was compared with a nonrandomized cohort of 947 men who were clinically diagnosed with prostate cancer in a geographically neighboring region. In both cohorts, cases were diagnosed with prostate cancer between January 1989 and December 1997. A partitioning of overall survival by variables associated with cancer onset such as age at diagnosis, stage at diagnosis, and grade at diagnosis was performed.

RESULTS

Age at diagnosis, tumor extent at diagnosis, and grade at diagnosis were significantly different between the screened and clinically diagnosed cohort. The 5- and 10-yr survival rates were higher in the screened cohort than in the clinically diagnosed cohort (88.8% vs. 52.4%, and 68.4% vs. 29.6%, respectively). Significant differences in survival were evident for all age, stage, and grade subgroups, except for metastatic disease at diagnosis.

CONCLUSIONS

Differences in overall survival favoring the screened population were observed for all baseline characteristics (age, stage, and grade of disease), and these variables may all explain differences in overall survival because screening achieves early diagnosis as well as a stage and grade shift. As observed survival rates in the screened population mirrored those within the general population, the contribution of lead time and overdiagnosis to final patient outcome is considered to be large as well.

Screening for prostate cancer

BACKGROUND

Any form of screening aims to reduce mortality and increase a person's quality of life. Screening for prostate cancer has generated considerable debate within the medical community, as demonstrated by the varying recommendations made by medical organizations and governed by national policies. Much of this debate is due to the limited availability of high quality research and the influence of false-positive or false-negative results generated by use of the diagnostic techniques such as the digital rectal examination (DRE) and prostate specific antigen (PSA) blood test.

OBJECTIVES

To determine whether screening for prostate cancer reduces prostate cancer mortality and has an impact on quality of life.

SEARCH STRATEGY

Electronic databases (PROSTATE register, CENTRAL the Cochrane Central Register of Controlled Trials, MEDLINE, EMBASE, CANCERLIT and the NHS EED) were searched electronically in addition to hand searching of specific journals and bibliographies in an effort to identify both published and unpublished trials.

SELECTION CRITERIA

All randomised controlled trials of screening versus no screening or routine care for prostate cancer were eligible for inclusion in this review.

DATA COLLECTION AND ANALYSIS

The search identified 99 potentially relevant articles that were selected for full text review. From these 99 citations, two randomised controlled trials were identified as meeting the review's inclusion criteria. Data from the trials were independently extracted by two authors.

MAIN RESULTS

Two randomised controlled trials with a total of 55,512 participants were included; however, both trials had methodological weaknesses. Re-analysis using intention-to-screen and meta-analysis of results from the two randomised controlled trials indicated no statistically significant difference in prostate cancer mortality between men randomised for prostate cancer screening and controls (RR 1.01, 95% CI: 0.80-1.29). Neither study assessed the effect of prostate cancer screening on quality of life, all-cause mortality or cost effectiveness.

AUTHORS' CONCLUSIONS

Given that only two randomised controlled trials were included, and the high risk of bias of both trials, there is insufficient evidence to either support or refute the routine use of mass, selective or opportunistic screening compared to no screening for reducing prostate cancer mortality. Currently, no robust evidence from randomised controlled trials is available regarding the impact of screening on quality of life, harms of screening, or its economic value. Results from two ongoing large scale multicentre randomised controlled trials that will be available in the next several years are required to make evidence-based decisions regarding prostate cancer screening.