This document discusses methods for adjusting for treatment switching in randomized controlled trials. It addresses the problem that intention-to-treat analysis may not accurately estimate treatment effectiveness when patients in the control group can switch to the experimental treatment. The document outlines several adjustment methods, including rank preserving structural failure time models and two-stage estimation, and discusses the issue of informative censoring due to treatment switching. It also presents a simulation study comparing the performance of adjustment methods with and without re-censoring to artificially censor patients who switched.
When designing a clinical study, a fundamental aspect is the sample size. In this article, we describe the rationale for sample size calculations, when it should be calculated and describe the components necessary to calculate it. For simple studies, standard formulae can be
used; however, for more advanced studies, it is generally necessary to use specialized statistical software programs and consult a biostatistician. Sample size calculations for non-randomized studies are also discussed and two clinical examples are used for illustration
When designing a clinical study, a fundamental aspect is the sample size. In this article, we describe the rationale for sample size calculations, when it should be calculated and describe the components necessary to calculate it. For simple studies, standard formulae can be
used; however, for more advanced studies, it is generally necessary to use specialized statistical software programs and consult a biostatistician. Sample size calculations for non-randomized studies are also discussed and two clinical examples are used for illustration
A power point on the various types of flaps and their respective indications. This presentation briefly describes the various flaps and how to care for flaps.
Methods of randomisation in clinical trialsAmy Mehaboob
Randomization is the process by which allocation of subjects to treatment groups is done by chance, without the ability to predict who is in what group. A randomized clinical trial is a clinical trial in which participants are randomly assigned to separate groups that compare different treatments.
Randomized trials are gold standard of study designs because the potential for bias (selection into treatment groups) is avoided.
This document includes the purpose, types, advantages and disadvantages of each type of randomisation.
Randomization is the process by which allocation of subjects to treatment groups is done by chance, without the ability to predict who is in what group. It is done in clinical trials. This presentation describes the methods of randmization used in clinical trials.
Preventive analgesia:
Broader definition of preemptive analgesia
Perioperative analgesic regimen that able to control pain-induced sensitization
Not the timing of the analgesic treatment but the duration and efficacy of an analgesic intervention are more important for an effective postoperative pain relief
Adequate preventive analgesia should include multimodal techniques and with a sufficient duration of tretment
A power point on the various types of flaps and their respective indications. This presentation briefly describes the various flaps and how to care for flaps.
Methods of randomisation in clinical trialsAmy Mehaboob
Randomization is the process by which allocation of subjects to treatment groups is done by chance, without the ability to predict who is in what group. A randomized clinical trial is a clinical trial in which participants are randomly assigned to separate groups that compare different treatments.
Randomized trials are gold standard of study designs because the potential for bias (selection into treatment groups) is avoided.
This document includes the purpose, types, advantages and disadvantages of each type of randomisation.
Randomization is the process by which allocation of subjects to treatment groups is done by chance, without the ability to predict who is in what group. It is done in clinical trials. This presentation describes the methods of randmization used in clinical trials.
Preventive analgesia:
Broader definition of preemptive analgesia
Perioperative analgesic regimen that able to control pain-induced sensitization
Not the timing of the analgesic treatment but the duration and efficacy of an analgesic intervention are more important for an effective postoperative pain relief
Adequate preventive analgesia should include multimodal techniques and with a sufficient duration of tretment
2nd CUTEHeart Workshop Manuel Gomes PresentationLBNicolau
Manuel Gomes from the London School of Hygiene and Tropical Medicine, UK gave valuable insights on "Non-compliance in randomised controlled trials comparing vascular and endovascular interventions for cardiovascular care" at the 2nd CUTEHeart Workshop in CPC2016.
Extending A Trial’s Design Case Studies Of Dealing With Study Design IssuesnQuery
About the webinar
As trials increase in complexity and scope, there is a requirement for trial designs to reflect this.
From dealing with non-proportional hazards in survival analysis to dealing with cluster randomization, we examine how to deal with study design issues of complex trials.
In this free webinar, you will learn about:
Dealing with study design issues
Practical worked examples of
Non-proportional Hazards
Cluster Randomization
Three Armed Trials
Non-proportional Hazards
Non-proportional hazards and complex survival curves have become of increasing interest, due to being commonly seen in immunotherapy development. This has led to interest in assessing the robustness of standard methods and alternative methods that better adapt to deviations.
In this webinar, we look at methods proposed for complex survival curves and the weighted log-rank test as a candidate model to deal with a delayed survival effect.
Cluster Randomization
Cluster-randomized designs are often adopted when there is a high risk of contamination if cluster members were randomized individually. Stepped-wedge designs are useful in cases where it is difficult to apply a particular treatment to half of the clusters at the same time.
In this webinar, we introduce cluster randomization and stepped-wedge designs to provide an insight into the requirements of more complex randomization schedules.
Three Armed Trials
Non-inferiority testing is a common hypothesis test in the development of generic medicine and medical devices. The most common design compares the proposed non-inferior treatment to the standard treatment alone but this leaves uncertain if the treatment effect is the same as from previous studies. This “assay sensitivity” problem can be resolved by using a three arm trial which includes placebo alongside the new and reference treatments for direct comparison.
In this webinar we show a complete testing approach to this gold standard design and how to find the appropriate allocation and sample size for this study.
Duration - 60 minutes
Speaker: Ronan Fitzpatrick, Head of Statistics, Statsols
Do height and BMI affect human capital formation? Natural experimental evidence from DNA. CHE seminar presentation by Neil Davies, University of Bristol 12 June 2020
Healthy Minds: A Randomised Controlled Trial to Evaluate PHSE Curriculum Deve...cheweb1
CHE Seminar presentation 16 January 2020, Alistair McGuire, Department of Health Policy, LSE. Evaluating the Healthy Minds program: The impact on adolescent’s health related quality of life of a change in a school curriculum
Baker what to do when people disagree che york seminar jan 2019 v2cheweb1
Public values, plurality and health care resource allocation: What should we do when people disagree? (..and should economists care about reasons as well as choices?) CHE Seminar 21 January 2019
Lung Cancer: Artificial Intelligence, Synergetics, Complex System Analysis, S...Oleg Kshivets
RESULTS: Overall life span (LS) was 2252.1±1742.5 days and cumulative 5-year survival (5YS) reached 73.2%, 10 years – 64.8%, 20 years – 42.5%. 513 LCP lived more than 5 years (LS=3124.6±1525.6 days), 148 LCP – more than 10 years (LS=5054.4±1504.1 days).199 LCP died because of LC (LS=562.7±374.5 days). 5YS of LCP after bi/lobectomies was significantly superior in comparison with LCP after pneumonectomies (78.1% vs.63.7%, P=0.00001 by log-rank test). AT significantly improved 5YS (66.3% vs. 34.8%) (P=0.00000 by log-rank test) only for LCP with N1-2. Cox modeling displayed that 5YS of LCP significantly depended on: phase transition (PT) early-invasive LC in terms of synergetics, PT N0—N12, cell ratio factors (ratio between cancer cells- CC and blood cells subpopulations), G1-3, histology, glucose, AT, blood cell circuit, prothrombin index, heparin tolerance, recalcification time (P=0.000-0.038). Neural networks, genetic algorithm selection and bootstrap simulation revealed relationships between 5YS and PT early-invasive LC (rank=1), PT N0—N12 (rank=2), thrombocytes/CC (3), erythrocytes/CC (4), eosinophils/CC (5), healthy cells/CC (6), lymphocytes/CC (7), segmented neutrophils/CC (8), stick neutrophils/CC (9), monocytes/CC (10); leucocytes/CC (11). Correct prediction of 5YS was 100% by neural networks computing (area under ROC curve=1.0; error=0.0).
CONCLUSIONS: 5YS of LCP after radical procedures significantly depended on: 1) PT early-invasive cancer; 2) PT N0--N12; 3) cell ratio factors; 4) blood cell circuit; 5) biochemical factors; 6) hemostasis system; 7) AT; 8) LC characteristics; 9) LC cell dynamics; 10) surgery type: lobectomy/pneumonectomy; 11) anthropometric data. Optimal diagnosis and treatment strategies for LC are: 1) screening and early detection of LC; 2) availability of experienced thoracic surgeons because of complexity of radical procedures; 3) aggressive en block surgery and adequate lymph node dissection for completeness; 4) precise prediction; 5) adjuvant chemoimmunoradiotherapy for LCP with unfavorable prognosis.
The prostate is an exocrine gland of the male mammalian reproductive system
It is a walnut-sized gland that forms part of the male reproductive system and is located in front of the rectum and just below the urinary bladder
Function is to store and secrete a clear, slightly alkaline fluid that constitutes 10-30% of the volume of the seminal fluid that along with the spermatozoa, constitutes semen
A healthy human prostate measures (4cm-vertical, by 3cm-horizontal, 2cm ant-post ).
It surrounds the urethra just below the urinary bladder. It has anterior, median, posterior and two lateral lobes
It’s work is regulated by androgens which are responsible for male sex characteristics
Generalised disease of the prostate due to hormonal derangement which leads to non malignant enlargement of the gland (increase in the number of epithelial cells and stromal tissue)to cause compression of the urethra leading to symptoms (LUTS
Title: Sense of Smell
Presenter: Dr. Faiza, Assistant Professor of Physiology
Qualifications:
MBBS (Best Graduate, AIMC Lahore)
FCPS Physiology
ICMT, CHPE, DHPE (STMU)
MPH (GC University, Faisalabad)
MBA (Virtual University of Pakistan)
Learning Objectives:
Describe the primary categories of smells and the concept of odor blindness.
Explain the structure and location of the olfactory membrane and mucosa, including the types and roles of cells involved in olfaction.
Describe the pathway and mechanisms of olfactory signal transmission from the olfactory receptors to the brain.
Illustrate the biochemical cascade triggered by odorant binding to olfactory receptors, including the role of G-proteins and second messengers in generating an action potential.
Identify different types of olfactory disorders such as anosmia, hyposmia, hyperosmia, and dysosmia, including their potential causes.
Key Topics:
Olfactory Genes:
3% of the human genome accounts for olfactory genes.
400 genes for odorant receptors.
Olfactory Membrane:
Located in the superior part of the nasal cavity.
Medially: Folds downward along the superior septum.
Laterally: Folds over the superior turbinate and upper surface of the middle turbinate.
Total surface area: 5-10 square centimeters.
Olfactory Mucosa:
Olfactory Cells: Bipolar nerve cells derived from the CNS (100 million), with 4-25 olfactory cilia per cell.
Sustentacular Cells: Produce mucus and maintain ionic and molecular environment.
Basal Cells: Replace worn-out olfactory cells with an average lifespan of 1-2 months.
Bowman’s Gland: Secretes mucus.
Stimulation of Olfactory Cells:
Odorant dissolves in mucus and attaches to receptors on olfactory cilia.
Involves a cascade effect through G-proteins and second messengers, leading to depolarization and action potential generation in the olfactory nerve.
Quality of a Good Odorant:
Small (3-20 Carbon atoms), volatile, water-soluble, and lipid-soluble.
Facilitated by odorant-binding proteins in mucus.
Membrane Potential and Action Potential:
Resting membrane potential: -55mV.
Action potential frequency in the olfactory nerve increases with odorant strength.
Adaptation Towards the Sense of Smell:
Rapid adaptation within the first second, with further slow adaptation.
Psychological adaptation greater than receptor adaptation, involving feedback inhibition from the central nervous system.
Primary Sensations of Smell:
Camphoraceous, Musky, Floral, Pepperminty, Ethereal, Pungent, Putrid.
Odor Detection Threshold:
Examples: Hydrogen sulfide (0.0005 ppm), Methyl-mercaptan (0.002 ppm).
Some toxic substances are odorless at lethal concentrations.
Characteristics of Smell:
Odor blindness for single substances due to lack of appropriate receptor protein.
Behavioral and emotional influences of smell.
Transmission of Olfactory Signals:
From olfactory cells to glomeruli in the olfactory bulb, involving lateral inhibition.
Primitive, less old, and new olfactory systems with different path
New Drug Discovery and Development .....NEHA GUPTA
The "New Drug Discovery and Development" process involves the identification, design, testing, and manufacturing of novel pharmaceutical compounds with the aim of introducing new and improved treatments for various medical conditions. This comprehensive endeavor encompasses various stages, including target identification, preclinical studies, clinical trials, regulatory approval, and post-market surveillance. It involves multidisciplinary collaboration among scientists, researchers, clinicians, regulatory experts, and pharmaceutical companies to bring innovative therapies to market and address unmet medical needs.
Tom Selleck Health: A Comprehensive Look at the Iconic Actor’s Wellness Journeygreendigital
Tom Selleck, an enduring figure in Hollywood. has captivated audiences for decades with his rugged charm, iconic moustache. and memorable roles in television and film. From his breakout role as Thomas Magnum in Magnum P.I. to his current portrayal of Frank Reagan in Blue Bloods. Selleck's career has spanned over 50 years. But beyond his professional achievements. fans have often been curious about Tom Selleck Health. especially as he has aged in the public eye.
Follow us on: Pinterest
Introduction
Many have been interested in Tom Selleck health. not only because of his enduring presence on screen but also because of the challenges. and lifestyle choices he has faced and made over the years. This article delves into the various aspects of Tom Selleck health. exploring his fitness regimen, diet, mental health. and the challenges he has encountered as he ages. We'll look at how he maintains his well-being. the health issues he has faced, and his approach to ageing .
Early Life and Career
Childhood and Athletic Beginnings
Tom Selleck was born on January 29, 1945, in Detroit, Michigan, and grew up in Sherman Oaks, California. From an early age, he was involved in sports, particularly basketball. which played a significant role in his physical development. His athletic pursuits continued into college. where he attended the University of Southern California (USC) on a basketball scholarship. This early involvement in sports laid a strong foundation for his physical health and disciplined lifestyle.
Transition to Acting
Selleck's transition from an athlete to an actor came with its physical demands. His first significant role in "Magnum P.I." required him to perform various stunts and maintain a fit appearance. This role, which he played from 1980 to 1988. necessitated a rigorous fitness routine to meet the show's demands. setting the stage for his long-term commitment to health and wellness.
Fitness Regimen
Workout Routine
Tom Selleck health and fitness regimen has evolved. adapting to his changing roles and age. During his "Magnum, P.I." days. Selleck's workouts were intense and focused on building and maintaining muscle mass. His routine included weightlifting, cardiovascular exercises. and specific training for the stunts he performed on the show.
Selleck adjusted his fitness routine as he aged to suit his body's needs. Today, his workouts focus on maintaining flexibility, strength, and cardiovascular health. He incorporates low-impact exercises such as swimming, walking, and light weightlifting. This balanced approach helps him stay fit without putting undue strain on his joints and muscles.
Importance of Flexibility and Mobility
In recent years, Selleck has emphasized the importance of flexibility and mobility in his fitness regimen. Understanding the natural decline in muscle mass and joint flexibility with age. he includes stretching and yoga in his routine. These practices help prevent injuries, improve posture, and maintain mobilit
Knee anatomy and clinical tests 2024.pdfvimalpl1234
This includes all relevant anatomy and clinical tests compiled from standard textbooks, Campbell,netter etc..It is comprehensive and best suited for orthopaedicians and orthopaedic residents.
HOT NEW PRODUCT! BIG SALES FAST SHIPPING NOW FROM CHINA!! EU KU DB BK substit...GL Anaacs
Contact us if you are interested:
Email / Skype : kefaya1771@gmail.com
Threema: PXHY5PDH
New BATCH Ku !!! MUCH IN DEMAND FAST SALE EVERY BATCH HAPPY GOOD EFFECT BIG BATCH !
Contact me on Threema or skype to start big business!!
Hot-sale products:
NEW HOT EUTYLONE WHITE CRYSTAL!!
5cl-adba precursor (semi finished )
5cl-adba raw materials
ADBB precursor (semi finished )
ADBB raw materials
APVP powder
5fadb/4f-adb
Jwh018 / Jwh210
Eutylone crystal
Protonitazene (hydrochloride) CAS: 119276-01-6
Flubrotizolam CAS: 57801-95-3
Metonitazene CAS: 14680-51-4
Payment terms: Western Union,MoneyGram,Bitcoin or USDT.
Deliver Time: Usually 7-15days
Shipping method: FedEx, TNT, DHL,UPS etc.Our deliveries are 100% safe, fast, reliable and discreet.
Samples will be sent for your evaluation!If you are interested in, please contact me, let's talk details.
We specializes in exporting high quality Research chemical, medical intermediate, Pharmaceutical chemicals and so on. Products are exported to USA, Canada, France, Korea, Japan,Russia, Southeast Asia and other countries.
Pulmonary Thromboembolism - etilogy, types, medical- Surgical and nursing man...VarunMahajani
Disruption of blood supply to lung alveoli due to blockage of one or more pulmonary blood vessels is called as Pulmonary thromboembolism. In this presentation we will discuss its causes, types and its management in depth.
NVBDCP.pptx Nation vector borne disease control programSapna Thakur
NVBDCP was launched in 2003-2004 . Vector-Borne Disease: Disease that results from an infection transmitted to humans and other animals by blood-feeding arthropods, such as mosquitoes, ticks, and fleas. Examples of vector-borne diseases include Dengue fever, West Nile Virus, Lyme disease, and malaria.
Flu Vaccine Alert in Bangalore Karnatakaaddon Scans
As flu season approaches, health officials in Bangalore, Karnataka, are urging residents to get their flu vaccinations. The seasonal flu, while common, can lead to severe health complications, particularly for vulnerable populations such as young children, the elderly, and those with underlying health conditions.
Dr. Vidisha Kumari, a leading epidemiologist in Bangalore, emphasizes the importance of getting vaccinated. "The flu vaccine is our best defense against the influenza virus. It not only protects individuals but also helps prevent the spread of the virus in our communities," he says.
This year, the flu season is expected to coincide with a potential increase in other respiratory illnesses. The Karnataka Health Department has launched an awareness campaign highlighting the significance of flu vaccinations. They have set up multiple vaccination centers across Bangalore, making it convenient for residents to receive their shots.
To encourage widespread vaccination, the government is also collaborating with local schools, workplaces, and community centers to facilitate vaccination drives. Special attention is being given to ensuring that the vaccine is accessible to all, including marginalized communities who may have limited access to healthcare.
Residents are reminded that the flu vaccine is safe and effective. Common side effects are mild and may include soreness at the injection site, mild fever, or muscle aches. These side effects are generally short-lived and far less severe than the flu itself.
Healthcare providers are also stressing the importance of continuing COVID-19 precautions. Wearing masks, practicing good hand hygiene, and maintaining social distancing are still crucial, especially in crowded places.
Protect yourself and your loved ones by getting vaccinated. Together, we can help keep Bangalore healthy and safe this flu season. For more information on vaccination centers and schedules, residents can visit the Karnataka Health Department’s official website or follow their social media pages.
Stay informed, stay safe, and get your flu shot today!
Pharynx and Clinical Correlations BY Dr.Rabia Inam Gandapore.pptx
Adjusting for treatment switching in randomised controlled trials
1. Adjusting for treatment switching in
randomised controlled trials
Nicholas Latimer, University of Sheffield, Sheffield, UK,
Reader in Health Economics,
NIHR Post Doctoral Research Fellow
Thanks to Ian White, Keith Abrams and Uwe Siebert who I
have been working with on this
2. 1. Background
2. Addressing the problem
3. Adjustment methods
4. Re-censoring (and health economics)
5. Simulation study
6. Conclusions
2
3. 1. Background
2. Addressing the problem
3. Adjustment methods
4. Re-censoring (and health economics)
5. Simulation study
6. Conclusions
3
4. Health economic evaluation
Economic evaluation aims to ensure the benefits of
programmes that are implemented exceed their opportunity
costs
Aim is to compare the new treatment to all relevant
comparators
Usually this involves a comparison of the costs and effects of
the new treatment and the standard treatment
Usually information to make these comparisons is taken from
randomised controlled trials
4
5. Treatment switching
What if patients randomised to the control group in a clinical
trial are permitted to switch onto the experimental treatment at
some point during the trial?
5
6. Survival time
Control Treatment
Intervention
Control Intervention
PFS
PFS
PFS
PPS
PPS
PPS
True OS difference
ITT OS
difference
Treatment switching: Patients randomised to the control
group are allowed to switch to the new intervention
• Common in oncology RCTs
PFS = Progression-free survival; PPS = Post-progression survival;
OS = Overall survival
7. Treatment switching: Patients randomised to the control
group are allowed to switch to the new intervention
• Common in oncology RCTs
ITT analysis likely to fail to
address the decision problem
Different analytical methods
are needed to estimate
effectiveness and cost-effectiveness
PFS = Progression-free survival; PPS = Post-progression survival;
OS = Overall survival
8. Treatment switching
Is an issue in over 55% of oncology technology assessments
Adjustment methods can change decisions
8
NICE TA321 Dabrafenib for melanoma
57% switched
ITT analysis: OS HR 0.76; ICER £95,225
Adjustment analysis: OS HR 0.55; ICER £49,019
Dabrafenib was recommended for use
9. • But…
• <50% TAs include adjustments for switching
• ≈60% of adjustment analyses rejected
Poor application of methodsLow decision-maker confidence in
methods
Treatment switching
- Methods make untestable
assumptions
- May lack face-validity
- Analysts have to make
many choices when
applying methods
- Concern that favourable
application decisions are
being made
- Or that methods have just
been used “badly”
10. • Reliance on ITT, or upon poor adjustment analyses, has
severe consequences
− Inappropriate recommendations
− Sub-optimal resource allocation
− Lost lives, lost QALYs
Treatment switching
11. 1. Background
2. Addressing the problem
3. Adjustment methods
4. Re-censoring (and health economics)
5. Simulation study
6. Conclusions
11
12. 1. Develop analytical
techniques/methods to address
barriers that restrict use of
adjustment methods
2. Establish techniques/methods in
practice
• Address barriers around
testability of methodological
assumptions
• Further examine application
choices
• Use case studies to
establish practical use of
methods/techniques
• Without this, adjustment
analyses likely to remain
under-used
Addressing the problem (my fellowship)
This is what I’m going to focus
on today
13. Application decisions
When applying adjustment methods, analysts must make
several decisions, e.g…
Which covariates to include in the analysis (which variables are
predictive of survival and influence the probability of switching?)
Duration of the treatment effect (is it likely to endure beyond
treatment discontinuation?)
Whether or not to re-censor
13
14. Motivating example
14
Trametinib vs chemotherapy for metastatic melanoma [67% of control group
patients switched onto trametinib]
From: Latimer NR et al. Cancer Medicine 2016; 5(5):806–815
Re-censoring had a
big impact on the
results of the
adjustment methods
- Why?
- Which analysis is
more reliable?
15. 1. Background
2. Addressing the problem
3. Adjustment methods
4. Re-censoring (and health economics)
5. Simulation study
6. Conclusions
15
16. Adjustment methods
• Two of the available adjustment methods involve estimating
counterfactual survival times using a counterfactual survival model
(Rank Preserving Structural Failure Time Model (RPSFTM) and two-
stage estimation (TSE))
𝑇𝑖 = 𝑇𝑜𝑓𝑓 𝑖
+ 𝑇𝑜𝑛 𝑖
(1)
𝑈𝑖 = 𝑇𝑜𝑓𝑓 𝑖
+ 𝑒 𝜓
𝑇𝑜𝑛 𝑖
(2)
𝑇𝑖 = observed survival time for individual i
𝑈𝑖 = untreated survival time for individual i
𝑇𝑜𝑓𝑓 𝑖
= Time spent off treatment for individual i
𝑇𝑜𝑛 𝑖
= Time spent on treatment for individual i
𝑒 𝜓 is a time ratio associated with treatment (inverse of the treatment effect)
• Estimate 𝜓 and plug into (2) to calculate untreated survival times
16
17. Estimating 𝝍 – RPSFTM
Use g-estimation to identify 𝝍
Two key assumptions:
a) Perfect randomisation – no treatment, equal average survival
b) Common treatment effect – no matter when treatment received
Counterfactual survival model: 𝑈𝑖 = 𝑇𝑜𝑓𝑓 𝑖
+ 𝑒 𝜓 𝑇𝑜𝑛 𝑖
We know 𝑇𝑜𝑓𝑓 𝑖
We know 𝑇𝑜𝑛 𝑖
We know (assume) that 𝑈𝑖 is equal between randomised groups
𝒆 𝝍 is the only unknown
Test lots of values of 𝝍 until we find one that results in equal average 𝑻𝒊 𝟎
between randomised groups (g-estimation)
17
18. Estimating 𝝍 – RPSFTM
Use g-estimation to identify 𝝍
Two key assumptions:
a) Perfect randomisation – no treatment, equal average survival
b) Common treatment effect – no matter when treatment received
Counterfactual survival model: 𝑈𝑖 = 𝑇𝑜𝑓𝑓 𝑖
+ 𝑒 𝜓 𝑇𝑜𝑛 𝑖
We know 𝑇𝑜𝑓𝑓 𝑖
We know 𝑇𝑜𝑛 𝑖
We know (assume) that 𝑇𝑖 0 is equal between randomised groups
𝒆 𝝍 is the only unknown
Test lots of values of 𝝍 until we find one that results in equal average 𝑻𝒊 𝟎
between randomised groups (g-estimation)
18
For each patient in the control and
experimental group, plug value for 𝜓
into:
𝑈𝑖 = 𝑇𝑜𝑓𝑓 𝑖
+ 𝑒 𝜓 𝑇𝑜𝑛 𝑖
Are untreated survival times equal?
… if not, try next value for 𝜓
20. Survival time
Control Non-switchers
Control Switchers
PFS
PFS
PPS
PPS
2. Estimate post-secondary
baseline treatment effect in
switchers compared to non-
switchers using an AFT model
1. Identify
secondary baseline
in control group
3. “Shrink” survival times in
switchers according to the
AF, deriving counterfactual
survival times
PPS
20
No unmeasured
confounding
Estimating 𝝍 – TSE
21. Censoring
RPSFTM and TSE both estimate counterfactual survival times
Assuming treatment is beneficial, survival times will be shrunken
Usually not everyone dies in cancer trials
For switchers who die, we estimate shrunken event times
For switchers who are censored, we estimate shrunken censoring times
This is a problem
For survival analysis to be unbiased, censoring times must be
independent of any prognostic variables
But here, we are censoring switchers at an earlier time-point
Whether and when patients switch is unlikely to be random
Shrinking censoring times for switchers but not non-switchers
constitutes informative censoring
21
22. 1. Background
2. Addressing the problem
3. Adjustment methods
4. Re-censoring (and health economics)
5. Simulation study
6. Conclusions
22
23. Re-censoring
Standard solution to this is re-censoring
Artificially censor everyone in the group(s) affected by switching
(irrespective of whether they switched) to their earliest possible
censoring time over all possible treatment strategies. This breaks the
relationship between prognosis and censoring time.
23
24. Re-censoring
Control arm (switchers)
Control arm (non-switchers)
24
𝐶𝑒𝑛𝑠𝑜𝑟𝑖𝑛𝑔 𝑜𝑛 𝑡ℎ𝑒 𝑇 𝑠𝑐𝑎𝑙𝑒𝑅𝑒 − 𝑐𝑒𝑛𝑠𝑜𝑟 ℎ𝑒𝑟𝑒
Survival and
censoring times
shrunk to adjust for
switching
Switchers are
censored before
non-switchers
S
S
S
1
2
3
4
5
6
25. Re-censoring and health economics
25
Re-censoring has long been accepted as a requirement when
estimating counterfactual survival times (Robins1989; White 1999)
It means we have estimates of treatment effects that are not subject
to informative censoring, up to the maximum re-censored time-point
For use in economic evaluation, is that enough? Is it useful?
26. Re-censoring and health economics
26
Economic evaluations usually need to extrapolate out to a life-time
Re-censoring involves censoring people artificially early
Results in lost longer term information
From: Latimer NR et al. Cancer Medicine 2016; 5(5):806–815
HR=0.43 HR=0.53
27. Re-censoring and health economics
27
Economic evaluations usually need to extrapolate out to a life-time
Re-censoring involves censoring people artificially early
Results in lost longer term information
From: Latimer NR et al. Cancer Medicine 2016; 5(5):806–815
HR=0.43 HR=0.53
• What if the treatment effect changes over
time?
• What if longer-term trends in the hazard
function haven’t become established in the
re-censored dataset?
What is more important – lost information, or
informative censoring?
28. 1. Background
2. Addressing the problem
3. Adjustment methods
4. Re-censoring (and health economics)
5. Simulation study
6. Conclusions
28
29. Simulation study
Simulate survival times, apply switching from control group (after
progression) that is related to prognosis
Apply adjustment methods with and without re-censoring
RPSFTM
RPSFTMnr
Compare bias (and empirical standard error, root mean squared error, coverage)
in estimation of control group restricted mean survival time (RMST)
at end of trial follow-up
Vary potentially important characteristics across scenarios
treatment effect size
complexity of survivor function
switch proportion
144 scenarios in total (only going to show you a subset of these)
29
TSE
TSEnr
common treatment effect
switcher prognosis
disease severity
30. Data generation
In the majority of scenarios used a mixture Weibull model to
simulate survival times
Allows complex survivor functions to be generated, which we
believe are likely to reflect reality better than a simple parametric
distribution
30
0.000.250.500.751.00
320 312 291 257 208 177 0Experimental group
180 169 138 106 65 17 0Control group
Number at risk
0 100 200 300 400 500 600
Analysis time (days)
Control group
Experimental group
.0005
.001
.0015
.002
.0025
Hazardrate
0 100 200 300 400 500 600
Analysis time (days)
Control group
Experimental group
31. Results holding complexity of the survivor function
(moderate), disease severity (low), prognosis of switchers
(good) constant
31
Switch proportion (low, moderate)
Treatment effect (low, high)
Common treatment effect (no CTE, CTE)
-10
-8-6-4-2
02468
10
1 2 3 4 5 6 7 8
Scenario
ITT
Low trt effect /
low switch %
Low trt effect /
high switch %High trt effect /
low switch %
High trt effect /
high switch %
32. 32
Switch proportion (low, moderate)
Treatment effect (low, high)
Common treatment effect (no CTE, CTE)
-10
-8-6-4-2
02468
10
1 2 3 4 5 6 7 8
Scenario
ITT
RPSFTM
Results holding complexity of the survivor function
(moderate), disease severity (low), prognosis of switchers
(good) constant
Low trt effect /
low switch %
Low trt effect /
high switch %High trt effect /
low switch %
High trt effect /
high switch %
33. 33
Switch proportion (low, moderate)
Treatment effect (low, high)
Common treatment effect (no CTE, CTE)
-10
-8-6-4-2
02468
10
1 2 3 4 5 6 7 8
Scenario
ITT
RPSFTM
Results holding complexity of the survivor function
(moderate), disease severity (low), prognosis of switchers
(good) constant
Low trt effect /
low switch %
Low trt effect /
high switch %High trt effect /
low switch %
High trt effect /
high switch %
Why does RPSFTM under-estimate RMST?
2. No common treatment effect
When switchers get a reduced effect RPSFTM over-adjusts
because the treatment effect in switchers is assumed to be just
as big as it is in the experimental group – RPSFTM does even
worse in these scenarios
.0005
.001
.0015
.002
.0025
Hazardrate
0 100 200 300 400 500 600
Analysis time (days)
Control group
Experimental group
Two sources of negative bias:
1. Simulated decreasing hazards over time
When re-censor, have to extrapolate
from re-censored dataset to estimate
RMST at the end of trial follow-up
Decreasing hazards haven’t established
before re-censoring time-point, so
extrapolation over-estimates long-term
hazards and under-estimates RMST
34. 34
Switch proportion (low, moderate)
Treatment effect (low, high)
Common treatment effect (no CTE, CTE)
-10
-8-6-4-2
02468
10
1 2 3 4 5 6 7 8
Scenario
ITT
RPSFTM
Results holding complexity of the survivor function
(moderate), disease severity (low), prognosis of switchers
(good) constant
Low trt effect /
low switch %
Low trt effect /
high switch %High trt effect /
low switch %
High trt effect /
high switch %
No CTE
CTE
No CTE
CTE No CTE
CTE
No CTE
CTE
35. 35
Results holding complexity of the survivor function
(moderate), disease severity (low), prognosis of switchers
(good) constant
Switch proportion (low, moderate)
Treatment effect (low, high)
Common treatment effect (no CTE, CTE)
-10
-8-6-4-2
02468
10
1 2 3 4 5 6 7 8
Scenario
ITT
TSE
RPSFTM
Low trt effect /
low switch %
Low trt effect /
high switch %High trt effect /
low switch %
High trt effect /
high switch %
36. 36
Results holding complexity of the survivor function
(moderate), disease severity (low), prognosis of switchers
(good) constant
Switch proportion (low, moderate)
Treatment effect (low, high)
Common treatment effect (no CTE, CTE)
-10
-8-6-4-2
02468
10
1 2 3 4 5 6 7 8
Scenario
ITT
TSE
RPSFTM
Low trt effect /
low switch %
Low trt effect /
high switch %High trt effect /
low switch %
High trt effect /
high switch %
No CTE
CTE
No CTE
CTE No CTE
CTE
No CTE
CTE
Why does TSE under-estimate RMST?
TSE does better than RPSFTM when there is not a common
treatment effect, and does similarly when there is a common
treatment effect
Note, for TSE and RPSFTM, size of treatment effect is biggest
influence on bias – increase effect size = more lost information
.0005
.001
.0015
.002
.0025
Hazardrate
0 100 200 300 400 500 600
Analysis time (days)
Control group
Experimental group
Same reason as RPSFTM:
1. Loss of longer-term information means
that re-censored data-set does not
contain information on the decreasing
hazards over time
BUT, TSE does not assume a common
treatment effect, so is not prone to this
additional negative bias
37. 37
Results holding complexity of the survivor function
(moderate), disease severity (low), prognosis of switchers
(good) constant
Switch proportion (low, moderate)
Treatment effect (low, high)
Common treatment effect (no CTE, CTE)
-10
-8-6-4-2
02468
10
1 2 3 4 5 6 7 8
Scenario
ITT
TSE
RPSFTM
Low trt effect /
low switch %
Low trt effect /
high switch %High trt effect /
low switch %
High trt effect /
high switch %
No CTE
CTE
No CTE
CTE No CTE
CTE
No CTE
CTE
38. 38
Results holding complexity of the survivor function
(moderate), disease severity (low), prognosis of switchers
(good) constant
Switch proportion (low, moderate)
Treatment effect (low, high)
Common treatment effect (no CTE, CTE)
-10
-8-6-4-2
02468
10
1 2 3 4 5 6 7 8
Scenario
ITT
TSE
RPSFTM
RPSFTMnr
Low trt effect /
low switch %
Low trt effect /
high switch %High trt effect /
low switch %
High trt effect /
high switch %
39. 39
Results holding complexity of the survivor function
(moderate), disease severity (low), prognosis of switchers
(good) constant
Switch proportion (low, moderate)
Treatment effect (low, high)
Common treatment effect (no CTE, CTE)
-10
-8-6-4-2
02468
10
1 2 3 4 5 6 7 8
Scenario
ITT
TSE
RPSFTM
RPSFTMnr
Low trt effect /
low switch %
Low trt effect /
high switch %High trt effect /
low switch %
High trt effect /
high switch %
No CTE
CTE
No CTE
CTE No CTE
CTE
No CTE
CTE
Why does RPSFTM nr over-estimate RMST?
Two competing sources of bias:
1. Informative censoring
Switchers are censored at earlier time-points
Non-switchers are not
Re-censoring affects the right-hand-side of the KM curve
Non-switchers who have long-term survival are by definition
people who have done well (perhaps they haven’t even had
disease progression yet)
Not re-censoring leaves the best performing patients with the
longest censoring times – hence inducing positive bias
2. Common treatment effect. When switchers get a reduced
effect the RPSFTM over-adjusts inducing negative bias
To some extent, these two biases cancel out
RPSFTMnr does better when there is not a CTE
40. 40
Results holding complexity of the survivor function
(moderate), disease severity (low), prognosis of switchers
(good) constant
Switch proportion (low, moderate)
Treatment effect (low, high)
Common treatment effect (no CTE, CTE)
-10
-8-6-4-2
02468
10
1 2 3 4 5 6 7 8
Scenario
ITT
TSE
RPSFTM
RPSFTMnr
Low trt effect /
low switch %
Low trt effect /
high switch %High trt effect /
low switch %
High trt effect /
high switch %
No CTE
CTE No CTE CTE
No CTE CTE
No CTE
CTE
41. 41
Results holding complexity of the survivor function
(moderate), disease severity (low), prognosis of switchers
(good) constant
Switch proportion (low, moderate)
Treatment effect (low, high)
Common treatment effect (no CTE, CTE)
-10
-8-6-4-2
02468
10
1 2 3 4 5 6 7 8
Scenario
ITT
TSE
TSEnr
RPSFTM
RPSFTMnr
Low trt effect /
low switch %
Low trt effect /
high switch %High trt effect /
low switch %
High trt effect /
high switch %
42. 42
Results holding complexity of the survivor function
(moderate), disease severity (low), prognosis of switchers
(good) constant
Switch proportion (low, moderate)
Treatment effect (low, high)
Common treatment effect (no CTE, CTE)
-10
-8-6-4-2
02468
10
1 2 3 4 5 6 7 8
Scenario
ITT
TSE
TSEnr
RPSFTM
RPSFTMnr
Low trt effect /
low switch %
Low trt effect /
high switch %High trt effect /
low switch %
High trt effect /
high switch %
No CTE
CTE
No CTE
CTE No CTE
CTE
No CTE
CTE
Why does TSEnr over-estimate RMST?
Same reason as RPSFTMnr
1. Informative censoring means that the best performing
patients have the longest censoring times – hence RMST is
over-estimated
BUT, TSEnr does not assume a common treatment effect, so
there is no competing negative bias
Some trend towards RPSFTMnr doing slightly better than
TSEnr when there is not a CTE
Note, for RPSFTMnr and TSEnr increased switch proportion
and (to a lesser extent) increased treatment effect were the
most important drivers of bias – both result in informative
censoring being a bigger problem (increased selection effect
and increased censoring disparity between switchers and
non-switchers)
43. 43
Results holding complexity of the survivor function
(moderate), disease severity (low), prognosis of switchers
(good) constant
Switch proportion (low, moderate)
Treatment effect (low, high)
Common treatment effect (no CTE, CTE)
-10
-8-6-4-2
02468
10
1 2 3 4 5 6 7 8
Scenario
ITT
TSE
TSEnr
RPSFTM
RPSFTMnr
Low trt effect /
low switch %
Low trt effect /
high switch %High trt effect /
low switch %
High trt effect /
high switch %
No CTE
CTE No CTE CTE
No CTE CTE
No CTE
CTE
44. 1. Background
2. Addressing the problem
3. Adjustment methods
4. Re-censoring (and health economics)
5. Simulation study
6. Conclusions
44
45. Conclusions (1)
Re-censoring and not re-censoring are both prone to bias when our
objectives are to estimate longer-term survival or long-term
treatment effects
Re-censoring is likely to lead to under-estimates of control group
survival
Not re-censoring is likely to lead to over-estimates of control group
survival
We should do both! (provides decision-maker with useful info)
45
46. Conclusions (2)
Should also assess hazard and survivor functions to attempt to
identify the likely impact of re-censoring
Does the hazard have a turning point or sudden change of slope? When?
And should assess the characteristics of long-term survivors to
identify the likely impact of informative censoring
Are there any long-term survivors who did not switch?
46
47. Conclusions (3)
Should consider which method is likely to produce least bias
Re-censoring methods are most prone to bias when the treatment
effect is high
Non-re-censoring methods are most prone to bias when the
switching proportion is high
Better consideration of all these issues leads to better informed
analyses, with less scope for unjustified “innocuous”
application choices
47
48. Conclusions (4)
Finally, neither method is perfect for the HTA context
Can we do better?
Inverse probability of censoring weighting (IPCW) represents a well-
known method for dealing with informative censoring
Perhaps we could not re-censor, and use IPCW to account for
informative censoring…?
48
50. Results
50
Switch proportion (low, moderate)
Severity (low, high)
Complexity of survivor function (simple, moderate, high)
Treatment effect (low, high)
Common treatment effect (no CTE, CTE)
Prognosis of switchers (good, poor)
Scenario9
Scenario13
Scenario57
Scenario61
-10
0
1020
1-24 25-48 49-72 73-96
Scenario
No switch
ITT
TSE
TSEnr
RPSFTM
RPSFTMnr
51. 51
Results holding complexity of the survivor function
(moderate), disease severity (low), prognosis of switchers
(good) constant
Switch proportion (low, moderate)
Treatment effect (low, high)
Common treatment effect (no CTE, CTE)
02468
10
1 2 3 4 5 6 7 8
Scenario
ITT
TSE
TSEnr
RPSFTM
RPSFTMnr
Low trt effect /
low switch %
Low trt effect /
high switch %High trt effect /
low switch %
High trt effect /
high switch %
52. 52
Results holding complexity of the survivor function
(moderate), disease severity (low), prognosis of switchers
(good) constant
Switch proportion (low, moderate)
Treatment effect (low, high)
Common treatment effect (no CTE, CTE)
048
12
RMSE
1 2 3 4 5 6 7 8
Scenario
ITT
TSE
TSEnr
RPSFTM
RPSFTMnr
Low trt effect /
low switch %
Low trt effect /
high switch %High trt effect /
low switch %
High trt effect /
high switch %
53. What is a realistic hazard
function in cancer?
53
Hazardrate
Time
Trial entrants
relatively fit
low hazard, but
increasing due
to disease
Over time patient mix
changes, long-term
survivors remain
hazard has a turning
point, and reduces
In the long-term,
hazard increases due
to old age hazard
has another turning
point, and increases
[May not observe this
in trial period]
54. What is a realistic hazard
function in cancer? Real case
54
All cause survival data
for 9,721 breast cancer
patients age<50,
diagnosed in England
and Wales between
1986 and end 1991
[adapted from Rutherford et al,
Journal of Statistical Computation
and Simulation, 2015;85;4:777-793]
55. • Ipilimumab plus dacarbazine compared to
dacarbazine for previously untreated metastatic
melanoma
• These data are reconstructed from the pivotal trial
publication [Robert et al, New England Journal of
Medicine. 2011; 364(26):2517-26].
Work done by Ash
Bullemont for his
MSc dissertation at
ScHARR
Issues raised by I-O therapy
Hazardrate
56. Re-censoring
Standard solution to this is re-censoring. Artificially censor everyone
(irrespective of whether they switched) to their earliest possible
censoring time over all possible treatment strategies. This breaks the
relationship between prognosis and censoring time.
Assuming the treatment is beneficial, the most favourable treatment
strategy for any patient would have been to receive treatment for the
entire duration of the study, i.e. 𝑇𝑜𝑛 𝑖
= 𝐶𝑖, where 𝐶𝑖 is the administrative
censoring time
Then, patient i’s censoring time would be shrunk to 𝑒 𝜓
𝐶𝑖
Under re-censoring,
For switchers: 𝑈𝑖 is replaced with 𝑒 𝜓
𝐶𝑖 if 𝑒 𝜓
𝐶𝑖 < 𝑈𝑖
For non-switchers: 𝑇𝑖 is replaced with 𝑒 𝜓 𝐶𝑖 if 𝑒 𝜓 𝐶𝑖 < 𝑇𝑖
If this is done for all patients in the trial, there is no longer any
relationship between switching/prognosis and censoring time 56