SlideShare ist ein Scribd-Unternehmen logo
1 von 71
Downloaden Sie, um offline zu lesen
REPEATED EVENTS ANALYSES
12/17/2012
1
Outline for Talk
 Recurrent Events
 Analyses for Count Outcomes
 Survival Analysis Approaches for Repeated Events
 Parametric Analyses for Counts
 Poisson Regression
 Negative Binomial
 Non-parametric Analyses for Counts
 Mean of Ratios (Q)
 Ratio of Means (R)
2
Recurrent Events
 Usually in survival studies subjects are followed until
they experience a single index event
 Death
 Cancer
 Return to employment
 But there are many events of interest that are
repeatable
 Seizures
 Infections
 Hospitalizations
3
Recurrent Events
 Most of the methods employed in survival analysis
are set up for a single event
 But, there have been methods developed for use
with repeated events
 Some of these methods come out of the area of
statistical methods developed for count data – these
focus on the count of the number of recurrent event
 Some methods come directly out of survival analysis
methods – these focus on the time duration between the
recurrent events
4
Recurrent Events
 Methods coming out of either area need to address
the domain of the other methods
 The count methods also need to address the time
component
 The survival methods need to address the number of
events as well as the gap times between events
5
Diabetes Control and Complications
Trial (DCCT)
 NIH-funded trial
 Launched in 1981
 To evaluate the effect of intensive blood glucose-
lowering on risk of albuminuria in diabetic subjects
 Subjects were randomized to either intensive blood
glucose lowering or conventional treatment
 Intensive glucose lowering used self-monitoring 4 or
5 times daily, multiple daily insulin injections or a
pump, diet and exercise
6
Diabetes Control and Complications
Trial (DCCT)
 A concern was that the intensive glucose lowering
could lead to hypoglycemia
 Dizzy spells
 Possible comas
 Seizures
 Hypoglycemia events were tracked as a secondary
outcome (DCCT 1997)
7
Diabetes Control and Complications
Trial (DCCT)
Intensive
Treatment
Conventional
Treatment
Subjects 363 352
Hypoglycemia
Events
1723 543
Person-Years of
Follow-Up
2598.5 2480.8
Rate per 100
Person-Years
66.3 21.9
8
Count Outcome Approaches
 The most basic analysis for count outcomes comes
from the Poisson distribution
 𝑃𝑃 𝑦𝑦|𝜆𝜆 =
𝑒𝑒−𝜆𝜆 𝜆𝜆 𝑦𝑦
𝑦𝑦!
 Here 𝜆𝜆 is the mean count
 y is the observed count (non-negative integer)
 One feature of the Poisson distribution is that
 Variance = Mean = 𝜆𝜆
9
Poisson Densities
10
Count Outcome Approaches
 The maximum likelihood estimator of the mean is
provided by the sample average
 ̂𝜆𝜆 =
𝑦𝑦1+𝑦𝑦2+⋯+𝑦𝑦𝑛𝑛
𝑛𝑛
 Often people are followed for somewhat different
lengths of time, giving some people more
opportunity to get larger event counts
 In this case 𝜆𝜆 is standardized to be per-unit of time
11
Count Outcome Approaches
 With unequal follow-up times 𝑡𝑡1, 𝑡𝑡2, … , 𝑡𝑡𝑛𝑛
 𝑃𝑃 𝑦𝑦|𝑡𝑡, 𝜆𝜆 =
𝑒𝑒−𝜆𝜆𝑡𝑡(𝜆𝜆𝑡𝑡)𝑦𝑦
𝑦𝑦!
 Where 𝑡𝑡 is the length of follow-up
 With unequal follow-up, the MLE for 𝜆𝜆 is provided
by
 ̂𝜆𝜆 =
𝑦𝑦1+𝑦𝑦2+⋯+𝑦𝑦𝑛𝑛
𝑡𝑡1+𝑡𝑡2+⋯+𝑡𝑡𝑛𝑛
=
�𝑦𝑦
̅𝑡𝑡
=
𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛 𝑜𝑜𝑜𝑜 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒
𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝−𝑡𝑡𝑡𝑡 𝑡𝑡𝑡𝑡 𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 𝑓𝑓−𝑢𝑢𝑢𝑢
12
Count Outcome Approaches
 With unequal follow-up times 𝑡𝑡1, 𝑡𝑡2, … , 𝑡𝑡𝑛𝑛
 𝑃𝑃 𝑦𝑦|𝑡𝑡, 𝜆𝜆 =
𝑒𝑒−𝜆𝜆𝑡𝑡(𝜆𝜆𝑡𝑡)𝑦𝑦
𝑦𝑦!
 Where 𝑡𝑡 is the length of follow-up
 With unequal follow-up, the MLE for 𝜆𝜆 is provided
by
 ̂𝜆𝜆 =
𝑦𝑦1+𝑦𝑦2+⋯+𝑦𝑦𝑛𝑛
𝑡𝑡1+𝑡𝑡2+⋯+𝑡𝑡𝑛𝑛
=
�𝑦𝑦
̅𝑡𝑡
=
𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛 𝑜𝑜𝑜𝑜 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒
𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝−𝑡𝑡𝑡𝑡 𝑡𝑡𝑡𝑡 𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 𝑓𝑓−𝑢𝑢𝑢𝑢
 We will see this again
13
Link From Poisson to Exponential
 If the count of events follows a Poisson distribution,
then the times between events follow an exponential
distribution
 𝑓𝑓 𝑡𝑡 𝜆𝜆 = 𝜆𝜆𝑒𝑒−𝜆𝜆𝜆𝜆
 A feature of exponential survival times is that the
hazard event rate is 𝜆𝜆 and is constant over time
 The mean time to an event is
1
𝜆𝜆
 Gives a way to estimate 𝜆𝜆 in a survival analysis
setting (rather than a count data setting)
14
Exponential Densities
15
Poisson Regression
 In Poisson regression we model the expected count
for an observation as a function of predictor
variables
 Poisson regression is a log-linear model so that the
log of the mean is set equal to a linear combination
of the predictors
16
Poisson Regression
 𝑙𝑙𝑙𝑙 𝑙𝑙 𝜆𝜆𝜆𝜆 = 𝛽𝛽0 + 𝛽𝛽1 𝑥𝑥
 This leads to
 𝑙𝑙𝑙𝑙 𝑙𝑙 𝜆𝜆 + 𝑙𝑙𝑙𝑙 𝑙𝑙 𝑡𝑡 = 𝛽𝛽0 + 𝛽𝛽1 𝑥𝑥
 So, we get
 𝑙𝑙𝑙𝑙 𝑙𝑙 𝜆𝜆 = 𝛽𝛽0 + 𝛽𝛽1 𝑥𝑥 − 𝑙𝑙𝑙𝑙 𝑙𝑙 𝑡𝑡
 The −𝑙𝑙𝑙𝑙 𝑙𝑙 𝑡𝑡 is called the offset and corrects for the
differences in follow-up time between subjects
 The model can be expressed 𝜆𝜆 = 𝑒𝑒 𝛽𝛽0+𝛽𝛽1 𝑥𝑥−𝑙𝑙𝑙𝑙𝑙𝑙 𝑡𝑡
17
Poisson Regression
 This model is fit using maximum likelihood
 SAS GENMOD
 The 𝛽𝛽 estimates for predictors are interpreted as
logs of incidence rate ratios
 A one-unit increase in 𝑥𝑥 gives a ratio of 𝑒𝑒𝑒𝑒𝑒𝑒 𝛽𝛽1 in the
expected count
 If 𝑥𝑥 codes for a group difference (0 = control, 1 =
treatment), then 𝑒𝑒𝑒𝑒𝑒𝑒 𝛽𝛽1 corresponds to the ratio of
counts in the treated group divided by counts in the
control group
18
Poisson Regression in DCCT
proc genmod;
model nevents = group
/ dist = poisson
link = log
offset = lnyears;
TITLE1 'Poisson regression models of risk
of hypoglycemia';
title2 'unadjusted treatment group effect';
19
Poisson Regression in DCCT
20
Analysis Of Parameter Estimates
Parameter DF Estimate
Standard
Error
Wald 95%
Confidence Limits Chi-Square Pr > ChiSq
Intercept 1 -1.5190 0.0429 -1.6031 -1.4349 1252.90 <.0001
group Exp 1 1.1081 0.0492 1.0117 1.2046 507.01 <.0001
group Std 0 0.0000 0.0000 0.0000 0.0000 . .
Scale 0 1.0000 0.0000 1.0000 1.0000
Poisson Regression in DCCT
21
Highly significant result!
But, is it based on a reasonable model
for the data?
Analysis Of Parameter Estimates
Parameter DF Estimate
Standard
Error
Wald 95%
Confidence Limits Chi-Square Pr > ChiSq
Intercept 1 -1.5190 0.0429 -1.6031 -1.4349 1252.90 <.0001
group Exp 1 1.1081 0.0492 1.0117 1.2046 507.01 <.0001
group Std 0 0.0000 0.0000 0.0000 0.0000 . .
Scale 0 1.0000 0.0000 1.0000 1.0000
Poisson Regression
 The big limitation for Poisson regression is
overdispersion
 The data are overdispersed when the variation in the
observed counts is greater that the mean value
 Under Poisson model these should match
 Variance in the counts much greater than the means
 Goodness of fit testing for overdispersion is
typically done in Poisson regression
22
Overdispersion in DCCT
23
Criteria For Assessing Goodness Of Fit
Criterion DF Value Value/DF
Deviance 713 3928.7828 5.5102
Scaled Deviance 713 3928.7828 5.5102
Pearson Chi-Square 713 5131.3429 7.1968
Scaled Pearson X2 713 5131.3429 7.1968
Log Likelihood 775.1804
Overdispersion in DCCT
24
Criteria For Assessing Goodness Of Fit
Criterion DF Value Value/DF
Deviance 713 3928.7828 5.5102
Scaled Deviance 713 3928.7828 5.5102
Pearson Chi-Square 713 5131.3429 7.1968
Scaled Pearson X2 713 5131.3429 7.1968
Log Likelihood 775.1804
If model fits well then Value/DF should be close to 1.0
• Large ratios indicate overdispersion or model
misspecification
Poisson Regression
 For a clinical trial, our primary goal is a good test
of treatment effect, so wouldn’t normally need to
include covariates
 But, could try to remove excess variation by
controlling for covariates
25
Poisson Regression in DCCT
proc genmod; class group;
model nevents = group
insulin duration female
adult bcval5 hbael hxcoma
/ dist = poisson
link = log
offset = lnyears
covb;
title2 'covariate adjusted treatment group
effect';
26
Poisson Regression in DCCT
27
Analysis Of Parameter Estimates
Parameter DF Estimate
Standard
Error
Wald 95%
Confidence Limits Chi-Square Pr > ChiSq
Intercept 1 -0.9568 0.2174 -1.3829 -0.5308 19.38 <.0001
group Exp 1 1.0845 0.0493 0.9879 1.1812 483.91 <.0001
group Std 0 0.0000 0.0000 0.0000 0.0000 . .
insulin 1 0.0051 0.0995 -0.1898 0.2000 0.00 0.9593
duration 1 0.0015 0.0006 0.0004 0.0026 6.79 0.0092
female 1 0.1794 0.0424 0.0963 0.2624 17.93 <.0001
adult 1 -0.5980 0.0656 -0.7265 -0.4694 83.13 <.0001
bcval5 1 -0.5283 0.3630 -1.2398 0.1833 2.12 0.1456
hbael 1 -0.0335 0.0151 -0.0631 -0.0038 4.89 0.0271
hxcoma 1 0.6010 0.0685 0.4669 0.7352 77.09 <.0001
Scale 0 1.0000 0.0000 1.0000 1.0000
Poisson Regression in DCCT
28
Analysis Of Parameter Estimates
Parameter DF Estimate
Standard
Error
Wald 95%
Confidence Limits Chi-Square Pr > ChiSq
Intercept 1 -0.9568 0.2174 -1.3829 -0.5308 19.38 <.0001
group Exp 1 1.0845 0.0493 0.9879 1.1812 483.91 <.0001
group Std 0 0.0000 0.0000 0.0000 0.0000 . .
insulin 1 0.0051 0.0995 -0.1898 0.2000 0.00 0.9593
duration 1 0.0015 0.0006 0.0004 0.0026 6.79 0.0092
female 1 0.1794 0.0424 0.0963 0.2624 17.93 <.0001
adult 1 -0.5980 0.0656 -0.7265 -0.4694 83.13 <.0001
bcval5 1 -0.5283 0.3630 -1.2398 0.1833 2.12 0.1456
hbael 1 -0.0335 0.0151 -0.0631 -0.0038 4.89 0.0271
hxcoma 1 0.6010 0.0685 0.4669 0.7352 77.09 <.0001
Scale 0 1.0000 0.0000 1.0000 1.0000
Still very significant, but does the model fit better?
Poisson Regression in DCCT
29
Criteria For Assessing Goodness Of Fit
Criterion DF Value Value/DF
Deviance 706 3707.7027 5.2517
Scaled Deviance 706 3707.7027 5.2517
Pearson Chi-Square 706 4792.7876 6.7887
Scaled Pearson X2 706 4792.7876 6.7887
Log Likelihood 885.7204
The fit is better with the covariates included, but still very
overdispersed.
Poisson Regression
 Another approach that can be used to remove
overdispersion is to use a variance correction
 Pearson correction
 Standard approach – has the actual variance equal the
modeled variance multiplied by an estimated overdispersion
parameter
 Corrects the standard errors and test results to account for
the overdispersion
 GEE correction based on robust sandwich variance
estimator
 Based on a robust sandwich-variance estimator
30
GEE Poisson Regression in DCCT
proc genmod;
class group subnum;
model nevents = group
/ dist = poisson
link = log
offset = lnyears
type3;
repeated subject=subnum / type=unstr;
title2 ‘GEE unadjusted treatment group
effect';
31
GEE Poisson Regression in DCCT
32
Analysis Of GEE Parameter Estimates
Empirical Standard Error Estimates
Parameter Estimate
Standard
Error
95% Confidence
Limits Z Pr > |Z|
Intercept -1.5190 0.1003 -1.7155 -1.3225 -15.15 <.0001
group Exp 1.1081 0.1256 0.8620 1.3543 8.82 <.0001
group Std 0.0000 0.0000 0.0000 0.0000 . .
GEE Poisson Regression in DCCT
33
Analysis Of GEE Parameter Estimates
Empirical Standard Error Estimates
Parameter Estimate
Standard
Error
95% Confidence
Limits Z Pr > |Z|
Intercept -1.5190 0.1003 -1.7155 -1.3225 -15.15 <.0001
group Exp 1.1081 0.1256 0.8620 1.3543 8.82 <.0001
group Std 0.0000 0.0000 0.0000 0.0000 . .
Still very significant treatment effect, but note that standard error
is much larger in this model.
Standard error was 0.0492 in Poisson model with only treatment.
Negative Binomial Regression
 The negative binomial distribution is another count
data distribution with more variance than the
Poisson
 Can be used to give the probability that n trials are
required in order to get m successes (m≤n)
 Model takes the form 𝜆𝜆 = 𝑒𝑒 𝛽𝛽0+𝛽𝛽1 𝑥𝑥−𝑙𝑙𝑙𝑙𝑙𝑙 𝑡𝑡 +𝜀𝜀
 Where 𝑒𝑒𝑒𝑒𝑒𝑒 𝜀𝜀 has a gamma distribution
 Can be considered to be a Poisson model with gamma-
distributed random effects
 Parametric approach to overdispersion
34
Negative Binomial Regression in
DCCT
proc genmod;
class group;
model nevents = group
/ dist = negbin
link = log
offset = lnyears;
title2 ‘Neg Binomial unadjusted
treatment group effect';
35
Negative Binomial Regression in
DCCT36
Analysis Of Parameter Estimates
Parameter DF Estimate
Standard
Error
Wald 95%
Confidence
Limits Chi-Square Pr > ChiSq
Intercept 1 -1.5510 0.0902 -1.728 -1.374 295.60 <.0001
group Exp 1 1.1173 0.1215 0.8791 1.3555 84.52 <.0001
group Std 0 0.0000 0.0000 0.0000 0.0000 . .
Dispersion 1 2.1863 0.1649 1.8631 2.5095
Negative Binomial Regression in
DCCT37
Analysis Of Parameter Estimates
Parameter DF Estimate
Standard
Error
Wald 95%
Confidence
Limits Chi-Square Pr > ChiSq
Intercept 1 -1.5510 0.0902 -1.728 -1.374 295.60 <.0001
group Exp 1 1.1173 0.1215 0.8791 1.3555 84.52 <.0001
group Std 0 0.0000 0.0000 0.0000 0.0000 . .
Dispersion 1 2.1863 0.1649 1.8631 2.5095
The results here are quite similar to those of the GEE Poisson
model, with almost the same standard errors.
DCCT Treatment Results
38
Survival Analyses
 Repeated events
 Cox model on time to first event
 Use Andersen-Gill approach and partition follow-up
time according to which event it applies
 Restart the follow-up time clock after an event
 Multiple observations per subject
 Need to be concerned about correlation
 Unobserved heterogeneity (due to the correlation) can bias
estimates downward while significance is overstated
39
Survival Analysis in DCCT
proc phreg data=three;
model stopday*event(0) =
intgroup/ risklimits;
title1 'Cox Model for First
Event';
run;
40
Survival Analysis in DCCT
41
Analysis of Maximum Likelihood Estimates
Variable DF
Parameter
Estimate
Standard
Error Chi-Square Pr > ChiSq
Hazard
Ratio
95% Hazard
Ratio
Confidence
Limits
INTGROUP 1 0.77252 0.10354 55.6711 <.0001 2.165 1.768 2.652
Survival Analysis in DCCT
proc phreg data=four;
model gaptime*event(0) =
intgroup priorgap /
risklimits;
title1 'Cox Model for
Second Event - Correlation';
run;
42
Survival Analysis in DCCT
43
Analysis of Maximum Likelihood Estimates
Variable DF
Parameter
Estimate
Standard
Error Chi-Square Pr > ChiSq
Hazard
Ratio
95% Hazard
Ratio
Confidence
Limits
INTGROUP 1 0.27790 0.12451 4.9814 0.0256 1.320 1.034 1.685
priorgap 1 -0.000523 0.000114 20.9317 <.0001 0.999 0.999 1.000
So, having a longer time to the first event reduces risk of the
second event. So correlation in these times to events is quite
substantial.
Survival Analysis in DCCT
proc phreg data=six
covsandwich(aggregate);
model gaptime*event(0) =
intgroup / risklimits;
id patient;
title1 'GEE Cox Model for
all Events';
run;
44
Survival Analysis in DCCT
45
Testing Global Null Hypothesis: BETA=0
Test Chi-Square DF Pr > ChiSq
Likelihood Ratio 74.8679 1 <.0001
Score (Model-Based) 72.4448 1 <.0001
Score (Sandwich) 51.8412 1 <.0001
Wald (Model-Based) 71.4253 1 <.0001
Wald (Sandwich) 40.2564 1 <.0001
Survival Analysis in DCCT
46
Analysis of Maximum Likelihood Estimates
Variable DF
Parameter
Estimate
Standard
Error
StdErr
Ratio Chi-Square Pr > ChiSq
Hazard
Ratio
95% Hazard
Ratio
Confidence
Limits
INTGROUP 1 0.67526 0.08988 1.808 56.4435 <.0001 1.965 1.647 2.343
So, accounting for the correlation increases the standard error by
about 80%. However, the resulting hazard ratio is much less than
the other estimates so far.
Nonparametric Analyses
 Work done with Jing Xu
 Motivated by his experiences with analysis of
repeated events at
 The SDAC for the AIDS Clinical Trials Group
 Later experience in the pharmaceutical industry
47
Nonparametric Analyses
 In follow-up studies of recurrent events, there were
studies taking a non-parametric approach to
analysis of two-arm trials by
 𝑞𝑞𝑖𝑖𝑖𝑖 =
𝑦𝑦𝑖𝑖𝑖𝑖
𝑡𝑡𝑖𝑖𝑖𝑖
 𝑦𝑦𝑖𝑖𝑖𝑖 is the number of events for subject i in arm j
(j is 0 or 1)
 𝑡𝑡𝑖𝑖𝑖𝑖 is the total length of follow-up for subject i in group j
 𝑞𝑞𝑖𝑖𝑖𝑖 represents a subject-specific event rate
48
Nonparametric Analyses
 𝑞𝑞𝑖𝑖𝑖𝑖 were analyzed without assuming any particular
distribution by standard two-sample tests
 Student t-test
 Wilcoxon test
 Van der Waerden test
49
Nonparametric Analyses
 For estimation, the means of the subject-specific
event rates were calculated for each group
 �𝑄𝑄𝑗𝑗 =
1
𝑛𝑛𝑗𝑗
∑𝑖𝑖=1
𝑛𝑛𝑗𝑗
𝑞𝑞𝑖𝑖𝑖𝑖 =
1
𝑛𝑛𝑗𝑗
∑𝑖𝑖=1
𝑛𝑛𝑗𝑗 𝑦𝑦𝑖𝑖𝑖𝑖
𝑡𝑡𝑖𝑖𝑖𝑖
 These �𝑄𝑄𝑗𝑗 could then be used to estimate ratios of
the rates and differences in the rates by group
 �𝑅𝑅 𝑅𝑅𝑄𝑄 =
�𝑄𝑄1
�𝑄𝑄0
 �𝑅𝑅 𝑅𝑅𝑄𝑄 = �𝑄𝑄1 − �𝑄𝑄0
50
Nonparametric Analyses
 Our concern about this 𝑞𝑞𝑖𝑖𝑖𝑖 approach was that
 There could be outliers – subjects having a large
number of events in a short amount of time who then
drop out of the study early
 The asymptotic consistency of the �𝑄𝑄𝑗𝑗 estimators
depends on both the numbers of events and the amount
of follow-up increasing within all subjects
51
Nonparametric Analyses
 A different approach had been suggested by L.J.
Wei
 �𝑅𝑅𝑗𝑗 =
𝑥𝑥𝑗𝑗
𝑡𝑡𝑗𝑗
= �
∑𝑖𝑖=1
𝑛𝑛𝑗𝑗
𝑥𝑥𝑖𝑖𝑖𝑖
∑𝑖𝑖=1
𝑛𝑛𝑗𝑗
𝑡𝑡𝑖𝑖𝑖𝑖
=
𝑥𝑥𝑗𝑗
�𝑡𝑡𝑗𝑗
 Ratio of the means, not the mean of the ratios
52
Nonparametric Analyses
 A different approach had been suggested by L.J.
Wei
 �𝑅𝑅𝑗𝑗 =
𝑥𝑥𝑗𝑗
𝑡𝑡𝑗𝑗
= �
∑𝑖𝑖=1
𝑛𝑛𝑗𝑗
𝑥𝑥𝑖𝑖𝑖𝑖
∑𝑖𝑖=1
𝑛𝑛𝑗𝑗
𝑡𝑡𝑖𝑖𝑖𝑖
=
𝑥𝑥𝑗𝑗
�𝑡𝑡𝑗𝑗
 Ratio of the means, not the mean of the ratios
 This is just the standard Poisson events per person-
time without assuming the distribution
53
Nonparametric Analyses
 Can use the �𝑅𝑅𝑗𝑗 to estimate rate ratio and difference
in rates between groups
 �𝑅𝑅𝑅𝑅𝑅𝑅 =
�𝑅𝑅1
�𝑅𝑅0
 �𝑅𝑅 𝑅𝑅𝑅𝑅 = �𝑅𝑅1 − �𝑅𝑅0
54
Nonparametric Analyses
 �𝑅𝑅𝑗𝑗 advantages
 Average out any outliers
 Makes consistency more reasonable as it uses the group
totals for number of events and total follow-up time
 �𝑅𝑅𝑗𝑗 disadvantages
 Measure is group-specific and not subject-specific
 Can’t use standard 2-sample tests
55
Nonparametric Analyses
 Asymptotic forms of the variance for the �𝑅𝑅
measures are in the Xu and LaValley paper in the
references
 Confidence intervals based on Fieller’s method and
resampling
 In this talk, I’ll use resampling methods to evaluate
the �𝑅𝑅𝑗𝑗 measures
 Permutation tests
 Bootstrap confidence intervals
56
57
Nonparametric Analyses
 In simulations, we found that the �𝑄𝑄𝑗𝑗 with a non-
parametric test and the �𝑅𝑅𝑗𝑗 methods maintained the
type-1 error and had comparable power over a
range of sample sizes
 The �𝑅𝑅𝑗𝑗 methods provided reasonable confidence
interval coverage if the sample sizes were at least
100 per group
58
Nonparametric Analyses
 In the paper we use the nonparametric methods on
a dataset for recurrence of bladder cancer in a
two-arm clinical trial of the drug thiotepa
59
Nonparametric Analyses
 Thiotepa trial results
 Results are uniformly non-significant, with rate ratios
from 0.71 to 0.84
60
Nonparametric Analyses
 Thiotepa trial results
 These are the results that I trust the most out of
these
61
Nonparametric Analyses in DCCT
62
Intensive
Treatment
Conventional
Treatment
𝑛𝑛 Subjects 363 352
�𝑦𝑦 Hypoglycemia
Events/Subject
4.75 1.54
̅𝑡𝑡 Person-Years of
Follow-Up/Subject
7.16 7.05
�
�𝑦𝑦
̅𝑡𝑡
�𝑅𝑅𝑗𝑗 0.66 0.22
�𝑄𝑄𝑗𝑗 0.65 0.21
Nonparametric Analyses in DCCT
63
Distribution
of qij by
group
Nonparametric Analyses in DCCT
64
Test Test Statistic P-value
t-test of qij 8.28 P < 0.0001
Van der Waerden
test of qij
8.77 P < 0.0001
Wilcoxon test of qij 8.57 P < 0.0001
Nonparametric Analyses in DCCT
65
Estimator Estimate Permutation Test
P-value*
Rate Difference (Q) 0.43681 P < 0.0005
Rate Difference (R) 0.44415 P < 0.0005
Rate Ratio (Q) 3.08059 P < 0.0005
Rate Ratio (R) 3.02872 P < 0.0005
*Based on 2000
permutations
Nonparametric Analyses in DCCT
66
Q versus R
measures of rate
difference across
2000 permuted
datasets
Nonparametric Analyses in DCCT
67
Estimator Estimate Bootstrap
Confidence Interval
Rate Difference (Q) 0.43681 (0.329, 0.549)
Rate Difference (R) 0.44415 (0.329, 0.561)
Rate Ratio (Q) 3.08059 (2.399, 3.979)
Rate Ratio (R) 3.02872 (2.333, 3.951)
Nonparametric Analyses in DCCT
68
Q versus R
measures of rate
difference across
2000 bootstrap
samples
Nonparametric Analyses
 Both Q and R measures work well for reasonable
sized datasets
 In these DCCT data, both are fine
 In the bladder cancer data, the R (ratio of means)
seems to have a slight edge
69
Conclusions
 There are a lot of good options for the analysis of
repeated events
 GEE Survival models
 Gee Poisson Regression
 Negative binomial regression
 Q and R measures – especially in clinical trial setting
 Worthwhile to work with several as secondary
analyses to verify consistency
70
Main References
 Allison PD. Survival Analysis Using SAS: a Practical
Guide, second edition. SAS Publishing, 2010.
 Lachin JM. Biostatistical Methods: the Assessment of
Relative Risks. Wiley, 2000.
 Xu J, LaValley M. One-sample and two-sample
analysis of heterogeneous person-time data in
clinical trials. Pharmaceutical Statistics 2012; 11:
194 – 203.
71

Weitere ähnliche Inhalte

Was ist angesagt?

ANAESTHESIA CONSIDERATIONS IN GERIATRIC PATEINTS
ANAESTHESIA CONSIDERATIONS IN  GERIATRIC PATEINTSANAESTHESIA CONSIDERATIONS IN  GERIATRIC PATEINTS
ANAESTHESIA CONSIDERATIONS IN GERIATRIC PATEINTSHimanshu Sharma
 
Principles of mechanical ventilation
Principles of mechanical ventilationPrinciples of mechanical ventilation
Principles of mechanical ventilationHimanshu Jangid
 
Pulmonary hypertension and anesthesia
Pulmonary hypertension and anesthesiaPulmonary hypertension and anesthesia
Pulmonary hypertension and anesthesiaWesam Mousa
 
Dual controlled modes of mechanical ventilation [onarılmış]
Dual controlled modes of mechanical ventilation [onarılmış]Dual controlled modes of mechanical ventilation [onarılmış]
Dual controlled modes of mechanical ventilation [onarılmış]tyfngnc
 
Cardiopulmonary bypass
Cardiopulmonary bypassCardiopulmonary bypass
Cardiopulmonary bypassarunsagar25
 
Respiratory physiology h.o.d.
Respiratory physiology h.o.d.Respiratory physiology h.o.d.
Respiratory physiology h.o.d.KGMU, Lucknow
 
Geriatric anaesthesia- Dr harsimran Walia
Geriatric  anaesthesia- Dr harsimran WaliaGeriatric  anaesthesia- Dr harsimran Walia
Geriatric anaesthesia- Dr harsimran Waliaharry11818a
 
Neuromuscular Monitoring
Neuromuscular MonitoringNeuromuscular Monitoring
Neuromuscular MonitoringMohtasib Madaoo
 
Scavenging system in operating room
Scavenging system in operating roomScavenging system in operating room
Scavenging system in operating roomDr Kumar
 
Multipara monitor – what is it and how does it help
Multipara monitor – what is it and how does it helpMultipara monitor – what is it and how does it help
Multipara monitor – what is it and how does it helpSilverline Meditech
 
Hemodynamic parameters & fluid therapy Asim
Hemodynamic parameters &  fluid therapy AsimHemodynamic parameters &  fluid therapy Asim
Hemodynamic parameters & fluid therapy AsimMuhammad Asim Rana
 
Cerebral oximetry brain protection in cardiac surgery
Cerebral oximetry brain protection in cardiac surgery Cerebral oximetry brain protection in cardiac surgery
Cerebral oximetry brain protection in cardiac surgery Wesam Mousa
 
High pressure system- Anaesthesia Machine
High pressure system- Anaesthesia MachineHigh pressure system- Anaesthesia Machine
High pressure system- Anaesthesia MachineDr.Daber Pareed
 
100 most popular anesthesiologists of india 2018
100 most  popular anesthesiologists of india 2018100 most  popular anesthesiologists of india 2018
100 most popular anesthesiologists of india 2018dr tushar chokshi
 
The basic anaesthesia machine
The basic anaesthesia machineThe basic anaesthesia machine
The basic anaesthesia machinehrishi bharali
 
anaesthesia Breathing circuits and its classification and functional analysis
anaesthesia Breathing circuits and its classification and functional analysisanaesthesia Breathing circuits and its classification and functional analysis
anaesthesia Breathing circuits and its classification and functional analysisprateek gupta
 
Low flow anesthesia
Low flow anesthesiaLow flow anesthesia
Low flow anesthesiaShreyas Kate
 

Was ist angesagt? (20)

ANAESTHESIA CONSIDERATIONS IN GERIATRIC PATEINTS
ANAESTHESIA CONSIDERATIONS IN  GERIATRIC PATEINTSANAESTHESIA CONSIDERATIONS IN  GERIATRIC PATEINTS
ANAESTHESIA CONSIDERATIONS IN GERIATRIC PATEINTS
 
Principles of mechanical ventilation
Principles of mechanical ventilationPrinciples of mechanical ventilation
Principles of mechanical ventilation
 
Pulmonary hypertension and anesthesia
Pulmonary hypertension and anesthesiaPulmonary hypertension and anesthesia
Pulmonary hypertension and anesthesia
 
Dual controlled modes of mechanical ventilation [onarılmış]
Dual controlled modes of mechanical ventilation [onarılmış]Dual controlled modes of mechanical ventilation [onarılmış]
Dual controlled modes of mechanical ventilation [onarılmış]
 
Cardiopulmonary bypass
Cardiopulmonary bypassCardiopulmonary bypass
Cardiopulmonary bypass
 
Respiratory physiology h.o.d.
Respiratory physiology h.o.d.Respiratory physiology h.o.d.
Respiratory physiology h.o.d.
 
Geriatric anaesthesia- Dr harsimran Walia
Geriatric  anaesthesia- Dr harsimran WaliaGeriatric  anaesthesia- Dr harsimran Walia
Geriatric anaesthesia- Dr harsimran Walia
 
Vaporizers 2015 spmc
Vaporizers 2015 spmcVaporizers 2015 spmc
Vaporizers 2015 spmc
 
Neuromuscular Monitoring
Neuromuscular MonitoringNeuromuscular Monitoring
Neuromuscular Monitoring
 
Scavenging system in operating room
Scavenging system in operating roomScavenging system in operating room
Scavenging system in operating room
 
Anesthesia workstation
Anesthesia workstationAnesthesia workstation
Anesthesia workstation
 
Multipara monitor – what is it and how does it help
Multipara monitor – what is it and how does it helpMultipara monitor – what is it and how does it help
Multipara monitor – what is it and how does it help
 
Hemodynamic parameters & fluid therapy Asim
Hemodynamic parameters &  fluid therapy AsimHemodynamic parameters &  fluid therapy Asim
Hemodynamic parameters & fluid therapy Asim
 
Cerebral oximetry brain protection in cardiac surgery
Cerebral oximetry brain protection in cardiac surgery Cerebral oximetry brain protection in cardiac surgery
Cerebral oximetry brain protection in cardiac surgery
 
ANES 1502 - M9 PPT: Hemodynamic Monitoring
ANES 1502 - M9 PPT: Hemodynamic MonitoringANES 1502 - M9 PPT: Hemodynamic Monitoring
ANES 1502 - M9 PPT: Hemodynamic Monitoring
 
High pressure system- Anaesthesia Machine
High pressure system- Anaesthesia MachineHigh pressure system- Anaesthesia Machine
High pressure system- Anaesthesia Machine
 
100 most popular anesthesiologists of india 2018
100 most  popular anesthesiologists of india 2018100 most  popular anesthesiologists of india 2018
100 most popular anesthesiologists of india 2018
 
The basic anaesthesia machine
The basic anaesthesia machineThe basic anaesthesia machine
The basic anaesthesia machine
 
anaesthesia Breathing circuits and its classification and functional analysis
anaesthesia Breathing circuits and its classification and functional analysisanaesthesia Breathing circuits and its classification and functional analysis
anaesthesia Breathing circuits and its classification and functional analysis
 
Low flow anesthesia
Low flow anesthesiaLow flow anesthesia
Low flow anesthesia
 

Ähnlich wie Repeated events analyses

Clinical Trials Versus Health Outcomes Research: SAS/STAT Versus SAS Enterpri...
Clinical Trials Versus Health Outcomes Research: SAS/STAT Versus SAS Enterpri...Clinical Trials Versus Health Outcomes Research: SAS/STAT Versus SAS Enterpri...
Clinical Trials Versus Health Outcomes Research: SAS/STAT Versus SAS Enterpri...cambridgeWD
 
Clinical Trials Versus Health Outcomes Research: SAS/STAT Versus SAS Enterpri...
Clinical Trials Versus Health Outcomes Research: SAS/STAT Versus SAS Enterpri...Clinical Trials Versus Health Outcomes Research: SAS/STAT Versus SAS Enterpri...
Clinical Trials Versus Health Outcomes Research: SAS/STAT Versus SAS Enterpri...cambridgeWD
 
Common statistical pitfalls & errors in biomedical research (a top-5 list)
Common statistical pitfalls & errors in biomedical research (a top-5 list)Common statistical pitfalls & errors in biomedical research (a top-5 list)
Common statistical pitfalls & errors in biomedical research (a top-5 list)Evangelos Kritsotakis
 
Eugm 2012 pritchett - application of adaptive sample size re-estimation in ...
Eugm 2012   pritchett - application of adaptive sample size re-estimation in ...Eugm 2012   pritchett - application of adaptive sample size re-estimation in ...
Eugm 2012 pritchett - application of adaptive sample size re-estimation in ...Cytel USA
 
Presentation to CCG - Capita Health Freakononics v3
Presentation to CCG - Capita Health Freakononics v3Presentation to CCG - Capita Health Freakononics v3
Presentation to CCG - Capita Health Freakononics v3Mike Thorogood
 
33151-33161.ppt
33151-33161.ppt33151-33161.ppt
33151-33161.pptdawitg2
 
Clinical trial bms clinical trials methodology 17012018
Clinical trial bms   clinical trials methodology 17012018Clinical trial bms   clinical trials methodology 17012018
Clinical trial bms clinical trials methodology 17012018SoM
 
Metanalysis Lecture
Metanalysis LectureMetanalysis Lecture
Metanalysis Lecturedrmomusa
 
Big Data Analytics for Healthcare
Big Data Analytics for HealthcareBig Data Analytics for Healthcare
Big Data Analytics for HealthcareChandan Reddy
 
4. Update on non-invasive prenatal testing
4. Update on non-invasive prenatal testing4. Update on non-invasive prenatal testing
4. Update on non-invasive prenatal testingPHEScreening
 
Eugm 2011 pocock - dm cs-and-adaptive-trials
Eugm 2011   pocock - dm cs-and-adaptive-trialsEugm 2011   pocock - dm cs-and-adaptive-trials
Eugm 2011 pocock - dm cs-and-adaptive-trialsCytel USA
 
Chapter 9Multivariable MethodsObjectives• .docx
Chapter 9Multivariable MethodsObjectives• .docxChapter 9Multivariable MethodsObjectives• .docx
Chapter 9Multivariable MethodsObjectives• .docxspoonerneddy
 
Affect of Metabolic Obesity and Body Mass Index in Coronary Artery Diseases
Affect of Metabolic Obesity and Body Mass Index in Coronary Artery DiseasesAffect of Metabolic Obesity and Body Mass Index in Coronary Artery Diseases
Affect of Metabolic Obesity and Body Mass Index in Coronary Artery DiseasesNikhil Gupta
 
Survival analysis on kidney failure of kidney transplant patients
Survival analysis on kidney failure of kidney transplant patientsSurvival analysis on kidney failure of kidney transplant patients
Survival analysis on kidney failure of kidney transplant patientsDwaipayan Mukhopadhyay
 
Survival Analysis On Kidney Failure of Kidney Tranplant Patients
Survival Analysis On Kidney Failure of Kidney Tranplant PatientsSurvival Analysis On Kidney Failure of Kidney Tranplant Patients
Survival Analysis On Kidney Failure of Kidney Tranplant PatientsDwaipayan Mukhopadhyay
 
Adjusting for treatment switching in randomised controlled trials
Adjusting for treatment switching in randomised controlled trialsAdjusting for treatment switching in randomised controlled trials
Adjusting for treatment switching in randomised controlled trialscheweb1
 

Ähnlich wie Repeated events analyses (20)

Clinical Trials Versus Health Outcomes Research: SAS/STAT Versus SAS Enterpri...
Clinical Trials Versus Health Outcomes Research: SAS/STAT Versus SAS Enterpri...Clinical Trials Versus Health Outcomes Research: SAS/STAT Versus SAS Enterpri...
Clinical Trials Versus Health Outcomes Research: SAS/STAT Versus SAS Enterpri...
 
Clinical Trials Versus Health Outcomes Research: SAS/STAT Versus SAS Enterpri...
Clinical Trials Versus Health Outcomes Research: SAS/STAT Versus SAS Enterpri...Clinical Trials Versus Health Outcomes Research: SAS/STAT Versus SAS Enterpri...
Clinical Trials Versus Health Outcomes Research: SAS/STAT Versus SAS Enterpri...
 
Common statistical pitfalls & errors in biomedical research (a top-5 list)
Common statistical pitfalls & errors in biomedical research (a top-5 list)Common statistical pitfalls & errors in biomedical research (a top-5 list)
Common statistical pitfalls & errors in biomedical research (a top-5 list)
 
Eugm 2012 pritchett - application of adaptive sample size re-estimation in ...
Eugm 2012   pritchett - application of adaptive sample size re-estimation in ...Eugm 2012   pritchett - application of adaptive sample size re-estimation in ...
Eugm 2012 pritchett - application of adaptive sample size re-estimation in ...
 
Survival analysis
Survival analysisSurvival analysis
Survival analysis
 
Quantitative Synthesis I
Quantitative Synthesis IQuantitative Synthesis I
Quantitative Synthesis I
 
Presentation to CCG - Capita Health Freakononics v3
Presentation to CCG - Capita Health Freakononics v3Presentation to CCG - Capita Health Freakononics v3
Presentation to CCG - Capita Health Freakononics v3
 
33151-33161.ppt
33151-33161.ppt33151-33161.ppt
33151-33161.ppt
 
Clinical trial bms clinical trials methodology 17012018
Clinical trial bms   clinical trials methodology 17012018Clinical trial bms   clinical trials methodology 17012018
Clinical trial bms clinical trials methodology 17012018
 
Metanalysis Lecture
Metanalysis LectureMetanalysis Lecture
Metanalysis Lecture
 
Big Data Analytics for Healthcare
Big Data Analytics for HealthcareBig Data Analytics for Healthcare
Big Data Analytics for Healthcare
 
Pancholy SB - AIMRADIAL 2014 Endovascular - Renal denervation
Pancholy SB - AIMRADIAL 2014 Endovascular - Renal denervationPancholy SB - AIMRADIAL 2014 Endovascular - Renal denervation
Pancholy SB - AIMRADIAL 2014 Endovascular - Renal denervation
 
4. Update on non-invasive prenatal testing
4. Update on non-invasive prenatal testing4. Update on non-invasive prenatal testing
4. Update on non-invasive prenatal testing
 
Eugm 2011 pocock - dm cs-and-adaptive-trials
Eugm 2011   pocock - dm cs-and-adaptive-trialsEugm 2011   pocock - dm cs-and-adaptive-trials
Eugm 2011 pocock - dm cs-and-adaptive-trials
 
Chapter 9Multivariable MethodsObjectives• .docx
Chapter 9Multivariable MethodsObjectives• .docxChapter 9Multivariable MethodsObjectives• .docx
Chapter 9Multivariable MethodsObjectives• .docx
 
Sampling distributions
Sampling distributionsSampling distributions
Sampling distributions
 
Affect of Metabolic Obesity and Body Mass Index in Coronary Artery Diseases
Affect of Metabolic Obesity and Body Mass Index in Coronary Artery DiseasesAffect of Metabolic Obesity and Body Mass Index in Coronary Artery Diseases
Affect of Metabolic Obesity and Body Mass Index in Coronary Artery Diseases
 
Survival analysis on kidney failure of kidney transplant patients
Survival analysis on kidney failure of kidney transplant patientsSurvival analysis on kidney failure of kidney transplant patients
Survival analysis on kidney failure of kidney transplant patients
 
Survival Analysis On Kidney Failure of Kidney Tranplant Patients
Survival Analysis On Kidney Failure of Kidney Tranplant PatientsSurvival Analysis On Kidney Failure of Kidney Tranplant Patients
Survival Analysis On Kidney Failure of Kidney Tranplant Patients
 
Adjusting for treatment switching in randomised controlled trials
Adjusting for treatment switching in randomised controlled trialsAdjusting for treatment switching in randomised controlled trials
Adjusting for treatment switching in randomised controlled trials
 

Mehr von Mike LaValley

Survival analysis for lab scientists
Survival analysis for lab scientistsSurvival analysis for lab scientists
Survival analysis for lab scientistsMike LaValley
 
Basic survival analysis
Basic survival analysisBasic survival analysis
Basic survival analysisMike LaValley
 
Using R Software for Statistics in Lab Science
Using R Software for Statistics in Lab ScienceUsing R Software for Statistics in Lab Science
Using R Software for Statistics in Lab ScienceMike LaValley
 
Statistics for Lab Scientists
Statistics for Lab ScientistsStatistics for Lab Scientists
Statistics for Lab ScientistsMike LaValley
 
Intent-to-Treat (ITT) Analysis in Randomized Clinical Trials
Intent-to-Treat (ITT) Analysis in Randomized Clinical TrialsIntent-to-Treat (ITT) Analysis in Randomized Clinical Trials
Intent-to-Treat (ITT) Analysis in Randomized Clinical TrialsMike LaValley
 

Mehr von Mike LaValley (7)

Survival analysis for lab scientists
Survival analysis for lab scientistsSurvival analysis for lab scientists
Survival analysis for lab scientists
 
Basic survival analysis
Basic survival analysisBasic survival analysis
Basic survival analysis
 
To p or not to p
To p or not to pTo p or not to p
To p or not to p
 
Using R Software for Statistics in Lab Science
Using R Software for Statistics in Lab ScienceUsing R Software for Statistics in Lab Science
Using R Software for Statistics in Lab Science
 
Statistics for Lab Scientists
Statistics for Lab ScientistsStatistics for Lab Scientists
Statistics for Lab Scientists
 
Against all odds
Against all oddsAgainst all odds
Against all odds
 
Intent-to-Treat (ITT) Analysis in Randomized Clinical Trials
Intent-to-Treat (ITT) Analysis in Randomized Clinical TrialsIntent-to-Treat (ITT) Analysis in Randomized Clinical Trials
Intent-to-Treat (ITT) Analysis in Randomized Clinical Trials
 

Kürzlich hochgeladen

原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档
原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档
原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档208367051
 
Multiple time frame trading analysis -brianshannon.pdf
Multiple time frame trading analysis -brianshannon.pdfMultiple time frame trading analysis -brianshannon.pdf
Multiple time frame trading analysis -brianshannon.pdfchwongval
 
Advanced Machine Learning for Business Professionals
Advanced Machine Learning for Business ProfessionalsAdvanced Machine Learning for Business Professionals
Advanced Machine Learning for Business ProfessionalsVICTOR MAESTRE RAMIREZ
 
Conf42-LLM_Adding Generative AI to Real-Time Streaming Pipelines
Conf42-LLM_Adding Generative AI to Real-Time Streaming PipelinesConf42-LLM_Adding Generative AI to Real-Time Streaming Pipelines
Conf42-LLM_Adding Generative AI to Real-Time Streaming PipelinesTimothy Spann
 
Semantic Shed - Squashing and Squeezing.pptx
Semantic Shed - Squashing and Squeezing.pptxSemantic Shed - Squashing and Squeezing.pptx
Semantic Shed - Squashing and Squeezing.pptxMike Bennett
 
Heart Disease Classification Report: A Data Analysis Project
Heart Disease Classification Report: A Data Analysis ProjectHeart Disease Classification Report: A Data Analysis Project
Heart Disease Classification Report: A Data Analysis ProjectBoston Institute of Analytics
 
Predictive Analysis for Loan Default Presentation : Data Analysis Project PPT
Predictive Analysis for Loan Default  Presentation : Data Analysis Project PPTPredictive Analysis for Loan Default  Presentation : Data Analysis Project PPT
Predictive Analysis for Loan Default Presentation : Data Analysis Project PPTBoston Institute of Analytics
 
Decoding the Heart: Student Presentation on Heart Attack Prediction with Data...
Decoding the Heart: Student Presentation on Heart Attack Prediction with Data...Decoding the Heart: Student Presentation on Heart Attack Prediction with Data...
Decoding the Heart: Student Presentation on Heart Attack Prediction with Data...Boston Institute of Analytics
 
Top 5 Best Data Analytics Courses In Queens
Top 5 Best Data Analytics Courses In QueensTop 5 Best Data Analytics Courses In Queens
Top 5 Best Data Analytics Courses In Queensdataanalyticsqueen03
 
How we prevented account sharing with MFA
How we prevented account sharing with MFAHow we prevented account sharing with MFA
How we prevented account sharing with MFAAndrei Kaleshka
 
毕业文凭制作#回国入职#diploma#degree美国加州州立大学北岭分校毕业证成绩单pdf电子版制作修改#毕业文凭制作#回国入职#diploma#de...
毕业文凭制作#回国入职#diploma#degree美国加州州立大学北岭分校毕业证成绩单pdf电子版制作修改#毕业文凭制作#回国入职#diploma#de...毕业文凭制作#回国入职#diploma#degree美国加州州立大学北岭分校毕业证成绩单pdf电子版制作修改#毕业文凭制作#回国入职#diploma#de...
毕业文凭制作#回国入职#diploma#degree美国加州州立大学北岭分校毕业证成绩单pdf电子版制作修改#毕业文凭制作#回国入职#diploma#de...ttt fff
 
Real-Time AI Streaming - AI Max Princeton
Real-Time AI  Streaming - AI Max PrincetonReal-Time AI  Streaming - AI Max Princeton
Real-Time AI Streaming - AI Max PrincetonTimothy Spann
 
FAIR, FAIRsharing, FAIR Cookbook and ELIXIR - Sansone SA - Boston 2024
FAIR, FAIRsharing, FAIR Cookbook and ELIXIR - Sansone SA - Boston 2024FAIR, FAIRsharing, FAIR Cookbook and ELIXIR - Sansone SA - Boston 2024
FAIR, FAIRsharing, FAIR Cookbook and ELIXIR - Sansone SA - Boston 2024Susanna-Assunta Sansone
 
办理(Vancouver毕业证书)加拿大温哥华岛大学毕业证成绩单原版一比一
办理(Vancouver毕业证书)加拿大温哥华岛大学毕业证成绩单原版一比一办理(Vancouver毕业证书)加拿大温哥华岛大学毕业证成绩单原版一比一
办理(Vancouver毕业证书)加拿大温哥华岛大学毕业证成绩单原版一比一F La
 
Data Analysis Project : Targeting the Right Customers, Presentation on Bank M...
Data Analysis Project : Targeting the Right Customers, Presentation on Bank M...Data Analysis Project : Targeting the Right Customers, Presentation on Bank M...
Data Analysis Project : Targeting the Right Customers, Presentation on Bank M...Boston Institute of Analytics
 
LLMs, LMMs, their Improvement Suggestions and the Path towards AGI
LLMs, LMMs, their Improvement Suggestions and the Path towards AGILLMs, LMMs, their Improvement Suggestions and the Path towards AGI
LLMs, LMMs, their Improvement Suggestions and the Path towards AGIThomas Poetter
 
IMA MSN - Medical Students Network (2).pptx
IMA MSN - Medical Students Network (2).pptxIMA MSN - Medical Students Network (2).pptx
IMA MSN - Medical Students Network (2).pptxdolaknnilon
 
专业一比一美国俄亥俄大学毕业证成绩单pdf电子版制作修改
专业一比一美国俄亥俄大学毕业证成绩单pdf电子版制作修改专业一比一美国俄亥俄大学毕业证成绩单pdf电子版制作修改
专业一比一美国俄亥俄大学毕业证成绩单pdf电子版制作修改yuu sss
 
Business Analytics using Microsoft Excel
Business Analytics using Microsoft ExcelBusiness Analytics using Microsoft Excel
Business Analytics using Microsoft Excelysmaelreyes
 
MK KOMUNIKASI DATA (TI)komdat komdat.docx
MK KOMUNIKASI DATA (TI)komdat komdat.docxMK KOMUNIKASI DATA (TI)komdat komdat.docx
MK KOMUNIKASI DATA (TI)komdat komdat.docxUnduhUnggah1
 

Kürzlich hochgeladen (20)

原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档
原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档
原版1:1定制南十字星大学毕业证(SCU毕业证)#文凭成绩单#真实留信学历认证永久存档
 
Multiple time frame trading analysis -brianshannon.pdf
Multiple time frame trading analysis -brianshannon.pdfMultiple time frame trading analysis -brianshannon.pdf
Multiple time frame trading analysis -brianshannon.pdf
 
Advanced Machine Learning for Business Professionals
Advanced Machine Learning for Business ProfessionalsAdvanced Machine Learning for Business Professionals
Advanced Machine Learning for Business Professionals
 
Conf42-LLM_Adding Generative AI to Real-Time Streaming Pipelines
Conf42-LLM_Adding Generative AI to Real-Time Streaming PipelinesConf42-LLM_Adding Generative AI to Real-Time Streaming Pipelines
Conf42-LLM_Adding Generative AI to Real-Time Streaming Pipelines
 
Semantic Shed - Squashing and Squeezing.pptx
Semantic Shed - Squashing and Squeezing.pptxSemantic Shed - Squashing and Squeezing.pptx
Semantic Shed - Squashing and Squeezing.pptx
 
Heart Disease Classification Report: A Data Analysis Project
Heart Disease Classification Report: A Data Analysis ProjectHeart Disease Classification Report: A Data Analysis Project
Heart Disease Classification Report: A Data Analysis Project
 
Predictive Analysis for Loan Default Presentation : Data Analysis Project PPT
Predictive Analysis for Loan Default  Presentation : Data Analysis Project PPTPredictive Analysis for Loan Default  Presentation : Data Analysis Project PPT
Predictive Analysis for Loan Default Presentation : Data Analysis Project PPT
 
Decoding the Heart: Student Presentation on Heart Attack Prediction with Data...
Decoding the Heart: Student Presentation on Heart Attack Prediction with Data...Decoding the Heart: Student Presentation on Heart Attack Prediction with Data...
Decoding the Heart: Student Presentation on Heart Attack Prediction with Data...
 
Top 5 Best Data Analytics Courses In Queens
Top 5 Best Data Analytics Courses In QueensTop 5 Best Data Analytics Courses In Queens
Top 5 Best Data Analytics Courses In Queens
 
How we prevented account sharing with MFA
How we prevented account sharing with MFAHow we prevented account sharing with MFA
How we prevented account sharing with MFA
 
毕业文凭制作#回国入职#diploma#degree美国加州州立大学北岭分校毕业证成绩单pdf电子版制作修改#毕业文凭制作#回国入职#diploma#de...
毕业文凭制作#回国入职#diploma#degree美国加州州立大学北岭分校毕业证成绩单pdf电子版制作修改#毕业文凭制作#回国入职#diploma#de...毕业文凭制作#回国入职#diploma#degree美国加州州立大学北岭分校毕业证成绩单pdf电子版制作修改#毕业文凭制作#回国入职#diploma#de...
毕业文凭制作#回国入职#diploma#degree美国加州州立大学北岭分校毕业证成绩单pdf电子版制作修改#毕业文凭制作#回国入职#diploma#de...
 
Real-Time AI Streaming - AI Max Princeton
Real-Time AI  Streaming - AI Max PrincetonReal-Time AI  Streaming - AI Max Princeton
Real-Time AI Streaming - AI Max Princeton
 
FAIR, FAIRsharing, FAIR Cookbook and ELIXIR - Sansone SA - Boston 2024
FAIR, FAIRsharing, FAIR Cookbook and ELIXIR - Sansone SA - Boston 2024FAIR, FAIRsharing, FAIR Cookbook and ELIXIR - Sansone SA - Boston 2024
FAIR, FAIRsharing, FAIR Cookbook and ELIXIR - Sansone SA - Boston 2024
 
办理(Vancouver毕业证书)加拿大温哥华岛大学毕业证成绩单原版一比一
办理(Vancouver毕业证书)加拿大温哥华岛大学毕业证成绩单原版一比一办理(Vancouver毕业证书)加拿大温哥华岛大学毕业证成绩单原版一比一
办理(Vancouver毕业证书)加拿大温哥华岛大学毕业证成绩单原版一比一
 
Data Analysis Project : Targeting the Right Customers, Presentation on Bank M...
Data Analysis Project : Targeting the Right Customers, Presentation on Bank M...Data Analysis Project : Targeting the Right Customers, Presentation on Bank M...
Data Analysis Project : Targeting the Right Customers, Presentation on Bank M...
 
LLMs, LMMs, their Improvement Suggestions and the Path towards AGI
LLMs, LMMs, their Improvement Suggestions and the Path towards AGILLMs, LMMs, their Improvement Suggestions and the Path towards AGI
LLMs, LMMs, their Improvement Suggestions and the Path towards AGI
 
IMA MSN - Medical Students Network (2).pptx
IMA MSN - Medical Students Network (2).pptxIMA MSN - Medical Students Network (2).pptx
IMA MSN - Medical Students Network (2).pptx
 
专业一比一美国俄亥俄大学毕业证成绩单pdf电子版制作修改
专业一比一美国俄亥俄大学毕业证成绩单pdf电子版制作修改专业一比一美国俄亥俄大学毕业证成绩单pdf电子版制作修改
专业一比一美国俄亥俄大学毕业证成绩单pdf电子版制作修改
 
Business Analytics using Microsoft Excel
Business Analytics using Microsoft ExcelBusiness Analytics using Microsoft Excel
Business Analytics using Microsoft Excel
 
MK KOMUNIKASI DATA (TI)komdat komdat.docx
MK KOMUNIKASI DATA (TI)komdat komdat.docxMK KOMUNIKASI DATA (TI)komdat komdat.docx
MK KOMUNIKASI DATA (TI)komdat komdat.docx
 

Repeated events analyses

  • 2. Outline for Talk  Recurrent Events  Analyses for Count Outcomes  Survival Analysis Approaches for Repeated Events  Parametric Analyses for Counts  Poisson Regression  Negative Binomial  Non-parametric Analyses for Counts  Mean of Ratios (Q)  Ratio of Means (R) 2
  • 3. Recurrent Events  Usually in survival studies subjects are followed until they experience a single index event  Death  Cancer  Return to employment  But there are many events of interest that are repeatable  Seizures  Infections  Hospitalizations 3
  • 4. Recurrent Events  Most of the methods employed in survival analysis are set up for a single event  But, there have been methods developed for use with repeated events  Some of these methods come out of the area of statistical methods developed for count data – these focus on the count of the number of recurrent event  Some methods come directly out of survival analysis methods – these focus on the time duration between the recurrent events 4
  • 5. Recurrent Events  Methods coming out of either area need to address the domain of the other methods  The count methods also need to address the time component  The survival methods need to address the number of events as well as the gap times between events 5
  • 6. Diabetes Control and Complications Trial (DCCT)  NIH-funded trial  Launched in 1981  To evaluate the effect of intensive blood glucose- lowering on risk of albuminuria in diabetic subjects  Subjects were randomized to either intensive blood glucose lowering or conventional treatment  Intensive glucose lowering used self-monitoring 4 or 5 times daily, multiple daily insulin injections or a pump, diet and exercise 6
  • 7. Diabetes Control and Complications Trial (DCCT)  A concern was that the intensive glucose lowering could lead to hypoglycemia  Dizzy spells  Possible comas  Seizures  Hypoglycemia events were tracked as a secondary outcome (DCCT 1997) 7
  • 8. Diabetes Control and Complications Trial (DCCT) Intensive Treatment Conventional Treatment Subjects 363 352 Hypoglycemia Events 1723 543 Person-Years of Follow-Up 2598.5 2480.8 Rate per 100 Person-Years 66.3 21.9 8
  • 9. Count Outcome Approaches  The most basic analysis for count outcomes comes from the Poisson distribution  𝑃𝑃 𝑦𝑦|𝜆𝜆 = 𝑒𝑒−𝜆𝜆 𝜆𝜆 𝑦𝑦 𝑦𝑦!  Here 𝜆𝜆 is the mean count  y is the observed count (non-negative integer)  One feature of the Poisson distribution is that  Variance = Mean = 𝜆𝜆 9
  • 11. Count Outcome Approaches  The maximum likelihood estimator of the mean is provided by the sample average  ̂𝜆𝜆 = 𝑦𝑦1+𝑦𝑦2+⋯+𝑦𝑦𝑛𝑛 𝑛𝑛  Often people are followed for somewhat different lengths of time, giving some people more opportunity to get larger event counts  In this case 𝜆𝜆 is standardized to be per-unit of time 11
  • 12. Count Outcome Approaches  With unequal follow-up times 𝑡𝑡1, 𝑡𝑡2, … , 𝑡𝑡𝑛𝑛  𝑃𝑃 𝑦𝑦|𝑡𝑡, 𝜆𝜆 = 𝑒𝑒−𝜆𝜆𝑡𝑡(𝜆𝜆𝑡𝑡)𝑦𝑦 𝑦𝑦!  Where 𝑡𝑡 is the length of follow-up  With unequal follow-up, the MLE for 𝜆𝜆 is provided by  ̂𝜆𝜆 = 𝑦𝑦1+𝑦𝑦2+⋯+𝑦𝑦𝑛𝑛 𝑡𝑡1+𝑡𝑡2+⋯+𝑡𝑡𝑛𝑛 = �𝑦𝑦 ̅𝑡𝑡 = 𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛 𝑜𝑜𝑜𝑜 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝−𝑡𝑡𝑡𝑡 𝑡𝑡𝑡𝑡 𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 𝑓𝑓−𝑢𝑢𝑢𝑢 12
  • 13. Count Outcome Approaches  With unequal follow-up times 𝑡𝑡1, 𝑡𝑡2, … , 𝑡𝑡𝑛𝑛  𝑃𝑃 𝑦𝑦|𝑡𝑡, 𝜆𝜆 = 𝑒𝑒−𝜆𝜆𝑡𝑡(𝜆𝜆𝑡𝑡)𝑦𝑦 𝑦𝑦!  Where 𝑡𝑡 is the length of follow-up  With unequal follow-up, the MLE for 𝜆𝜆 is provided by  ̂𝜆𝜆 = 𝑦𝑦1+𝑦𝑦2+⋯+𝑦𝑦𝑛𝑛 𝑡𝑡1+𝑡𝑡2+⋯+𝑡𝑡𝑛𝑛 = �𝑦𝑦 ̅𝑡𝑡 = 𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛 𝑜𝑜𝑜𝑜 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝−𝑡𝑡𝑡𝑡 𝑡𝑡𝑡𝑡 𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 𝑓𝑓−𝑢𝑢𝑢𝑢  We will see this again 13
  • 14. Link From Poisson to Exponential  If the count of events follows a Poisson distribution, then the times between events follow an exponential distribution  𝑓𝑓 𝑡𝑡 𝜆𝜆 = 𝜆𝜆𝑒𝑒−𝜆𝜆𝜆𝜆  A feature of exponential survival times is that the hazard event rate is 𝜆𝜆 and is constant over time  The mean time to an event is 1 𝜆𝜆  Gives a way to estimate 𝜆𝜆 in a survival analysis setting (rather than a count data setting) 14
  • 16. Poisson Regression  In Poisson regression we model the expected count for an observation as a function of predictor variables  Poisson regression is a log-linear model so that the log of the mean is set equal to a linear combination of the predictors 16
  • 17. Poisson Regression  𝑙𝑙𝑙𝑙 𝑙𝑙 𝜆𝜆𝜆𝜆 = 𝛽𝛽0 + 𝛽𝛽1 𝑥𝑥  This leads to  𝑙𝑙𝑙𝑙 𝑙𝑙 𝜆𝜆 + 𝑙𝑙𝑙𝑙 𝑙𝑙 𝑡𝑡 = 𝛽𝛽0 + 𝛽𝛽1 𝑥𝑥  So, we get  𝑙𝑙𝑙𝑙 𝑙𝑙 𝜆𝜆 = 𝛽𝛽0 + 𝛽𝛽1 𝑥𝑥 − 𝑙𝑙𝑙𝑙 𝑙𝑙 𝑡𝑡  The −𝑙𝑙𝑙𝑙 𝑙𝑙 𝑡𝑡 is called the offset and corrects for the differences in follow-up time between subjects  The model can be expressed 𝜆𝜆 = 𝑒𝑒 𝛽𝛽0+𝛽𝛽1 𝑥𝑥−𝑙𝑙𝑙𝑙𝑙𝑙 𝑡𝑡 17
  • 18. Poisson Regression  This model is fit using maximum likelihood  SAS GENMOD  The 𝛽𝛽 estimates for predictors are interpreted as logs of incidence rate ratios  A one-unit increase in 𝑥𝑥 gives a ratio of 𝑒𝑒𝑒𝑒𝑒𝑒 𝛽𝛽1 in the expected count  If 𝑥𝑥 codes for a group difference (0 = control, 1 = treatment), then 𝑒𝑒𝑒𝑒𝑒𝑒 𝛽𝛽1 corresponds to the ratio of counts in the treated group divided by counts in the control group 18
  • 19. Poisson Regression in DCCT proc genmod; model nevents = group / dist = poisson link = log offset = lnyears; TITLE1 'Poisson regression models of risk of hypoglycemia'; title2 'unadjusted treatment group effect'; 19
  • 20. Poisson Regression in DCCT 20 Analysis Of Parameter Estimates Parameter DF Estimate Standard Error Wald 95% Confidence Limits Chi-Square Pr > ChiSq Intercept 1 -1.5190 0.0429 -1.6031 -1.4349 1252.90 <.0001 group Exp 1 1.1081 0.0492 1.0117 1.2046 507.01 <.0001 group Std 0 0.0000 0.0000 0.0000 0.0000 . . Scale 0 1.0000 0.0000 1.0000 1.0000
  • 21. Poisson Regression in DCCT 21 Highly significant result! But, is it based on a reasonable model for the data? Analysis Of Parameter Estimates Parameter DF Estimate Standard Error Wald 95% Confidence Limits Chi-Square Pr > ChiSq Intercept 1 -1.5190 0.0429 -1.6031 -1.4349 1252.90 <.0001 group Exp 1 1.1081 0.0492 1.0117 1.2046 507.01 <.0001 group Std 0 0.0000 0.0000 0.0000 0.0000 . . Scale 0 1.0000 0.0000 1.0000 1.0000
  • 22. Poisson Regression  The big limitation for Poisson regression is overdispersion  The data are overdispersed when the variation in the observed counts is greater that the mean value  Under Poisson model these should match  Variance in the counts much greater than the means  Goodness of fit testing for overdispersion is typically done in Poisson regression 22
  • 23. Overdispersion in DCCT 23 Criteria For Assessing Goodness Of Fit Criterion DF Value Value/DF Deviance 713 3928.7828 5.5102 Scaled Deviance 713 3928.7828 5.5102 Pearson Chi-Square 713 5131.3429 7.1968 Scaled Pearson X2 713 5131.3429 7.1968 Log Likelihood 775.1804
  • 24. Overdispersion in DCCT 24 Criteria For Assessing Goodness Of Fit Criterion DF Value Value/DF Deviance 713 3928.7828 5.5102 Scaled Deviance 713 3928.7828 5.5102 Pearson Chi-Square 713 5131.3429 7.1968 Scaled Pearson X2 713 5131.3429 7.1968 Log Likelihood 775.1804 If model fits well then Value/DF should be close to 1.0 • Large ratios indicate overdispersion or model misspecification
  • 25. Poisson Regression  For a clinical trial, our primary goal is a good test of treatment effect, so wouldn’t normally need to include covariates  But, could try to remove excess variation by controlling for covariates 25
  • 26. Poisson Regression in DCCT proc genmod; class group; model nevents = group insulin duration female adult bcval5 hbael hxcoma / dist = poisson link = log offset = lnyears covb; title2 'covariate adjusted treatment group effect'; 26
  • 27. Poisson Regression in DCCT 27 Analysis Of Parameter Estimates Parameter DF Estimate Standard Error Wald 95% Confidence Limits Chi-Square Pr > ChiSq Intercept 1 -0.9568 0.2174 -1.3829 -0.5308 19.38 <.0001 group Exp 1 1.0845 0.0493 0.9879 1.1812 483.91 <.0001 group Std 0 0.0000 0.0000 0.0000 0.0000 . . insulin 1 0.0051 0.0995 -0.1898 0.2000 0.00 0.9593 duration 1 0.0015 0.0006 0.0004 0.0026 6.79 0.0092 female 1 0.1794 0.0424 0.0963 0.2624 17.93 <.0001 adult 1 -0.5980 0.0656 -0.7265 -0.4694 83.13 <.0001 bcval5 1 -0.5283 0.3630 -1.2398 0.1833 2.12 0.1456 hbael 1 -0.0335 0.0151 -0.0631 -0.0038 4.89 0.0271 hxcoma 1 0.6010 0.0685 0.4669 0.7352 77.09 <.0001 Scale 0 1.0000 0.0000 1.0000 1.0000
  • 28. Poisson Regression in DCCT 28 Analysis Of Parameter Estimates Parameter DF Estimate Standard Error Wald 95% Confidence Limits Chi-Square Pr > ChiSq Intercept 1 -0.9568 0.2174 -1.3829 -0.5308 19.38 <.0001 group Exp 1 1.0845 0.0493 0.9879 1.1812 483.91 <.0001 group Std 0 0.0000 0.0000 0.0000 0.0000 . . insulin 1 0.0051 0.0995 -0.1898 0.2000 0.00 0.9593 duration 1 0.0015 0.0006 0.0004 0.0026 6.79 0.0092 female 1 0.1794 0.0424 0.0963 0.2624 17.93 <.0001 adult 1 -0.5980 0.0656 -0.7265 -0.4694 83.13 <.0001 bcval5 1 -0.5283 0.3630 -1.2398 0.1833 2.12 0.1456 hbael 1 -0.0335 0.0151 -0.0631 -0.0038 4.89 0.0271 hxcoma 1 0.6010 0.0685 0.4669 0.7352 77.09 <.0001 Scale 0 1.0000 0.0000 1.0000 1.0000 Still very significant, but does the model fit better?
  • 29. Poisson Regression in DCCT 29 Criteria For Assessing Goodness Of Fit Criterion DF Value Value/DF Deviance 706 3707.7027 5.2517 Scaled Deviance 706 3707.7027 5.2517 Pearson Chi-Square 706 4792.7876 6.7887 Scaled Pearson X2 706 4792.7876 6.7887 Log Likelihood 885.7204 The fit is better with the covariates included, but still very overdispersed.
  • 30. Poisson Regression  Another approach that can be used to remove overdispersion is to use a variance correction  Pearson correction  Standard approach – has the actual variance equal the modeled variance multiplied by an estimated overdispersion parameter  Corrects the standard errors and test results to account for the overdispersion  GEE correction based on robust sandwich variance estimator  Based on a robust sandwich-variance estimator 30
  • 31. GEE Poisson Regression in DCCT proc genmod; class group subnum; model nevents = group / dist = poisson link = log offset = lnyears type3; repeated subject=subnum / type=unstr; title2 ‘GEE unadjusted treatment group effect'; 31
  • 32. GEE Poisson Regression in DCCT 32 Analysis Of GEE Parameter Estimates Empirical Standard Error Estimates Parameter Estimate Standard Error 95% Confidence Limits Z Pr > |Z| Intercept -1.5190 0.1003 -1.7155 -1.3225 -15.15 <.0001 group Exp 1.1081 0.1256 0.8620 1.3543 8.82 <.0001 group Std 0.0000 0.0000 0.0000 0.0000 . .
  • 33. GEE Poisson Regression in DCCT 33 Analysis Of GEE Parameter Estimates Empirical Standard Error Estimates Parameter Estimate Standard Error 95% Confidence Limits Z Pr > |Z| Intercept -1.5190 0.1003 -1.7155 -1.3225 -15.15 <.0001 group Exp 1.1081 0.1256 0.8620 1.3543 8.82 <.0001 group Std 0.0000 0.0000 0.0000 0.0000 . . Still very significant treatment effect, but note that standard error is much larger in this model. Standard error was 0.0492 in Poisson model with only treatment.
  • 34. Negative Binomial Regression  The negative binomial distribution is another count data distribution with more variance than the Poisson  Can be used to give the probability that n trials are required in order to get m successes (m≤n)  Model takes the form 𝜆𝜆 = 𝑒𝑒 𝛽𝛽0+𝛽𝛽1 𝑥𝑥−𝑙𝑙𝑙𝑙𝑙𝑙 𝑡𝑡 +𝜀𝜀  Where 𝑒𝑒𝑒𝑒𝑒𝑒 𝜀𝜀 has a gamma distribution  Can be considered to be a Poisson model with gamma- distributed random effects  Parametric approach to overdispersion 34
  • 35. Negative Binomial Regression in DCCT proc genmod; class group; model nevents = group / dist = negbin link = log offset = lnyears; title2 ‘Neg Binomial unadjusted treatment group effect'; 35
  • 36. Negative Binomial Regression in DCCT36 Analysis Of Parameter Estimates Parameter DF Estimate Standard Error Wald 95% Confidence Limits Chi-Square Pr > ChiSq Intercept 1 -1.5510 0.0902 -1.728 -1.374 295.60 <.0001 group Exp 1 1.1173 0.1215 0.8791 1.3555 84.52 <.0001 group Std 0 0.0000 0.0000 0.0000 0.0000 . . Dispersion 1 2.1863 0.1649 1.8631 2.5095
  • 37. Negative Binomial Regression in DCCT37 Analysis Of Parameter Estimates Parameter DF Estimate Standard Error Wald 95% Confidence Limits Chi-Square Pr > ChiSq Intercept 1 -1.5510 0.0902 -1.728 -1.374 295.60 <.0001 group Exp 1 1.1173 0.1215 0.8791 1.3555 84.52 <.0001 group Std 0 0.0000 0.0000 0.0000 0.0000 . . Dispersion 1 2.1863 0.1649 1.8631 2.5095 The results here are quite similar to those of the GEE Poisson model, with almost the same standard errors.
  • 39. Survival Analyses  Repeated events  Cox model on time to first event  Use Andersen-Gill approach and partition follow-up time according to which event it applies  Restart the follow-up time clock after an event  Multiple observations per subject  Need to be concerned about correlation  Unobserved heterogeneity (due to the correlation) can bias estimates downward while significance is overstated 39
  • 40. Survival Analysis in DCCT proc phreg data=three; model stopday*event(0) = intgroup/ risklimits; title1 'Cox Model for First Event'; run; 40
  • 41. Survival Analysis in DCCT 41 Analysis of Maximum Likelihood Estimates Variable DF Parameter Estimate Standard Error Chi-Square Pr > ChiSq Hazard Ratio 95% Hazard Ratio Confidence Limits INTGROUP 1 0.77252 0.10354 55.6711 <.0001 2.165 1.768 2.652
  • 42. Survival Analysis in DCCT proc phreg data=four; model gaptime*event(0) = intgroup priorgap / risklimits; title1 'Cox Model for Second Event - Correlation'; run; 42
  • 43. Survival Analysis in DCCT 43 Analysis of Maximum Likelihood Estimates Variable DF Parameter Estimate Standard Error Chi-Square Pr > ChiSq Hazard Ratio 95% Hazard Ratio Confidence Limits INTGROUP 1 0.27790 0.12451 4.9814 0.0256 1.320 1.034 1.685 priorgap 1 -0.000523 0.000114 20.9317 <.0001 0.999 0.999 1.000 So, having a longer time to the first event reduces risk of the second event. So correlation in these times to events is quite substantial.
  • 44. Survival Analysis in DCCT proc phreg data=six covsandwich(aggregate); model gaptime*event(0) = intgroup / risklimits; id patient; title1 'GEE Cox Model for all Events'; run; 44
  • 45. Survival Analysis in DCCT 45 Testing Global Null Hypothesis: BETA=0 Test Chi-Square DF Pr > ChiSq Likelihood Ratio 74.8679 1 <.0001 Score (Model-Based) 72.4448 1 <.0001 Score (Sandwich) 51.8412 1 <.0001 Wald (Model-Based) 71.4253 1 <.0001 Wald (Sandwich) 40.2564 1 <.0001
  • 46. Survival Analysis in DCCT 46 Analysis of Maximum Likelihood Estimates Variable DF Parameter Estimate Standard Error StdErr Ratio Chi-Square Pr > ChiSq Hazard Ratio 95% Hazard Ratio Confidence Limits INTGROUP 1 0.67526 0.08988 1.808 56.4435 <.0001 1.965 1.647 2.343 So, accounting for the correlation increases the standard error by about 80%. However, the resulting hazard ratio is much less than the other estimates so far.
  • 47. Nonparametric Analyses  Work done with Jing Xu  Motivated by his experiences with analysis of repeated events at  The SDAC for the AIDS Clinical Trials Group  Later experience in the pharmaceutical industry 47
  • 48. Nonparametric Analyses  In follow-up studies of recurrent events, there were studies taking a non-parametric approach to analysis of two-arm trials by  𝑞𝑞𝑖𝑖𝑖𝑖 = 𝑦𝑦𝑖𝑖𝑖𝑖 𝑡𝑡𝑖𝑖𝑖𝑖  𝑦𝑦𝑖𝑖𝑖𝑖 is the number of events for subject i in arm j (j is 0 or 1)  𝑡𝑡𝑖𝑖𝑖𝑖 is the total length of follow-up for subject i in group j  𝑞𝑞𝑖𝑖𝑖𝑖 represents a subject-specific event rate 48
  • 49. Nonparametric Analyses  𝑞𝑞𝑖𝑖𝑖𝑖 were analyzed without assuming any particular distribution by standard two-sample tests  Student t-test  Wilcoxon test  Van der Waerden test 49
  • 50. Nonparametric Analyses  For estimation, the means of the subject-specific event rates were calculated for each group  �𝑄𝑄𝑗𝑗 = 1 𝑛𝑛𝑗𝑗 ∑𝑖𝑖=1 𝑛𝑛𝑗𝑗 𝑞𝑞𝑖𝑖𝑖𝑖 = 1 𝑛𝑛𝑗𝑗 ∑𝑖𝑖=1 𝑛𝑛𝑗𝑗 𝑦𝑦𝑖𝑖𝑖𝑖 𝑡𝑡𝑖𝑖𝑖𝑖  These �𝑄𝑄𝑗𝑗 could then be used to estimate ratios of the rates and differences in the rates by group  �𝑅𝑅 𝑅𝑅𝑄𝑄 = �𝑄𝑄1 �𝑄𝑄0  �𝑅𝑅 𝑅𝑅𝑄𝑄 = �𝑄𝑄1 − �𝑄𝑄0 50
  • 51. Nonparametric Analyses  Our concern about this 𝑞𝑞𝑖𝑖𝑖𝑖 approach was that  There could be outliers – subjects having a large number of events in a short amount of time who then drop out of the study early  The asymptotic consistency of the �𝑄𝑄𝑗𝑗 estimators depends on both the numbers of events and the amount of follow-up increasing within all subjects 51
  • 52. Nonparametric Analyses  A different approach had been suggested by L.J. Wei  �𝑅𝑅𝑗𝑗 = 𝑥𝑥𝑗𝑗 𝑡𝑡𝑗𝑗 = � ∑𝑖𝑖=1 𝑛𝑛𝑗𝑗 𝑥𝑥𝑖𝑖𝑖𝑖 ∑𝑖𝑖=1 𝑛𝑛𝑗𝑗 𝑡𝑡𝑖𝑖𝑖𝑖 = 𝑥𝑥𝑗𝑗 �𝑡𝑡𝑗𝑗  Ratio of the means, not the mean of the ratios 52
  • 53. Nonparametric Analyses  A different approach had been suggested by L.J. Wei  �𝑅𝑅𝑗𝑗 = 𝑥𝑥𝑗𝑗 𝑡𝑡𝑗𝑗 = � ∑𝑖𝑖=1 𝑛𝑛𝑗𝑗 𝑥𝑥𝑖𝑖𝑖𝑖 ∑𝑖𝑖=1 𝑛𝑛𝑗𝑗 𝑡𝑡𝑖𝑖𝑖𝑖 = 𝑥𝑥𝑗𝑗 �𝑡𝑡𝑗𝑗  Ratio of the means, not the mean of the ratios  This is just the standard Poisson events per person- time without assuming the distribution 53
  • 54. Nonparametric Analyses  Can use the �𝑅𝑅𝑗𝑗 to estimate rate ratio and difference in rates between groups  �𝑅𝑅𝑅𝑅𝑅𝑅 = �𝑅𝑅1 �𝑅𝑅0  �𝑅𝑅 𝑅𝑅𝑅𝑅 = �𝑅𝑅1 − �𝑅𝑅0 54
  • 55. Nonparametric Analyses  �𝑅𝑅𝑗𝑗 advantages  Average out any outliers  Makes consistency more reasonable as it uses the group totals for number of events and total follow-up time  �𝑅𝑅𝑗𝑗 disadvantages  Measure is group-specific and not subject-specific  Can’t use standard 2-sample tests 55
  • 56. Nonparametric Analyses  Asymptotic forms of the variance for the �𝑅𝑅 measures are in the Xu and LaValley paper in the references  Confidence intervals based on Fieller’s method and resampling  In this talk, I’ll use resampling methods to evaluate the �𝑅𝑅𝑗𝑗 measures  Permutation tests  Bootstrap confidence intervals 56
  • 57. 57
  • 58. Nonparametric Analyses  In simulations, we found that the �𝑄𝑄𝑗𝑗 with a non- parametric test and the �𝑅𝑅𝑗𝑗 methods maintained the type-1 error and had comparable power over a range of sample sizes  The �𝑅𝑅𝑗𝑗 methods provided reasonable confidence interval coverage if the sample sizes were at least 100 per group 58
  • 59. Nonparametric Analyses  In the paper we use the nonparametric methods on a dataset for recurrence of bladder cancer in a two-arm clinical trial of the drug thiotepa 59
  • 60. Nonparametric Analyses  Thiotepa trial results  Results are uniformly non-significant, with rate ratios from 0.71 to 0.84 60
  • 61. Nonparametric Analyses  Thiotepa trial results  These are the results that I trust the most out of these 61
  • 62. Nonparametric Analyses in DCCT 62 Intensive Treatment Conventional Treatment 𝑛𝑛 Subjects 363 352 �𝑦𝑦 Hypoglycemia Events/Subject 4.75 1.54 ̅𝑡𝑡 Person-Years of Follow-Up/Subject 7.16 7.05 � �𝑦𝑦 ̅𝑡𝑡 �𝑅𝑅𝑗𝑗 0.66 0.22 �𝑄𝑄𝑗𝑗 0.65 0.21
  • 63. Nonparametric Analyses in DCCT 63 Distribution of qij by group
  • 64. Nonparametric Analyses in DCCT 64 Test Test Statistic P-value t-test of qij 8.28 P < 0.0001 Van der Waerden test of qij 8.77 P < 0.0001 Wilcoxon test of qij 8.57 P < 0.0001
  • 65. Nonparametric Analyses in DCCT 65 Estimator Estimate Permutation Test P-value* Rate Difference (Q) 0.43681 P < 0.0005 Rate Difference (R) 0.44415 P < 0.0005 Rate Ratio (Q) 3.08059 P < 0.0005 Rate Ratio (R) 3.02872 P < 0.0005 *Based on 2000 permutations
  • 66. Nonparametric Analyses in DCCT 66 Q versus R measures of rate difference across 2000 permuted datasets
  • 67. Nonparametric Analyses in DCCT 67 Estimator Estimate Bootstrap Confidence Interval Rate Difference (Q) 0.43681 (0.329, 0.549) Rate Difference (R) 0.44415 (0.329, 0.561) Rate Ratio (Q) 3.08059 (2.399, 3.979) Rate Ratio (R) 3.02872 (2.333, 3.951)
  • 68. Nonparametric Analyses in DCCT 68 Q versus R measures of rate difference across 2000 bootstrap samples
  • 69. Nonparametric Analyses  Both Q and R measures work well for reasonable sized datasets  In these DCCT data, both are fine  In the bladder cancer data, the R (ratio of means) seems to have a slight edge 69
  • 70. Conclusions  There are a lot of good options for the analysis of repeated events  GEE Survival models  Gee Poisson Regression  Negative binomial regression  Q and R measures – especially in clinical trial setting  Worthwhile to work with several as secondary analyses to verify consistency 70
  • 71. Main References  Allison PD. Survival Analysis Using SAS: a Practical Guide, second edition. SAS Publishing, 2010.  Lachin JM. Biostatistical Methods: the Assessment of Relative Risks. Wiley, 2000.  Xu J, LaValley M. One-sample and two-sample analysis of heterogeneous person-time data in clinical trials. Pharmaceutical Statistics 2012; 11: 194 – 203. 71