Skip to main content

Table 2 Summary of experts' choices and their additional suggestions if not in concordance with A, B or C

From: UKCTOCS update: applying insights of delayed effects in cancer screening trials to the long-term follow-up mortality analysis

Expert

Expertise

Choice

Additional suggestions

EX1

Biostatistics, public health

A

Suggests only include cancers diagnosed from period of intervention.

EX2

Biostatistics, clinical trials and cancer research

A

 

EX3

Statistics

A

Ticked “alternative” but suggested hybrid of A for testing and C for estimation—interpreted as A

EX4

Cancer epidemiology, prevention and screening

Change analysis

Suggested “number needed to screen”.

EX5

Biostatistics, cancer epidemiology

Change analysis

Did not complete form but indicated choice by email, test based on difference of restricted mean survival time (RMST).

EX6

Biostatistics and epidemiology

Change analysis

Suggested splitting data into yearly bins and assess HR in each, possibly with smoothing. Avoid single HR.

EX7

Biostatistics, clinical trials and cancer research

C

Did not complete form but indicated choice by email. Prefers more parsimonious model with interpretable parameters.

EX8

Biostatistics, clinical trials

C

 

EX9

Biostatistics, public health

C

Prefers more parsimonious model with interpretable parameters.

EX10

Cancer epidemiology, public health

C

Also suggests “versatile weighted log-rank test”

EX11

Statistics, public policy

C

 

EX12

Biostatistics

Did not respond within timeframe