Actually none of the CDR-COA (Continuity of attention) domain components were reported. The dataset reported did include 2 components of the CDR-POA (power of attention) domain.
CDR-COA was the primary endpoint. In the PR and presentation, Missling reported that Anavex 2-73 had significance in 2 components of a post-hoc domain called the "Cognitive Domain of Attention", apparently combining the CDR-COA and CDR-POA endpoints.
These links help explain the tests and the domains,. As Investor pointed out CDR-COA and CDR-POA (and Quality of episodic memory) are not tests they are composite scores derived from several tests during the CDR computerized battery. So how are they derived?
There are 3 tests in the CDR Battery relevant to attention. From these 3 tests you get 6 scores, numbered 1-6 below:
Test 1: Simple Reaction Time: The volunteer is instructed to press the 'YES' response button, as quickly as possible, every time the word 'YES' was presented on the screen. Fifty stimuli are presented with a varying inter-stimulus interval of between 1 and 3.5 seconds. The task lasts approximately 1.5 minutes.
From this SRT test you get 1 score: 1. Simple Reaction Time (ms) [mean reaction time of responses]
Test 2: Digit Vigilance: A target digit is pseudo-randomly selected and constantly displayed to the right of the screen. A series of 450 digits is then presented in the centre of the screen at the rate of 150 per minute. The volunteer is required to press the 'YES' button, as quickly as possible, every time a digit in the series matches the target digit. The task lasts for 3 minutes.
From this DV test you get 3 scores: 2. Digit Vigilance Targets Detected (%) [percentage of correct responses] 3. Digit Vigilance Speed (ms) [mean reaction time of correct responses] 4. Digit Vigilance False Alarms (#) [number of errors]
Test 3: Choice Reaction Time: The volunteer is required to respond to the words 'YES' and 'NO' as they appear on screen by pressing the corresponding button as quickly as possible. There are 50 trials during which each stimulus word is chosen randomly with equal probability; there is a varying inter-stimulus interval of between 1 and 3.5 seconds. The task lasts approximately 1.5 minutes.
From this CRT test you get 2 scores 5. Choice Reaction Time Accuracy (%) [percentage of correct responses] 6. Choice Reaction Time (ms) [mean reaction time of correct responses]
Now from these 6 scores, you derive the two attention domains (CDR-COA and CDR-POA):
CDR-POA (Power of attention) Domain: Power of Attention is reported in milliseconds (ms)
The formula is: CDR-POA = Simple Reaction Time + Digit Vigilance Speed + Choice Reaction Time (subscores 1 + 3 + 6).
CDR-COA (Continuity of attention ) Domain (the primary endpoint): Continuity of Attention is reported as a number (#) It measures the ability to sustain accurate attention
The formula is: CDR-COA = (Digit Vigilance Targets Detected *0.45) + (Choice Reaction Time Accuracy *0.5) – Digit Vigilance False Alarms
Hence it is derived from subscores 2, 4 and 5 above).
So, bottom-line, Anavex presented 0/3 of the outcomes of the CDR-COA domain and 2/3 of the outcomes of the CDR-POA domain. They also presented positive data from another Domain -- Quality of Episodic Memory which were significant. This is a positive. Perhaps if Anavex had pre-specified CDR-QEM as the primary endpoint, the SP might have doubled. Post-hoc, almost all studies can find some significant values.
As I posted in the past, this was always a phase 2 both in endpoints and size. So they have some useful data to help design a phase 3. Most likely, the primary endpoint will be ADAS-Cog. I don't recall any dementia studies that used a computerized test as the primary outcome in a phase 3 that will be judged by the regulatory agencies. They are fine, howver, for phase 2's.
There is some correlation between CDR-QEM and ADAS-Cog (r = 0.7 according to the Anavex presentation) ---- but there is much tighter correlation between MoCA and ADAS-Cog (r= 0.85 see figure 1)-- will the PDD Moca data be presented?
I look forward to seeing what the Phase 3 design will be, assuming they move forward after the FDA meeting.
Back to the original point: Anavex reported two attention measures DV and CRT in ms. So treated patients were mildly faster in processing (measured in ms) --- but did this increase in speed come at the cost of accuracy which is why CDR-COA did not reach significance? Does Anavex increase impulsive decisions? If the wrong decision is going to be made, is it favorable that it be made a few ms earlier? ___________________________________________________________
News 
Market Data 
Discover 
