I find the RBANS to be an extremely useful, brief assessment of the neurocognitive performance among patients with dementia, including the vascular /subcortical vascular dementias. The subtest within the attention index should allow you to assess visual attention objectively.

Thank you Vigneswaran, thats really helpful, I shall take a look at the RBANS. 

I agree with Vigneswaran. I worked with a neuropsychologist at the University of Iowa characterizing the cognitive performance of people with vascular disease (in advance of dementia) and we found the RBANS to be the best measure. We found deficits in visual attention and working memory as well as in processing speed that were specific to older adults with vascular disease, compared with healthy elderly controls. 

http://atvb.ahajournals.org/content/27/1/141.full.pdf

The RBANS is useful as a brief screening device, but note that each area was normed on an n of 90.  For many measures in an older cohort, MOANS (Mayo Older Adult Normative Studies) is more exact and diagnostic.  At least one study showed that the MoCA had a standard deviation of 8 points on a 30-point test in a population based group over age 55. The MMSE is a very brief screen, useful with norms from a 1993 JAMA article.  Diagnosing with a screening test is unreliable and highly problematic.

Rossetti, H.C., Lacritz, L.H., Cullum, C.M., & Weiner, M.F. (2011.) Normative data for the Montreal Cognitive Assessment (MoCA) in a population-based sample.. Neurology® 77:1272–1275.

The MoCA outperform the MMSE as screening tool for the detection of vascular cognitive impairment (Dong et al. 2010; Pendlebury et al. 2010). Larner in his book on cognitive screening instruments (2013) also recommend the Brief Memory and Executive test (Brookes et al. 2012) as a bedside examination for small vessel disease, or the R-CAMCOG for evaluating post-stroke dementia.

Obviously, as said here before, screening measures do have important limitations, and a further assessment using the RBANS or other tailored battery focusing on executive/attention and visuospatial domains would be desirable. Along with the previously cited meta-analyses, see this publication available in researchgate for examples of specific tests to be used (Graham et al. 2004):

Distinctive cognitive profiles in Alzheimer's disease and subcortical vascular dementia

Article Distinctive cognitive profiles in Alzheimer's disease and su...

There are a couple of issues here. One is that brief screening tasks like the MMSE or MoCa are best used to help make decisions for further testing, and have confidence intervals that are way to wide to be used for diagnostic purposes. Even the RBANS, which is a more extensive, but still a very brief instrument, has rather wide CIs for any individual domain. Also to my knowledge, no brief screeners, and in fact few comprehensive neuropsych assessments can by themselves differentiate a neurodegenerative-based dementia fronm Vascular etiologies. Having said that, a solid exam across multiple cognitive domains provides very helpful information that can then be used along with rule-out data (MRI data, history of difficulties)  to pretty effectively differentiate the two.

As several people have pointed out, there is no ideal test to differentiate the dementias. Our team has been using the ACE- III for a few years (and the ACE-r before that). It is a useful tool and has cut down our referrals for neuropsychometric testing significantly. In pure AD scores are often low on memory and fluency. The problem is that "pure" AD seems to be quite rare, as many patients have a combination of cerebral atrophy(+/- hippocampal atrophy) and cerebrovascular disease on brain imaging. These patients tend to have more patchy deficits on the ACE-III. Significant visuospatial deficits would suggest Lewy Body disease or Posterior Cortical Atrophy). Visual attention can be reduced in a number of conditions e.g. PDD and I can't say it is the most useful measure for subtyping dementia. Gait disturbance and urinary symptoms seem to favour a more vascular picture (where there is no evidence of normal pressure hydrocephalus).

Overall, I would prefer the ACE-III to the MoCA as a relatively brief screening tool (the MMSE should be assigned to the trashcan).

Personally, once testing (or even simple conversations with the person) has identified that all is not right in memory/judgement/decision-making I find that collateral from family/carers frequently indicates whether the changes have been gradual over time or if there have been 'steps down' in function. Especially if they are told how TIAs can appear to an observer. Hindsight, as they say, gives us all 20:20 vision and when looking back family members will frequently recall a day when their relative may have been exceptionally sleepy, hard to wake, confused but following a good night in bed were recovered, although not quite back to where they were before.

Don't forget that people with vascular dementia often display moments of insight which is different to that displayed by suffers of AD.

Also, as we age I firmly believe that we all experience vascular events - smoking, diet, cholesterol etc combine to fur up those arteries and eventually there will be small blockages in the brain.  It's all about reducing the risk of a 'major' event which could alter our personality and/or performance.

My advice - do tests by all means, but get good collateral, use your gumption and look after your arteries!