Computerized Cognitive Assessments: opportunities and concerns

You know your weight. And your phys­i­cal fit­ness. And a vari­ety of health-relat­ed metrics.

What about your brain fitness?

Two recent announce­ments bring out how the assess­ment of cog­ni­tive abil­i­ties, or brain func­tions, is increas­ing­ly being done thanks to new com­put­er­ized options:

1) Last week, OptumHealth announced an exclu­sive 3‑year agree­ment (esti­mat­ed at $18m) with the Aus­tralian com­pa­ny Brain Resource. OptumHealth will be embed­ding the Brain Resource plat­form into their over­all Behav­ioral Solu­tions program.

- OptumHealth Behav­ioral Solu­tions will work with Brain Resource to pro­vide clin­i­cians with a Web-based assess­ment that mea­sures gen­er­al cog­ni­tion (how peo­ple process infor­ma­tion) and social cog­ni­tion (how peo­ple man­age their emo­tions). This 40-minute assess­ment is based on well-known and val­i­dat­ed tests of mem­o­ry, atten­tion, exec­u­tive func­tion, and response speed, and mood, social skills and emo­tion­al resilience.

- When used by trained clin­i­cians as a tool that is part of the total spec­trum of health care, this unique infor­ma­tion can be help­ful in the treat­ment deci­sion-mak­ing process in sev­er­al ways.

2) A few weeks ago, we read that U.S. Troops To Get Cog­ni­tive Screening

- The mil­i­tary will begin giv­ing cog­ni­tive tests this sum­mer to troops head­ing to war, in an effort to get a base­line mea­sure of their reac­tion time, mem­o­ry, con­cen­tra­tion and oth­er brain func­tions, which could be ref­er­enced in case they are injured.
— Assis­tant Defense Sec­re­tary S. Ward Cass­cells recent­ly direct­ed mil­i­tary lead­ers to begin pre-deploy­ment screen­ing of troops by late-July, using a com­put­er-based test known as the Auto­mat­ed Neu­ropsy­cho­log­i­cal Assess­ment Met­rics, or ANAM, a Depart­ment of Defense spokes­woman con­firmed in writ­ten respons­es to The Courant.

- The test­ing, which takes about 15 to 20 min­utes, will “allow for greater lev­els of accu­ra­cy when mak­ing assess­ments fol­low­ing injury,” said the spokes­woman, Cyn­thia Smith.
— Smith said the new test­ing is not intend­ed as a diag­nos­tic tool for mild trau­mat­ic brain injury, but instead would enable clin­i­cians “to com­pare a per­son to their own ‘norms’ or base­line scores” in the event of an injury.

I see these instru­ments as a crit­i­cal part in the brain fit­ness puz­zle. Neu­roimag­ing tech­niques such as MRI and fMRI are very impor­tant to sup­port clin­i­cal and research work, but are not mature/ scal­able enough to help mea­sure brain func­tions in mil­lions of healthy indi­vid­u­als. Neu­ropsy­cho­log­i­cal test­ing is still today often done with pen and paper, admin­is­tered by a trained expert, and very resource-intensive.

Com­put­er­ized cog­ni­tive assess­ments can start offer­ing val­ue in many con­texts that nei­ther neu­roimag­ing nor tra­di­tion­al neu­ropsy­cho­log­i­cal test­ing can reach.

The media is start­ing to take note. We recent­ly reviewed a recent arti­cle that explained,

- “Cog­ni­tive Drug Research is one a hand­ful of busi­ness­es, most of them out­side of the U.S., that work with phar­ma­ceu­ti­cal com­pa­nies to test how new drugs for every­thing from nico­tine addic­tion to Alzheimer’s dis­ease affect the mind’s abil­i­ty to remem­ber things, make deci­sions, and ana­lyze information.”

- “Cog­ni­tive tests have been around for a cen­tu­ry as exam­i­na­tions tak­en with paper and pen­cil. In the 1970s and ’80s the tests shift­ed to com­put­ers, Cog­ni­tive Drug Research founder Kei­th Wesnes says.

In fact, one of the key high­lights from the mar­ket report we released in March was that “Large-scale, ful­ly-auto­mat­ed cog­ni­tive assess­ments are being used in a grow­ing num­ber of clin­i­cal tri­als. This opens the way for the devel­op­ment of inex­pen­sive con­sumer-fac­ing, base­line cog­ni­tive assess­ments.” And we pro­filed a few lead­ing com­pa­nies in the space: Brain Resource Com­pa­ny, Cog­ni­tive Drug Research, CNS Vital Signs and CogState.

Brain sci­en­tists don´t rec­og­nize one over­all “brain age” or “intel­li­gence”. We can view our brain func­tions or cog­ni­tive abil­i­ties as a vari­ety of skills, some more per­cep­tion-relat­ed, some more mem­o­ry-relat­ed, some more lan­guage-relat­ed, some more visu­al, some more abstract-think­ing and plan­ning ori­ent­ed. There is no gen­er­al “brain age” that can be mea­sured or trained in a mean­ing­ful way.

We explored this in more detail in our mar­ket report, say­ing that

-“A major bot­tle­neck in the use and refine­ment of cog­ni­tive train­ing tools for the appro­pri­ate groups today and in the future is the time and eco­nom­ic invest­ment involved in most­ly man­u­al neu­ropsy­cho­log­i­cal assessments.

- “Poten­tial­ly, these (new, com­put­er­ized) assess­ments could be repur­posed to help estab­lish a cog­ni­tive base­line, assess men­tal func­tion­ing before and after clin­i­cal con­di­tions, track the con­se­quences of aging, inden­ti­fy pri­or­i­ties for cog­ni­tive train­ing, and mea­sure progress inde­pen­dent from the train­ing itself”

Now, the use of these new tech­nolo­gies also rais­es con­cerns, nd not just about their reli­a­bil­i­ty and validity.

John Moore of Chilmark Research just com­ment­ed on the Brain Resource-OptumHealth announce­ment in his post Will a Men­tal Eval­u­a­tion be a Part of Your Next HRA?, point­ing out that “it is fair­ly well-known that many chron­ic dis­eases have a high comor­bid­i­ty fac­tor with men­tal health, OptumHealth’s part­ner­ship with Brain Resource, and its suc­cess (or lack there­of) will be an inter­est­ing one to fol­low. And while I applaud this effort, it also rais­es some pret­ty scary pri­va­cy con­cerns. How will these assess­ments be used beyond the con­fines of the clin­i­cian’s office?, What access will OptumHealth have to the data? And what about the employ­ers who have OptumHealth’s par­ent, Unit­ed Health Group as an insur­er for their employ­ees? Will employ­ers have access to this data, par­tic­u­lar­ly if they start embed­ding it with­in HealthAtoZ?”.

These are excel­lent ques­tions. Humana, a health insur­ance com­pa­ny, recent­ly announced that they were dis­con­tin­u­ing their agree­ment with Posit Sci­ence under which they had been offer­ing the Posit Sci­ence Brain Fit­ness Pro­gram to their Medicare mem­bers. Where­as a num­ber of rea­sons were offered for that deci­sion (rang­ing from low uptake rates of the pro­mo­tions giv­en the legal com­plex­i­ties of reach­ing out to Medicare users, to low uti­liza­tion of the prod­uct), anoth­er con­cern was men­tioned to us dur­ing a set of inter­views with Humana mem­bers: they were con­cerned about whether a pro­gram that had been giv­en to them for free by their insur­ance com­pa­ny would some­how trans­mit data back on the men­tal per­for­mance of the user.

Fur­ther­more, we can expect clear pub­lic pol­i­cy impli­ca­tions in this area. Art Kramer recent­ly explained that “the NIH is prepar­ing an NIH Tool­box to pro­vide valid, reli­able instru­ments to researchers and clin­i­cians, to solve the prob­lem that exists today, name­ly, the lack of uni­for­mi­ty among many mea­sures used. The ini­tia­tive was launched in 2006, and it is a 5‑year effort, so we’ll need to wait to see results”.

As with any new tool, we´ll need the define the rules of the road.

1) First of all, we´ll need to make sure it mea­sures what it is sup­posed to, and with high degrees of reliability.

2) Sec­ond, there need to be clear poli­cies in place as to whom can access which data and for which purpose.

3) Final­ly, we expect the assess­ments will lead into action­able per­son­al­ized rec­om­men­da­tions to improve if not help main­tain cog­ni­tive functions.

We will con­tin­ue to pay close atten­tion to this emerg­ing, and very promis­ing, field.