I can't speak specifically for ACN but I have set up QA & UR programs 
in the
past and this sounds like one of the levels of statistical review.

IOW, the MCO tracks all of the stats for a treating doc and compares 
to the
bell curve generated for the entire physician/patient population. 
Docs more
than 1.0 standard deviation units outside the curve (either high or 
low) are
looked at more closely.  Generally docs 1-1.5 SD units high or low 
are just
monitored for a period of time to see if this is an anomaly or if it 
remains
outside.  Those that are 2+ SD out are looked at more closely over a 
shorter
time.  With a large enough sampling size the doctor's utilization rate
stabilizes and then the plan looks at the doctor's practice habits and 
are
they comparable to the rest of the doctor population.  If there still
appears to be difference then the doctor is "interviewed" to see why 
they
are high or low.

Some of the acceptable reasons why they are: a doc that sees very 
few plan
enrollees and have 1 or 2 new patients requiring an extremely high
utilization subsequently skewing the stats; patients coming from a 
higher
risk demographic such as all elderly, all from a specific occupation, 
etc.

Unacceptable reasons are: a technique requires this level of 
utilization, my
practice management group says so, I get everyone else's failures, 
etc.

The plan then looks at the doctor's responses and has a panel of docs
evaluate the responses.  If there appear to be quality of care 
problems then
the doctor is "counseled" to see if he/she is amenable to looking at a
literature search or summation provided by the review panel, have a 
dialogue
about it and provide literature in response.

The bottom line is, generally a larger plan has a huge database 
including
docs in the same geographic and demographic categories to compare 
to.  If
these docs have more appropriate utilization then the final remedy 
is
de-credentialing.  That is BAD1  VERY BAD!  With this mark on your 
file you
probably won't credential for another plan for a long time.

When I was doing these reviews we saw the majority of problems 
with
technique "addicts" who: performed full spine and multiple extremity
adjustments for limited diagnoses like a headache; took serial x-
rays to
evaluate curves, etc,; refused to refer even when the patient was not
improving; obtained full spine x-rays on every new patient and re-
activate
regardless of area of complaint; performed more than 2 passive 
therapies for
an extended time.  These were the most common.

However, we did see a limited number of "1 visit wonder" docs.  
Their
patient satisfaction scores were usually very low and we were 
concerned
about missing problems that only show up after a trial of treatment.

This type of system was required by NCQA and URAC at the time I 
was doing
this work and I assume it hasn't changed.  By far, the majority of the 
docs
we spoke with were very professional and we were able to outline a 
plan to
improve clinical utilization that was understandable and amenable 
to all of
us.  There were a very few who took the "my way or the highway" 
approach and
refused to talk much less change. The plan dropped them

Please understand that a group like ACN has an enormous database 
and they
know very well what utilization they can expect.  They docs on their 
panels
are very knowledgeable about the literature and they have tons of 
practice
experience.  The 2 panels I was on ranged from 10 to 30 years 
clinical
experience.

This should get the discussion rolling and I will be happy to respond 
to
specific questions.