176 Forensic dentistry
unfortunately, these types of statistics are improperly applied, misleading, and
should be avoided.”^16
Steadman et al. in 2006 stated, “Since individual dental characteristics are
not independent (e.g., the chance of having a restored molar is not the same as
the chance of having a restored canine), the resulting dental patterns formed
by missing, filled, and unrestored teeth are not completely random events.
If this were the case, all conceivable dental patterns would be equiprobable in
the population, a trend that is certainly not valid.”^17
The assumption that dental treatment and loss of teeth occur randomly
throughout the mouth is incorrect. If that assumption were true, the maxil-
lary central incisors would be missing with the same frequency as first molars
or third molars and dental restorations would be seen in the facial surfaces
of incisors as frequently as in the occlusal surfaces of molars. Also, inter-
proximal restorations would occur with frequencies independent of occlusal
restorations in the same tooth. These assumptions are obviously improbable
in clinical dentistry.
While it is undoubtedly possible to develop a statistical analysis that
considers all of the variables involved in predicting the frequency of occur-
rence of dental features, the method would be very complex. Perhaps a better
method already exists.
The Joint POW/MIA Accounting Command/Central Identification
Laboratory, Hawaii (JAPAC/CILHI) was developed to account for Americans
lost during past U.S. conflicts. CILHI has developed a program to assist in
the assessment of patterns of dental conditions and treatment. The program,
called OdontoSearch, is described in a paper published by Adams in 2003.^18
Use of OdontoSearch allows investigators to input dental patterns, then
the program provides a frequency value for that pattern in the databases.
The program is available online at no cost.^19
In an OdontoSearch analysis of the probability of dental feature occur-
rences a formula is used that is conceptually similar to the Keiser-Neilsen for-
mula: nCr
n
rnr=
−!
!( )!. In this equation n is the overall sample size or number
of teeth being considered and r is the number of occurrences of the feature
being considered. In the case of twenty-eight teeth being considered and con-
sidering five missing teeth the equation would be^285
28
5285C = 98 280
−=!
!( )!,.Unlike the pure mathematical approach offered by earlier methods, the data
in OdontoSearch analyze the frequency of occurrence for dental and dental
treatment patterns in populations. According to CILHI, “The program works
by comparing an individual’s pattern of missing, filled, and unrestored teeth
to a large, representative sample of the U.S. population. The methodology
and rationale behind the OdontoSearch 2.0 program is very similar to the