SPECIAL ARTICLES

The Probability Index

James F. Gramlingt
Jonesboro, Ark.

The clinical study of the treatment of a sample of Class II malocclusions was made. This information
from both successful and unsuccessful Class II malocclusion correction was recorded. The
objective of the study was to determine whether there are predictive characteristics of Class II
cases that could give, with reasonable accuracy, an indication of those Class II malocclusions that
would more readily respond to treatment and those that would have less chance of conventional
treatment being successful. This information could give the orthodontist differential diagnostic
guidance before starting treatment to determine whether alternative treatment strategy should be
considered. The study used five very important cranial and dental cephalometric measurements
that individually had significance, but no predictive value. However, these measurements, when
combined, and given a weighted value that reflected a relative importance of each criteria, were
found to collectively give a predictive capability in determining whether the case was favorable for
Class II correction. The five angles selected were (1) the Frankfort mandibular plane angle (FMA);
(2) the point A nasion point B, (ANB) angle; (3) the occlusal plane, Frankfort plane angle; (4) the
Frankfort-mandibular incisor (FMIA) angle; (5) the sella-nasion-point B, (SNB) angle. The Probability
Index is the sum of the weighted values of the five cranial, and dental angles and seems to have
significant predictive value in the differential diagnosis and treatment planning of the Class II
malocclusion. (AM J ORTHODDENTOFACORTHOP1995; 107:165-71 .)

Jim Gramling passed away in June, 1993, after suf- a study of Class II malocclusions. Instead, it was a
fering a cerebral hemorrhage. He was 58. At the time of study of Class II orthodontic treatment, a big
his death he was practicing in Jonesboro, Ark. He was difference.
also serving orthodontic education and research as a
professor of orthodontics at the University of Tennessee, PAST RESEARCH
Memphis, Tenn., and as Director of Research of the
A necessary prelude to the presentation of the
Charles H. Tweed International Foundation.
This article was being prepared for submission to the Tweed Foundation research is a brief summary of
AJO/DO at the time of Jim's death. The final editing for the past research projects of the Tweed Founda-
its submission was accomplished by Levern Merrifield, tion. The first of three previous research projects
Ponca City, Okla., and James L. Vaden, Cookeville, was a study of Charles Tweed's Class II treatment. 1
Tenn. The words and thoughts belong to Jim Gramling. A random sample of 54 Class II malocclusions were
selected from the Tweed library for a statistical
T h e research program of the Charles H. investigation. The results of this study were pre-
Tweed International Foundation began in the sented to the Charles H. Tweed International
1970s. In those early years, progress was slow; Foundation at its thirteenth biennial meeting in
indeed we were groping for a direction. Finally, 1980. This study showed that Dr. Tweed corrected
after a great deal of thought and after many in- Class II malocclusions 40 years ago 2'3 about as
sights gained from observation of consistent fail- effectively as they are corrected today. Some Class
ures during the orthodontic treatment of Class II II malocclusions Tweed corrected quite well; oth-
malocclusions, Levern Merrifield, the Chairman of ers, not so well. Even Charles Tweed experienced
the Board of the Charles H. Tweed International varying degrees of success. His success rate was
Foundation, suggested a study of Class II orth- infinitely greater than other orthodontists of his
odontic treatment. It is important to make the time; in part because he was a master technician,
distinction that this research was clinical research. but largely because he pursued excellence with an
Therefore the study that was embarked on was not incomparable zeal and enthusiasm. The pursuit of
that quest for excellence in orthodontic treatment
remains open this day to all orthodontists.
tDeceased.
Copyright © 1995 by the American Association of Orthodontists. The second research project followed closely
0889-5406/95/$3.00 + 0 8/1/51981 and was a study of 150 difficult Class II malocclu-
165
166 Gramling American Journal of Orthodontics and Dentofacial Orthopedics
February 1995

sions that had been treated successfully by member THE PROBABILITY INDEX
orthodontists of the Charles H. Tweed Interna- This investigation defined the Probability Index.
tional Foundation.4 The paper resulting from this Forty successfully corrected difficult Class II real-
study was presented to the Foundation at its fif- occlusions and a like number of unsuccessfully
teenth biennial meeting in 1984. The sample was corrected difficult Class II malocclusions were com-
composed of data from successfully corrected dif- pared. The purpose was to search for a means of
ficult Class II malocclusions. It is important to predicting the relative success or failure in Class II
underscore that this study was not treatment of just malocclusion treatment. As the data is presented
any Class II malocclusion; it was a study of the and the evidence revealed, a judgment of the effi-
orthodontic treatment of difficult Class II malocclu- cacy of the Probability Index can be made.
sions, all of which had been successfully corrected. The elements of the Probability Index are five
The study revealed essentially the same findings as key cephalometric angles. When properly inte-
the study of Tweed's orthodontic treatment. That grated, they appear to be reliable in predicting the
is, a wide variety of Class II malocclusions were prognosis of a given orthodontic treatment.
corrected, some better than others. The question The first key is the very familiar Frankfort
naturally arose as to which Class II malocclusions mandibular plane angle,~ long recognized as one of
can be successfully corrected and which cannot; the most important of cephalometric criteria in
and whether there are distinctive features of a diagnosis, treatment planning, and prognosis. This
Class II malocclusion that might provide a clue to angle, more than any other, delineates the direc-
this very important question. To seek a clue, a third tional quality of facial growth and the interrelation
study was instituted. of the vertical and the horizontal dimensions of the
The third study was another investigation of the face.
orthodontic treatment of difficult Class II maloc- The second key is the ANB angle,7 again a
clusions, but this study included only unsuccessfully criteria familiar to every orthodontist. This is the
corrected Class II malocclusions.5 The results were angle that specifically classifies a malocclusion.
presented to the Charles H. Tweed International There are other angles similar to this angle, but
Foundation at its sixteenth biennial meeting in none quite so direct in the expression of the inter-
1986. Statistical analysis of the results of these relation of the maxilla to the mandible.
unsuccessful Class II treatments began to show The occlusal plane measured to Frankfort hori-
trends. Parallels were evident in the failures that zontals has long been considered as the most direct
were not seen in the successful Class II corrections. determinant of the quality of orthodontic forces,
It appeared that there were some cephalometric and is the third key. It is equally important as a
keys for prognosis. These emerging keys were not determinant of the difficulty of an orthodontic
cephalometric angles being arbitrarily chosen, they correction because the malocclusion is corrected
were the angles that statistical analysis was showing along the occlusal plane. In the second study of 150
to be more reliable in predicting the success or successfully treated difficult Class II malocclusions,
failure of any Class II correction. it was the Class II malocclusions with high occlusal
Prompted by the reality that there are certain plane angles that proved to be the most difficult to
Class II malocclusions that can successfully be correct in many perspectives.
corrected, and that there are certain Class II mal- The fourth key is the Frankfort mandibular
occlusions that cannot be successfully corrected, a incisor angle,9 the third angle of the Tweed tri-
fourth research project was designed; the subject of angle. It is the most meaningful of the angles
this article. It is an effort to discover predictive depicting the relative protrusion of the mandibular
elements of success or failure in Class II orthodon- incisor. The FMIA not only relates the protrusion
tic treatment by statistically analyzing pretreatment of the mandibular incisor relative to the mandible,
and posttreatment cephalometric data. but also it relates the protrusion of the mandibular
Any prediction is extremely difficult and no incisor relative to the face.
prediction can be perfectly accurate. However, a The fifth angle used in the Probability Index is
predictive capability in Class II orthodontic treat- the SNB angle. 1° It most clearly and most precisely
ment would be of sufficient value to the clinical represents the spatial relationship of the mandible
orthodontist to make this a worthy research to the cranium. From earlier studies of more than
project. Indeed a reliable method of predicting the 400 cases, the SNB angle did not change from
success or failure in Class II orthodontic correction before to after treatment in virtually all cases.
would be of enormous value. Please note that the predictive capability of
AmericanJournalof Orthodonticsand DentofacialOrthopedics Gramling 167
Volume107,No. 2

Table I. Comparison of the FMA of successfully Table Ill. Comparison of the occlusal plane of
and unsuccessfully corrected successfully and unsuccessfully corrected
Class II malocclusions Class II malocclusions
I successful I Unsuccessful Successful Unsuccessful
Below20 5 5 7 and below 8 6
20-30 29 15 8 and above 32 34
Above 30 6 20

Table IV. Comparison of the FMIA of successfully

Table II. Comparison of the ANB of successfully
and unsuccessfully corrected
and unsuccessfully corrected
Class II malocclusions
Class II malocclusions
Successful Unsuccessful
l Successful I Unsuccessful
60 and above 14 13
6 and below 25 21 59 and below 26 27
7 and above 15 19

these angles is not valid when each is considered Table V. Comparison of the SNB of successfully
separately. The predictive nature only seems to and unsuccessfully corrected
materialize when these angles are considered col- Class II malocclusions
lectively. The following is an explanation and dis- Successful Unsuccessful
cussion in support of this hypothesis.
Table I shows data of the sample grouped only Below 75 14 23
by the Frankfort mandibular plane angle. There 75-80 20 16
Above 80 6 1
was a moderate trend; there were more cases that
had medium and high Frankfort mandibular plane
angles in the unsuccessful sample, but the trend
was not clearly decisive. A conclusive judgment of most equally as well as Class II malocclusions with
the prognosis of a Class II malocclusion on the high Frankfort mandibular plane angles.
basis of the FMA alone could not be made. Finally, the sample was divided by the value of
Table II shows the sample when dMded by the the SNB angle as seen in Table V. This grouping
ANB angle and separated into high and low ANB did reveal some minor trends similar to the FMA
angle groups. This data shows very little difference groupings. The trends that were observed were that
in the successfully treated and the unsuccessfully there were more unsuccessful corrections toward
treated Class II samples. Again, the size of the the lower SNB angle cases. However, the observa-
ANB angle alone was not a reliable predictor of the tions were only trends, inclinations, proclivities;
success or failure of Class II correction. there was no conclusive evidence, precisely predic-
Table III shows the sample divided into high tive in nature, for the correction of a given Class II
and low occlusal plane angle cases. This data re- malocclusion.
vealed even less difference between the success- The preceding data confirmed the suspicion
fully treated and the unsuccessfully treated sample. that these key cephalometric indicators separately
Again, on the basis of the value of the occlusal were of little predictive value. Therefore the crite-
plane angle alone, no valid prediction could be ria were combined in anticipation that their collec-
made concerning the correctability of a given Class tive value might be more reliable in predicting
II malocclusion. success of failure of a given Class II orthodontic
Table IV shows the data when the sample was treatment. An index was formulated in such a
divided by the FMIA and separated into high and manner that values could be numerically added and
low FMIA. These samples were virtually identical. then expressed as a single mathematical entity, the
The size of the FMIA alone did not provide any Probability Index.
prognosis of the correctability of a given Class II The angular numerical difference that existed
malocclusion. Class II malocclusions with low between these cephalometric criteria made it nec-
Frankfort mandibular incisor angles corrected al- essary to assign an arithmetic factor to the variance
168 Gramling American Journal of Orthodontics and Dentofacial Orthopedics
February 1995

Table Vh The probability index

Point value CephaIometric value [ Probabilityindex

F M A 20-30 5
ANB 6 or less 15
F M I A 60 or more 2
OCC PL 7 or less 3
SNB 80 or more 5
Total

Table VIh The probability index

Point value Cephalometric value I Probabilityindex

FMA20-30 5 35 25
ANB 6 or less 15 8 30
F M I A 6 0 or more 2 54 12
O C C P L 7 or less 3 10 9
S N B 8 0 or more 5 75 25
Total 101

Table VIIh The PI distribution of successfully corrected difficult Class II malocclusions

Before treatment After treatment

Over 100 0 cases 0 cases

90-99 3 cases 0 cases
80-89 3 cases 1 cases
70-79 6 cases 2 cases
60-69 6 cases 3 cases
50-59 7 cases 4 cases
40-49 3 cases 6 cases
30-39 6 cases 4 cases
20-29 3 cases 10 cases
10-20 2 cases 9 cases
0-10 1 cases 1 cases

in degrees so that the relative importance of each using the Probability Index for a sample Class II
criteria would be correctly interrelated or weighted malocclusion. The arithmetic is simple. The FMA
in importance. The mathematic factor shown in of 35 ° is 5° outside the correctable range. Five
Table VI was determined by considering the ana- degrees multiplied by the point value of 5, gives the
tomic importance of each cephalometric angle and Probability Index for the FMA of 25. The other
the arithmetic value of that angle. The ranges of variables are calculated in the same manner and
successful correctability appear following each totaled, in this instance, to yield a Probability Index
cephalometric crtieria. When the key cephalomet- of 101.
ric angles of a given Class II malocclusion fall
PROBABILITY INDEX OF
within these limits, that Class II malocclusion falls
within the favorable range for successful Class II SUCCESSFUL TREATMENTS
correction. The amount by which the key cephalo- This study was composed of a random selection
metric angles of a given Class II malocclusion fall of 40 unsuccessfully corrected difficult Class II
outside these limits will be the varying degree of malocclusions and 40 successfully corrected diffi-
difficulty encountered in the correction of a given cult Class II malocclusions taken from previous
Class II malocclusion. research. Table VIII shows the distribution of the
Table VII shows an example of the method of Probability Index in the sample of 40 Class II
American Journal of Orthodontics and Dentofacial Orthopedics Gramling 169
Volume 107, No. 2

Table IX. The PI distribution of successfully Table X. The probability index of successfully
corrected difficult Class II malocclusions corrected difficult Class II malocclusions
Over 100 - 0 c a s e s Before treatment After treatment
Over 90 - 3 c a s e s 7.5%
Over 80 - 6 c a s e s 15% High 95 (37) 89 (81)
Over 70 - 12 c a s e s 30% Low 7 (14) 9 (30)
Over 60 - 18 c a s e s 45% Average 54.18 35.5
Over 50 - 25 c a s e s 62.5%
Over 40 - 28 c a s e s 70%
Over 30 - 34 c a s e s 85%
Over 20 - 37 c a s e s 92.5% sample had a Probability Index of greater than 80
and more than half had an index of greater than 90.
The unsuccessful case with the highest Probability
malocclusions that were successfully corrected. It is Index was 222 compared with only 95 in the suc-
noteworthy that none of the cases had a Probability cessful sample. The lowest Probability Index in this
Index of greater than 100. Three of the cases had unsuccessful sample was 46 compared with 7 in the
an index of greater than 90 and only six had an successful sample. By any standards, these differ-
index greater than 80. ences were profound. Finally, and perhaps most
In Table IX the data appears in a slightly revealing, this sample of unsuccessfully corrected
different manner, the Probability Index distribution difficult Class II malocclusions had an average
was expressed in percentages. Only 7.5% of the Probability Index of 98 compared with only 54 in
sample had an index greater than 90, whereas/5% the successfully corrected sample.
had a Probability Index of greater than 80. In short, One other distinctive observation was that the
85% of the difficult Class II malocclusions were average Probability Index in the successful sample
successfully corrected when the Probability Index was changed by orthodontic treatment from 54 to
was less than 80. 35, a 65% correction. In the unsuccessful sample
Table X shows the data from a third perspec- the average pretreatment Probability Index was 98
tive. The highest pretreatment Probability Index in and the average posttreatment was 96, almost no
those Class II malocclusions that were successfully correction at all as a result of orthodontic treat-
corrected was 95, the lowest was 7. However, the ment. The Probability Index is not only of value in
most significant observation was that the average predicting the correctability of a given Class II
Probability Index of this successfully corrected malocclusion, but might also be of some value in
Class II sample was only 54. It was also of some evaluating an orthodontist's performance in Class
interest to note the changes that took place in the II orthodontic treatment. Simply stated, the greater
Probability Index as a result of orthodontic treat- the reduction of the Probability Index of a given
ment. The pretreatment Probability Index of 95 Class II malocclusion, the better are the treatment
was corrected to 37, denoting excellent orthodontic methods. Perhaps, through the use of the Probabil-
treatment. However, the most significant finding ity Index there will be a means by which an objec-
here was in the averages. The average Probability tive evaluation of orthodontic treatment results
Index was reduced by orthodontic treatment from submitted by other orthodontists can be made.
54 to 35. In Table XIII are some comparisons that may
help in further evaluating the reliability of the
PROBABILITY INDEX OF
Probability Index. In the successfully corrected
UNSUCCESSFUL TREATMENTS
sample only six cases (15%) had a Probability Index
TaMe XI shows the Probability Index distribu- of greater than 80, whereas in the unsuccessfully
tion in the unsuccessfully corrected Class II corrected Class II sample 25 cases (62.5%) had a
sample. There were 16 cases greater than 100. In Probability Index of greater than 80.
the successfully treated Class II sample there were
only three cases greater than 90, compared with 21 RESULTS
in this grouping. In the successful sample there The Probability Index not only indicates the
were only six cases greater than 80, compared with correctability of a Class II malocclusion, and bears
25 cases with a Probability Index of greater than 80 some reflection of performance in orthodontic
in unsuccessfully treated cases. Table XII shows treatment; but perhaps it is also an expression of
almost two-thirds of the cases in the unsuccessful the growth potential of a given Class II malocclu-
170 Gramling Amerwan Journalof Orthodonticsand DentofacialOrthopedics
February 1995

Table XI. The PI distribution of unsuccessfully corrected difficult Class II malocclusions

Before treatment [ After treatment
Over 200 1 cases 4 cases
150-199 4 cases 4 cases
125-149 5 cases 3 cases
100-124 6 cases 5 cases
90-99 5 cases 2 cases
80-89 4 cases 2 cases
70-79 4 cases 3 cases
60-69 5 cases 6 cases
50-59 5 cases 2 cases
40-49 1 cases 5 cases
30-39 0 cases 1 cases
20-29 0 cases 2 cases
10-19 0 cases 0 cases
0-9 0 cases 1 cases

Table XII. The PI distribution of unsuccessfully Table XIV. Probability index

corrected difficult Class 1I malocclusions
Over 100 Impossible prognosis
Over 100 - 16 cases 40% 90-99 Very poor prognosis
Over 90 - 21 cases 52% 80-89 Poor prognosis
Over 80 - 25 cases 62.5% 70-79 Fair prognosis
Over 70 - 29 cases 75% 60-69 Good prognosis
Over 60 - 34 cases 85% 50 and below Excellent prognosis
Over 50 - 39 cases 97.5%

success it will be of the utmost importance that the

Table XIII. Comparison of successful and
orthodontist exercise excellent intrusive force
unsuccessful Class II sample
control.
I Successful Unsuccessful W h e n the Probability Index falls between 70 to
79, Class II correction becomes much easier. How-
Over 200 0 1
150-199 0 4 ever, excellent appliance control is still a prerequi-
125-149 0 5 site to excellent orthodontic results.
100-124 0 6 W h e n the Probability Index falls below 69,
90-99 3 5 prognosis becomes good. Usually a minimum effort
80-89 3=6 4=25
will result in an excellent correction of a Class II
malocclusion.

CONCLUSIONS
sion. When it is greater than 100, successful cor-
The following points are offered as potential
rection of a Class II malocclusion is virtually im-
uses of the Probability Index:
possible without adjunctive orthognathic surgery.
When the Probability Index (Table XIV) is 1. To aid in identifying those Class II malocclu-
between 90 and 99, the prognosis is very poor. The sions severe enough to consider maxillofacial
treatment plan will be borderline surgery. Success- surgery as an adjunct to Class II orthodontic
ful orthodontic correction of these types of Class II correction.
malocclusions will almost certainly require addi- 2. To aid in identifying those very difficult Class
tional extractions. Even then, successful treatment II malocclusions that may require alternate
will be elusive. treatment methods, such as extraction of the
W h e n the Probability Index is between 80 and maxillary first or second molars in addition
89, the prognosis remains poor, but successful Class to the extraction of premolars.
II correction is more likely. However, to attain this 3. To aid in predicting more accurately the
American Journal of Orthodontics and Dentofacial Orthopedics Gramling 171
Volume t07, No. 2

treatment time necessary to correct a given 2. Tweed CH. Indications for the extraction of teeth in orth-
Class II malocclusion and to thereby enable odontic procedures. AM J ORTHOD ORAL SURG 1944;30:
405-28.
an orthodontist to assign a fairer and more
3. Tweed CH. A philosophy of orthodontic treatment. AM J
appropriate fee. ORTHOD ORAL SURG 1945;31:74-103.
4. To evaluate previously treated Class II mal- 4. Gramting JF. A cephalometric appraisal of the results of
occlusions for critical review of treatment orthodontic treatment on one hundred fifty successfully
methods and subsequent revision of these corrected difficult Class II malocclusions. J Charles Tweed
Found 1987;15:102-11.
methods to improve Class II treatment pro-
5. Gramling JF. A cephalometric appraisal of the results of
cedures. orthodontic treatment on fifty-five unsuccessfully corrected
5. To indicate possible growth potential of a difficult Class II malocclusions. J Charles Tweed Found
given Class II malocclusion. 1987;15:112-24.
6. To evaluate the performance of a clinician in 6. Tweed CH. The Frankfort-mandibular plane angle in orth-
odontic diagnosis, classification, treatment planning, and
orthodontic treatment.
prognosis. AM J ORTHOD 1946;32:175-221.
The Probability Index is far from being proven; 7. Reidel R. The relation of maxillary structures to cranium in
malocclusion and in normal occlusion. Angle Orthod 1952;
this article is only an introduction. Every clinician is
22:142-5.
invited to test its validity and reliability. The 8. Downs WB. The role of cephalometrics in orthodontics case
Charles H. Tweed International Foundation will analysis and diagnosis. AM J ORTHOD 1952;38:162-82.
continue to test this index until its reliability has 9. Tweed CH. The Frankfort-mandibular incisor angle
been conclusively proven or disproven. Contingent (FMIA) in orthodontic diagnosis, treatment planning and
prognosis. AM J ORTHOD 1954;24:121-69.
on the results, it can then either be added to
10. Reidel R. The relation of maxillary structures to cranium in
orthodontic diagnosis and treatment planning, or it malocclusion and in normal occlusion. Angle Orthod 1952;
can be discarded. 22:142-5.
Reprint requests to:
REFERENCES Dr. James Vaden
308 East First St.
1. Gramling JF. A study of Tweed's Class II correction. Un-
published paper presented to the thirteenth biennial meet- Cookeville, TN 38501
ing of the Charles H. Tweed International Foundation,
Memphis, Tennessee, October 3, 1980.