ORIGINAL ARTICLE

Predictors for Class II treatment duration

Kurt Popowich,a Brian Nebbe,b Giseon Heo,c Kenneth E. Glover,d and Paul W. Majore
St. Albert and Edmonton, Alberta, Canada

Background: The purpose of this study was to identify clinical factors that predict treatment length for
patients with Class II malocclusions. Methods: A sample of 237 active retention patients representing 3
observational groups (Angle Class I nonextraction, and Class II Division 1 extraction and nonextraction),
based on specific selection criteria, was obtained from 3 private offices. From the patient records, data were
collected in these categories: (1) patient information, (2) model information, (3) pretreatment cephalogram
information, and (4) treatment information. Two regression analyses were completed, with total treatment
time as the dependent variable for both models. Results: The first regression analysis (Class I and Class II
patients) indicated significant predictors for the patient, model, and cephalometric variables. These
predictors included age, pretreatment overjet, and pretreatment ANB angle. The second regression analysis
(Class II treatment variables) identified the following factors as significantly associated with treatment length:
(1) type of Class II appliance, (2) number of months of Class II appliance wear, (3) number of months of
interarch elastic wear, (4) maxillary expansion, (5) number of debonds, and (6) average time (weeks) between
appointments. Conclusions: Six variables explained 56.7% of the variation in Class II treatment length.
Further research is required to help explain more of the variance associated with treatment duration. (Am J
Orthod Dentofacial Orthop 2005;127:293-300)

I
t is difficult to accurately predict orthodontic treat-
ment times for specific malocclusions. Shia1
stressed the need for orthodontists to evaluate their
own practices to determine what factors are responsible
for treatment overruns. Because orthodontists have
access to various types of information during the
treatment planning stage (models, radiographs, clinical
examinations, photos), it would be very beneficial to
know which factors could be used as predictors of
treatment time. Unfortunately, due to limited research
in this field, there seems to be no consensus as to what
predictors, if any, can be used to reliably predict
treatment duration. Accurate predictions would not
only help orthodontists establish a time-related fee
guide for specific malocclusions, but also increase
patient satisfaction and overall practice success.1,2
Turbill et al3 evaluated the effects of patient char-
acteristics, practitioners, malocclusions, and treatment
a
variables on total treatment length. Based on analysis of
15 variables, they could explain 41% of the variance in
treatment time. Increased treatment duration was asso-
ciated with use of fixed appliances, multiple stages of
treatment, premolar extraction, anteroposterior buccal
occlusion, patient age, and the practitioner’s level of
qualification. Beckwith et al4 also evaluated possible
factors related to treatment duration and explained over
half of the variation with 6 variables. Increased treat-
ment length was associated with missed appointments,
number of replaced brackets, poor oral hygiene, head-
gear, and variation between offices. Fink and Smith5
evaluated patient characteristics, treatment characteris-
tics, and office variables for connections to treatment
length. They found a significant association between
treatment time and pretreatment ANB angle, pretreat-
ment mandibular plane angle, extraction of premolars,
and number of broken appointments.
Several studies have tried to determine the effect of

b
Private practice, St. Albert, Alberta. extractions on treatment duration. O’Brien et al6 fo-
Clinical assistant professor of Orthodontics, Department of Dentistry, Faculty
of Medicine and Dentistry, University of Alberta, Edmonton, Alberta. cused on Class II Division 1 patients and concluded that
c
Assistant professor of Statistics, Orthodontics, Department of Dentistry, orthodontic treatment involving extractions leads to
Faculty of Medicine and Dentistry, University of Alberta, Edmonton, Alberta.
d
longer treatment times. Vig et al7 determined that
Professor of Orthodontics, Department of Dentistry, Faculty of Medicine and
Dentistry, University of Alberta, Edmonton, Alberta. extraction treatment takes an additional 5 months, and
e
Professor and director of Orthodontic Graduate Program, Department of Fink and Smith5 found that treatment time increased by
Dentistry, Faculty of Medicine and Dentistry, University of Alberta. .9 of a month per extracted premolar. In addition,
Reprint requests to: Dr Paul W. Major, Faculty of Medicine and Dentistry.
Room 4051b, Dentistry/Pharmacy Centre, University of Alberta, Edmonton, Turbill et al3 and Alger8 both found increased treatment
Alberta, Canada T6G 2N8; e-mail, major@ualberta.ca. times with extractions. Contrary to those studies, Vig et
Submitted, July 2003; revised and accepted, December 2003. al9 found no statistically significant treatment time
0889-5406/$30.00
Copyright © 2005 by the American Association of Orthodontists. difference between extraction and nonextraction pa-
doi:10.1016/j.ajodo.2003.12.025 tients when they combined the data from 5 offices.
293
294 Popowich et al
Table I. Selection criteria for patients
Complete records including treatment notes, pretreatment lateral
cephalograms, and pretreatment and posttreatment models
Age of 10 to 15.99 years at start of treatment
Class II Division 1 and Class I patients adhering to Angle’s
definition of occlusion, with diagnosis clearly indicated in chart
by orthodontist
No history of active orthodontic treatment before full fixed
banding (patients were included if they had deciduous tooth
extractions to assist in general eruption sequence or nonactive
appliances such as lower or upper holding arch
No missing teeth before treatment that would result in unusual
extraction pattern (patients were included if missing tooth was
part of desired extraction sequence or if space for missing tooth
was being maintained for prosthetic restoration)
No premature removal of braces
No impacted teeth (requiring surgical uncovering and forced
eruption)
No surgery in conjunction with orthodontic treatment
No Class II Division 2 incisor patterns
Table II. Selection criteria for observational groups
Class I nonextraction Angle Class I molar bilaterally
group Treated without extractions
Class II Division 1 ⱖ 5 mm overjet or ANB angle of ⱖ 5°
nonextraction Minimum end-to-end or greater molar
group relationship, bilaterally
No extractions
Class II Division 1 ⱖ 5 mm overjet or ANB angle of ⱖ 5°
extraction group Minimum end-to-end or greater molar
relationship, bilaterally
Extractions of 2 or 4 premolars
Similarly, Beckwith et al4 found no significant differ-
ence in treatment duration between extraction and
nonextraction patients. The effect of extractions on
treatment time is still somewhat controversial because
of a lack of well-controlled studies.
Some investigations have concentrated on treat-
ment lengths associated with specific malocclusions.
O’Brien et al6 focused on Class II Division 1 patients
and found that certain variables affect treatment time.
Vig et al7 also assessed the duration, occlusal outcome,
and relative improvement in malocclusion severity for
Classes I, II, and III patients. They concluded that Class
II treatment lasts 5 months longer than Class I treat-
ment. In addition, the pretreatment peer assessment
rating (PAR) was higher in the Class II group than in
the Class I group, reinforcing the notion that it takes
longer to correct the buccal occlusion and overjet in
orthodontic patients.3,10
Attempts have been made to determine whether
cephalometric values could be used to predict orth-
odontic treatment duration. Kim et al11 studied the
American Journal of Orthodontics and Dentofacial Orthopedics
March 2005
predictive value of common cephalometric parameters
on pretreatment severity and treatment outcomes. With
41 cephalometric parameters studied, only the maxil-
lary and mandibular incisors to N-A point and N-B
point were significant. Similarly, Fink and Smith5
found only pretreatment ANB and mandibular plane
angles significant for treatment time. Beckwith et al4
looked at saddle angle, articular angle, gonial angle,
and the total sum of these measurements and found no
significant effects.
The purpose of this study was to identify any
clinical variables (patient, model, cephalometric, and
treatment) that might predict treatment length for Class
I (nonextraction) and Class II Division 1 (extraction
and nonextraction) malocclusions. If orthodontists have
a better understanding of the variables associated with
treatment length, they can be more accurate in treat-
ment estimations.
MATERIAL AND METHODS
Approval to conduct this study was granted by the
University of Alberta Health Research Ethics Board.
This study was conducted in 3 private orthodontic
offices in the greater Edmonton area of Alberta, Can-
ada. These offices were selected because of their
philosophies to treat most patients with 1 phase of
treatment rather than with 2 or more phases. The
orthodontists in these offices had a minimum 10 years
of private practice experience. The active retention
patients from these offices represented 3 observational
groups (Class I nonextraction, Class II Division 1
nonextraction, and Class II Division 1 extraction). The
overall criteria for patient selection are shown in Table
I, and the selection criteria for the specific observa-
tional groups are given in Table II.
The design of this study was unique; therefore, it
was not possible to perform a sample size calculation
by using previously published literature. Furthermore,
because of a limited number of patient records avail-
able, a pilot study to determine sample size was not
feasible. As a result, based on a publication by Norman
and Streiner,12 the goal for sample size in each obser-
vational group was set at 30.
Due to a limited number of active retention files and
the rigid inclusion criteria, it was difficult to gather
large numbers of patient files for each observation
group. As a result, the sample size reflected all avail-
able patient files that fit our selection criteria. Because
of the inherent restrictions associated with achieving a
large sample size; randomization of these study sub-
jects was not possible.
During data collection, information was recorded
American Journal of Orthodontics and Dentofacial Orthopedics
Volume 127, Number 3
from patient treatment records (charts, models, and
lateral cephalograms).
1. Patient information—age and sex.
2. Model information— overjet and overbite.
3. Pretreatment lateral cephalogram information—
ANB angle, palatal plane (ANS-PNS) to SN, max-
illary incisor to palatal plane, mandibular incisor to
mandibular plane (Go-Me), and mandibular plane
angle to SN.
4. Treatment information—start and finish dates of
active orthodontic treatment, type of Class II appli-
ance (headgear, Herbst, Jasper jumper, pendulum),
number of months of Class II appliance wear,
number of months interarch elastics were used, use
of expansion appliance (yes/no), months into treat-
ment when second molars were banded, premolar
extractions (yes/no), closing mechanics (sliding or
closing loops), number of debonds, number of
missed appointments, and average time between
appointments (weeks). A cancelled appointment
was not considered missed if it was rescheduled
within a week. Repositioning a bracket to change
tooth position was included in the total number of
debonds.
Before data collection, names on charts, pretreat-
ment lateral cephalograms, and models were covered
with a tab indicating an assigned patient number. This
procedure ensured blinding during the data-collection
stage. The patient records were then measured in a
randomized order that was predetermined by the Excel
(Microsoft, Redmond, Wash) software.
By using a digitizer (Kurta, Phoenix Az) and a
customized software program (University of Alberta,
Edmonton) called CEPH ID, all pretreatment cephalo-
metric values were measured to the nearest .001°.
Overjet was measured on the dental study models
with a straight ruler to the nearest 0.5 mm from the
labial surface of the most protrusive maxillary incisor
to the corresponding mandibular incisor. Overbite
value was obtained with the same straight ruler to the
nearest 0.5 mm, by measuring the mandibular incisal
edge to a pencil line drawn parallel to the occlusal
plane, at the point of greatest overlap.
The data set for this investigation was part of a
larger study.13
Ten sets of records were randomly selected from a
nonstudy orthodontic office in Edmonton. Intraexam-
iner reliability was determined by the principal inves-
tigator randomly measuring the cephalometric values,
overjet, and overbite on 10 pretreatment lateral cepha-
lograms and models on 3 separate occasions with a
Popowich et al 295
Table III. Description of sample
Number of
Class Office patients
Class I (nonextraction) A 26
B 26
C 25
Class II (nonextraction) A 26
B 30
C 25
Class II (extraction) A 26
B 28
C 25
Total 237
week between sessions. All models and cephalograms
were blinded.
Statistical analysis
All data was viewed graphically to ensure normal
distribution. The independent variables were grouped
into specific categories: patient, model, cephalometric,
and Class II treatment variables. Separate stepwise
multiple regressions were used to analyze each cate-
gory to determine which variables are significantly
associated with total treatment length (dependent vari-
able). Because patient, model, and cephalometric vari-
ables are not specific to Class II treatment, their
regressions were calculated from a data set that in-
cluded the Class I and Class II patients from all 3
offices. Conversely, when the treatment variables were
examined, only the Class II patients from the 3 offices
were included. A P value of ⱕ .05 was considered
statistically significant. In addition, ANOVA and t tests
were conducted to determine the difference in treatment
duration (months) between the different Class II appli-
ances, and between maxillary expansion and nonexpan-
sion.
RESULTS
The total sample size from the 3 orthodontic offices
was 237; Table III gives a breakdown of patients for
each observational group. Graphic evaluation of the
data demonstrated normal distribution. Descriptive sta-
tistics indicating the values for the patient, model, and
cephalometric variables between the various observa-
tion groups are shown in Table IV.
Based on ANOVA calculations, it was deter-
mined that there were no significant differences for
any model or cephalometric measurement between
the 3 sessions. These results indicated excellent
intraexaminer reliability.
Descriptive statistics indicating the values for the
Class II treatment variables between the Class II
296 Popowich et al American Journal of Orthodontics and Dentofacial Orthopedics
March 2005

Table IV. Descriptive statistics for patient, model, and cephalometric variables
Class I Class II Class II
Independent variables (patient, model, and cephalometric) nonextraction nonextraction extraction

Male, n (%) 34 (44.2) 38 (46.9) 34 (43)

Female, n (%) 43 (55.8) 43 (53.1) 45 (57)
Office A, n (%) 26 (33.8) 26 (32.1) 26 (32.9)
Office B, n (%) 26 (33.8) 30 (37.0) 28 (35.4)
Office C, n (%) 25 (32.5) 25 (30.9) 25 (31.6)
Age (y), mean (SD) 13.25 (1.32) 12.4 (1.28) 12.57 (1.31)
Treatment duration (mo), mean (SD) 20.25 (5.96) 25.7 (6.78) 24.97 (5.48)
Treatment duration (appointments), mean (SD) 14.91 (4.29) 19.2 (5.0) 17.52 (3.64)
Overjet (mm), mean (SD) 3.38 (1.40) 7.00 (2.40) 7.59 (2.22)
Overbite (mm), mean (SD) 3.46 (2.01) 4.01 (1.37) 3.63 (1.57)
ANB angle (°), mean (SD) 3.27 (2.07) 5.81 (1.89) 6.35 (1.95)
PP to SN (°), mean (SD) ⫺8.69 (4.04) ⫺7.27 (3.36) ⫺8.02 (3.42)
Mx inc to PP (°), mean (SD) 113.33 (6.22) 114.14 (5.97) 114.48 (7.88)
Md inc to MP (°), mean (SD) 92.17 (6.28) 96.84 (7.47) 98.34 (9.41)
SN to MP (°), mean (SD) 35.64 (6.01) 33.13 (5.39) 35.88 (5.70)

Table V. Descriptive statistics for Class II treatment variables

Class II Class II Combined
Independent variables (Class II treatment) nonextraction extraction Class II groups

Mx expansion appliance: yes, n (%) 24 (29.6) 6 (7.6) 30 (18.8)

Patients with second molars banded, n (%) 21 (25.9) 23 (29.1) 44 (27.5)
Patients with premolar extractions, n (%) 0 (0) 79 (100) 79 (49.4)
Sliding mechanics, n (%) 65 (80.2) 40 (50.6) 105 (65.6)
Closing loops, n (%) 16 (19.8) 39 (49.4) 55 (34.4)
Class II appliance type, n (%)
Headgear 41 (50.6) 35 (44.3) 76 (47.5)
Herbst 25 (30.9) 4 (5.1) 29 (18.1)
Jasper jumper 1 (1.2) 0 (0) 1 (0.6)
Pendulum 1 (1.2) 0 (0) 1 (0.6)
Appliance wear (mo), mean (SD) 12.50 (5.45) 12.38 (7.54) 12.46 (6.23)
Interarch elastics wear (mo), mean (SD) 10.05 (5.98) 13.42 (7.34) 11.86 (6.93)
Treatment length (mo), mean (SD) 25.7 (6.78) 24.97 (5.48) 25.32 (6.16)
Treatment length appointment (n), mean (SD) 19.2 (5.0) 17.52 (3.64) 18.38 (4.45)
Debonds (n), mean (SD) 2.95 (3.07) 1.52 (1.80) 2.24 (2.62)
Missed appointments (n), mean (SD) 0.80 (1.08) 1.11 (1.44) 0.96 (1.28)
Average length between appointments, weeks (SD) 5.76 (0.81) 6.15 (0.82) 5.95 (0.84)

observation groups (nonextraction and extraction) are
shown in Table V.
A multiple regression analysis was done to deter-
mine which patient variables (age and sex) were sig-
nificantly related to total treatment duration. For anal-
ysis of patient variables, the Class I and Class II
patients from all 3 offices were grouped together. Based
on the regression analysis only patient age was signif-
icantly related to orthodontic treatment duration. Age
had a P value less than .001 and an R2 value of 2.9%
(Table VI).
Based on the regression analysis, overjet (P ⬍ .001,
R2 value of 10.6%) was the only model variable found
to be significantly associated with total treatment dura-
tion (Table VII). The only cephalometric variable
significantly associated with treatment duration was
pretreatment ANB angle (P ⬍ .001, R2 value of 6%)
(Table VIII).
For analysis of treatment variables, just the Class II
patients from the 3 offices were included. Because only
a few patients had their second molars banded, this
variable was not included. Similarly, the Jasper jumper
and the pendulum appliances (Class II appliances) were
used only once each in the whole patient sample;
therefore, they were also not factored into the regres-
sion model. Based on the Class II patients, a multiple
American Journal of Orthodontics and Dentofacial Orthopedics Popowich et al 297
Volume 127, Number 3

Table VI. Results from multiple regression of patient Table IX. Results of multiple regression for treatment
variables variables of Class II patients
Regression Overall Regression Overall
Variable P value coefficient power Variable P value coefficient power

Age* .009* ⫺.827 0.751 Maxillary expansion* .020* — .649

Class II appliance .0001* — .990
R ⫽ .029
2
type*
*P value ⬍ .05 is significant. Months of elastics .0001* .324 .990
wear*
Table VII. Results from multiple regression of model Months of Class II .003* .206 .840
variables appliance wear*
Total debonds* .006* .378 .797
Regression Overall Average time .0001* 2.04 .997
Variable P value coefficient power between
appointments*
Overjet* .0001* .769 .990
R2 for the group ⫽ .576
R2 ⫽ .106 *P value ⬍ .05 is significant.
*P value ⬍ .05 is significant.

Table VIII.Results from multiple regression of cepha-
lometric variables
Regression
Variable P value coefficient
Pretreatment ANB angle* .0001* .675
R ⫽ .060
2
*P value ⬍ .05 is significant.
regression was conducted, and 6 variables were found
to be significantly associated with Class II treatment
time. The analysis also indicated that average time
(weeks) between appointments, total debonds, and
months of elastic and Class II appliance wear had
positive regression coefficients. The R2 value for this
regression model was 57.6%, and any P values ⱕ .05
were considered statistically significant (Table IX).
A Bonferroni pairwise comparison test (ANOVA)
was used to analyze some significant variables identi-
fied in Table IX. The type of Class II appliance had a
significant affect on total treatment time. The average
treatment time for Class II patients varied, depending
on the type of Class II appliance (Table X). The group
treated without Class II appliances consisted mainly of
Class II extraction patients, whose Class II corrections
were achieved with controlled space closure and inter-
arch elastics. Furthermore, in the Class II extraction
patients, when only the maxillary premolars were
removed, the Class II molar relationship was not
corrected.
Class II treatment involving maxillary expansion
also had a significant association with total treatment
time. The patients in the maxillary expansion group
were treated on average 2.6 months longer than those in
the nonexpansion group. Based on a pairwise compar-
ison test (ANOVA), this result was statistically signif-
icant at P ⫽ .02 (Table XI).
Overall To ensure that the regression model was robust, the
power significant variables from the 4 previous regressions
.971
were incorporated into a combined regression model
(these included age, overjet, pretreatment ANB angle,
maxillary expansion, Class II appliance type, months of
elastics, months of Class II appliance wear, total
debonds, and average time between appointments). In
the combined Class II regression model, age, overjet,
and pretreatment ANB angle were not significant,
supporting the results of the original Class II regression
model.
Power calculations of the statistics used in this
study indicated that the sample size for this investiga-
tion was adequate.
DISCUSSION
Age was found to be the only patient variable
significantly associated with treatment length. The
negative regression coefficient meant that treatment
duration decreases as patient age increases. A possible
explanation for this finding could be related to the start
of orthodontic treatment before all permanent teeth
have erupted. Another explanation could be based on
the assumption that patients mature with age. Mature
patients might be better cooperators, and this could lead
to decreased treatment times for specific malocclusions.
The small R2 value (2.9%) indicated that only a small
percentage of the variance could be explained.
When we assessed the effect of model variables on
treatment time, only overjet and overbite were evalu-
ated. Based on our results, overjet was significantly
associated with total treatment length; however, with an
R2 value of 10.6%, only a small portion of the treatment
298 Popowich et al
Table X. Treatment time difference (mo) between Class II appliances
Class II appliance (a) Class II appliance (b)
Herbst Headgear
Herbst No Class II appliance
No Class II appliance Headgear
*Mean difference is significant at .05 level.
Table XI. Treatment time difference (mo) for maxillary expansion
Maxillary expansion Maxillary expansion
appliance (a) appliance (b)
Yes No
*Mean difference is significant at .05 level.
variance could be explained. Furthermore, when we
evaluated the cephalometric variables, only pretreat-
ment ANB angle had a significant association with
treatment time. It too had a small R2 value and
explained only 6% of the treatment variance. A larger
overjet and ANB angle are often associated with Class
II Division 1 malocclusions, which, based on our
investigation, were found to have a longer treatment
duration than Class I malocclusions. These results are
supported by Vig et al,7 who determined that Class II
treatment on average takes 5 months longer than Class
I treatment. Fink and Smith5 also demonstrated a
significant association of pretreatment ANB angle with
treatment length. Consequently, these results support
the notion that it takes longer to correct the buccal
occlusion and overjet in orthodontic patients.3,10 Other
cephalometric results from this study are contrary to
Fink and Smith,5 who found a significant effect for
mandibular plane angle on treatment duration. With the
exception of pretreatment ANB angle, there appears to
be no consensus in the literature about cephalometric
predictive values for treatment duration.
The primary objective of this study was to assess
treatment predictors by focusing on Class II treatment
duration in the 3 orthodontic offices. The following
treatment variables were found to be significantly
related to treatment time: type of Class II appliance,
maxillary expansion, number of months Class II appli-
ance was used, number of months of interarch elastic
wear, total debonds, and average time between appoint-
ments (weeks).
Class II appliance type and use of maxillary expan-
sion were both related to treatment duration. Use of the
Herbst appliance rather than headgear increased treat-
ment time by an average of 9.097 months. A possible
explanation for the difference between these 2 appli-
American Journal of Orthodontics and Dentofacial Orthopedics
March 2005
Mean difference Standard Overall
(mo) (a – b) error P value power
9.097* 1.325 .0001*
6.864* 1.436 .0001* .990
2.233 1.193 .063
Mean difference Standard Overall
(a – b) error P value power
2.642* 1.120 .020* .649
ances is that maxillary expansion is routinely required
before the Herbst appliance is placed. When maxillary
expansion was incorporated into the treatment plan, the
treatment length was 2.642 months longer than for
patients treated without maxillary expansion. These
differences were found to be statistically significant
after accounting for the other 6 treatment variables
(significant) in the Class II regression model.
The number of months of Class II appliance and
interarch elastic wear were significantly linked to treat-
ment length in Class II patients. These results seem
logical, because one would expect to use interarch
elastics and Class II appliances (growth modification)
longer in Class II patients who are more severe (greater
discrepancy in buccal occlusion and overjet). Also,
these results could reflect patient cooperation. For
patients who do not wear their appliances (ie, headgear
or interarch elastics) for the prescribed time, correction
will take longer. Shia1 concurred with these findings;
he found poor patient cooperation to be a primary
reason for treatment overrun.
Total number of debonds also had a significant
association with increased treatment length. The num-
ber of debonds included occasions when the bracket
was knocked off the tooth and required replacement.
Also included was the occasional repositioning of a
bracket to improve tooth angulation or alignment. In
both situations, the orthodontist might have slowed
down treatment progress by temporarily decreasing
wire size for realigning the teeth. Furthermore, the need
to replace brackets that have been knocked off because
of unfavorable activities or eating habits could reflect
poor patient cooperation.
In this study, the average time between appoint-
ments was found to be significantly associated with
increased treatment duration. Even though this associ-
American Journal of Orthodontics and Dentofacial Orthopedics
Volume 127, Number 3
ation was statistically significant, the R2 value of 11.6%
indicated a weak association; therefore, the clinical
relevance of this association might be limited. In 1988,
Alger8 found no significant treatment duration differ-
ences between longer (6 week) and shorter (4 week)
appointment intervals, but Alger did not have a control
sample of patients treated on a 4-week recall on which
to base his conclusions.
Previous investigations examining different vari-
ables with multiple regressions4,5 have claimed that
missed appointments are significantly associated with
longer treatment times. This study does not support
their claims, because missed appointments had no
significant effect on treatment times for the Class II
patients of this study. Furthermore, Cadman et al14
found that Canadian First Nations patients missed more
appointments than non-First Nations patients; however,
when assessing the treatment times between these 2
groups, no significant differences were found. Cadman
et al found no significant differences in posttreatment
PAR scores between First Nations and non-First Na-
tions patients.
The present study supports the findings of Beckwith
et al4 and Vig et al9 that extractions do not significantly
influence the duration of treatment. These results are
contrary to other studies, which found that extractions
significantly affected treatment length.3,5,7,8 Vig et al7
found that extractions increased treatment time by 5
months, and Fink and Smith5 determined that treatment
length increased by 0.9 of a month per extracted
premolar. The effect of extractions on treatment length
is still controversial.
Collectively, the significant Class II treatment vari-
ables in this study explained 57.6% of the variance in
treatment length. This result was higher than the vari-
ance explained by previous studies that also looked at
treatment length predictors.3-5 A possible explanation
for this result could be our study design, which made
the Class II groups as homogeneous as possible. All
Class II patients met a strict selection criterion, to
ensure that they represented similar malocclusions
among the 3 offices.
For the overall R2 value of 57.6%, much of the
Class II treatment length variance remains unexplained.
Some factors that have been left out of this investiga-
tion include the variable rates associated with jaw
growth and bone remodeling between patients. In
addition, the application of different tooth forces com-
bined with different materials (ie, heat-activated wires,
beta-titanium and nickel-titanium coils) might also
have contributed to the variance in treatment duration.
Because some of these factors would be extremely
challenging to measure, a complete explanation of all
Popowich et al 299
treatment variance associated with Class II malocclu-
sions could be difficult.
A limitation of this study was the retrospective
study design. A prospective study would have been
advantageous in reducing bias, but this was impossible
because of limited time restraints and resources. This
study was based on patients from 3 offices, and caution
should be used in applying the results.
CONCLUSIONS
Based on the results of this study with patients from
3 orthodontic offices, the following can be concluded:
1. For Class I (nonextraction) and Class II (extraction
and nonextraction) subjects, age, overjet, and pre-
treatment ANB angle were the only variables that
significantly contributed to treatment length.
2. After assessing specific treatment variables for
Class II subjects, we found the following to signif-
icantly increase treatment time: type of Class II
appliance, number of months of Class II appliance
wear, number of months of interarch elastic wear,
maxillary expansion, number of debonds, and av-
erage time between appointments.
3. When maxillary expansion was done, the average
treatment length significantly increased by 2.6
months.
4. Use of the Herbst appliance rather than a headgear
resulted in a significant average increase of 9.1
months in treatment time.
5. The 6 significant treatment variables for Class II
patients explained 56.7% of the variance in treat-
ment length.
REFERENCES
1. Shia GJ. Treatment overruns. J Clin Orthod 1986;20:602-4.
2. O’Connor PJ. Patients’ perceptions before, during, and after
orthodontic treatment. J Clin Orthod 2000;34:591-2.
3. Turbill EA, Richmond S, Wright JL. The time-factor in orth-
odontics: what influences the duration of treatments in National
Health Service practices? Community Dent Oral Epidemiol
2001;29:62-72.
4. Beckwith FR, Ackerman RJ Jr, Cobb CM, Tira DE. An evalua-
tion of factors affecting duration of orthodontic treatment. Am J
Orthod Dentofacial Orthop 1999;115:439-47.
5. Fink DF, Smith RJ. The duration of orthodontic treatment. Am J
Orthod Dentofacial Orthop 1992;102:45-51.
6. O’Brien KD, Robbins R, Vig KWL, Vig PS, Shnorhokian H,
Weyant R. The effectiveness of Class II, Division 1 treatment.
Am J Orthod Dentofacial Orthop 1995;107:329-34.
7. Vig KW, Weyant R, Vayda D, O’Brien K, Bennett E. Orthodon-
tic process and outcome: efficacy studies—strategies for devel-
oping process and outcome measures: a new era in orthodontics.
Clin Orthod Res 1998;1:147-55.
300 Popowich et al
8. Alger DW. Appointment frequency versus treatment time. Am J
Orthod Dentofacial Orthop 1988;94:436-9.
9. Vig PS, Weintraub JA, Brown C, Kowalski CJ. The duration of
orthodontic treatment with and without extractions: a pilot study
of five selected practices. Am J Orthod Dentofacial Orthop
1990;97:45-51.
10. Robb SI, Sadowsky C, Schneider BJ, BeGole EA. Effectiveness
and duration of orthodontic treatment in adults and adolescents.
Am J Orthod Dentofacial Orthop 1998;114:383-6.
11. Kim JC, Mascarenhas AK, Joo BH, Vig KW, Beck FM, Vig PS.
American Journal of Orthodontics and Dentofacial Orthopedics
March 2005
Cephalometric variables as predictors of Class II treatment
outcome. Am J Orthod Dentofacial Orthop 2000;118:636-40.
12. Norman GR, Streiner DL. Biostatistics: the bare essentials.
Hamilton, Ontario, Canada: B.C. Decker; 1998.
13. Popowich K. Orthodontic treatment duration in Class II patients
[thesis]. Edmonton, Alberta, Canada: University of Alberta; 2003.
14. Cadman KC, Glover KE, Heo G, Warren S, Major PW. Orth-
odontic treatment outcome in a First Nations population in
Alberta, Canada: a comparative study. Am J Orthod Dentofacial
Orthop 2002;121:396-402.