ORIGINAL ARTICLE

Patient-centered evaluation of orthodontic care: A longitudinal

cohort study of children’s and parents’ attitudes

Lucete M. Fernandes, DDS,a Lisen Espeland, DDS, DrOdont,b and Arild Stenvik, DDS, LicOdont,
DrOdontc
Oslo, Norway

As health services are adapted to meet consumers’ needs, patient-centered evaluation of quality of care
as well as informed consent to treatment decisions become increasingly important concepts in
orthodontics. In an attempt to assess the orthodontic service in a region, this study focused on attitudes
among children and their parents. The attitudes were recorded both before and after the period in which
orthodontic treatment is usually carried out. Changes in children’s and parents’ attitudes were applied to
measure outcome of care, and to evaluate the relevance of informed consent in decisions about
treatment. Seventy-nine family units were interviewed with the use of questionnaires when the child was
11 years of age and again at 16 years. Both orthodontically treated and untreated subjects were
included. Responses to questions about satisfaction with dental appearance and desire for treatment
were transformed to a score for orthodontic concern. A significant decrease in the concern score was
observed over the 5-year period among the treated subjects, and the care system apparently identified
and provided treatment to the majority of those children concerned at 11 years. Intra-unit disagreement
in concern was observed among 25% of the child/parent units at the 11-year stage, whereas at 16 years
nearly all units agreed. Informed consent as a tool to ensure patients’ autonomy when decisions about
treatment are made did not appear to be negatively affected by conflicting attitudes between children
and their parents. (Am J Orthod Dentofacial Orthop 1999;115:227-32)

Over the last half of this century patient
autonomy and quality of care have been increasingly
focused in provision of health services. With increasing
consumerism in health services and development of
quality assurance programs, there has been a gradual
shift of attention to a more patient-centered approach.
Gradually, the professionally oriented approach to
decisions about treatment and evaluation of outcome
has become supplemented by more patient-centered
procedures for decision-making, monitoring, and
assessment of treatment.
Universally accepted guidelines like the Nürnberg-
code (1947), the United Nations declaration of human
rights (1948), and the Helsinki declaration (1964) have
had a strong impact on legislative development and
professional code of conduct for the health services.1-3
The paternalistic approach to provision of health care
has been replaced by a more interactive process in
which the patients are involved in decisions about treat-
aResident.
bAssociate Professor.
cChairman.
Reprint requests to: Arild Stenvik, Dental Faculty, University of Oslo, Postbox
1109 Blindern, N-0317 Oslo, Norway; e-mail: stenvik@odont.uio.no
Copyright © 1999 by the American Association of Orthodontists.
0889-5406/99/$5.00 + 0 8/1/90395
ment.4,5 Among the consequences, a formalized
process for obtaining informed consent about treatment
has been established in some branches of health care.
Although welcomed by the public, this approach con-
fronts health professionals with new challenges; one of
them is to prioritize autonomy in situations of conflict-
ing interests. For the orthodontic professional, this
problem may have to be addressed when the child is in
need of treatment, and the child and parents have oppo-
site views whether treatment should be initiated.
The quality of treatment outcome has traditionally
been assessed by applying professionally established
metric or categorical scales and indexes using mea-
surements obtained from casts, radiographs, and clini-
cal examination. The scope of quality development
has, over the last two decades in addition to outcome-
quality, also included structure-quality and process-
quality.6 As health services exist primarily to benefit
the patient, an important variable for measuring out-
come would be overall patient satisfaction with the
care provided. Patients are the ultimate authorities on
the criteria for good care.7
In the present study the same children and their par-
ents were approached before and after the period in
which orthodontic counseling and treatment usually
are carried out. The study represents an attempt to eval-
227
228 Fernandes, Espeland, and Stenvik
Table I. Distribution of children and parents according to their score on the concern variable (absolute frequencies)
11-Year-old stage
Children
Level of concern Concern score T UT
No 0 17 21
Slight 1 16 2 18
Moderate 2 10 5 9
Marked+Great 3+4 7 1
T = 50 treated individuals; UT = 29 untreated individuals.
Significant differences between treated and untreated groups (dichotomized data):
Children: 11-year-olds: Chi-square = 10.849, df = 1, P < .01.
16-year-olds: Fisher exact test NS.
Parents: 11-year-olds: Chi-square = 15.504, df = 1, P < .01.
16-year-olds: Chi-square = 2.549, df = 1 NS.
uate an orthodontic service by applying patients’ atti-
tudes to measure outcome, both for the individuals hav-
ing received treatment and for those who were left
untreated. Another objective was to utilize longitudinal
data from both the children and their parents in an
effort to study whether decisions about treatment coin-
cided with attitudes over a 5-year period. This was
done by assessing whether conflicting attitudes among
children and their parents were prevalent or persisted.
Disagreement within the family units may be convolut-
ing to the decision-making process and thus affect the
relevance of informed consent.
SUBJECTS AND METHODS
Subjects Studied in 1990 (11-year-olds)
In 1990, a sample of fourth-grade children was
established by selecting four classes from three schools
in the city of Drammen, Norway.8 To assure a geo-
graphic and socioeconomic spread, schools located in
different areas of the city were selected. The children
and one parent of each child were invited by letter to
participate in the study. Of 104 families invited, 99
children (48 girls, 51 boys; mean age, 10.7 years) and
93 accompanying parents presented themselves at a
public dental clinic for a clinical examination of the
child and an interview of both the child and the parent.
Follow-up Study in 1996 (16-year-olds)
In 1996, 94 of the 99 children examined in 1990
could be traced. The children were invited to reattend
for a clinical examination and an interview; a consent
form for participating in the study together with a ques-
tionnaire were sent to the parents. After the initial invi-
tation, 83 children responded; an additional two
responded after one reminder. A total of 85 children
attended a public dental clinic. Two individuals still in
American Journal of Orthodontics and Dentofacial Orthopedics
March 1999
16-Year-old stage
Parents Children Parents
T UT T UT T UT
15 22 49 25 47 4
7 1 4 2 4
0 0 0 1 1
8 0 0 0 0 0
treatment were excluded, and the sample consisted of
83 children (37 girls, 46 boys; mean age, 16.3 years);
52 had received treatment and 31 were untreated.
Because three parents’ questionnaires were missing in
1990 and another one was missing in 1996, complete
longitudinal data were available from 79 family units.
At both examinations, impressions for study casts were
obtained from all the children for assessment of
occlusal morphology.
Attitudes to Dental Appearance and Need
for Orthodontic Treatment
Both in 1990 and 1996, the interviews were per-
formed with the use of questionnaires that included
questions about the children’s satisfaction with their
own dental arrangement and desire for orthodontic
treatment (Appendix). The parents were asked to
assess their attitudes to the child’s occlusal status and
need for treatment. Separate forms were presented to
the children and their parents. In 1990, the children and
the parents filled in the questionnaires independently in
separate rooms. In 1996, the children completed the
questionnaires at the public dental clinic, whereas for
practical reasons, the parental form was mailed to the
home address. The questions presented in 1990 and
1996 were similar and adapted from other studies.9,10
Each question had four fixed alternative answers that
range from a strongly positive to a strongly negative
response.
Statistical Procedures
According to a method adapted from Helm et al,9
scores from 0 to 2 were assigned to the answers (see
Appendix). Score 0 was given to the positive answers,
whereas scores 1 and 2 were given to the negative and
strongly negative answers, respectively. Association
American Journal of Orthodontics and Dentofacial Orthopedics Fernandes, Espeland, and Stenvik 229
Volume 115, Number 3

between answers to the questions was analyzed by cal-

culating Spearman rank-order correlation coefficients.
By adding the scores of each of the two questions, we
constructed a composite measure representing “ortho-
dontic concern” for the children and the parents sepa-
rately. The scores on the concern variable thus ranged
from 0 to 4, with 0 indicating no concern, 1 slight, 2
moderate, 3 marked, and 4 great concern. For parts of
the analysis, the concern variable was dichotomized
into no concern (score 0) and concern (score 1, 2, 3, 4).
Differences in concern between genders and between
orthodontically treated and untreated subgroups were
analyzed by the Chi-square test or Fisher exact test
after dichotomizing the concern variable. The Mann-
Whitney test was used to compare the children’s and
parents’ responses. For longitudinal comparisons the
Wilcoxon signed rank test was applied.

RESULTS
The answers to the questions about satisfaction with
dental arrangement and desire for orthodontic treatment
were significantly related to each other, both for the chil-
dren’s and the parents’ responses at both stages. Spear-
man rank-order correlation coefficients of .433 (P <
.001) and .284 (P < .01) emerged for the children; corre-
lation coefficients of .670 (P < .001) and .421 (P < .001)
were found for the parents, at the first and second exam-
ination, respectively. The baseline data at 11 years of age
for the dropouts and those who remained through age 16
years were compared. The distribution of answers to the
questions about satisfaction with dental arrangement and
desire for treatment, as well as the scores on the concern
variable showed no statistically significant difference
between the two groups, either for the children’s or par-
ents’ responses (Chi-square test, df = 1).
The children’s and the parents’ scores on the con-
cern variable at both examinations (11 and 16 years)
are given in Table I, and rates of agreement between
child’s and parent’s responses are presented in Fig 1.
Although the children were orthodontically untreated
at the time of the first examination, the results are also
presented separately for those who later received treat-
ment and those who remained untreated in order to
relate the individuals’ attitudes to the treatment deci-
sion. As no statistical differences were observed
between the genders, or between the parents of girls
and boys, a pooled analysis is presented.
Concern in the Whole Sample and in Subgroups
Total sample. At 11 years of age, 52% of the children
and 53% of the parents expressed concern for their
own/child’s occlusal condition. However, the majority of
subjects could be characterized as slightly or moderate-
Fig 1. Relative frequencies of 79 children and their par-
ents who agreed or disagreed in their concern for own/
child’s occlusal condition for the total sample (A), for
orthodontically treated (B), and for untreated (C) indi-
viduals. The left part of the illustration represents the 11-
year-old stage and the right part the 16-year-old stage.
The connecting dotted lines delineate the proportion of
children (upper part) and parents (lower part) that
agreed in their orthodontic concern (red or gray sectors
coinciding), and the proportion of children and parents
that disagreed (the child is concerned and the parent is
not, and vice versa: delineated sectors have different
colors).
ly concerned (score 1 or 2) (Table I). At the age of 16
years, most children (94%) and their parents (90%) were
not concerned (score 0), and none had concern scores
indicating marked or great concern (scores 3 and 4).
Treated versus untreated. At the first examination,
relatively more individuals in the group that later
received treatment expressed concern compared with
the untreated group (children and parents both P <
0.01) (Table I). At 16 years, there was no significant
difference between previously treated and untreated
subjects or their parents.
230 Fernandes, Espeland, and Stenvik
Table II. Longitudinal comparisons between expressed concern at both stages among children and parents (absolute frequencies). P values refer
to difference in concern at 11 and 16 years (Wilcoxon signed rank test).
No concern
T UT
Response at
11 years Children Parents Children
No concern 16 14 19
Concern 33 33 6 5
Total 49 47 25
Treated (T): Children, P < .001; parents, P < .001.
Untreated (UT): Children, NS; parents, NS.
T = 50 treated individuals, UT = 29 untreated individuals.
The change in relative frequency of concerned sub-
jects over the 5-year period is illustrated in Fig 1 both
for the whole sample and the subgroups (dichotomized
responses). A decrease was apparent in all groups, and
the greatest change was observed among those having
had treatment.
Agreement Between Children and Parents
The majority of family units agreed in their respons-
es as expressed by the concern variable (concern/no
concern); overall agreement was 74% at the age of 11
years and 91% at 16 years. No statistical differences
were observed between child and parental scores on the
concern variable for the total sample and subgroups
both at 11 and at 16 years (Mann-Whitney test). In Fig
1, the relative frequencies of agreements and disagree-
ments within family units are illustrated as expressed by
the concern score. Conflicting opinions between chil-
dren and parents varied from 24% to 30% in the whole
sample and subgroups at 11 years, and persisted only in
8% to 10% of the family units in the total sample and
subgroups at 16 years. Fig 1 also illustrates the direction
in which the conflicting opinions emerged.
Longitudinal Comparisons
Table II presents longitudinal comparisons between
the responses at 11 and 16 years (dichotomized
responses). Among the children who did not receive
orthodontic treatment, the majority was unconcerned at
both 11 and 16 years, and only two of the eight children
who expressed concern as 11-year-olds remained con-
cerned. The majority of the children who had received
orthodontic treatment had changed their response and
were not concerned at the second examination. One
third of the treated 16-year-olds were unconcerned at
both stages. The change in concern was statistically
significant only for the treated group (children and par-
American Journal of Orthodontics and Dentofacial Orthopedics
March 1999
Response at 16 years
Concern
T UT
Parents Children Parents Children Parents
19 1 1 2 3
0 2 2 2
24 1 3 4 5
ents both P < .001). Parents and children paralleled
each other in the change in responses.
The longitudinal comparison between the agree-
ments and disagreements within the 79 family units
over the 5-year period appears from Fig 1. A decrease
in disagreements was observed among the treated and
untreated individuals. Among the treated individuals
the decrease was statistically significant (P < .05).
DISCUSSION
Of the 94 children who could be traced in 1996, the
attendance rate was 90% (85 of 94), and 79 (85%) of
the initial 93 family units were included in the follow-
up. Over a 5-year period, a dropout rate of this magni-
tude or higher is to be expected. The nonattendants did
not appear to have biased the results as comparison
between the dropouts and those who remained in the
study revealed no differences in responses at 11 years.
The concern variable, which is a construction based
on the individual’s satisfaction with dental arrangement
and desire for treatment, is a general estimate of an
overall attitude to malocclusion and orthodontic care
and has been discussed previously.9,11 The concept “sat-
isfaction” is in itself ambiguous as discussed by ter
Heege et al12 but is among the most frequently used in
studies of laypersons’ attitudes to health care, and there-
fore appears to be a well-established term that general-
ly is incorporated in patient-centered evaluation of need
and outcome. Desire for treatment was assessed without
providing any information about treatment factors such
as expenses, duration, and inconvenience. Such infor-
mation might have produced a different response,
although the child and parent generations represented in
the study are assumed to be familiar with orthodontics
as the service has been well organized in the area over
the last decades. The orthodontic concern scores in this
study are considered to categorize attitudes on an ordi-
American Journal of Orthodontics and Dentofacial Orthopedics
Volume 115, Number 3
nal scale, although the labels “slight,” “moderate,”
“marked,” and “great concern” are only indicative as
they have not been validated.
No information has been available about the ortho-
dontic history of the respondents in the interim period
between 11 and 16 years of age, except whether
orthodontic treatment had been undertaken or not. The
orthodontic screening, referral, and advice about treat-
ment that had been provided to the individual child-
parent units are not known. In the present study treatment,
no treatment has been looked on as the independent vari-
able. It might be argued, however, that treatment results
primarily from patient’s concerns, and that concern is the
independent variable. Selection of orthodontic concern as
the dependent or outcome variable in this study has been
made in order to focus on the patient and parent perspec-
tive of care.
The significant difference in concern among treated
versus nontreated at the 11-year-old stage indicated
that the care system generally identified those with a
perceived need for treatment. The explanation may be
that the individual’s expressed attitudes had been taken
into account or there are more severe malocclusions
among those receiving treatment. This will be analyzed
in a separate study. The outcome studied comprises
both the effectiveness (whether the right decisions were
made) and efficiency (the quality of the treatment) of
the care system from a patient perspective.
Incongruities in concern among child and parent
were observed in about 25% of the family units at the
11-year-old stage. This might indicate that for some,
the question of whom to give priority, child or parent,
could arise during counseling and advice in the process
of obtaining informed consent about treatment. No par-
ticular pattern appeared as to whether the child’s or
parent’s initial concern most frequently was related to
the decision to start treatment. Conflicting attitudes
among child and parent may have been resolved during
orthodontic counseling. A substantial agreement with-
in the family units at the 16-year-old stage could be
looked on as support for an assumption that intra-unit
resolution really took place during counseling and
treatment. On the other hand, a significant agreement at
16 years of age is to be expected because very few in
this sample were concerned at that stage. Disagreement
at the 11-year-old stage did not prevail in this sample
long term, regardless of whether treatment/no treat-
ment coincided with child’s or parent’s initial concern.
The agreement within the family units observed in
the present study may in part be due to the inherent
family characteristic that children and parents share
basic values and priorities.13 Parents know their chil-
dren well, and know when to deny or accept treatment
Fernandes, Espeland, and Stenvik 231
in situations of conflicting concerns. The general
absence of disagreement between the 16-year-olds and
their parents may also be due to the high uptake of
treatment in the sample studied. In situations with low
treatment frequency because of limitations in access to
treatment or restricted funding or inability to pay, a
more diverse child/parent relationship might result and
accordingly confront the orthodontist with a more chal-
lenging consent process.
SUMMARY AND CONCLUSIONS
Longitudinal data on subjective need for orthodon-
tic treatment among a cohort including 79 children and
their parents showed:
1. The care system identified and provided treatment
to those with an orthodontic concern.
2. Most individuals who had received treatment during
the observation period as well as those who were left
untreated were satisfied with their occlusal condition
at the follow-up examination.
3. Disagreement between children and parents about
need was observed among one of four family units
before orthodontic consultation (11-year-old stage).
At follow-up (16-year-old stage), more than 90% of
the units agreed. Among units disagreeing, scores at
16 years represented slight and moderate concern.
Informed consent as a tool to make decisions about
orthodontic treatment therefore did not seem to be
negatively affected by conflicting attitudes among
children and their parents.
We wish to acknowledge Chief Dental Officers
Turid Album Alstad and Per Hauge and their staff for
assistance in collecting the material.
REFERENCES
1. Reiser J, Dyck AJ, Curran WJ. Ethics in medicine: historical perspectives and con-
temporary concerns. London: MIT Press, 1978.
2. Faden RR, Beauchamp TL. A history and theory of informed consent. New York:
Oxford University Press, 1986.
3. Rothman DJ. Strangers at the bedside: a history of how law and bioethics transformed
medical decision making. New York: Basic Books Inc; 1991.
4. Nash DA. Professional ethics and esthetic dentistry. J Am Dent Assoc 1988;115:7E-9E.
5. Ackerman JE. Ethics and risk management in orthodontics. In: Ghafari JG, Moorrees
CFA, eds. Orthodontics at Crossroads. Boston: Harvard Soc Adv Orthod; 1991. p 49-60.
6. Donabedian A. The definition of quality and approaches to its assessment. Ann Arbor:
Health Administration Press, University of Michigan; 1980. p 118-20.
7. Vuori H. Patient satisfaction: an attribute or indicator of the quality of care? Qual Rev
Bull 1987;13:106-8.
8. Espeland LV, Ivarsson K, Stenvik A. A new Norwegian index of orthodontic treatment
need related to orthodontic concern among 11-year-olds and their parents. Communi-
ty Dent Oral Epidemiol 1992;20:274-9.
9. Helm S, Petersen PE, Kreiborg S, Solow B. Effect of separate malocclusion traits on
concern for dental appearance. Community Dent Oral Epidemiol 1986;14:217-20.
10. Espeland LV, Stenvik A. Perception of personal dental appearance in young adults:
relationship between occlusion, awareness and satisfaction. Am J Orthod Dentofacial
Orthop 1991;100:234-41.
11. Espeland LV. An appraisal of non-professional perspectives on occlusal anomalies and
orthodontic care [Thesis]. University of Oslo, 1993. ISBN 82-992492-3-6.
12. ter Heege G, Prahl-Andersen B, Stenvik A, Williams R. Orthodontic services: a con-
sumers view. Proc Studieweek 1995 p. 244-54. Nederlandse Vereniging voor Ortho-
dontische Studie, 1995. ISBN 90 8006 472 6.
13. Albino JE, Tedesco LA. The role of perception in treatment of impaired facial appear-
ance. In: Alley TR, ed. Social and applied aspects of perceiving faces. Hillsdale, NJ:
Lawrence Erlbaum Associates; 1988. p 217-37.
232 Fernandes, Espeland, and Stenvik American Journal of Orthodontics and Dentofacial Orthopedics
March 1999

APPENDIX Yes, I think so (1)

Questionnaire form (translated from Norwegian). Yes, very much (2)
Questions with alternative answers presented to the
PARENT’S FORM (1990 AND 1995/96):
children and the parents in 1990 and in 1995/96.
(Scores assigned to the answers are in parentheses.) • Are you satisfied with the arrangement of your child’s
teeth?
Very satisfied (0)
CHILD’S FORM (1990 AND 1995/96): Satisfied (0)
• Are you satisfied with the arrangement of your teeth? Rather dissatisfied (1)
Very satisfied (0) Very dissatisfied (2)
Satisfied (0) • Do you think that your child needs orthodontic
Rather dissatisfied (1) treatment?
Very dissatisfied (2) No, definitely not (0)
• Do you want to have your teeth straightened? No, I do not think so (0)
No, not at all (0) Yes, I think so (1)
No, I do not think so (0) Yes, definitely (2)