Periodontal Risk Assessment Model in a Sample of Regular and Irregular

Volume 83 • Number 3
Periodontal Risk Assessment Model
in a Sample of Regular and Irregular
Compliers Under Maintenance Therapy:
A 3-Year Prospective Study
Fernando Oliveira Costa,* Luı´s Ota´vio Miranda Cota,* Eugeˆnio Jose´ Pereira Lages,*
Ana Paula Lima Oliveira,* Sheila Cavalca Cortelli,† Jose´ Roberto Cortelli,†
Telma Campos Medeiros Lorentz,* and Jose´ Eusta´quio Costa*
Background: The purpose of this study is to investigate the
association of the periodontal risk assessment (PRA) model
with the recurrence of periodontitis and tooth loss during periodontal maintenance therapy (PMT).
Methods: In a prospective PMT program, 75 regular complier (RC) and 89 erratic complier (EC) patients were selected.
A periodontal examination and PRA were performed after active periodontal therapy and after 3 years of PMT. Risk profiles
(low, moderate, or high) of participants were evaluated, and
the recurrence of periodontitis and tooth loss were analyzed
using univariate and multivariate analyses.
Results: RCs showed less recurrence of periodontitis and
tooth loss than ECs (P <0.05). Rates of periodontitis recurrence in RCs and ECs were 2.7% and 3.4%, respectively, for
the moderate-risk profile and 6.7% and 11.2%, respectively,
for the high-risk profile. During PMT, 49 teeth (0.65 – 1.4
teeth per participant) were lost in the RC group, and 70 teeth
(0.78 – 2.1 teeth per participant) were lost in the EC group.
High-risk profile participants showed more recurrence of
periodontitis and lost significantly more teeth than did participants with moderate- or low-risk profiles in RC and EC
groups (P <0.05).
Conclusion: The risk profile influenced the recurrence of
periodontitis and tooth loss. RCs had less recurrence of periodontitis and less tooth loss. The PRA model can be useful in
particularizing the risk of patients and adjusting recall intervals. J Periodontol 2012;83:292-300.
KEY WORDS
Compliance, patient; maintenance; periodontitis; risk factor;
tooth loss.
* Department of Periodontology, Dentistry School, Federal University of Minas Gerais, Belo
Horizonte, MG, Brazil.
† Department of Dentistry, Periodontics Research Division, University of Taubate´, Taubate´,
SP, Brazil.
T
he most usual problem in periodontal maintenance therapy (PMT) is
the compliance and return of patients at regular intervals. It was well documented that patients may or may not at
all comply with suggested maintenance
regimens.1-7
Because treated patients are not equally
susceptible to periodontal disease progression,8,9 some patients may attend
periodontal maintenance visits at shorter
intervals than less-susceptible patients.10
This criterion can be useful to adjust intervals and minimize difficulties in the compliance of patients to PMT.7
The creation of multifactorial riskassessment models that include relevant
risk factors for future disease progression was proposed to identify the susceptibility of patients for the recurrence of
periodontitis.6,10-13 Lang and Tonetti12
proposed the periodontal risk assessment (PRA), which is a functional diagram
composed of six vectors based on six
parameters (clinical, systemic, and environmental factors) to evaluate the risk
of recurrence of periodontitis at a patient
level, classifying patients in a low-risk
(LR), moderate-risk (MR), or high-risk
(HR) profile. The combined evaluation
of these factors provides an individualized
total risk profile after active periodontal
therapy (APT).
doi: 10.1902/jop.2011.110187
292
Costa, Cota, Lages, et al.
J Periodontol • March 2012
Only some longitudinal studies6,10,14-16 attempted
to validate the effect and reproducibility of this approach. A recent study16 also evaluated the PRA model,
and the authors included a limited number of patients
(n = 20) treated for severe periodontitis and concluded
that the proposed model overestimated the risk for disease recurrence. Leininger et al.6 used the PRA model
to evaluate 30 patients who were reexamined after APT
and reported that erratic compliers (EC) patients lost
more teeth than did regular compliers (RC) patients.
At present, a great challenge for periodontitis is to
use instruments that can bring benefits to periodontal
practice. In this sense, the use of multifunctional risk
models can be a powerful tool to monitor the risk of
the recurrence of periodontitis, optimize clinical decisions, influence the pattern of compliance and the
adherence to PMT, improve oral health, and reduce
treatment costs.13
Hence, the aims of this prospective longitudinal
study were to: 1) evaluate and classify the patient’s
risk through the PRA model proposed by Lang
and Tonetti12 in patients enrolled in a PMT program
during a 3-year interval; and 2) assess and compare
the association of the proposed PRA model with the
recurrence of periodontitis and tooth loss according
to the pattern of compliance (regular or erratic).
MATERIALS AND METHODS
All participants under PMT from a cohort study that
comprised 238 patients (96 males and 142 females;
age range: 18 to 62 years) who were monitored in
a private dental clinic in Belo Horizonte, Brazil, from
November 2006 to November 2009 were determined
to be eligible to participate in the present study. Patients with good general health who underwent basic
periodontal therapy (comprised of non-surgical
and/or surgical procedures) were included in the
sample. In addition, these patients also presented
the following criteria: 1) diagnosis of chronic moderate–advanced periodontitis before the APT with ‡4
sites with probing depths (PDs) ‡5 mm and clinical attachment loss (AL) ‡3 mm, bleeding on probing
(BOP) and/or suppuration (SUP), and radiographic
evidence of bone loss;17 2) completion of APT in a period of <4 months before entry into the PMT program;
and 3) ‡14 teeth in the oral cavity.18 All participants
provided written informed consent. The present study
was approved by the Ethics Research Committee,
Federal University of Minas Gerais (protocol #060/05).
Patient exclusion criteria were: 1) pregnancy
(n = 3), 2) debilitating diseases that could impair
the immune system (such as human immunodeficiency virus/acquired immune deficiency syndrome,
cancer, and autoimmune diseases; n = 4), 3) drug-induced gingival hyperplasia (n = 6), 4) systemic anti-
biotics use during APT (n = 14), 5) >2 osseointegrated
dental implants (n = 10), 6) <14 teeth present (n = 19),
and 7) refusal to participate (n = 18).
A sample consisting of 164 volunteer participants
was selected and classified according to the criteria
proposed by Demirel and Efeodlu.19 Participants were
considered: 1) RCs (n = 75) if they presented 100%
cooperation with recall visits, or 2) EC (n = 89) if they
missed any of the scheduled recall visits but continued
to appear irregularly. Data obtained after APT were
determined at baseline and compared to data obtained after 3 years of follow-up, which was
determined as the final examination (T2). T2 was
performed after nine recall visits for RC patients and
four recall visits for EC patients. The mean – SD time
between recalls in the RC group was 3.3 – 0.5 months
and 8.1 – 1.2 months in EC group (no individual had
<4 recall visits in the study period). A maximum 4month interval time was proposed for all patients
independent of their risk profiles.
Data related to the following characteristics were
collected from all patients: sex, age, family income,
education level, plaque index (PI),5 number of teeth,
smoking (non-smokers/ex-smokers and smokers:
10 to 19 or >19 cigarettes per day, respectively),12
and the presence of diabetes (glycemic values >110
mg/dL).20
In all periodontal clinical examinations during monitoring visits, data comprising the PD, clinical AL, furcation involvement, BOP, PI, and SUP were recorded
for each patient using the methodology proposed by
Lorentz et al.5 Data of interest (clinical parameters,
radiographs, and systemic and behavioral variables)
collected after APT (baseline) and at T2 were used in
this study to construct the PRA model proposed by
Lang and Tonetti.12
Radiographs
Intraoral periapical and bite-wing radiographs were
taken using a long-cone paralleling technique. Alveolar bone loss was measured at both proximal surfaces (mesial and distal) with a negatoscope‡ that
magnified nine times the original image size. The distance from the cemento-enamel junction to the alveolar bone crest was measured using a pachymeter.§
The bone loss variable was recorded according to
Lang and Tonetti.12
PRA Model
On the basis of patient data at APT and T2, the risk
for the recurrence of periodontitis was calculated for
all 164 patients using the PRA model.12 A summary
of definitions of LR, MR, and HR profiles for the recurrence of periodontitis is presented in Table 1. Patients
‡ Ampligraf, Castells, Sa˜o Paulo, SP, Brazil.
§ Mitutoyo Sul Americana, Suzano, SP, Brazil.
293
Periodontal Risk Assessment in Maintenance Therapy
Volume 83 • Number 3
Table 1.
Classification of Risk According to the Parameters of the PRA Model Proposed
by Lang and Tonetti12
BOP
Sites With PD ‡5 mm
Tooth Loss
Bone Loss/
Age Ratio (mm)*
Smoking
Genetic Factor/
Systemic
LR
0% to 9%
0 to 4
0 to 4
0 to 0.5
Non-smoker/ex-smoker
Negative
MR
10% to 25%
5 to 8
5 to 8
>0.5 to 1.0
10- to 19- cigarettes/day
–
HR
>25%
>8
>8
>1.0
>19 cigarettes/day
Positive
Risk Profile
– = not applicable.
* Bone loss was estimated from the percentage of the root length at the worst site in the posterior region on periapical radiographs or on bitewing radiographs
where 1 mm was equivalent to 10% bone loss.
determined to be LR yielded all parameters in the LR
area or, at the most, one parameter in the MR area.
Patients determined to be MR yielded ‡2 parameters
in the MR area and, at the most, one parameter in
the HR area. Patients determine to be HR yielded ‡2
parameters in the HR area.
Recurrence of Periodontitis
The progression of periodontitis was defined as an
interproximal clinical AL ‡3 mm in ‡2 teeth between
two different observation points according to the definition of the fifth European Workshop of Periodontology.21 In the present study, this definition was adapted
to define the recurrence of periodontitis between the
end of APT and T2.10
Intraexaminer and Interexaminer Reliability
Interviews, examinations, and periodontal procedures
were conducted by two calibrated and trained periodontists (FOC and EJPL) who were masked to the
degree of compliance. Measurements of PD and clinical AL were recorded and repeated within a 1-week
interval for 10 patients who were randomly selected
from both examinations (APT and T2). Results
showed satisfactory intraexaminer and interexaminer
k values for PD and clinical AL (k ‡ 0.81). Additionally,
intraclass correlation coefficients ‡0.82 were attained
in both evaluations. Bone-loss measures and calculations were performed by only one researcher (FOC)
who was masked to clinical data. The intraexaminer
k test was performed for 20 radiographs within a
15-day interval (k = 0.95).
Statistical Analyses
Statistical analyses included a characterization of the
sample and descriptive analysis of variables of interest.
Parametric and non-parametric tests (x2, KruskalWallis, Friedman, Fisher-exact, Student t, and MannWhitney U tests) were used when appropriate. To
determine the significance between groups, the correction of Bonferroni was applied.
294
PRA models were constructed twice for each of the
164 patients. Data were compared through the nonparametric Kruskal-Wallis test in relation to the risk
profile.
A logistic regression analysis was performed to
investigate the association among all risk variables
and the recurrence of periodontitis and tooth loss.
The first model, in which RCs and ECs were separately
analyzed in relation to the recurrence of periodontitis,
included the following variables: age, sex, number of
missing teeth, BOP in >30% of sites, PD >5 mm in 10%
of sites, smoking, and diabetes. Subsequently, a second multivariate analysis was separately performed
for the recurrence of periodontitis and tooth loss including the continuous variables of the PRA model
(as proposed by Matuliene et al.10) percentage of
BOP (unit increase of 1% of sites with BOP), number
of periodontal pockets with PD ‡5 mm (increase of
one pocket), number of missing teeth (increase of
one missing tooth), bone loss in relation to age (increase of 1% of bone loss), smoking (yes/no), and diabetes (yes/no). Odds ratio (OR) values and
respective 95% confidence intervals (CIs) were reported. All tests were performed using statistical
software.i Results were considered significant if a P
value <5% was attained (P <0.05).
RESULTS
The characterization of the sample dichotomized in
RC and EC groups is shown in Table 2. Variables such
as sex, marital status, smoking, diabetes, age (except
for the age group from 31 to 40 years,) and average
time of completion of active therapy showed no significant differences between groups, which demonstrated great homogeneity between groups.
Table 3 presents the periodontal status of RC and
EC groups after active APT and T2. RCs and ECs
showed no significant differences regarding the clinical
parameters PD, BOP, clinical AL, and SUP after APT
i SPSS v.16.0, IBM, Chicago, IL.
Costa, Cota, Lages, et al.
J Periodontol • March 2012
Table 2.
Characterization of RC and EC Patients According to Variables of Interest (N = 164)
Characteristic
RCs (n = 75)
ECs (n = 89)
P
Sex (n [%])
Females
Males
54 (72)
21 (28)
59 (66)
30 (34)
Age (range: 18 to 62 years) (years; n [%])
£30
31 to 40
41 to 50
>50
9
19
24
23
7
25
27
30
Marital status (n [%])
With companion
Without companion
51 (68.0)
24 (32.0)
57 (64.1)
32 (36.0)
NS*
NS*
Smoking (n [%])
Non-smoker/ex-smoker
Smoker
46 (61.3)
29 (38.7)
55 (61.8)
34 (38.2)
NS*
NS*
8 (10.7)
14 (15.7)
NS*
Number of recall visits (time between
recalls [months; mean – SD])
9 (3.3 – 0.5)
4 (8.1 – 1.2)
0.002†
Time of completion of active therapy
(months; average – SD)
39.2 – 1.9
40.3 – 1.7
NS†
Diabetes (n [%])
(12.0)
(25.3)
(32.0)
(30.7)
(7.9)
(28.1)
(30.3)
(33.7)
NS*
NS*
NS*
0.048*
NS*
NS*
NS = not significant.
2
* x test.
† Student t test for independent samples.
(baseline) (P >0.05). However, great differences regarding these parameters at T2 revealed a worse periodontal condition among EC patients.
The classification of periodontal risk for RC and EC
patients according to the PRA model after APT and at
T2 as well as changes in risk classification in the PRA
model from after APT to T2 are shown in Table 4. It
was observed that RC and EC patients showed no significant differences in risk classifications after APT.
However, the EC group presented a higher incidence
of HR patients at T2 (an increase of 6.7%) and a lower
incidence of MR patients (a reduction of 12.4%) compared to the changes in the RC group (an increase of
4% in HR and a reduction of 17.4% in MR). In both
groups, there was an increase of LR patients (RCs:
an increase of 13.3%; ECs: an increase of 5.6%).
It was observed at T2 that the EC group had a significantly greater number of patients presenting a recurrence of periodontitis (n = 15 [16.8%]) and tooth loss
(n = 27 [30.3%]) compared to the RC group (n = 8
[10.6%] and n = 17 [22.6%], respectively) (P =
0.027). In addition, when analyzing HR patients, it
was observed that the recurrence of periodontitis
and tooth loss among ECs (11.2% and 21.3%, respectively) was significantly higher than among RCs (6.7%
and 18.6%, respectively) (P <0.02). The reasons for
tooth loss were considered dichotomously (i.e., tooth
loss from periodontal disease and other reasons altogether [cariogenic, prosthetic, and endodontic reasons and root fracture). In the RC and EC groups,
respectively, 35 and 55 teeth (71.4% and 78.6%) were
lost due to periodontal disease, and 14 and 15 teeth
(28.6% and 21.4%) were lost because of other reasons. In both groups, the loss due to periodontitis
was significantly higher than for reasons other than
periodontal disease (P <0.001). Thus, at the end of
the 3-year period, 119 teeth were lost (0.72 – 1.9 teeth
per patient). The RC group had 49 lost teeth (0.65 –
1.4 teeth per patient), and the EC group had 70 lost
teeth (0.78 – 2.1 teeth per patient).
The following results were reported for the occurrence of tooth loss among RC and EC patients at T2
according to the PRA profile: In the RC group, three individuals (4%) at MR lost 10 teeth, and 14 individuals
(18.6%) at HR lost 39 teeth; no tooth loss was observed in LR individuals. In the EC group, one individual (1.1%) at LR lost four teeth, seven individuals
(7.9%) at MR lost 21 teeth, and 19 individuals at HR
(21.3%) lost 45 teeth.
The multivariate model for the recurrence of periodontitis among RC patients at T2 included smoking,
BOP in >30% of the sites, and number of missing teeth.
295
Periodontal Risk Assessment in Maintenance Therapy
Volume 83 • Number 3
Table 3.
Periodontal Status of RC and EC Patients After APT and at T2
T2†
APT (baseline)*
RC
Periodontal Parameters
EC
RC
EC
Mean
SD
Mean
SD
Mean
SD
Mean
SD
2.3
0.9
– 1.2
– 0.2
2.7
1.1
– 1.5
– 0.6
2.5A
1.2A
– 3.3
– 0.5
4.7B
1.8B
– 3.7
– 0.9
Clinical AL (% affected sites)
4 to 5 mm
11.6
‡ 6 mm
7.1
– 1.3
– 0.9
12.3
7.3
– 1.8
– 1.3
12.3A
7.8A
– 1.7
– 1.2
13.6B
9.7B
– 2.3
– 1.8
BOP (% affected sites)
– 3.3
25.7
– 4.4
27.1A
– 4.1
32.9B
– 5.8
PD (% affected sites)
4 to 5 mm
‡ 6 mm
Suppuration (%
affected sites)
24.1
0.05A
– 0.07
0.07
– 0.13
0.11A
– 0.04
0.19B
– 0.25
PI
30.5
– 9.2
33.6
– 10.1
32.9A
– 4.8
37.8B
– 7.7
Mean number of teeth
per patient
24.6
–
23.4
–
22.7A
–
20.05B
–
– = not applicable.
Total sample = 164 patients (75 RC and 89 EC patients). APT = After active periodontal therapy.
* Comparisons between RC and EC patients after APT (baseline) and at T2 (P >0.05).
† Comparisons between RC and EC patients at T2: values followed by different capital letters in rows were significantly different (P <0.05).
Student t test was used for independent samples; Bonferroni correction was used for multiple comparisons (P <0.005).
Among EC patients, the multivariate final model included BOP in >30% of sites, smoking, PD >5 mm in
10% of sites, diabetes, and the number of missing
teeth (Table 5).
A second multivariate logistic regression analysis
was performed separately for tooth loss and the recurrence of periodontitis that occurred between APT and
T2 considering sample characteristics and parameters of the PRA model. Odds ratio values and respective 95% confidence intervals are shown in Table 6.
The multivariate model, considering the parameters
of the PRA model as continuous variables, included
BOP, the number of missing teeth, bone loss/age
ratio, diabetes, and smoking for tooth loss and BOP,
diabetes, and smoking for the recurrence of periodontitis.
DISCUSSION
The present prospective study presents two distinct
objectives: 1) to evaluate and classify patient risk
using the PRA model proposed by Lang and Tonetti,12 and 2) to determine the association of this
risk classification with the recurrence of periodontitis
and tooth loss according to the pattern of the compliance level of the 164 patients during 3 years of
PMT.
EC and RC groups had a significantly greater number of individuals classified as LR at T2 (RCs: 17.3%;
296
ECs: 10.3%). This finding showed a beneficial effect of
PMT in both groups. Thus, the pattern of compliance
and regular recall visits during PMT strongly influenced the risk profiles of patients. This finding was
in accordance with previous studies reporting that
compliance and adherence to PMT strongly reduced
the risk of disease recurrence and tooth loss over
time.10,22-25
A great problem during PMT is the adherence to follow-up visits. Many studies reported low rates of compliance.1,2,4,9,24,26-28 In relation to the average time
between recalls, the present study showed an important difference between the RC group (3.3 months)
and EC group (8.1 months). This difference was
strongly related to the worse periodontal condition
observed among EC patients.
Thus, in the present study, the range of 4 months is
ideally recommended. However, this time is very difficult to establish in clinical practice. The results of
the present and other studies10,24,26-28 suggested
that RCs, who complied with suggested maintenance
schedules, substantially reduced the risk for a recurrence of periodontitis and tooth loss compared to ECs.
In this sense, the use of the PRA model as a way to
individualize risk and customize recall visits can be
useful to optimize periodontal care, reduce treatment
costs, avoid treatment overestimation, and minimize
the recurrence of periodontitis and tooth loss. The
Costa, Cota, Lages, et al.
J Periodontol • March 2012
Table 4.
Risk Classification and Risk Profile Changes for RC and EC Patients (n [%]) After APT
(baseline examination) and at T2
Individuals With Changes in PRA Model From APT to T2
No Risk Change
Risk Increase†
RC
EC
RC
4 (4.5)
0 (0)
19 (25.4)
30 (33.7)
27 (36)
33 (37)
Risk Decrease*
Risk Profile
RC
EC
LR
10 (13.3)
5 (5.6)
MR
13 (17.4)
11 (12.4)
HR
0 (0)
3 (4)
0 (0)
APT
T2
RC
(n = 75)
EC
(n = 89)
RC
(n = 75)
EC
(n = 89)
0 (0)
3 (4)
4 (4.5)
13 (17.3)
9 (10.1)
0 (0)
0 (0)
45 (60)
52 (58.3)
32 (42.8)
41 (46,1)
3 (4)
6 (6.7)
27 (36)
33 (37.2)
30 (40)
39 (43.7)
EC
* Lower PRA profile at T2 in relation to after APT.
† Higher PRA profile at T2 in relation to after APT.
Table 5.
Multivariate Logistic Regression Model
for the Recurrence of Periodontitis for RC
and EC Patients at T2
Recurrence of Periodontitis
RC patients (n = 75)
Smoking
BOP >30% of sites
Number of missing teeth
Constant (coefficient: 0.027)
ECs patients (n = 89)
Smoking
BOP >30% of sites
PD >5 mm in 10% of sites
Diabetes
Number of missing teeth
Constant (coefficient: 0.08)
OR (95% CI)
P
4.5 (1.36 to 9.71)
2.9 (1.02 to 10.1)
1.82 (1.04 to 6.45)
–
0.001
0.032
0.041
0.000
6.3
3.4
3.6
1.82
3.46
<0.001
0.038
0.025
0.031
0.002
0.000
(1.23
(1.27
(1.02
(1.02
(1.19
–
to
to
to
to
to
11.3)
10.9)
5.21)
2.72)
5.89)
– = not applicable.
PRA model showed that EC patients classified as HR
had a significantly greater recurrence of periodontitis
and tooth loss (ECs: 11.2% and 21.3%, respectively;
RCs: 6.7% and 18.6%, respectively). Previous studies10,16 using the PRA model in PMT programs
reported significantly higher rates of disease recurrence and tooth loss than those reported in our
study.
In a retrospective cohort, Matuliene et al.,10 performed PRAs of 160 patients after periodontal therapy
and after 9.5 years of PMT. The authors reported rates
of periodontitis recurrence of 18.2% in LR, 42.2% in
MR, and 49.2% in HR patients. During PMT, 1.61 teeth
per patient were lost. Despite using the same criteria
proposed in our study, a longer period of study and
a greater interval time between recalls may have contributed to this difference.
Jansson and Norderyd,16 using a PRA model in 20
patients treated for severe periodontitis, concluded that
the proposed model overestimated the risk for disease
recurrence. However, additional comparisons with this
study16 are discouraged because the sample included
a very small number of patients (n = 20).
Various retrospective6,23,29-32 and prospective
studies5,10,15,32 of PMT reported a higher prevalence
of tooth loss among EC compared to RC patients. The
present study reported a high rate of tooth loss of a total of 119 teeth; 49 teeth were lost among RC patients,
and 70 teeth were lost among EC patients, which corresponded to an annual average of 0.216 teeth per
patient and 0.260 teeth per patient, respectively.
The higher rates of tooth loss were observed among
HR patients (79% in the RC group and 64.3% in the
EC group). A greater number of missing teeth was
concentrated in a small number of individuals. Previous studies10,15 using the PRA model showed different
annual rates of tooth loss: Eickholz et al.15 reported
0.049 for regular and 0.257 for irregular compliers;
Matuliene et al.10 reported 1.08 for regular and 3.12
for irregular compliers in 9.5 years, which corresponded to 0.13 and 0.30 annually, respectively.
Therefore, ECs lost a higher number of teeth compared to RCs in both groups when considering HR
patients. Differences observed between these groups
may be related to lower rates of a recurrence of periodontitis, lower values of PDs and clinical ALs, a
greater compliance and adherence with the PMT program, and, especially, a higher professional vigilance
for patients classified as HR patients. The number
of lost teeth was extensively reported as a reliable
indicator of the risk for the recurrence of periodontitis.6,10,23,29-31
Individuals with >2 dental implants were excluded
from the present study because the peri-implant clinical data were not suited for inclusion in the PRA
model. Thus, there is a need for risk-assessment
297
Periodontal Risk Assessment in Maintenance Therapy
Volume 83 • Number 3
Table 6.
Multivariate Logistic Regression Models for Tooth Loss and Recurrence of Periodontitis
During PMT in Relation to Patient Characteristics and Parameters Determining Risk
Profiles at APT
Recurrence of Periodontitis
OR (95% CI)
Variable
Tooth Loss OR (95% CI) (P)
Sample characteristic
RCs versus ECs
2.35 (1.02 to 5.23) (0.015)
1.64 (1.08 to 3.62) (0.001)
Parameters of PRA model
BOP
Sites with PD ‡5 mm
Number of missing teeth
Bone loss/age ratio
Diabetes (yes versus no)
Smoking (yes versus no)
2.23
1.81
2.21
2.73
1.92
3.41
1.88
1.04
1.69
1.21
1.66
3.78
(1.02
(0.96
(1.13
(1.04
(1.01
(1.26
to
to
to
to
to
to
5.68) (0.021)
1.94) (0.361)
5.31) (0.022)
4.92) (<0.001)
7.28) (0.026)
11.41) (<0.001)
(1.01
(0.64
(0.79
(0.42
(1.21
(1.32
to
to
to
to
to
to
5.46)
1.89)
2.72)
3.46)
4.62)
7.28)
(0.03)
(0.483)
(0.582)
(0.662)
(0.004)
(<0.003)
Significant P values are shown in bold.
models that include the evaluation of implants, especially in patients partially rehabilitated with implants.
The multivariate logistic model for the recurrence
of periodontitis included classic risk factors such as
smoking, BOP in >30% of the sites, and the number
of lost teeth for RC and EC groups. In addition, PDs
>5 mm in 10% of the sites and diabetes were also retained in the EC model. The inclusion of behavioral
and biologic variables together with clinical parameters was similar to a previous study.10 In a multivariate
logistic regression analysis, Matuliene et al.10 reported that an HR profile according to the PRA model
was associated with the recurrence of periodontitis.
Another significant factor for the recurrence of periodontitis was a PMT duration >9.5 years.
The second multivariate logistic model analyzing
characteristics of the sample showed that ECs had
a 2.3-times higher odds of tooth loss and a 1.6-times
higher odds of a recurrence of periodontitis compared
to RCs. In the analysis for tooth loss, when including
the continuous variables of the PRA model, a higher
number of PRA parameters were retained as significant variables (BOP, tooth loss, bone loss/age ratio,
and smoking). In the analysis for the recurrence of periodontitis, the model included BOP as an indicator of inflammation and risk for further clinical AL,33,34 and
PRA parameters determining an HR profile such as
smoking and diabetes.12
A number of studies7,10,18,35,36 related smoking to
a higher risk for clinical AL and tooth loss. According
to Matuliene et al.,37 among the six parameters used
to determine periodontal risk in the PRA model, only
smoking was considered a good predictor for the recurrence of periodontitis.
The high frequency of BOP among patients under
PMT in the present study can lead to some questions
298
regarding the adherence of patients to treatment and
their plaque-control abilities. Percentages of BOP from
20% to 30% determined an HR for the recurrence of
periodontitis in studies from Badersten et al.38 and
Claffey et al.33 In addition, bleeding gums can be a
self-perceived sign and alert subjects of the need for
a PMT visit.7,34
Similar results were reported by Jansson and Norderyd16 using a PRA model in 20 patients treated for
severe periodontitis. When they16 compared all patients by only using the BOP mean prevalence of
20% as a cutoff, 15 patients were categorized as having a low–moderate risk (75%) for periodontitis progression, and five patients (25%) were categorized
as having a HR for disease progression.
Studies have indicated that diabetes mellitus affects
the susceptibility, severity, and recurrence of periodontitis.7,28,35,39,40 Our results showed that diabetes
was associated with the recurrence of periodontitis in
the EC group. Patients with diabetes had a 1.66-times
higher chance of a recurrence of periodontitis and
a 1.92-times higher chance of tooth loss. These findings reaffirmed that factors related to the host susceptibility should be carefully monitored during PMT.
However, in the RC group, this association was not observed. This finding could have reflected the beneficial
effect of PMT during regular recall visits. One limitation
of the present study is the inclusion of numerous variables in the multivariate models. Thus, in small samples, the statistical power can be reduced. However,
these analysis strategies were often reported in general literature, and longitudinal studies in larger samples are difficult to obtain and require great cost and
logistics.5 In addition, all variables included in the final
logistic models were recognized indicators or risk factors for the progression of periodontitis and tooth loss.
Costa, Cota, Lages, et al.
J Periodontol • March 2012
CONCLUSIONS
It was concluded that the risk profile predicated the recurrence of periodontitis and tooth loss, and RCs had
a lower recurrence of periodontitis and tooth loss.
These findings reflect the need for an individual risk
classification in an attempt to optimize adherence
and customize recall intervals in PMT programs.
The use of a PRA model can be an important tool in
monitoring individual risk variables in relation to the
recurrence of periodontitis and tooth loss.
ACKNOWLEDGMENTS
This study was supported by grants from the Research
Support Foundation of Minas Gerais, Belo Horizonte,
Brazil (project #10137) and the National Research
Council, Brası´lia, Brazil (project #471616/2007-9).
The authors thank Dr. Leonardo Costa Lima and
dental hygienists Milene Aparecida Mendes Rocha,
and Wilciane Alessandra Machado, private practice,
Belo Horizonte, Brazil, for their assistance during the
monitoring period of the study. The authors report
no conflicts of interest related to this study.
11.
12.
13.
14.
15.
16.
17.
REFERENCES
1. Mendoza AR, Newcomb GM, Nixon KC. Compliance
with supportive periodontal therapy. J Periodontol
1991;62:731-736.
2. Wilson TG Jr., Hale S, Temple R. The results of efforts
to improve compliance with supportive periodontal
treatment in a private practice. J Periodontol 1993;64:
311-314.
3. Demetriou N, Tsami-Pandi A, Parashis A. Compliance
with supportive periodontal treatment in private periodontal practice. A 14-year retrospective study.
J Periodontol 1995;66:145-149.
4. Soolari A, Rokn AR. Adherence to periodontal maintenance in Tehran, Iran. A 7-year retrospective study.
Quintessence Int 2003;34:215-219.
5. Lorentz TC, Cota LO, Cortelli JR, Vargas AM, Costa FO.
Prospective study of complier individuals under periodontal maintenance therapy: Analysis of clinical periodontal parameters, risk predictors and the progression
of periodontitis. J Clin Periodontol 2009;36:58-67.
6. Leininger M, Tenenbaum H, Davideau JL. Modified
periodontal risk assessment score: Long-term predictive value of treatment outcomes. A retrospective
study. J Clin Periodontol 2010;37:427-435.
7. Costa FO, Cota LO, Lages EJ, et al. Progression of
periodontitis in a sample of regular and irregular
compliers under maintenance therapy: A 3-year follow-up study. J Periodontol 2011; 82:1279-1287.
¨ m MK, Socransky SS,
8. Rosling B, Serino G, Hellstro
Lindhe J. Longitudinal periodontal tissue alterations
during supportive therapy. Findings from subjects with
normal and high susceptibility to periodontal disease.
J Clin Periodontol 2001;28:241-249.
9. Famili P, Short E. Compliance with periodontal maintenance at the University of Pittsburgh: Retrospective
analysis of 315 cases. Gen Dent 2010;58:e42-e47.
10. Matuliene G, Studer R, Lang NP, et al. Significance of
Periodontal Risk Assessment in the recurrence of peri-
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
odontitis and tooth loss. J Clin Periodontol 2010;37:
191-199.
Tonetti MS, Muller-Campanile V, Lang NP. Changes in
the prevalence of residual pockets and tooth loss in
treated periodontal patients during a supportive maintenance care program. J Clin Periodontol 1998;25:
1008-1016.
Lang NP, Tonetti MS. Periodontal risk assessment
(PRA) for patients in supportive periodontal therapy
(SPT). Oral Health Prev Dent 2003;1:7-16.
Trombelli L, Farina R, Ferrari S, Pasetti P, Calura G.
Comparison between two methods for periodontal risk
assessment. Minerva Stomatol 2009;58:277-287.
Persson GR, Matuliene´ G, Ramseier CA, Persson RE,
Tonetti MS, Lang NP. Influence of interleukin-1 gene
polymorphism on the outcome of supportive periodontal therapy explored by a multi-factorial periodontal
risk assessment model (PRA). Oral Health Prev Dent
2003;1:17-27.
Eickholz P, Kaltschmitt J, Berbig J, Reitmeir P, Pretzl
B. Tooth loss after active periodontal therapy. 1:
Patient-related factors for risk, prognosis, and quality
of outcome. J Clin Periodontol 2008;35:165-174.
Jansson H, Norderyd O. Evaluation of a periodontal
risk assessment model in subjects with severe periodontitis. A 5-year retrospective study. Swed Dent
J 2008;32:1-7.
American Academy of Periodontology. Parameters on
comprehensive periodontal examination. J Periodontol 2000;71:847-883.
Papantonopoulos GH. Effect of periodontal therapy in
smokers and non-smokers with advanced periodontal
disease: Results after maintenance therapy for a minimum of 5 years. J Periodontol 2004;75:839-843.
Demirel K, Efeodlu A. Retrospective evaluation of
patient compliance with supportive periodontal treatment. J Nihon Univ Sch Dent 1995;37:131-137.
American Diabetes Association. Diagnosis and classification of diabetes mellitus. Position statement. Diabetes Care 2005;28(Suppl.1):S37-S42.
Tonetti MS, Claffey N; European Workshop in Periodontology group C. Advances in the progression of
periodontitis and proposal of definitions of a periodontitis case and disease progression for use in risk
factor research. Group C consensus report of the 5th
European Workshop in Periodontology. J Clin Periodontol 2005;32(Suppl. 6):210-213.
¨ nig J, Dzierzon U, Sawaf H, Plagmann
Kocher T, Ko
H-C. Disease progression in periodontally treated and
untreated patients – A retrospective study. J Clin
Periodontol 2000;27:866-872.
¨ nig J, Plagmann HC, Langenfeld N, Kocher T.
Ko
Retrospective comparison of clinical variables between compliant and non-compliant patients. J Clin
Periodontol 2001;28:227-232.
Fardal O, Johannessen AC, Linden GJ. Compliance in
a Norwegian periodontal practice. Oral Health Prev
Dent 2003;1:93-98.
Fardal O. Interviews and assessments of returning
non-compliant periodontal maintenance patients.
J Clin Periodontol 2006;33:216-220.
Novaes AB Jr., Novaes AB, Bustamanti A, Villavicencio
JJ, Muller E, Pulido J. Supportive periodontal therapy
in South America. A retrospective multi-practice study
on compliance. J Periodontol 1999;70:301-306.
Cohen RE; Research, Science and Therapy Committee, American Academy of Periodontology. Position
299
Periodontal Risk Assessment in Maintenance Therapy
28.
29.
30.
31.
32.
33.
34.
300
paper: Periodontal maintenance. J Periodontol 2003;
74:1395-1401.
Costa FO, Cota LO, Lages EJ, et al. Oral impact on
daily performance, personality traits and compliance
in periodontal maintenance therapy. J Periodontol
2011 Feb 2. [Epub ahead of print].
Checchi L, Montevecchi M, Gatto MR, Trombelli L.
Retrospective study of tooth loss in 92 treated
periodontal patients. J Clin Periodontol 2002;29:
651-656.
Chambrone LA, Chambrone L. Tooth loss in wellmaintained patients with chronic periodontitis during
long-term supportive therapy in Brazil. J Clin Periodontol 2006;33:759-764.
Miyamoto T, Kumagai T, Jones JA, Van Dyke TE,
Nunn ME. Compliance as a prognostic indicator:
Retrospective study of 505 patients treated and
maintained for 15 years. J Periodontol 2006;77:223232.
Chambrone L, Chambrone D, Lima LA, Chambrone
LA. Predictors of tooth loss during long-term periodontal maintenance: A systematic review of observational studies. J Clin Periodontol 2010;37:675-684.
Claffey N, Nylund K, Kiger R, Garrett S, Egelberg J.
Diagnostic predictability of scores of plaque, bleeding,
suppuration and probing depth for probing attachment
loss. 3 1/2 years of observation following initial
periodontal therapy. J Clin Periodontol 1990;17:108114.
Oliveira Costa F, Cota LO, Costa JE, Pordeus IA.
Periodontal disease progression among young subjects with no preventive dental care: A 52-month
follow-up study. J Periodontol 2007;78:198-203.
Volume 83 • Number 3
35. Heitz-Mayfield LJA. Disease progression: Identification
of high-risk groups and individuals for periodontitis.
J Clin Periodontol 2005;32(Suppl. 6):196-209.
36. Fisher S, Kells L, Picard J-P, et al. Progression of
periodontal disease in a maintenance population of
smokers and non-smokers: A 3-year longitudinal
study. J Periodontol 2008;79:461-468.
37. Matuliene G, Pjetursson BE, Salvi GE, et al. Influence
of residual pockets on progression of periodontitis and
tooth loss: Results after 11 years of maintenance.
J Clin Periodontol 2008;35:685-695.
38. Badersten A, Nilve´us R, Egelberg J. Scores of plaque,
bleeding, suppuration and probing depth to predict
probing attachment loss. 5 years of observation
following nonsurgical periodontal therapy. J Clin
Periodontol 1990;17:102-107.
¨ stalo PV, Knuuttila ML. Confounding and effect
39. Ylo
modification: Possible explanation for variation in the
results on the association between oral and systemic
diseases. J Clin Periodontol 2006;33:104-108.
40. Bandyopadhyay D, Marlow NM, Fernandes JK, Leite
RS. Periodontal disease progression and glycaemic
control among Gullah African Americans with type2 diabetes. J Clin Periodontol 2010;37:501-509.
Correspondence: Dr. Fernando Oliveira Costa, Department of Periodontology, Federal University of Minas
ˆ nio Carlos Avenue, 6627, Pampulha, P.O.
Gerais, Anto
Box 359, Belo Horizonte, MG 31270-901, Brazil. Fax: 5531-3282-6787; e-mail: focperio@uol.com.br.
Submitted March 23, 2011; accepted for publication June
1, 2011.