Abstract
Objectives
The aim of this retrospective practice-based study was to investigate the longevity of direct restorations placed by a group of general dental practitioners (GDPs) and to explore the effect of practice/operator, patient, and tooth/restoration related factors on restoration survival.
Methods
Electronic Patient Files of 24 general dental practices were used for collecting the data for this study. From the patient files, longevity of 359,548 composite, amalgam, glass-ionomer and compomer placed in 75,556 patients by 67 GDPs between 1996 and 2011 were analyzed. Survival was calculated from Kaplan-Meier statistics.
Results
A wide variation in annual failure rate (AFR) exists between the different dental practices varying between 2.3% and 7.9%. Restorations in elderly people (65 years and older, AFR 6.9%) showed a shorter survival compared to restorations placed in patients younger than 65 years old (AFR 4.2%–5.0%). Restorations in molar teeth, multi-surface restorations and restorations placed in endodontically treated teeth seemed to be more at risk for re-intervention.
Conclusion
The investigated group of GDPs place restorations with a satisfactory longevity (mean AFR 4.6% over 10 years), although substantial differences in outcome between practitioners exist. Several potential risk factors on practice/operator, patient, and tooth/restoration level have been identified and require further multivariate investigation.
1
Introduction
Placing and replacing of restorations is the main work of most general dental practitioners (GDPs). The longevity of the restorations can be seen as an indicator for the quality of care delivered. Factors that have been identified as affecting the restoration performance are the filling material and their properties as well as the dental piece itself and the patient (e.g., socio-economic status caries risk) and dentist characteristics (experience) . The results of these reviews are however rather inconclusive. Some studies found a better performance of amalgam restorations compared to other restorative materials , while others showed a comparable survival of composite and amalgam restorations . An increased number of restoration surfaces was shown to result in a higher re-intervention rate , and molar teeth and endodontically treated teeth have been reported to have a higher risk for early re-intervention . Socioeconomic status of the patient has been shown to affect the longevity of restorations , probably because the prevalence of dental caries is associated with social determinants . Also the influence of caries risk of patients on restoration longevity has been demonstrated . With respect to age and gender, some studies reported that restorations in older patients and male patients have a lower survival , while other studies failed to demonstrate this effect . A paper from the UK, based on an insurance dataset showed that operator and practice related factors, notably changing dentists, influenced the longevity of restorations . Another study, comparing different types of indirect restorations demonstrated a clear operator effect on survival . However, the influence of the dentist on the results is not always obvious , as was also shown in the review of Beck et al. Overall, this is the level least investigated. This is not surprising, as most scientific research is not carried out in general dental practice, and if it is, it is not common that many operators are included and taken into account as a factor. The number of longitudinal studies on longevity of restorations placed by GDPs is limited to studies related to isolated dental practices and public health care in Scandinavian adolescents . On a larger scale, several longitudinal analyses have been made based on data from the NHS insurance database in the UK , but larger databases from GDPs have not been analyzed yet. Therefore there is need for a longitudinal practice based study, with at least a 5 year follow up time, and a multi factorial approach. In the Netherlands, dentists generally have a large group of listed patients who are loyal to the practice and show up regularly for checkups over a longer period of time. Moreover, most practices have electronic patient files. This offers the unique opportunity to investigate the longevity of restorations placed by a large group of dentists.
The aim of this retrospective practice-based study is to investigate the longevity of direct restorations placed by a large group of GDPs and to explore the effect of practice/operator, patient, and tooth/restoration related factors on restoration survival.
2
Materials and methods
2.1
Inclusion and data collection
General practices were recruited from the Nijmegen dental practice based research network. Within these practices, all individual dentists were included that contributed with a minimum of 300 restorations. Within these practices, all patients were included that visited the practices for regular checkups. Data from all direct restorations placed in permanent teeth in the years 1996 to 2011 were collected from the Electronic Medical Files (EMF) of the patients. Restorations with missing variables and uncertainties were excluded from the dataset. Design and protocol were approved by the local ethics committee, METC (CMO file nr. 2013/483). Data were digitally extracted, rendered anonymous and sent to the research group by the dentists using an application designed by the two involved software firms that provided the EMF software (Exquise ® , Kwadijk, NL, starting 1999; Complan ® , Heerhugowaard, NL, starting 1996). The application transformed all data on the placed direct restorations into excel data files.
2.2
Outcome parameters
From all direct restorations, dates of restoration placement, last check-up visit of the patient and dates of re-intervention were recorded. The restoration was considered as failed if a restoration was replaced or repaired, the tooth was extracted, or in case of an endodontic or prosthetic treatment. Replacement or repair was defined as an intervention when a new restoration was placed in the same tooth and one or more surfaces already involved in the previous restoration. An exception was made for mesial-occlusal (MO) and distal-occlusal (DO) class II restorations in molars and premolars. When a MO restoration was placed as the first restoration and the intervention treatment was a DO restoration, analysis for the initial restoration was censored, because many MO and DO restorations in posterior teeth are two independent (box type) restorations, and it would not be appropriate to qualify them all as failed. In anterior teeth, the same exception was made for DB/MB and DP/MP class III restorations. When no intervention was performed on the teeth during the evaluation period, and the tooth was still in function at the last check-up visit, the restoration was considered as successful and censored at that date.
2.3
Independent variables
On the practice level the following variables were recorded: urban (towns with >40.00 inhabitants) or rural location, practice type (solo, small (2 or 3 dentists) or larger group (>3 dentists)), practice size (small; placing <1000 restorations per year, larger; placing >1000 restorations per year) deprived working area (based on practice ZIP-codes and a standard conversion table provided by the Dutch Ministry of Health, Welfare and Sport), and experience expressed by the year of graduation of the GPDs (graduated before 1981; graduated between 1981 and 1990 and graduated in 1991 or later).
On a patient level, gender, age and the presence of a removable denture were recorded. Regarding age, patients were divided into 5 groups; 5–15 years (children), 16–25 years (adolescents), 26–45 years (young adults), 46–65 years (adults) and 66–95 years (elderly). Removable denture presence was grouped into three categories: no denture present, partial denture present, and full denture present in opposing jaw.
On the tooth/restoration level tooth number (FDI system), number of included restored surfaces (1, 2, 3, ≥4), applied restorative materials (amalgam, composite, glass-ionomer and compomer) and whether a tooth was endodontically treated (yes/no), was recorded. Subgroups were made by quadrants, jaw (upper/lower), tooth group (anterior, premolar, molar) and tooth number in the arch (1 to 8).
2.4
Statistical analysis
Statistical analyses were performed with SPSS 20. To explore the effect of variables on longevity, Kaplan-Meier analyses were used to create survival tables and curves. Out of the survival tables, mean Annual Failure Rate (AFR) over 10 years was calculated according to the formula: <SPAN role=presentation tabIndex=0 id=MathJax-Element-1-Frame class=MathJax style="POSITION: relative" data-mathml='1−1−x10′>1−1−x−−−−−√101−1−x10
1 − 1 − x 10
, in which ‘ y ’ expresses the mean AFR, and ‘ x ’ the total re-interventions. As most patients in the study contributed with multiple restorations, the method described by Chuang et al. , to produce statistically valid standard errors for the estimates of survival, was performed.
The composition of the dataset dealing with different types of restorations, notably anterior and posterior restorations, rendered statistical testing of perceived effects of independent variables, using univariate log-rank tests or multivariate analysis like a cox-regression, unsuitable. Such analyses will be performed at a later stage on more homogeneous subgroups and reported separately.
2
Materials and methods
2.1
Inclusion and data collection
General practices were recruited from the Nijmegen dental practice based research network. Within these practices, all individual dentists were included that contributed with a minimum of 300 restorations. Within these practices, all patients were included that visited the practices for regular checkups. Data from all direct restorations placed in permanent teeth in the years 1996 to 2011 were collected from the Electronic Medical Files (EMF) of the patients. Restorations with missing variables and uncertainties were excluded from the dataset. Design and protocol were approved by the local ethics committee, METC (CMO file nr. 2013/483). Data were digitally extracted, rendered anonymous and sent to the research group by the dentists using an application designed by the two involved software firms that provided the EMF software (Exquise ® , Kwadijk, NL, starting 1999; Complan ® , Heerhugowaard, NL, starting 1996). The application transformed all data on the placed direct restorations into excel data files.
2.2
Outcome parameters
From all direct restorations, dates of restoration placement, last check-up visit of the patient and dates of re-intervention were recorded. The restoration was considered as failed if a restoration was replaced or repaired, the tooth was extracted, or in case of an endodontic or prosthetic treatment. Replacement or repair was defined as an intervention when a new restoration was placed in the same tooth and one or more surfaces already involved in the previous restoration. An exception was made for mesial-occlusal (MO) and distal-occlusal (DO) class II restorations in molars and premolars. When a MO restoration was placed as the first restoration and the intervention treatment was a DO restoration, analysis for the initial restoration was censored, because many MO and DO restorations in posterior teeth are two independent (box type) restorations, and it would not be appropriate to qualify them all as failed. In anterior teeth, the same exception was made for DB/MB and DP/MP class III restorations. When no intervention was performed on the teeth during the evaluation period, and the tooth was still in function at the last check-up visit, the restoration was considered as successful and censored at that date.
2.3
Independent variables
On the practice level the following variables were recorded: urban (towns with >40.00 inhabitants) or rural location, practice type (solo, small (2 or 3 dentists) or larger group (>3 dentists)), practice size (small; placing <1000 restorations per year, larger; placing >1000 restorations per year) deprived working area (based on practice ZIP-codes and a standard conversion table provided by the Dutch Ministry of Health, Welfare and Sport), and experience expressed by the year of graduation of the GPDs (graduated before 1981; graduated between 1981 and 1990 and graduated in 1991 or later).
On a patient level, gender, age and the presence of a removable denture were recorded. Regarding age, patients were divided into 5 groups; 5–15 years (children), 16–25 years (adolescents), 26–45 years (young adults), 46–65 years (adults) and 66–95 years (elderly). Removable denture presence was grouped into three categories: no denture present, partial denture present, and full denture present in opposing jaw.
On the tooth/restoration level tooth number (FDI system), number of included restored surfaces (1, 2, 3, ≥4), applied restorative materials (amalgam, composite, glass-ionomer and compomer) and whether a tooth was endodontically treated (yes/no), was recorded. Subgroups were made by quadrants, jaw (upper/lower), tooth group (anterior, premolar, molar) and tooth number in the arch (1 to 8).
2.4
Statistical analysis
Statistical analyses were performed with SPSS 20. To explore the effect of variables on longevity, Kaplan-Meier analyses were used to create survival tables and curves. Out of the survival tables, mean Annual Failure Rate (AFR) over 10 years was calculated according to the formula: <SPAN role=presentation tabIndex=0 id=MathJax-Element-2-Frame class=MathJax style="POSITION: relative" data-mathml='1−1−x10′>1−1−x−−−−−√101−1−x10
1 − 1 − x 10
, in which ‘ y ’ expresses the mean AFR, and ‘ x ’ the total re-interventions. As most patients in the study contributed with multiple restorations, the method described by Chuang et al. , to produce statistically valid standard errors for the estimates of survival, was performed.
The composition of the dataset dealing with different types of restorations, notably anterior and posterior restorations, rendered statistical testing of perceived effects of independent variables, using univariate log-rank tests or multivariate analysis like a cox-regression, unsuitable. Such analyses will be performed at a later stage on more homogeneous subgroups and reported separately.
Stay updated, free dental videos. Join our Telegram channel
VIDEdental - Online dental courses
