Tooth loss is still an unfortunate reality for many patients, especially for the elderly. ¹ Edentulism has consequences such as reduction of the lower third of the face, decrease of vertical dimension, loss or reduction of masticatory movement, poor esthetics and phonetic problems. ² Dietary restrictions and difficulty to eat certain foods are also mentioned by edentulous patients. ³ , ⁴ , ⁵ Typically, preference is given to foods that are easier to crush, which can compromise the nutritional needs of the individual, and thus affect general health. ⁶ , ⁷ Those alterations can impact oral health-related quality of life (OHRQoL) and compromise the psychosocial behavior of the individual. ⁸

Osseointegrated implants have been used as a treatment for dental loss with high success rates. However, this treatment modality is not available for all patients due to general health, cost, and/or anatomical problems. ⁹ In spite of removable complete dentures (CDs) being a viable treatment option for the edentulous, they require an adequate bone ridge height to allow the retention and stability, thus efficiently recovering masticatory function. ³

It is possible to notice a positive change in the behavior of these individuals after CDs oral rehabilitation with fully adapted, comfortable and aesthetic removable CDs. Patients regain self-esteem and general well-being, fit satisfactorily back into social esthetic standards and recover lost nutritional capacity. ¹⁰ , ¹¹

Thus, the objective of this systematic review and meta-analysis was to evaluate whether the scientific evidence of the long-term use of CD into promotes significant changes in the OHRQoL in edentulous patients.

This systematic review was recorded on the systematic reviews database PROSPERO (CRD: 42016038907). The written was performed according to Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) guidelines ( http://www.prisma-statement.org) ¹² [Appendix 1] and checked according to a Measurement Tool to Assess Systematic Reviews 2 (AMSTAR-2) ¹³ [Appendix 2].[INLINE:1][INLINE:2]

The following focused question was outlined according to the population, intervention, comparisons and outcomes (PICO): Do CDs influence the edentulous patients' OHRQoL in long-term?

Search strategy

The process to search primary studies was done up to June 28, 2020. The following electronic databases were assessed: Pubmed, Web of Science, Scopus and Cochrane Library. The search strategy included appropriate MeSH terms, keywords, and other free terms followed the syntax rules of each database. It was used Boolean operators (OR, AND) to combine searches Table 1. The grey literature was consulted through SIGLE (System of Information on Grey Literature) ( http://www.opengrey.eu). To find additional studies, a hand search was performed on the reference lists of the retrieved studies.{Table 1}

Inclusion and exclusion criteria

The selection of studies was made by analysis of titles and abstracts that met the inclusion criteria. There was no restriction on language or year of publication. The inclusion criteria outlines articles according to the PICO and study design as follows:

Population (P): Edentulous patients (both arches)

The following the exclusion criteria were considered: (i) case reports, review articles, book chapters; (ii) studies in patients with medical conditions such as systemic diseases, syndromes and craniofacial anomalies, or who have special needs or were hospitalized; (iii) studies that used nonvalidated questionnaires; (iv) absence of a baseline evaluation or a baseline was not used to compare with the follow up; (v) absence of follow up; (vii) without results per groups; (vii) studies out theme proposed records.

Study selection

Two independent reviewers analyzed all articles (LAAA and LSG). To assess the agreement between authors, 10% of the publications were random selected in this literature research, and their classification was compared. Kappa statistic was employed and demonstrated good inter-examiner agreement (K = 0.90). Duplicate studies were excluded. If the title and abstract were not clear, the article was read in full. If doubt remained, authors were contacted. If disagreements occurred, a third author (LSA) was called, aiming for a consensus.

Data extraction (qualitative data)

Two independent reviewers (LSG and AMCM) extracted relevant data presented in the articles. To characterize and demonstrate the methodological design, we presented the following in detail: Author/year of publication, country where the research was carried out, age of subjects, sample size, social dental index (questionnaire) used to assess the OHRQoL, form of application, type of study, groups compared and the time of follow up.

Another data extracted from the elected articles was average impact for the total scale and subscales before (baseline) and after the CD installation and its association with OHRQoL.

Evaluation criteria of study risk of bias

Methodological quality and risk of bias control were evaluated in accordance to the guidelines “Quality Assessment Tool for Before-After (Pre-Post) Studies with No Control Group” described by the National Institutes of Health. ¹⁴ This quality assessment tool allows classifying before-after studies with no control group and provides a standardized approach for evaluating the quality. The tools were designed to assist reviewers in focusing on concepts that are key for critical appraisal of the internal validity of a study.

Two reviewers (LSG and LAAA) independently assessed the quality of the included studies, which quality reviewers could select “yes,” “no,” or “cannot determine/not reported/not applicable” in response to each item on the tool. For each item in which “no” was selected, reviewers were instructed to consider the potential risk of bias that could be introduced by that flaw in the study design or implementation. “Cannot determine” and “not reported” were also noted as representing potential flaws. In general terms, a “good” study has the least risk of bias, and results are considered to be valid. A “fair” study is susceptible to some bias deemed not sufficient to invalidate its results. The fair quality category is likely to be broad, so studies with this rating will vary in their strengths and weaknesses. A “poor” rating indicates significant risk of bias. So, we established as “good” studies those that presented up to 3 answers “no”; “fair” studies that presented from 3 to 5 answers of “no”; and “poor” studies that presented more than 5 answers of “no.”

Meta-analysis (quantitative data)

For the meta-analysis, we pooled and extracted the mean and the standard deviation (continuous data) from the included studies. Subgroups were established prior to the overall analysis of the outcome, according to the time of follow-up of OHRQoL questionnaire. Each study was included in the analysis only once.

RevMan 5.4 software (Cochrane Central Executive Team, St Albans House, 57-59 Haymarket, London, United Kingdom) was used to analyze the data for heterogeneity and produce a graphical display of results. For forest plots, 95% confidence intervals (CIs) and P values were calculated. Heterogeneity among the results of studies and the quantification of inconsistency were evaluated using the I ²test. ¹⁵ Values of I ²>50%, I ²25%–50% and I ²<25% were considered high, moderate and low, respectively. ¹⁵ In the Forest plot, P < 0.05 was used to test for overall effect.

Co-variables that influence in the stability of the main outcomes of meta-analisys will be treated with sensibility analysis or meta-regression. Meta-regression consists of a form of sensitivity analysis in covariable meta-analysis. In meta-regression, the number of covariates to be included is limited to the number of studies considered in the meta-analysis. Ideally, one covariate should be used for every ten studies. If the sum of included studies of an outcome exceeded 10, funnel plots can also be generated to analyze the publication bias test. ¹⁶

Grading of recommendations assessment, development and evaluation

Two reviewers (LSG and LAAA) independently analyzed the quality of the evidence (certainty in the estimates of effect) using the grading of recommendations assessment, development and evaluation (GRADE) approach. The domains evaluated in clinical studies are risk of bias, inconsistency, indirectness, imprecision and publication bias. The GRADE defines the quality of scientific evidence more clearly and objectively and can be classified as high, moderate, low or very low. ¹⁷

Flowchart recommended by PRISMA guidelines ¹¹ Figure 1describes the number of articles identified in each step of the study. A total of 2452 articles were found, of which 928 were duplicate articles and were removed. Of the 1524 remaining articles, 1460 were excluded after the application of eligibility criteria. Sixty-four articles were accessed in full and of these, 24 were elected for evaluation of methodological. No studies were found through the manual search in the references of the articles. Figure 1

Flowchart for the search process in articles and selection using the Preferred Reference Items for Systematic Reviews (PRISMA). ¹¹

Tooth loss is a major problem for people worldwide because tooth replacement does not always meet the basic needs of these patients. The consequences of edentulism can impact OHRQoL ⁴¹ and to compromise social life. ⁶ , ⁷ Also, there is a preference for soft foods, which compromises the overall health of these patients through inadequate ingestion of nutrients.

Even though osseous implants present a great success rate, many patients are not able to be subjected to this type of treatment for many reasons. ⁹ Thus, CDs are a viable option of treatment for these cases. These prostheses recover the main functions of the stomatognathic system, ¹⁰ but it is necessary to present good retention and stability. ² Evidence-based dentistry is important to provide a basis of solid evidence for all professionals who are committed to offering the best treatment option for their patients.

In this systematic review, the articles selected used diferent instruments to detect if new CDs were able to improve patients' OHRQoL. On qualitative analysis, excepting two articles, ³¹ , ³⁵ 22 papers ⁸ , ¹⁸ , ¹⁹ , ²⁰ , ²¹ , ²² , ²³ , ²⁴ , ²⁵ , ²⁶ , ²⁷ , ²⁸ , ²⁹ , ³⁰ , ³² , ³³ , ³⁴ , ³⁶ , ³⁷ , ³⁸ , ³⁹ , ⁴⁰ concluded that the use CD improved the OHRQoL. CDs have been studied for many years, so a significant number of articles involving total prostheses and quality of life were found. A previous systematic ¹¹ review selected 6 articles to evaluate whether treatment with new CDs improves OHRQoL in elderly patients. The present systematic review selected 24 articles. So, based on increased number of publications on this important clinical evaluation, an update a systematic review needs to be done. ⁴² This fact makes us realize the importance that this therapeutic option still presents in the dentistry scenario.

The addition of new synthesis methods, such as GRADE, improved the quality of the analysis and the clarity of the findings to answer the question if the new CD improves de OHRQoL. Added to it, this research was carried out in the most important databases, in the grey literature and manually in the bibliographic references of the selected articles. We also used common MeSH terms and keywords from articles published in the same field in order to minimize the possibility of not finding potentially eligible studies. Thus, the likelihood of risk of bias from this systematic review is low as also observed by AMSTAR-2 checklist.

The meta-analysis detected that greater follow-up (6 months) improved impact on OHRQoL in the long-term. These findings emphasizes that studies with greater follow-up are necessary to obtain an improvement in the long-term impact of OHRQoL. The study with longest time of follow-up was 5 years ³⁹ followed by 1 study that follow-up for 2 years ²² and 4 studies ²³ , ²⁶ , ³² , ³⁷ that follow-up for 12 months. The methodological design from the majority of the excluded papers presented no evaluation of the baseline or presented short or unspecified follow-up periods. The lack of baseline in many studies probably occurred due to the lack of use of total prosthesis by the volunteers at the initial time of the study. Early evaluation of the use of new prostheses may compromise the outcome, due to patient's neuromuscular adaptation. ⁴³ Therefore, studies with a follow-up of <3 months were excluded.

The aim of this study was to search all available literature reporting the impact of new CD on OHRQoL. The possibility of combining patients' needs and desires with the professional's personal expertise in oral rehabilitation treatment planning should always be carried out based on the best scientific evidence available. Thus, it is important to evaluate the quality of evidence demonstrated by articles that propose to detect changes in OHRQoL after oral interventions.

Studies that met the eligibility criteria were submitted to a risk of bias analysis with a qualifier (“Quality Assessment Tool for Before-After (Pre-Post) Studies with No Control Group”). ¹⁴ The qualifier items most frequently missing in the selected articles were sample size calculation (question 3 and 5) and the evaluation of the instruments' psychometric properties (question 7).

Sample size calculation in clinical trials is of great importance to ensure that the number of participants is large enough to have a high probability of detecting true and clinically significant differences between groups or treatments. In this systematic review, ten studies ⁸ , ²⁰ , ²¹ , ²⁴ , ²⁶ , ³⁰ , ³¹ , ³³ , ³⁷ , ⁴⁰ performed the sample size calculation, which indicates the need for greater care in future, research in relation to this important question.

In addition to the methodological problems found in articles, some studies presented their results in a nonreproducible way considering the evaluation of psychometric properties of instruments for evaluation of OHRQoL. Psychometric properties are essential requirements for measuring instruments. The main psychometric properties of a measuring instrument are validity, reliability and in the studies analyzing before and after a treatment, the responsiveness. Seven studies realized some of these evaluations. ¹⁸ , ²¹ , ²⁶ , ²⁸ , ²⁹ , ³² , ³⁸ Validity of an instrument can be defined as its ability to actually measure what it proposes to measure. The validity as mentioned above was guaranteed in all selected studies since all of them used validated instruments, including validation for the languages of their respective populations. Reliability is the first characteristic that an instrument must present. This refers to the degree to which the repeated application of an instrument on the same subject produces equal results, that is, indicates the reproducibility of a measure. Reliability should be contextualized in terms of stability and internal consistency.

Only four studies ²¹ , ²⁶ , ²⁸ , ²⁹ realized test-retest of the OHRQoL instruments applied in their population. In this procedure, the same measuring instrument is applied at two times to the same group of people after a period of time to confirm the reliability of the instrument.

Internal consistency is perhaps the most widely used approach. It is understood as the degree of homogeneity in which the items designated to measure the same concept are interrelated. The most commonly used measurement for internal consistency is Cronbach's Alpha Coefficient. Three articles performed this measurement and the results were satisfactory. ¹⁸ , ²¹ , ²⁸

Responsiveness is the psychometric property that assesses the instrument's ability to detect changes and is used primarily in clinical work to test QoL changes during or after treatment. Responsiveness is an important characteristic of OHRQoL instruments, which are used as evaluative measures to assess the change pre-and post-treatment. This property is not well established in many studies that have measured OHRQoL, which is a significant omission given the increasing tendency to use OHRQoL measurements as outcomes in clinical trials and evaluation studies. ⁴⁴ The absence of evaluation of this property is a worrying fact. In the present systematic review, only four articles ²⁸ , ²⁹ , ³² , ³⁸ applied this measurement. This fact corroborates with Antunes et al. ⁴⁴ in their systematic review evaluating changes in the OHRQoL of children and adolescents under 14 years old after oral health interventions, a moderate level of evidence was observed. One factor responsible for this level of evidence was that there were no evaluations of psychometric properties such as responsiveness.

To perform the meta-analysis was a challenge in this study. Results expressed as graphs and frequency, absence of information examiners calibration, made the comparison of the data of some articles impossible. It is important to emphasize that we tried to contact the authors, but we did not receive an answer. The difficulty to perform the meta-analysis was also especially high for the included studies that did not use the same quality-of-life assessment instrument. So, we chose to analyze in subgroups when it is possible to compile results from the same instrument at different follow-up times, as commonly performed in quality of life systematic review studies. ⁴⁵ Despite these difficulties, the meta-analysis compiled the results of 12 included studies related to the OHIP-14, ³⁶ , ³⁸ OHIP-20, ³⁰ , ³¹ , ³² OHIP-EDENT ²¹ , ²⁴ , ²⁷ , ²⁸ and GOHAI ²⁵ , ²⁶ , ²⁹ instrument.

There was a diversity of instruments used in the articles included in this systematic review. However, there is a specific instrument validated for elderly patients (OHIP-EDENT), which, if standardized for this type of study, would allow a comparison between the results obtained by several studies. This study observed twelve studies (50%) using OHIP-EDENT ⁸ , ¹⁸ , ¹⁹ , ²⁰ , ²¹ , ²² , ²³ , ²⁴ , ²⁷ , ²⁸ , ³⁹ , ⁴⁰ as the questionnaire tool. We also observed an increasing tendency on use of this instrument once in the last 5 years from 9 studies, ⁸ , ²¹ , ²² , ²³ , ²⁴ , ²⁷ , ³⁸ , ³⁹ , ⁴⁰ 8 papers ⁸ , ²¹ , ²² , ²³ , ²⁴ , ²⁷ , ³⁹ , ⁴⁰ used the OHIP-EDENT. Despite, the meta-analysis confirmation of an improve on OHRQoL using different instruments, we can perceive that the lack of standardization of the instrument hinders a more objective and efficient analysis of the results.

The meta-analysis of this study to affirm a favorable outcome for the use of CD on improving OHRQoL in long-term; however very low certainty of evidence was observed in the GOHAI and OHIP-20 questionnaires analysis, and low certainty of evidence in the subgroup analysis for the GOHAI instrument and in OHIP-EDENT. It can be explained by the heterogeneity presented by some studies: Small follow-up periods, ⁸ , ¹⁸ , ¹⁹ , ³⁴ , ³⁵ , ³⁶ , ³⁸ , ⁴⁰ applied the instrument by mail, did not explain how the questionnaire was applied ²² , ²³ , ³¹ , ³⁹ , ⁴⁰ or did not use an expressive sample size ⁸ , ²² , ²³ , ²⁹ , ³⁰ , ³¹ , ³⁷ , ³⁹ for this type of therapeutic option. The results of this review suggest that the exchange of unsatisfactory CDs for new ones has the strong potential to contribute to OHRQoL. However, based on the heterogeneity, risk of bias and low certainty of the evidence that some studies presented, well-designed studies are necessary due to the importance that CD still present in the contemporary dentistry.

CD has the strong potential to contribute to oral health-related quality of life in long-term.

Financial support and sponsorship

This study was supported by grant (process E-26/010.002501/2016) from Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ).

Conflicts of interest

The authors of this manuscript declare that they have no conflicts of interest, real or perceived, financial or nonfinancial in this article.