This systematic review assessed the effectiveness of ceramic inlays compared with other forms of posterior dental restorations, and found no significant differences between them. However, this finding was based on only three small studies and, although the authors detailed the review methods used, it was unclear why these studies were included. It is possible that this was a biased sample, thus the results of the review may be unreliable.

To assess the effectiveness of ceramic inlays in comparison with other forms of posterior restorations.

MEDLINE, EMBASE and the Cochrane Controlled Trials Register were searched from 1990 to 2001; the search terms were listed. The following key journals were identified by experts and handsearched from 1990 to 2001: Acta Odontologica Scandinavica, British Dental Journal, International Dental Journal, Journal of the American Dental Association, Journal of Dentistry, Journal of Dental Research, Japanese Journal of Conservative Dentistry, Journal of Prosthetic Dentistry, Operative Dentistry, and Quintessence International. Reference lists of relevant studies, review articles and authors' personal reprint collections were screened.

Study designs of evaluations included in the review

Randomised controlled trials (RCTs), controlled trials and case series were eligible for inclusion in the review.

Specific interventions included in the review

Studies of class I and class II ceramic inlay restorations in premolar and permanent molar teeth were eligible for inclusion. All types of ceramic inlays were included: fired, milled-CAD/CAM (computer-assisted design/computer-assisted machining) and copy milled, cast and heat-pressed. Studies of laminate veneers, ceramic inlays and ceramic crowns were excluded. The control groups in the included studies were other forms of posterior restorations.

Participants included in the review

No details of the inclusion criteria for the participants were reported. There were also no details of the actual participants in the included studies.

Outcomes assessed in the review

The outcomes of interest appeared to be failure rate, post-operative pain or discomfort, and aesthetic qualities. Failure was defined as an indication for replacement, or an occurrence of endodontic problems. Restorations evaluated as clinically unacceptable using U.S. Public Health Service or California Dental Association criteria were also considered failures. Post-operative pain or discomfort was defined as the presence or absence of sensitivity to temperature or occlusal loading within 1 month of restoration.

How were decisions on the relevance of primary studies made?

Two reviewers independently conducted the validity assessment after a calibration stage using a sample paper. Any disagreements were identified and discussed, and a final score for each paper was decided by consensus. Inter- and intra-rater reliability were assessed.

Validity was assessed using a checklist of 24 items: clear description of objective, setting, population, inclusion and exclusion criteria, treatments, outcomes; report and justification of sample size; concurrent control group; random allocation; blind, objective, validated outcome measurement; statistical methods described; follow-up rate stated, follow-up less than 80%, drop-outs described; findings described; results given in actual numbers; confidence intervals given; adverse events reported; conclusions given. Two reviewers independently conducted the validity assessment after a calibration stage using a sample paper. Any disagreements were identified and discussed, and a final score for each paper was decided by consensus. Inter- and intra-rater reliability were assessed.

Two reviewers independently extracted the data using a specially designed data extraction forms, after an initial piloting phase. Data from figures and graphs were extracted where possible, but were only included if both reviewers independently extracted the same result. Any disagreements were discussed and a third reviewer was involved if necessary. Where papers or abstracts reported the same trial at different stages, only the longer term study was included in the review. Authors of selected studies were contacted for details of the study design and missing data.

How were the studies combined?

The studies were combined in a narrative. The results of RCTs were presented separately from the results of controlled clinical trials and case series.

How were differences between studies investigated?

Differences between the studies were not investigated.

The authors stated that 46 studies were included (the number of participants was unclear as some studies reported the number of teeth). Five were RCTs, 15 were controlled trials and 26 were case series. However, only 3 studies were included in the 'Results' section: one RCT and 2 controlled trials.

Clinical effectiveness of ceramic inlays.

Three studies (one RCT with 80 teeth in 20 patients and 2 controlled trials with 92 teeth) were included out of a possible eight. All 3 studies showed no significant differences in failure rates between ceramic inlays and other restorations.

Post-operative sensitivity.

The same 3 studies were included. One controlled trial found no restoration with post-operative sensitivity in either treatment group, while the other found no significant difference in the incidence of post-operative sensitivity between the groups (odds ratio 0.81, 95% confidence interval: 0.25, 2.68). In the RCT, no ceramic inlays were associated with post-operative sensitivity and one gold inlay was found to give symptoms 1 week after placement.

The aesthetic qualities of ceramic inlays could not be considered since all three of the included studies used metallic materials as a comparison group.

No significant differences existed in longevity or post-operative sensitivity between ceramic and other posterior restorations over assessment periods of up to 1 year. Greater attention to the design and reporting of studies is required if the quality of trials of ceramic inlays is to be improved.

In this review, inclusion criteria relating to the interventions were clearly stated but those relating to the participants, outcomes and study designs were not. Details of the review process (e.g. how many reviewers selected the studies) were reported, and these seemed adequate for minimising bias at this stage. However, few details of the included studies were reported. It was even unclear how many studies were included in the review: the authors stated that 46 studies were included, but that only three were of sufficient quality to be included in the evaluation of clinical effectiveness; it was unclear in what way the other 43 studies were included. The validity assessment criteria were reported in great detail, but the authors did not report where the cut-off was placed in terms of how many criteria had to be met before a study was included in the evaluation of clinical effectiveness. Therefore, it was not clear how this decision was made. Without knowing this it is not possible to judge whether the authors' conclusions are based on a biased set of studies.

The authors also made the erroneous statement that a meta-analysis was not possible as no studies showed significant differences between the groups. The lack of a statistically significant difference might have indicated that the sample size was too small to detect a difference, but this would not prevent a meta-analysis; the authors did not discuss this. The authors' conclusions may be unreliable.

Practice: The authors did not report any implications for practice.

Research: The authors stated that large practice-based studies with sophisticated design, and which consider cost-benefits and tooth preservation for life, are required in addition to clinical trials that follow the CONSORT (Consolidated Standards of Reporting Trials) requirements.

Japan Society for the Promotion of Science, grant number 14571812.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.