This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Alternative EndoCrowns Designs for Chairside CAD/CAM (NCT04604314), a non-randomized na study. The primary outcome is Restoration Failure.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT04604314 |
| Phase | NA |
| Condition | Endodontically Treated Teeth |
| Allocation | Non-Randomized |
| Design model | Parallel |
| Masking | None (open label) |
| Enrollment | 69 |
| Arms in the primary outcome | n/a |
| Lead sponsor | University of Michigan |
| Results posted | Yes |
2. Clinical Question
Population
The study concerns endodontically treated teeth restored with chairside CAD/CAM endocrowns. The registry reports an enrollment of 69. It does not provide further participant or tooth characteristics in the information available for this analysis.
Intervention
The restorations were made with CEREC Tessera, an advanced lithium disilicate material. The study describes adhesive bonding with a selective enamel etch technique, a universal adhesive, and dual-cure resin adhesive cement. The preparation category called Endocrown Onlay requires intact buccal and lingual walls with specified height and thickness characteristics.
Comparator
The other preparation category is Endocrown Shoulder. It applies when buccal and/or lingual walls are less than half the original occlusal-gingival height, or when those surfaces were previously prepared axially due to a prior restoration. Allocation was nonrandomized, so these groups are not a like-for-like control comparison.
Primary question
The primary outcome is the number of crowns that fail through the 3-year recall. Failure is defined as visual loss of the crown from the tooth requiring replacement with a new crown between delivery and three years. The broader study aim is to examine how preparation design relates to restoration adaptation and longer-term clinical performance.
3. Trial Design
This is a completed, nonrandomized, parallel-group clinical trial listed under the title “Alternative EndoCrowns Designs for Chairside CAD/CAM.” The sponsor is the University of Michigan. The registry lists enrollment of 69 and the condition as endodontically treated teeth. The intervention is CEREC Tessera, and the groups are defined by endocrown preparation design rather than by different restorative materials.
In the Endocrown Onlay category, the buccal and lingual walls are intact, their occlusal-gingival height is at least half the original crown height, and remaining buccal and lingual walls maintain a thickness of at least 2.0 mm. The Endocrown Shoulder category includes teeth with buccal and/or lingual walls less than half the original height, or surfaces previously prepared axially due to a prior restoration. These definitions describe clinically different starting conditions. Because assignment was not randomized, differences between groups may reflect those underlying tooth or preparation characteristics as well as any association with design.
The study description says restorations were to be followed over five years of clinical service. The listed primary endpoint covers delivery through the 3-year recall, so it addresses a shorter interval than the overall planned follow-up. The registry lists a start date of 2020-12-10, primary completion on 2025-05-08, and status as completed. Masking is listed as none. The registry record does not report a randomization procedure, allocation concealment, or further details of recruitment and follow-up.
4. Participant Flow and Accounting
The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.
Before assignment: Tree subjects received two endo onlays, which reduced the number of subjects that received an endo onlay to 35. Three subjects received two endo crowns, which reduced the number of subjects that received an endo crown to 36. The total number of subjects is 69, not 71,because two subjects received and endo onlay and an endo crown and thus are double counted.
Baseline
| Stage | Endocrown Onlay Restoration | Endocrown Shoulder Restorations |
|---|---|---|
| STARTED | 35 | 36 |
| COMPLETED | 35 | 36 |
| NOT COMPLETED | 0 | 0 |
6 month
| Stage | Endocrown Onlay Restoration | Endocrown Shoulder Restorations |
|---|---|---|
| STARTED | 35 | 36 |
| COMPLETED | 23 | 26 |
| NOT COMPLETED | 12 | 10 |
Reasons for not completing this period:
| Reason | Endocrown Onlay Restoration | Endocrown Shoulder Restorations |
|---|---|---|
| Lost to Follow-up | 12 | 10 |
2 years
| Stage | Endocrown Onlay Restoration | Endocrown Shoulder Restorations |
|---|---|---|
| STARTED | 35 | 36 |
| COMPLETED | 21 | 25 |
| NOT COMPLETED | 14 | 11 |
Reasons for not completing this period:
| Reason | Endocrown Onlay Restoration | Endocrown Shoulder Restorations |
|---|---|---|
| Death | 1 | 0 |
| Lost to Follow-up | 13 | 11 |
3 years
| Stage | Endocrown Onlay Restoration | Endocrown Shoulder Restorations |
|---|---|---|
| STARTED | 34 | 36 |
| COMPLETED | 16 | 24 |
| NOT COMPLETED | 18 | 12 |
Reasons for not completing this period:
| Reason | Endocrown Onlay Restoration | Endocrown Shoulder Restorations |
|---|---|---|
| Lost to Follow-up | 18 | 12 |
Analysis population for the primary outcome, as defined in the registry: Number of crowns that failed through the 3-year recall.
5. Baseline Characteristics
The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.
Population: Three participants received two endo onlays, which reduced the number of participants that received an endo onlay to 35. Three participants received two endo crowns, which reduced the number of participants that received an endo crown to 36. Two participants received an endo onlay and an endo crown.
| Characteristic | Unit | Endocrown Onlay Restoration | Endocrown Shoulder Restorations | Endocrown Onlay and Endocrown Shoulder Restorations | Total |
|---|---|---|---|---|---|
| Participants analysed | 33 | 34 | 2 | 69 | |
| Age, Customized: age between 18 and 19 years | Participants | 2 | 0 | 0 | 2 |
| Age, Customized: age between 20 and 29 years | Participants | 8 | 4 | 1 | 13 |
| Age, Customized: age between 30 and 39 years | Participants | 10 | 8 | 0 | 18 |
| Age, Customized: age between 40 and 49 years | Participants | 6 | 5 | 1 | 12 |
| Age, Customized: age between 50 and 59 years | Participants | 4 | 7 | 0 | 11 |
| Age, Customized: age between 60 and 69 years | Participants | 2 | 10 | 0 | 12 |
| Age, Customized: age 70 and older | Participants | 1 | 0 | 0 | 1 |
| Sex: Female, Male: Female | Participants | 18 | 22 | 1 | 41 |
| Sex: Female, Male: Male | Participants | 15 | 12 | 1 | 28 |
| Race and Ethnicity Not Collected | Participants | — | — | — | 0 |
6. Primary Endpoint
The registry lists the primary outcome as Restoration Failure, measured over from delivery of the crown up to 3 years.
| Element | Registry entry |
|---|---|
| Result type | Number |
| Measure of spread | not applicable |
| Unit | crown |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint is restoration failure, expressed as the number of crowns that failed through the 3-year recall. The registry defines failure as visual loss of the crown from the tooth that requires replacement of the study crown with a new crown, at any time from delivery through three years. This definition focuses on a consequential event that can be identified clinically and that leads to replacement, rather than on a minor change in appearance or fit.
A failure count can communicate how many observed crowns met that definition during the specified follow-up period. For comparisons, however, counts need context: the number of crowns followed, the duration and completeness of observation, and how the analysis handles participants who leave follow-up or whose status is unknown. Those analysis details and the primary outcome results are not reported in the registry record.
The endpoint is relevant to the study’s stated interest in long-term clinical performance, but it captures only the failure definition specified by the registry. It does not by itself quantify internal or marginal adaptation, which are also named as study interests, nor does it establish why a crown failed. The registry record lists the primary time frame as delivery through three years, while the study description refers to follow-up over five years of clinical service.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The registry describes a comparison between two preparation-design categories, but it does not report the primary failure counts, a treatment-effect estimate, a confidence interval, or a statistical test for that outcome. Consequently, the posted registry information does not establish whether observed failure differed between the categories, or how precisely such a difference may have been estimated. The endpoint definition and follow-up window are available; the primary outcome findings are not reported in the registry record.
Interpretation also needs to account for the nonrandomized allocation. The registry explicitly notes that the arms are not a like-for-like control comparison. Preparation categories are defined by tooth-wall height, thickness, and prior preparation history. Such characteristics may relate to the condition of the tooth and could also relate to its subsequent clinical course. Without random allocation or a reported adjustment strategy, a difference between categories could not automatically be attributed to preparation design itself. The registry record does not report an adjusted analysis.
The registry does list adverse-event figures by group: both groups have 0 serious adverse events and 0 other adverse events recorded, while deaths are listed as 1 for Endocrown Onlay Restoration and 0 for Endocrown Shoulder Restorations. These are descriptive registry entries. The record does not give details about the death, its timing, relationship to the intervention, or how adverse events were collected, so those figures do not support a causal interpretation.
The study’s broader description refers to internal and marginal adaptation and long-term clinical performance, with five years of planned clinical service. The primary endpoint in the record concerns crown failure through the 3-year recall. The record lists secondary outcomes of restoration loss of retention and margin staining, but does not provide their results. Nor does it report results for adaptation measures in the information. The completed status therefore should not be read as evidence that the relevant outcomes were favorable or unfavorable. It indicates the registry status, while the outcome estimates and their uncertainty remain unreported in the registry record.
Any planning illustration based on an observed effect must be distinguished from the trial’s own design justification and from the observed result. The registry characterizes such illustrations as hypothetical, using stated assumptions, and not as observed power. The record does not report a sample-size justification for the trial. Without reported outcome estimates, uncertainty measures, and analysis details, readers cannot assess the magnitude or precision of a design association from the registry information alone.
9. Statistical Methods Explained
What does a failure count describe?
A count records how many crowns meet the endpoint definition during the specified follow-up. It is a descriptive measure, not by itself a comparison of risk or a statement about the size of an effect. To interpret a count, readers need to know how many crowns were observed, how long they were followed, whether follow-up was complete, and how missing outcome status was handled. The registry reports 69 at risk in each group in the adverse-event information, but it does not report the primary failure counts or the number contributing to a primary-outcome analysis.
When follow-up duration varies, a simple count may not fully represent the experience of each group. Time-to-event methods can account for when failures occur and for periods during which outcome status is known. Such methods also rely on assumptions about censoring and the event process. The registry does not report a time-to-event analysis for the primary endpoint, so it is not possible to tell whether the endpoint was analyzed as a count, as a proportion, or with a time-to-event approach.
What would an effect estimate and confidence interval add?
An effect estimate expresses the direction and magnitude of a comparison using a defined scale. A confidence interval describes the uncertainty around that estimate under the method’s assumptions. A broad interval indicates that a range of effects remains compatible with the data; a narrow interval indicates greater precision, but does not remove concerns about study design or bias. Neither an estimate nor an interval can be interpreted without knowing the comparison, outcome definition, analysis population, and statistical method. The registry record does not report an estimate or confidence interval for restoration failure.
A p-value, if reported, would address compatibility between the observed data and a specified null model. It would not measure the clinical importance of an effect, the probability that a hypothesis is true, or the likelihood that results will reproduce. The registry record lists no analyses for the outcomes in the information. It does not report a p-value for the primary endpoint, a prespecified hypothesis test, or a model used to compare the groups.
Why does the nonrandomized design matter?
Random allocation can help balance measured and unmeasured prognostic characteristics across groups, on average. In this trial allocation was nonrandomized, and the registry itself cautions that the arms are not a like-for-like control comparison. The preparation definitions identify differences in tooth-wall condition and prior preparation. These features may be related to prognosis, making a direct unadjusted comparison difficult to interpret as an effect of preparation design.
Statistical adjustment can account for measured characteristics if suitable data are collected and a defensible model is specified. Adjustment cannot ensure removal of bias from characteristics that were not measured or inadequately represented. The registry record does not report an adjusted analysis, the covariates considered, or the rationale for any model. It is therefore not possible to assess whether confounding was addressed in the outcome analysis.
How should planning calculations be read?
The registry describes sample-size assumptions for a planning illustration based on the observed effect, with a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. It says the effect to detect is a stated fraction of the observed effect. It also distinguishes illustrations for time-to-event outcomes, which give total events, from other illustrations, which give participants per arm. These assumptions describe a planning exercise, not the trial’s own design justification. The registry record does not report a sample-size justification for the enrolled study.
Such illustrations do not represent observed power. Power is a design-stage quantity tied to assumed effects, variability, allocation, and analysis choices; calculating it after observing the data does not clarify how precise the result is. For that purpose, the observed effect estimate and its confidence interval are more informative. The actual planning values and a primary outcome effect are not reported in the registry record, so no trial-specific planning conclusion can be drawn from the assumptions alone.
What does over-dispersion mean in general?
For count outcomes, some statistical models assume that the variability in counts follows a particular relationship to the average count. Over-dispersion means the observed variability is greater than that assumption allows. If it is ignored, standard errors may be understated and confidence intervals may be too narrow. Alternative models or robust uncertainty estimates can address this in appropriate settings, but the choice depends on the data structure and endpoint. This is a general statistical consideration, not a reported feature of this trial: the registry does not provide primary outcome counts or describe a count model, dispersion assessment, or uncertainty method.
10. Multiplicity, Missing Data and Analysis Populations
The registry lists restoration failure as the primary outcome and names restoration loss of retention and margin staining as secondary outcomes. It also describes internal and marginal adaptation as study interests. Considering multiple outcomes or multiple tests can increase the chance of obtaining an apparently notable result by chance if no analysis strategy accounts for that multiplicity. Possible approaches include designating a primary hypothesis, controlling error across a defined family of tests, or treating secondary findings as exploratory. The registry record does not report a multiplicity strategy, the number of tests conducted, or results for the listed secondary outcomes.
Missing outcome data matter because the observed crowns may not represent all crowns intended for follow-up. If crowns with unknown status differ systematically from those observed, a complete-case analysis can be biased. Good interpretation requires reporting the number assessed at each time, reasons for missingness, and how missing outcomes were handled. The registry record does not report these details for the primary endpoint, nor does it describe imputation, sensitivity analyses, or censoring rules.
Analysis populations also affect interpretation. A report might define participants or crowns according to allocation, treatment received, or availability of outcome data, but the appropriate population depends on the design and question. The registry does not report an analysis population or explain how any deviations from the intended preparation or restorative procedure were handled. It also does not report an analysis plan or methods for accounting for the relationship between crowns and participants. These omissions limit assessment of the reported endpoint, independently of whether a numerical result is eventually posted.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Endocrown Onlay Restoration | 0 | 69 | 0.0% |
| Endocrown Shoulder Restorations | 0 | 69 | 0.0% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Endocrown Onlay Restoration | 1 of 69 |
| Endocrown Shoulder Restorations | 0 of 69 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 0% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Endocrown Onlay Restoration | 0 of 69 |
| Endocrown Shoulder Restorations | 0 of 69 |
The registry’s adverse-event entries list 0 serious adverse events and 0 other adverse events in each preparation group. It lists deaths as 1 for Endocrown Onlay Restoration and 0 for Endocrown Shoulder Restorations. The record reports 69 at risk in each group for these entries. These values should be read as registry counts, not as proof that no events of any kind occurred or that the intervention caused or prevented an event.
The registry does not provide a description of the death, its cause or timing, its relationship to treatment, or the procedures used to collect and classify adverse events. It also does not report event-level details, follow-up completeness for safety, or a statistical comparison of safety between groups. Without those details, the death count cannot be attributed to the restoration or preparation design. Likewise, zero recorded adverse events does not establish absence of risk; it indicates only that zero events are listed in these categories in the registry record.
Safety counts require context about observation time and ascertainment. The study description refers to clinical service follow-up over five years, but the adverse-event entries alone do not indicate how much follow-up contributed for each group or whether event capture was complete. The registry record therefore supports a narrow descriptive reading of the listed figures, not a comparative or causal conclusion.
12. What the Result Does, and Does Not, Mean
The registry record identifies a clinically defined endpoint for crown failure and describes two preparation categories used for chairside CAD/CAM endocrowns. It also records a nonrandomized, parallel design and states that the arms are not a like-for-like control comparison. These details help clarify what the study set out to examine and why any comparison would need cautious interpretation.
The record does not report the primary failure count, a comparison between preparation categories, an effect estimate, or an interval describing precision. It also does not report results for the listed secondary outcomes. Accordingly, the registry information does not establish the frequency of failure through the 3-year recall, whether failure differed by preparation category, or how uncertain a comparison would be. Completed study status does not fill this results gap.
Even if a difference were reported, the nonrandomized design would limit causal interpretation. The categories are defined by tooth-wall dimensions and prior preparation history, so the groups may differ in clinically relevant ways before restoration. A statistical association would not, by itself, show that preparation design caused the difference. The registry record does not report adjusted results or enough analysis details to assess how such factors were handled.
The adverse-event entries are also limited: they list no serious or other adverse events in either group and a death in the Endocrown Onlay group, but provide no event details or causal assessment. Those entries should not be treated as a comparative safety conclusion. Overall, the registry record supplies study definitions and design information, but not the outcome estimates needed to judge the magnitude, direction, or precision of clinical results.
13. Why This Trial Matters Statistically
This trial raises a statistical question that is important whenever clinicians compare treatment approaches in groups formed without random allocation: how much of an observed difference reflects the approach, and how much reflects the characteristics that led to a particular approach being used? Here, preparation categories are defined by the remaining tooth walls and prior preparation history. That makes the design clinically meaningful, but it also makes confounding a central consideration in interpreting a comparison.
The endpoint is stated in a way that emphasizes a consequential clinical event: visual loss of the crown requiring replacement. Clear operational definitions help readers understand what counts as failure and distinguish it from other outcomes, such as loss of retention or margin staining, which the registry lists separately. The stated primary time frame also makes clear that the failure question concerns delivery through the 3-year recall, whereas the study description refers to longer clinical service follow-up.
At the same time, statistical interpretation depends on more than a well-defined question. Outcome counts, follow-up completeness, effect estimates, uncertainty intervals, and analysis methods are needed to understand the evidence. These are not reported in the registry information available for this analysis. The absence of those results means that the design question can be described, but the strength and precision of any association cannot be assessed.
The case also illustrates why status labels and planning calculations should not substitute for results. A completed status does not indicate the direction or quality of findings, and hypothetical sample-size illustrations are not observed power or evidence of an effect. Transparent reporting of methods and estimates is necessary for readers to distinguish descriptive findings from causal claims and to judge how much uncertainty remains.
14. Limitations
- The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
- Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.
The registry record does not report the primary failure count or a statistical comparison, so the central clinical outcome cannot be assessed from the available information. It also does not report outcome estimates, confidence intervals, or the analysis method. The listed secondary outcomes have no reported results in the information available for this analysis.
Nonrandomized allocation is a major limitation for causal interpretation. The preparation categories reflect differences in remaining tooth-wall height, thickness, and prior axial preparation. Such differences may be associated with later outcomes. The registry record does not describe adjusted analyses, measured prognostic characteristics, or methods to address confounding, so the extent to which group differences were considered is unknown.
The study description refers to five years of clinical service, while the primary failure outcome is defined through the 3-year recall. The registry does not report the primary endpoint’s follow-up completeness, numbers assessed over time, or handling of missing outcome status. It does not provide details about participant or crown characteristics, the analysis population, adherence to the planned procedure, or how multiple outcomes were addressed.
Masking is listed as none. The record does not explain how clinical outcome assessment was conducted or whether assessors were independent of treatment delivery. The adverse-event figures lack event descriptions and collection details. These limitations mean that the registry information supports interpretation of the study’s question and design, but not a complete assessment of its clinical findings or their reliability.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Alternative EndoCrowns Designs for Chairside CAD/CAM — NCT04604314.
- Registry record last updated 2026-07-13; status last verified 2026-06.
- Results first posted 2026-07-13.
- Data retrieved from ClinicalTrials.gov on 2026-11-09.
- Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
NCT04604314 is a completed, nonrandomized, parallel clinical trial sponsored by the University of Michigan. It enrolled 69 and studied CEREC Tessera chairside CAD/CAM endocrowns in endodontically treated teeth. The arms are defined by Endocrown Onlay and Endocrown Shoulder preparation criteria. The registry states that allocation was nonrandomized and cautions that the arms are not a like-for-like control comparison.
The primary outcome is the number of crowns that failed through the 3-year recall, with failure defined as visual loss requiring replacement. The study description refers to five years of clinical service. Restoration loss of retention and margin staining are listed as secondary outcomes. The registry record does not report primary or secondary outcome results, an analysis method, or a sample-size justification. Adverse-event entries list 0 serious and 0 other events in both groups, and deaths of 1 and 0, respectively, without event details or causal assessment.