Sinus Floor Augmentation or Short Implants in the Atrophic Posterior Maxilla: A Critical Review of Grafted and Graftless Strategies and a Height-Stratified Non-Inferiority Trial Protocol With a Ten-Year Horizon
- Paper type
- Critical review with an embedded randomized non-inferiority trial protocol
- Status
- Original Work · Working Paper
- Keywords:
- maxillary sinus floor elevation · short implants · graftless sinus lift · residual bone height · non-inferiority trial · marginal bone loss · crown-to-implant ratio · posterior maxilla
Contents
- 1Introduction and Clinical Significance
- 2Critical Review of the Established Literature
- 3Identification of the Specific Research Gap
- 4Scientific Objective and Hypothesis
- 5Theoretical Foundation
- 6Methodology
- 7Evidence Synthesis and Explicitly Labeled Theoretical Predictions
- 8Discussion: Mechanisms and Competing Explanations
- 9Implications for Implant Surgery and Restorative Dentistry
- 10Technical and Clinical Limitations
- 11Proposed Clinical Validation
- 12Future Research Directions
- 13Evidence-Calibrated Conclusion
Abstract
Background: Sinus floor augmentation with placement of standard-length implants is well documented, with implant survival above 90 percent in systematic reviews and with membrane perforation as its commonest complication. Short implants avoid augmentation. Randomized trials to five years report similar survival and fewer complications with short implants, with a tendency toward more late failures in some meta-analyses. Trials enroll a range of residual bone heights and analyze them together.
Objective: To review the evidence for grafted and graftless approaches, to examine why outcomes might diverge with time and with residual height, and to design a trial that tests non-inferiority of short implants at ten years within defined height strata.
Methods: Critical narrative review of systematic reviews, consensus statements, and randomized trials. A reserve model was formulated in which time to loss of a minimum supporting length depends on initial osseointegrated length and the rate of marginal bone loss. A stratified randomized non-inferiority trial protocol was written with prespecified margin and analysis.
Findings: Lateral and transcrestal augmentation are effective, with complication profiles that differ. New bone forms beneath an elevated membrane without graft material, though the height gained is variable. Short implants perform comparably to longer implants in augmented sinuses over one to five years and are associated with less morbidity, cost, and treatment time. Evidence beyond five years is limited, and pooled late-failure estimates are imprecise. We hypothesize that 6 mm implants are non-inferior to longer implants with augmentation at ten years, within a margin of 5 percentage points in implant-level survival per patient, where residual height is 5 to 7 mm, and that non-inferiority is not demonstrated where it is 4 to 5 mm. A confidence bound crossing the margin in the first stratum would refute the first part.
Conclusions: The choice between augmentation and short implants should depend on residual bone height and on the time horizon of the evidence. Present data justify short implants as an alternative in the medium term and leave the long term open. The reserve calculations are illustrative.
Keywords: maxillary sinus floor elevation; short implants; graftless sinus lift; residual bone height; non-inferiority trial; marginal bone loss; crown-to-implant ratio; posterior maxilla
1. Introduction and Clinical Significance
After loss of the maxillary posterior teeth, the alveolar process resorbs and the sinus may expand into it; radiographic study has documented pneumatization of the sinus floor following extraction, most pronounced after loss of several adjacent teeth [1]. The bone remaining beneath the sinus is frequently less than the 8 to 10 mm that conventional implant lengths require. Two responses are available. The first creates bone: the sinus membrane is elevated and the space beneath it is filled, by a lateral window approach introduced in the 1980s [2, 3] or through the implant osteotomy with osteotomes [4]. The second accepts the bone that is there and places an implant short enough to fit, or elevates the membrane and relies on the blood clot and the implant apex to maintain space without graft material [5].
The decision is common, and its consequences are concrete for the patient. Lateral sinus augmentation is a second surgical site with its own morbidity, cost, healing time of months before or after implant placement, and risk of sinus complications. A short implant can often be placed in a single, shorter procedure. Against this stand two long-held reservations about short implants in the posterior maxilla, where bone is soft and occlusal forces high: that a smaller bone-implant interface and a long crown increase mechanical risk, and that a short implant has less bone to lose before it fails.
Evidence from randomized trials has accumulated to the point where consensus statements regard short implants as a valid alternative [6]. Most of that evidence extends to five years. Implants are intended to serve for decades, and the mechanisms by which a short implant might do worse, progressive bone loss and fatigue of a heavily loaded structure, are slow. Whether the equivalence seen at five years persists at ten, and whether it holds across the range of residual bone heights that clinicians encounter, are the questions examined here.
2. Critical Review of the Established Literature
Lateral window augmentation
A systematic review of lateral-approach sinus floor elevation estimated an annual implant failure rate corresponding to three-year survival of roughly 90 percent overall, higher for rough-surfaced implants and when a membrane covered the window, and identified membrane perforation as the most frequent surgical complication [7]. A meta-analysis of membrane perforation reported that it occurs in a substantial minority of procedures and is associated with other complications [8]; a large clinical series described its prevalence and management and found that implant survival fell with increasing perforation size [9]; and an earlier series reported surgical complications of the sinus graft procedure and their limited effect on implant survival when managed [10]. Sinus physiology is generally preserved after augmentation in patients without preexisting disease [11]. A systematic review and meta-analysis of long-term effectiveness, prepared for a European consensus workshop, found high implant survival in augmented sinuses beyond five years across graft materials [12], and a meta-analysis of histomorphometric outcomes showed that the proportion of new bone varies with graft material and healing time, autogenous bone yielding the most early [13]. The Cochrane review of sinus augmentation procedures found no clear evidence favoring one material or technique; for residual bone heights of 4 to 9 mm it identified four small trials of short implants (5 to 8.5 mm) as an alternative to sinus lift and judged the moderate-quality evidence insufficient to determine which approach had fewer prosthesis or implant failures up to one year after loading [14].
Transcrestal elevation
The osteotome technique elevates the floor through the implant site, avoiding a lateral window [4]. A systematic review estimated three-year implant survival of about 92 percent for the transalveolar technique and found that survival was lower when residual bone height was less than 5 mm [15]. In a randomized comparison, the transcrestal approach was associated with less postoperative morbidity than the lateral approach [16]. The technique gains less height and gives the surgeon no direct view of the membrane.
Membrane elevation without graft
Lundgren and colleagues reported bone formation in the space created by elevating the sinus membrane and placing implants as tent poles, without any graft [5]. Systematic reviews of graft-free sinus floor elevation, by lateral and transcrestal routes, report implant survival comparable to that of grafted procedures over short to medium follow-up, with less radiographic bone gain [17–19]. A randomized trial of osteotome elevation with and without grafting in severely atrophic maxillae, using short implants, found bone gain in both groups at one year, greater with grafting, and all implants in function [20]; at five years implants in both groups remained in function, and the authors concluded that grafting was unnecessary to achieve a mean of several millimeters of bone around the implant apex [21]. The evidence comes from small trials and case series in selected patients.
Short implants versus augmentation: randomized evidence
Several randomized trials compare short implants in native bone with longer implants in augmented sinuses. A multicenter trial of 6 mm implants versus 11 to 15 mm implants with sinus grafting reported patient-reported outcomes and clinical results at one year [22, 23] and at five years found similar implant survival in the two groups, with treatment time, cost, and morbidity favoring short implants [24]. A three-year trial of 6 mm implants versus sinus elevation with implants of 10 mm or more found comparable survival and fewer complications with short implants [25]. A five-year trial of single crowns on 6 mm implants versus 11 mm implants with sinus elevation reported similar outcomes [26].
Meta-analyses pool these and related trials. A systematic review for a consensus conference concluded that short implants and longer implants with sinus lifting gave similar survival at one year or more, with more complications and higher cost for grafting [27]. A meta-analysis of five-year randomized data found a non-significant tendency toward more implant loss with short implants and less marginal bone loss around them [28]. Another reported no significant difference in survival [29], and a time-stratified meta-analysis examined how the comparison changes with follow-up [30]. More broadly, a meta-analysis of implants of 6 mm or less against longer implants in posterior jaws found slightly lower survival for short implants over one to five years with wide variation between studies [31]; a meta-analysis of randomized trials of extra-short versus longer implants found similar survival to three years with fewer biological complications and lower cost for short implants [32]; and other reviews found no significant difference [33] or highlighted that most data on short implants concern the partially edentulous patient over limited follow-up [34]. The consensus report based on several of these reviews accepted short implants as a treatment option while stating that longer-term data were needed [6].
Mechanical considerations
A short implant under a crown of normal occlusal height in a resorbed ridge has a high crown-to-implant ratio. A systematic review found no evidence that crown-to-implant ratio, within the ranges studied, affects implant survival or marginal bone loss [35]. The bending moment on an implant depends on the height at which lateral force is applied above the crest and not on the length of implant within bone; implant length affects how that moment is distributed to bone and how much supporting bone remains if the crest recedes. Marginal bone loss around short implants in the trials is, if anything, slightly less than around longer implants in grafted bone [28].
Appraisal
The comparison is among the better-studied questions in implant dentistry: several randomized trials, consistent in direction. Its limits are those of the trials. Samples are small, typically some tens of patients per arm, so that implant loss, a rare event, is estimated imprecisely. Follow-up is five years at most in the randomized comparisons. Residual bone height at enrollment spans several millimeters within and between trials, and results are not reported by height. Operators are experienced. Outcomes are sometimes analyzed per implant where patients received more than one. And the designs are superiority comparisons interpreted as showing equivalence when no difference is found, which a non-significant result in a small trial does not establish.
3. Identification of the Specific Research Gap
Established: lateral and transcrestal sinus augmentation support implants with high survival over medium to long follow-up; perforation and other complications are not rare; graft material is not required for some bone formation beneath an elevated membrane; short implants in native bone perform similarly to longer implants in augmented sinuses over one to five years, with less morbidity and cost.
Uncertain: whether short implants remain comparable at ten years and beyond, and in which residual-height range. The clinical decision is most difficult where residual height is 4 to 6 mm: enough to stabilize a 6 mm implant with little or no elevation, and little enough that an operator may prefer to augment.
Existing methods are inadequate in four respects. Trials are not powered or designed for non-inferiority, so the absence of a significant difference is uninformative about how much worse short implants could be. Follow-up stops where slow mechanisms would begin to matter. Residual height, the variable most likely to modify the comparison, is not used for stratification or subgroup analysis. And the outcome most reported, implant survival, does not capture the loss of reserve that precedes failure.
The specific gap is a randomized comparison with a prespecified non-inferiority margin, stratified by residual bone height, followed for ten years, analyzed with the patient as the unit, and reporting remaining osseointegrated length as well as survival.
4. Scientific Objective and Hypothesis
The objective is to determine whether 6 mm implants placed without grafting are non-inferior to implants of 10 mm or more placed with sinus floor augmentation at ten years after loading, separately for two strata of residual bone height.
Hypothesis H15: in sites with residual bone height of 5 to 7 mm, the ten-year proportion of patients with all study implants surviving in the short-implant group is not lower than that in the augmentation group by more than 5 percentage points; in sites with residual bone height of at least 4 and less than 5 mm, non-inferiority within the same margin is not demonstrated.
The first part is falsified if the upper limit of the two-sided 95% confidence interval for the difference in ten-year failure proportion (short minus augmentation) exceeds 5 percentage points in the 5 to 7 mm stratum. The second part is falsified if non-inferiority is demonstrated in the 4 to 5 mm stratum, that is, if the corresponding upper limit lies below 5 points; this would show short implants to be acceptable even at the lower heights. The margin of 5 percentage points is a judgment about the largest loss of long-term survival that the avoidance of grafting would justify; it should be confirmed with patient and clinician input before the trial begins, and the conclusions would change if a different margin were chosen.
A mechanistic secondary hypothesis, H15b, is that the hazard of failure in the short-implant group relative to the augmentation group increases with time, being close to 1 in the first five years and greater thereafter, and that late failures in the short group are preceded by marginal bone loss. H15b is falsified by a hazard ratio that is constant or declining over time.
5. Theoretical Foundation
Interface area
The lateral surface of a cylindrical implant of diameter d and length L in bone, with threads increasing the area by a factor λ, is approximately
A = λ π d L (1)
A 6 mm implant of 5 mm diameter has about the same nominal area as a 7.3 mm implant of 4.1 mm diameter, and half that of a 12 mm implant of 5 mm diameter. Area is a crude index, since stress concentrates in the crestal few millimeters whatever the implant length, which is the mechanical argument that length beyond a certain point contributes little to load bearing under axial and lateral forces.
Lever arm
A lateral force component Flat applied at crown height hc above the bone crest produces a moment at the crest
M = F_(lat) h_c (2)
which is independent of the implant length in bone. The crown-to-implant ratio hc/L is larger for a short implant because L is smaller, with hc and M unchanged for the same prosthetic space. If the crest recedes by an amount b, the lever arm becomes hc + b and the supporting length L − b, so that the same bone loss both raises the moment and reduces the support, proportionally more for the shorter implant.
Bone reserve
Let L0 be the osseointegrated length after initial remodeling and Lmin the minimum length compatible with stable function, a quantity that is not known and is treated here as a parameter. If marginal bone is lost at a rate m millimeters per year, the time until support falls to the minimum is
T = (L_0 - L_(min))/m (3)
The reserve L0 − Lmin is what distinguishes long from short implants. With stable bone, m near zero, T is effectively unlimited for both, and implant length is irrelevant to survival. With progressive loss, as in peri-implantitis, T is finite and proportional to the reserve.
As an illustration with assumed values, take Lmin as 3 mm. A 6 mm implant with 5 mm of osseointegrated length after remodeling has a reserve of 2 mm; a 10 mm implant with 9 mm has a reserve of 6 mm. At a loss rate of 0.05 mm per year the times are 40 and 120 years, both beyond any clinical horizon. At 0.5 mm per year they are 4 and 12 years. The model thus predicts no difference in survival while bone is stable and an earlier failure of short implants in the subset of patients who develop progressive bone loss, with the difference emerging after several years. It predicts that the comparison depends on the prevalence of progressive bone loss in the trial population and on the length of follow-up, and that trials of five years in well-maintained patients would show equivalence even if a long-term disadvantage existed.
Why residual height should modify the comparison
In a site with 6 to 7 mm of residual bone, a 6 mm implant lies wholly in native bone. In a site with 4 to 5 mm, the same implant must engage or elevate the sinus floor and has 1 to 2 mm of its length beyond native bone, supported at best by bone that forms under the membrane; its effective reserve is smaller from the outset, and its primary stability depends on a thinner cortex. The augmentation arm is less sensitive to residual height, since the graft supplies the missing bone, although lateral augmentation with simultaneous placement also relies on residual bone for primary stability. Residual height is therefore expected to modify the difference between strategies, and pooling over it obscures the range in which each is preferable.
Non-inferiority framework
Let pS and pA be the ten-year probabilities that a patient loses one or more study implants with short implants and with augmentation. Non-inferiority of short implants with margin Δ is concluded if the upper limit of the confidence interval for pS − pA lies below Δ:
U_(95)(p_S - p_A) < Δ (4)
A result may show non-inferiority, show inferiority (lower limit above zero), show both (a difference that is real and smaller than the margin), or be inconclusive (interval spanning the margin). Reporting follows the extension of the CONSORT statement for non-inferiority trials [36]. Because time to failure is available, the comparison is also made on the survival scale with a restricted mean survival time difference, which does not assume proportional hazards and so accommodates the late divergence that H15b predicts.
6. Methodology
Review methods
The review is narrative. Survival and complication statements are drawn from the cited systematic reviews and trials; pooled values are quoted qualitatively unless confirmed from the abstracts. No new meta-analysis was performed.
Trial design
The proposed study is a multicenter, parallel-group, randomized non-inferiority trial with stratification by residual bone height, with a protocol conforming to SPIRIT [37] and reporting according to CONSORT and its non-inferiority extension [36, 38]. The unit of randomization and analysis is the patient. Centers include specialist and general practices to improve applicability.
Participants
Adults with one or two adjacent missing posterior maxillary teeth, residual bone height at the planned sites of at least 4 and no more than 7 mm on cone-beam computed tomography, and ridge width sufficient for a 4 to 5 mm diameter implant without horizontal augmentation are eligible. Residual height is measured on standardized cross-sections at the planned implant axis by two examiners. Exclusions are acute or untreated chronic sinus disease, heavy smoking, uncontrolled diabetes, and untreated periodontitis. Patients are stratified as 5.0 to 7.0 mm or 4.0 to 4.9 mm by the lowest height among their study sites.
Interventions
Short-implant arm: implants of 6 mm length and 4 to 5 mm diameter placed in native bone; where the floor is reached, limited transcrestal elevation without graft material is permitted and recorded. Augmentation arm: sinus floor augmentation by the lateral approach with a specified bone substitute and membrane, or by the transcrestal approach with graft where height permits, with implants of 10 mm or more placed simultaneously or after healing according to primary stability; the technique used is recorded and its choice follows a written algorithm. Both arms use the same implant system and connection, submerged or non-submerged healing by the same rule, and screw-retained crowns, splinted where two implants are adjacent. Maintenance is standardized.
Outcomes
The primary outcome is failure of one or more study implants in a patient by ten years after loading, defined as implant removal for any reason. Key secondary outcomes are: time to first implant failure; remaining osseointegrated length at each radiographic examination, computed as implant length minus the distance from the implant shoulder to the first bone contact, on standardized radiographs; marginal bone level change; peri-implantitis by the 2017 workshop definition; technical complications; surgical and sinus complications, including membrane perforation; number and duration of surgical procedures; time from enrollment to functional loading; patient-reported pain, swelling, and oral health-related quality of life; and direct cost. Outcome examiners are independent of the treating clinicians; blinding to implant length on radiographs is not possible.
| Outcome | Time points | Purpose |
|---|---|---|
| Patient-level implant failure (primary) | 10 years after loading | Non-inferiority test |
| Time to first failure; restricted mean survival time | Continuous | Detect late divergence (H15b) |
| Remaining osseointegrated length; marginal bone level | Loading; 1, 3, 5, 7, 10 years | Loss of reserve before failure |
| Peri-implantitis | Annually | Mechanism of late loss |
| Surgical and sinus complications | Perioperative; 1 year | Harm specific to augmentation |
| Technical complications | Annually | Mechanical consequences of crown height |
| Treatment time, number of procedures, cost | To loading; cumulative | Benefit of avoiding grafting |
| Patient-reported outcomes | 1 week; loading; 1, 5, 10 years | Patient-centered comparison |
Randomization
Allocation is by a central web-based system with permuted blocks of varying size within strata defined by center and height category, concealed until after eligibility is confirmed on tomography.
Statistical analysis
The primary analysis estimates, within each height stratum, the difference in ten-year cumulative incidence of patient-level failure between arms by the Kaplan-Meier method, with a two-sided 95% confidence interval, and compares the upper limit with the margin. Both intention-to-treat and per-protocol analyses are reported, since in non-inferiority trials protocol deviations tend to bias toward a finding of no difference; non-inferiority is concluded only if both agree. Deaths and withdrawals are censored, with a sensitivity analysis treating death as a competing event and a tipping-point analysis for losses to follow-up. The interaction between arm and height stratum is estimated on the risk-difference scale. H15b is examined by a time-varying hazard ratio and by the difference in restricted mean survival time at five and ten years. Remaining osseointegrated length is modeled longitudinally with patient-level random effects, and its value at five years is examined as a predictor of failure between five and ten years.
Sample size
For a non-inferiority comparison of two proportions with margin Δ, one-sided type I error α, and power 1 − β, assuming true proportions pS and pA, the number of patients per arm is approximately
n = ((z_(1-α) + z_(1-β))^2 [p_S(1 - p_S) + p_A(1 - p_A)])/((Δ - (p_S - p_A))^2) (5)
With ten-year patient-level failure of 5 percent in both arms, an assumption for planning, a margin of 5 percentage points, one-sided α of 0.025, and power of 0.80, n is about 298 per arm in each stratum. Allowing for 20 percent loss over ten years raises this to about 373 per arm, or roughly 1,490 patients across both strata. If the true failure proportion in the short arm were 7 percent, the requirement in that stratum would rise nearly threefold. These numbers explain why no existing trial can address the question and indicate that a trial of this kind is feasible only as a large multicenter or practice-network undertaking with long-term funding. A design powered for the 5 to 7 mm stratum alone, with the 4 to 5 mm stratum treated as exploratory, would halve the requirement.
7. Evidence Synthesis and Explicitly Labeled Theoretical Predictions
Synthesis
Table 2 summarizes the evidence for each strategy. Certainty labels are the authors' qualitative judgments.
| Strategy | Evidence | Advantages | Uncertainties |
|---|---|---|---|
| Lateral augmentation, standard implants [7, 12] | Systematic reviews; long-term cohorts | Longest record; applicable at very low residual height | Perforation and sinus complications; time, cost, morbidity |
| Transcrestal augmentation [15, 16] | Systematic review; randomized trial on morbidity | Less invasive than lateral window | Limited height gain; lower survival below 5 mm residual height |
| Membrane elevation without graft [5, 17, 21] | Small trials; case series; meta-analyses | No graft material or donor site | Variable bone gain; selected patients |
| Short implants in native bone [6, 24–26] | Randomized trials to 5 years; consensus | Less surgery, time, cost; fewer complications | Outcome beyond 5 years; performance at 4 to 5 mm; reserve under bone loss |
Theoretical predictions
The following are predictions and have not been tested.
Prediction 1. In the 5 to 7 mm stratum, ten-year patient-level failure will differ between arms by less than the margin, and non-inferiority will be shown.
Prediction 2. In the 4 to 5 mm stratum, the short-implant arm will have more early failures related to primary stability and more late failures, and the confidence interval will include or exceed the margin.
Prediction 3. Hazards will be similar in the first five years and will diverge afterward, with late failures in the short arm concentrated among implants that had lost marginal bone by year five (H15b).
Prediction 4. Mean marginal bone loss will be similar or lower in the short arm; the difference in outcome will arise from the consequence of a given loss, not from its amount.
Prediction 5. Surgical complications, time to loading, number of procedures, and cost will favor the short arm in both strata, and early patient-reported morbidity will be lower.
Prediction 6. Technical complications will be somewhat more frequent in the short arm in the lower stratum, where crown height is greatest.
8. Discussion: Mechanisms and Competing Explanations
The reserve argument is simple and may be wrong in several ways. It assumes that a minimum supporting length exists and is similar for short and long implants. If load is carried almost entirely by crestal bone, a long implant that has lost its crestal support is not much better off than a short one, and the difference in reserve is smaller than lengths suggest. It assumes that bone loss, when it occurs, proceeds at the same rate around both. The trials report slightly less marginal bone loss around short implants in native bone than around long implants in grafted bone [28], for which plausible reasons include the absence of a graft-bone interface at the crest, less surgical trauma, and the fact that grafted sites were more atrophic to begin with; if native bone resists peri-implantitis better than augmented bone, the short implant's disadvantage in reserve could be offset by a lower incidence of progressive loss. And it ignores retreatment: a failed short implant leaves a site that can still be augmented, whereas a failed implant in a grafted sinus presents a more difficult problem. On a lifetime view, the strategy that fails more often but more benignly may be preferable.
A competing position is that the question is settled in favor of short implants and that a ten-year trial is unnecessary. The consistent five-year results and the consensus endorsement support that view [6, 24]. The counter-argument is statistical and biological. Statistically, the trials are small, and the pooled five-year estimates include a tendency toward more losses with short implants that does not reach significance [28]; an effect of that size, if real, would matter over decades. Biologically, the mechanisms that would disadvantage short implants are slow. Neither argument shows that a long-term disadvantage exists; together they show that the available evidence could not have detected one.
The opposite position, that augmentation remains the standard because of its longer record, underweights its costs. Membrane perforation, sinusitis, graft infection, longer treatment, and higher expense are certain and immediate, whereas the hypothetical late disadvantage of short implants is uncertain and remote. The Cochrane review's observation of more complications with sinus lift [14] and trial data on patient-reported morbidity [16, 22] are part of the comparison, and a non-inferiority design is the appropriate way to weigh a possible small loss of long-term survival against them.
Residual height as an effect modifier is supported by indirect evidence. Survival after transcrestal elevation was lower with less than 5 mm of residual bone [15], and the Cochrane comparison with short implants was confined to heights of 4 to 9 mm and to one year of loading [14]. Below 4 mm a 6 mm implant has more of its length beyond native bone than within it, and neither the model nor existing trials support its use without augmentation; the proposed trial excludes that range. Extra-short implants of 4 to 5 mm length extend the same question further and are beyond the present scope.
Graftless membrane elevation occupies an intermediate position and complicates the dichotomy. If bone forms reliably around the apex of an implant that tents the membrane [5, 21], a 6 mm implant in 4 mm of bone may come to have 6 mm of support, and the distinction between short-implant and augmentation strategies blurs. The protocol permits limited graft-free elevation in the short arm and records it, so that its contribution can be examined, but the trial is not designed to compare grafted with graft-free elevation.
The patient as the unit of analysis, and the definition of the primary outcome as loss of any study implant, are deliberate. They avoid the clustering problem and express the outcome as patients experience it. They also mean that patients with two implants have two chances to fail; stratifying or adjusting for the number of study implants handles this.
Finally, the long horizon is itself a threat to validity. Over ten years, patients move, die, or change dentists; implant systems and prosthetic materials change; and differential loss to follow-up between arms could bias the comparison. Design features, registry linkage for vital status, remote follow-up, and prespecified sensitivity analyses, mitigate the problem without solving it.
9. Implications for Implant Surgery and Restorative Dentistry
On current evidence, for a posterior maxillary site with about 5 to 7 mm of residual bone in a patient who can be maintained, a 6 mm implant without grafting is a reasonable first choice, supported by randomized trials to five years and by consensus [6, 24], and it spares the patient a sinus procedure. The patient should be told that evidence beyond five years is limited. For sites with less than about 4 mm, augmentation, or remote anchorage in the severely atrophic arch, remains the documented approach [7, 12]. Between 4 and 5 mm the evidence does not clearly favor either, and the choice may reasonably turn on sinus health, bone quality, the patient's tolerance for additional surgery, and the operator's experience with each technique.
Where short implants are used, measures that protect marginal bone protect the implant's limited reserve: careful plaque control and maintenance, attention to prosthesis contour and cleansability, and early investigation of any bone loss. Splinting adjacent short implants and limiting lateral occlusal contacts are mechanically sensible in view of crown height, although evidence that crown-to-implant ratio affects outcome is lacking [35].
Where augmentation is chosen, the transcrestal approach is associated with less morbidity than the lateral approach when residual height permits [16], and preoperative assessment of the sinus for disease and for anatomical features that predispose to perforation is advisable [8].
10. Technical and Clinical Limitations
The review is narrative and relies on abstracts and summary statements for pooled results. The reserve model is a simplification with an unknown minimum length and a constant loss rate; real bone loss is episodic and, in peri-implantitis, may accelerate. The trial requires a very large sample and a decade of follow-up, and its feasibility is uncertain. Implant systems may be superseded during the trial. Operators cannot be blinded, and examiners cannot be blinded to implant length on radiographs. Allowing two augmentation techniques in the comparator arm adds heterogeneity, justified by the wish to reflect practice. The margin is a value judgment. Exclusion of heavy smokers, uncontrolled diabetes, and untreated periodontitis removes the patients most likely to lose bone, in whom the reserve argument would be most relevant, a tension between internal validity and the question of interest that might be addressed by a parallel observational cohort. Results will apply to 6 mm implants of the system used and may not transfer to other lengths or designs.
11. Proposed Clinical Validation
The trial is the definitive test. Before it, and at far lower cost, three analyses of existing data could narrow the question. First, long-term follow-up of the existing randomized cohorts to ten years, with pooled individual-patient analysis stratified by baseline residual height; the investigators of those trials hold the necessary data, and the combined sample, though small, would give a first estimate of late divergence and of effect modification by height. Second, registry or practice-network cohorts of short implants in the posterior maxilla with at least ten years of follow-up, analyzed by remaining osseointegrated length over time, would test whether late failures are preceded by loss of reserve. Third, a time-stratified re-analysis of published data, extending existing work [30], would show whether the relative risk changes with follow-up. If these analyses showed no hint of late divergence, the case for a large trial would weaken; if they did, it would strengthen. Refutation criteria for H15 and H15b are given in Section 4.
12. Future Research Directions
Several related questions deserve attention. The minimum osseointegrated length for stable function under defined loads is unknown and could be approached by finite element modeling validated against clinical failures and by analysis of implants that continue to function after substantial bone loss. The reliability of bone formation after graft-free elevation, and the factors that predict it, need larger controlled studies. Extra-short implants of 4 to 5 mm require their own long-term evidence. Patient preferences regarding the trade-off between a more invasive procedure now and a possibly higher risk of failure later should be elicited formally to set non-inferiority margins. And decision-analytic models that incorporate retreatment after failure would allow strategies to be compared over a lifetime and would indicate which uncertainties most affect the choice.
13. Evidence-Calibrated Conclusion
Sinus floor augmentation with standard-length implants is effective, with moderate certainty from systematic reviews and long-term cohorts, and carries surgical complications, of which membrane perforation is the most frequent. Short implants in native bone give similar implant survival to five years in several randomized trials, with less morbidity, time, and cost; certainty is moderate for that horizon and low beyond it. Graft-free membrane elevation leads to bone formation in small studies. Whether short implants remain comparable at ten years, and down to what residual height, has not been tested. A simple reserve model suggests that any disadvantage would appear late and only in patients with progressive bone loss, which existing trials could not have detected. The stratified non-inferiority trial outlined here would answer the question; it is large, and analyses of existing long-term data should precede it. The model's numerical examples are illustrative, and no patient data are reported.
Cite this paper
Jain S, Gupta A. Sinus Floor Augmentation or Short Implants in the Atrophic Posterior Maxilla: A Critical Review of Grafted and Graftless Strategies and a Height-Stratified Non-Inferiority Trial Protocol With a Ten-Year Horizon. Working paper. Center for Implant Dentistry; 2026. Available from: https://bayareaimplantdentistry.com/clinician-research-and-publications/sinus-augmentation-versus-graftless-alternatives
References
- Sharan A, Madjar D. Maxillary sinus pneumatization following extractions: a radiographic study. Int J Oral Maxillofac Implants. 2008;23(1):48-56. PMID: 18416412 (opens in a new tab).
- Boyne PJ, James RA. Grafting of the maxillary sinus floor with autogenous marrow and bone. J Oral Surg. 1980;38(8):613-6. PMID: 6993637 (opens in a new tab).
- Tatum H Jr. Maxillary and sinus implant reconstructions. Dent Clin North Am. 1986;30(2):207-29. PMID: 3516738 (opens in a new tab).
- Summers RB. A new concept in maxillary implant surgery: the osteotome technique. Compendium. 1994;15(2):152, 154-6, 158 passim; quiz 162. PMID: 8055503 (opens in a new tab).
- Lundgren S, Andersson S, Gualini F, Sennerby L. Bone reformation with sinus membrane elevation: a new surgical technique for maxillary sinus floor augmentation. Clin Implant Dent Relat Res. 2004;6(3):165-73. PMID: 15726851 (opens in a new tab).
- Jung RE, Al-Nawas B, Araujo M, Avila-Ortiz G, Barter S, Brodala N, et al. Group 1 ITI Consensus Report: The influence of implant length and design and medications on clinical and patient-reported outcomes. Clin Oral Implants Res. 2018;29 Suppl 16:69-77. doi:10.1111/clr.13342 (opens in a new tab) PMID: 30328189 (opens in a new tab).
- Pjetursson BE, Tan WC, Zwahlen M, Lang NP. A systematic review of the success of sinus floor elevation and survival of implants inserted in combination with sinus floor elevation. J Clin Periodontol. 2008;35(8 Suppl):216-40. doi:10.1111/j.1600-051X.2008.01272.x (opens in a new tab) PMID: 18724852 (opens in a new tab).
- Al-Dajani M. Incidence, Risk Factors, and Complications of Schneiderian Membrane Perforation in Sinus Lift Surgery: A Meta-Analysis. Implant Dent. 2016;25(3):409-15. doi:10.1097/ID.0000000000000411 (opens in a new tab) PMID: 26974034 (opens in a new tab).
- Hernández-Alfaro F, Torradeflot MM, Marti C. Prevalence and management of Schneiderian membrane perforations during sinus-lift procedures. Clin Oral Implants Res. 2008;19(1):91-8. doi:10.1111/j.1600-0501.2007.01372.x (opens in a new tab) PMID: 17961185 (opens in a new tab).
- Schwartz-Arad D, Herzberg R, Dolev E. The prevalence of surgical complications of the sinus graft procedure and their impact on implant survival. J Periodontol. 2004;75(4):511-6. doi:10.1902/jop.2004.75.4.511 (opens in a new tab) PMID: 15152813 (opens in a new tab).
- Timmenga NM, Raghoebar GM, Boering G, van Weissenbruch R. Maxillary sinus function after sinus lifts for the insertion of dental implants. J Oral Maxillofac Surg. 1997;55(9):936-9;discussion 940. doi:10.1016/s0278-2391(97)90063-x (opens in a new tab) PMID: 9294502 (opens in a new tab).
- Raghoebar GM, Onclin P, Boven GC, Vissink A, Meijer HJA. Long-term effectiveness of maxillary sinus floor augmentation: A systematic review and meta-analysis. J Clin Periodontol. 2019;46 Suppl 21:307-318. doi:10.1111/jcpe.13055 (opens in a new tab) PMID: 30624789 (opens in a new tab).
- Corbella S, Taschieri S, Weinstein R, Del Fabbro M. Histomorphometric outcomes after lateral sinus floor elevation procedure: a systematic review of the literature and meta-analysis. Clin Oral Implants Res. 2016;27(9):1106-22. doi:10.1111/clr.12702 (opens in a new tab) PMID: 26452326 (opens in a new tab).
- Esposito M, Felice P, Worthington HV. Interventions for replacing missing teeth: augmentation procedures of the maxillary sinus. Cochrane Database Syst Rev. 2014;2014(5):CD008397. doi:10.1002/14651858.CD008397.pub2 (opens in a new tab) PMID: 24825543 (opens in a new tab).
- Tan WC, Lang NP, Zwahlen M, Pjetursson BE. A systematic review of the success of sinus floor elevation and survival of implants inserted in combination with sinus floor elevation. Part II: transalveolar technique. J Clin Periodontol. 2008;35(8 Suppl):241-54. doi:10.1111/j.1600-051X.2008.01273.x (opens in a new tab) PMID: 18724853 (opens in a new tab).
- Farina R, Franceschetti G, Travaglini D, Consolo U, Minenna L, Schincaglia GP, et al. Morbidity following transcrestal and lateral sinus floor elevation: A randomized trial. J Clin Periodontol. 2018;45(9):1128-1139. doi:10.1111/jcpe.12985 (opens in a new tab) PMID: 29992594 (opens in a new tab).
- Duan DH, Fu JH, Qi W, Du Y, Pan J, Wang HL. Graft-Free Maxillary Sinus Floor Elevation: A Systematic Review and Meta-Analysis. J Periodontol. 2017;88(6):550-564. doi:10.1902/jop.2017.160665 (opens in a new tab) PMID: 28168901 (opens in a new tab).
- Silva LD, de Lima VN, Faverani LP, de Mendonça MR, Okamoto R, Pellizzer EP. Maxillary sinus lift surgery-with or without graft material? A systematic review. Int J Oral Maxillofac Surg. 2016;45(12):1570-1576. doi:10.1016/j.ijom.2016.09.023 (opens in a new tab) PMID: 27765427 (opens in a new tab).
- Pérez-Martínez S, Martorell-Calatayud L, Peñarrocha-Oltra D, García-Mira B, Peñarrocha-Diago M. Indirect sinus lift without bone graft material: Systematic review and meta-analysis. J Clin Exp Dent. 2015;7(2):e316-9. doi:10.4317/jced.51716 (opens in a new tab) PMID: 26155353 (opens in a new tab).
- Nedir R, Nurdin N, Khoury P, Perneger T, Hage ME, Bernard JP, et al. Osteotome sinus floor elevation with and without grafting material in the severely atrophic maxilla. A 1-year prospective randomized controlled study. Clin Oral Implants Res. 2013;24(11):1257-64. doi:10.1111/j.1600-0501.2012.02569.x (opens in a new tab) PMID: 22925088 (opens in a new tab).
- Nedir R, Nurdin N, Abi Najm S, El Hage M, Bischof M. Short implants placed with or without grafting into atrophic sinuses: the 5-year results of a prospective randomized controlled study. Clin Oral Implants Res. 2017;28(7):877-886. doi:10.1111/clr.12893 (opens in a new tab) PMID: 27296955 (opens in a new tab).
- Thoma DS, Haas R, Tutak M, Garcia A, Schincaglia GP, Hämmerle CH. Randomized controlled multicentre study comparing short dental implants (6 mm) versus longer dental implants (11-15 mm) in combination with sinus floor elevation procedures. Part 1: demographics and patient-reported outcomes at 1 year of loading. J Clin Periodontol. 2015;42(1):72-80. doi:10.1111/jcpe.12323 (opens in a new tab) PMID: 25418606 (opens in a new tab).
- Schincaglia GP, Thoma DS, Haas R, Tutak M, Garcia A, Taylor TD, et al. Randomized controlled multicenter study comparing short dental implants (6 mm) versus longer dental implants (11-15 mm) in combination with sinus floor elevation procedures. Part 2: clinical and radiographic outcomes at 1 year of loading. J Clin Periodontol. 2015;42(11):1042-51. doi:10.1111/jcpe.12465 (opens in a new tab) PMID: 26425812 (opens in a new tab).
- Thoma DS, Haas R, Sporniak-Tutak K, Garcia A, Taylor TD, Hämmerle CHF. Randomized controlled multicentre study comparing short dental implants (6 mm) versus longer dental implants (11-15 mm) in combination with sinus floor elevation procedures: 5-Year data. J Clin Periodontol. 2018;45(12):1465-1474. doi:10.1111/jcpe.13025 (opens in a new tab) PMID: 30341961 (opens in a new tab).
- Bechara S, Kubilius R, Veronesi G, Pires JT, Shibli JA, Mangano FG. Short (6-mm) dental implants versus sinus floor elevation and placement of longer (≥10-mm) dental implants: a randomized controlled trial with a 3-year follow-up. Clin Oral Implants Res. 2017;28(9):1097-1107. doi:10.1111/clr.12923 (opens in a new tab) PMID: 27402427 (opens in a new tab).
- Guljé FL, Raghoebar GM, Vissink A, Meijer HJA. Single crowns in the resorbed posterior maxilla supported by either 11-mm implants combined with sinus floor elevation or 6-mm implants:A 5-year randomised controlled trial. Int J Oral Implantol (Berl). 2019;12(3):315-326. PMID: 31535100 (opens in a new tab).
- Thoma DS, Zeltner M, Hüsler J, Hämmerle CH, Jung RE. EAO Supplement Working Group 4 - EAO CC 2015 Short implants versus sinus lifting with longer implants to restore the posterior maxilla: a systematic review. Clin Oral Implants Res. 2015;26 Suppl 11:154-69. doi:10.1111/clr.12615 (opens in a new tab) PMID: 25997901 (opens in a new tab).
- Wang M, Liu F, Ulm C, Shen H, Rausch-Fan X. Short Implants versus Longer Implants with Sinus Floor Elevation: A Systemic Review and Meta-Analysis of Randomized Controlled Trials with a Post-Loading Follow-Up Duration of 5 Years. Materials (Basel). 2022;15(13). doi:10.3390/ma15134722 (opens in a new tab) PMID: 35806845 (opens in a new tab).
- Mester A, Onisor F, Di Stasio D, Piciu A, Cosma AM, Bran S. Short Implants versus Standard Implants and Sinus Floor Elevation in Atrophic Posterior Maxilla: A Systematic Review and Meta-Analysis of Randomized Clinical Trials with ≥5 Years' Follow-Up. J Pers Med. 2023;13(2). doi:10.3390/jpm13020169 (opens in a new tab) PMID: 36836403 (opens in a new tab).
- Grunau O, Terheyden H. Lateral augmentation of the sinus floor followed by regular implants versus short implants in the vertically deficient posterior maxilla: a systematic review and timewise meta-analysis of randomized studies. Int J Oral Maxillofac Surg. 2023;52(7):813-824. doi:10.1016/j.ijom.2022.11.015 (opens in a new tab) PMID: 36529571 (opens in a new tab).
- Papaspyridakos P, De Souza A, Vazouras K, Gholami H, Pagni S, Weber HP. Survival rates of short dental implants (≤6 mm) compared with implants longer than 6 mm in posterior jaw areas: A meta-analysis. Clin Oral Implants Res. 2018;29 Suppl 16:8-20. doi:10.1111/clr.13289 (opens in a new tab) PMID: 30328206 (opens in a new tab).
- Ravidà A, Wang IC, Barootchi S, Askar H, Tavelli L, Gargallo-Albiol J, et al. Meta-analysis of randomized clinical trials comparing clinical and patient-reported outcomes between extra-short (≤6 mm) and longer (≥10 mm) implants. J Clin Periodontol. 2019;46(1):118-142. doi:10.1111/jcpe.13026 (opens in a new tab) PMID: 30362137 (opens in a new tab).
- Lemos CA, Ferro-Alves ML, Okamoto R, Mendonça MR, Pellizzer EP. Short dental implants versus standard dental implants placed in the posterior jaws: A systematic review and meta-analysis. J Dent. 2016;47:8-17. doi:10.1016/j.jdent.2016.01.005 (opens in a new tab) PMID: 26804969 (opens in a new tab).
- Telleman G, Raghoebar GM, Vissink A, den Hartog L, Huddleston Slater JJ, Meijer HJ. A systematic review of the prognosis of short (<10 mm) dental implants placed in the partially edentulous patient. J Clin Periodontol. 2011;38(7):667-76. doi:10.1111/j.1600-051X.2011.01736.x (opens in a new tab) PMID: 21564158 (opens in a new tab).
- Meijer HJA, Boven C, Delli K, Raghoebar GM. Is there an effect of crown-to-implant ratio on implant treatment outcomes? A systematic review. Clin Oral Implants Res. 2018;29 Suppl 18(Suppl Suppl 18):243-252. doi:10.1111/clr.13338 (opens in a new tab) PMID: 30306696 (opens in a new tab).
- Piaggio G, Elbourne DR, Pocock SJ, Evans SJ, Altman DG, CONSORT Group. Reporting of noninferiority and equivalence randomized trials: extension of the CONSORT 2010 statement. JAMA. 2012;308(24):2594-604. doi:10.1001/jama.2012.87802 (opens in a new tab) PMID: 23268518 (opens in a new tab).
- Chan AW, Tetzlaff JM, Altman DG, Laupacis A, Gøtzsche PC, Krleža-Jerić K, et al. SPIRIT 2013 statement: defining standard protocol items for clinical trials. Ann Intern Med. 2013;158(3):200-7. doi:10.7326/0003-4819-158-3-201302050-00583 (opens in a new tab) PMID: 23295957 (opens in a new tab).
- Schulz KF, Altman DG, Moher D, CONSORT Group. CONSORT 2010 statement: updated guidelines for reporting parallel group randomised trials. BMJ. 2010;340:c332. doi:10.1136/bmj.c332 (opens in a new tab) PMID: 20332509 (opens in a new tab).
- Reference verification note. 38 of 38 references are PubMed-indexed; their bibliographic records (authors, title, journal, year, volume, pages, DOI, PMID) were retrieved programmatically from the PubMed database on November 17, 2026. Database retrieval confirms that each record exists as cited; it does not replace full-text reading by the responsible authors of every statement attributed to a source.

