Search Strategy
The data for this systematic review and meta-analysis were obtained and
processed following PRISMA (Preferred Reporting Items for Systematic
Reviews and Meta-Analyses) principles (Moher et al., 2010).
Relevant articles published up to December 30th, 2020 were searched
using the relevant keywords and respective Boolean logic operators (AND,
OR, NOT) in the above-mentioned databases: PubMed, EMBASE, Ovid MEDLINE,
Web of Science. The relevant keywords were combined as follow for the
search: ((jaw, edentulous, partially) OR (partially edentulous) OR
(partial edentulism) OR (full edentulism) OR (fully edentulous) OR
(atrophic maxilla) OR (posterior maxilla)) OR (Lateral AND (“maxillary
sinus lift” OR “sinus lift” OR “maxillary sinus floor elevation” OR
“sinus floor elevation” OR “maxillary sinus floor augmentation” OR
“sinus floor augmentation” OR steroids OR steroid OR corticosteroids
OR corticosteroid)) OR (lateral AND (“maxillary sinus lift” OR “sinus
lift” OR “maxillary sinus floor elevation” OR “sinus floor
elevation” OR “maxillary sinus floor augmentation” OR “sinus floor
augmentation”)) AND (pain OR pain reduction OR trismus OR swelling OR
complications OR inflammation OR outcome).
Four independent reviewers (NAV, GLD, GP and LM) first screened all
study titles, then read abstracts, and lastly assessed the full texts of
included articles. Disagreements were resolved by discussion among
reviewers. The final selections according to the inclusion and exclusion
criteria were made by the same authors, as some articles were excluded
only after the full text analysis.
Studies were included if the following criteria were met:
• Studies specifically referring to lateral sinus lift
• RCTs
• Prospective cohort studies
• Case series
• Studies published in the English language
• Explicit reference to the peri-operative (pre and/or post)
pharmacological prescriptions (antibiotics, anti-inflammatories,
analgesics)
• Details of at least one of the post-operative inflammatory parameters
reported (swelling, suppuration, symptoms of acute sinusitis, trismus,
pain, wound dehiscence)
• Crestal/vertical sinus lift
• Less than 20 patients or 20 sinus lift surgeries
• Case reports
• Retrospective studies
• In vitro studies
• Animal studies
• Articles with same cohort of patients
• Post-operative inflammatory parameters not considered in the analysis
(not mentioned in the materials and methods, or the analysis of the
parameters were not clearly identifiable in the results)
• Inflammatory parameters not discernable as values for a meta-analysis
Two authors (NAV, LM) independently assessed the studies in terms of the
inclusion, relevance, eligibility, and risk of bias following the
Cochrane Collaboration tool for RCTs and the Newcastle-Ottawa tool for
prospective cohort studies; any disagreement was resolved by consensus
of reviewers and statistics researcher (ZN).
Data Extraction and Collection Process
Following the screening process, four reviewers (NAV, GLD, GP and LM)
independently extracted the data of the selected articles using data
tables. All extracted data were reviewed, and any conflict was resolved
among the authors and confirmed by the statistician. The following
information was extracted from each included trial: year of publication,
study design, number of patients, number of patients at the end of the
study, number of implants, dropouts, mean age of patients, age range,
mean initial bone height, type of biomaterial used, type of membrane
used, single or bilateral sinus augmentation, dosage and timing of
antibiotics prescribed, type of antibiotic, timing and dosage of
corticosteroids prescribed, other anti-inflammatory medications
prescribed, type and number of intra-operative complications, and type
and number of post-operative symptoms at one and two weeks.
The primary (swelling, pain, infection) and secondary outcomes (active
suppuration, bleeding, wound dehiscence, trismus, hematoma, early
implant failure) were classified as present or absent, as clearly
reported and numbered by the authors of the selected articles. If an
article did not directly provided the number of outcome occurrences,
they were extrapolated from the specific scales or tables, by counting
the number of patients that still reported that outcome during the first
two weeks post-op; as an example, we counted the number of patients that
still reported pain in a VAS pain scale after 7 days. In this case, the
pain was considered to be medium-high or high.
A meta-analysis was performed using the CMA software (Comprehensive Meta
Analysis Version 2.0) for each group separate. A random effects model
was performed to estimate the event rate. Heterogeneity was checked
based on methods, design, and type of complications.