Surgical procedures were performed according to a lateral approach technique at the edentulous region distal to the position of the first premolar. A mucoperiosteal buccal flap was elevated, exposing the lateral bony wall of the sinus antrum. A round diamond bur, 2 mm in diameter, was used to outline the demarcation of the lateral window, which was removed, thus completely exposing the underlying Schneiderian membrane. The membrane was separated from the housing bone, and a tension-free reflection exposing the sinus walls was achieved by gently pushing it away using a large flat curette (Kramer-Nevins, Hu-Friedy, Chicago, IL). The established voids were then filled with 100% AAB (Bio-Oss) followed by sealing the open lateral window by the placement of a collagen membrane (Bio Gide) and primary soft tissue closure using Vicryl 4.0 sutures (Vicryl, Ethicon J&J International, Sint-Stevens-Woluwe, Belgium).

Surgical procedures were performed according to a lateral approach technique at the edentulous region distal to the position of the first premolar. A mucoperiosteal buccal flap was elevated, exposing the lateral bony wall of the sinus antrum. A round diamond bur, 2 mm in diameter, was used to outline the demarcation of the lateral window, which was removed, thus completely exposing the underlying Schneiderian membrane. The membrane was separated from the housing bone, and a tension-free reflection exposing the sinus walls was achieved by gently pushing it away using a large flat curette (Kramer-Nevins, Hu-Friedy, Chicago, IL). The established voids were then filled with 50% AAB (Bio-Oss) plus 50% autologous bone harvested locally with a bone scraper followed by sealing the open lateral window by the placement of a collagen membrane (Bio Gide) and primary soft tissue closure using Vicryl 4.0 sutures (Vicryl, Ethicon J&J International, Sint-Stevens-Woluwe, Belgium).

All patients were evaluated clinically and their medical histories were recorded. Preliminary screening, including the acquisition of intraoral and panoramic radiographs, was performed to evaluate potential patients' eligibility. Patients who met the selection criteria received oral hygiene instructions and debridement, if required, after which bone volumes were analysed using cone-beam computed tomography (CBCT; Imaging Sciences International, Hatfield, PA, USA).

A staged approach was carried out at all sites. Thus, sinus lift grafting procedures were followed by implant placement 7-8 months later.

Both groups received tapered implants with an anodised surface.

The removal of implants was dictated by instability, progressive marginal bone loss, infection, or implant fracture. The stability of individual implants was measured by the prosthodontist at the time of definitive crown delivery (5 months after implant placement) by applying 35 Ncm of removal torque. After 1 year, implant stability was tested manually with two dental mirror handles.

Peri-implant marginal bone levels were evaluated on intraoral digital radiographs taken with the parallel technique. A blinded radiologist (FG), unaffiliated with the study centre, interpreted all radiographs. The distances from the mesial and distal interproximal bone to the reference point (the horizontal interface between the implant and abutment) were measured with a software tool (NIH Scion Image, ver. 4.0.2, Frederick, MD, USA) calibrated against the space between two threads to the nearest 0.1 mm, and the mean of these two measurements was calculated for each implant. The measurements were recorded with reference to the implant axis.

Probing pocket depth (PPD) and bleeding on probing (BOP) were measured by a blinded operator (AD) with a periodontal probe (PCP-UNC 15, Hu-Friedy Manufacturing) at 6 and 12 months. Three vestibular and three lingual values were collected for each implant by the same dentist.