At the time of the bone augmentation procedure, two patients (14%) in the PRF group and one (8%) patient in the control group were smokers. Patients were partially edentulous due to trauma (n = 22), agenesis (n = 3) or marginal periodontitis [2]. Two patients were unavailable for the final follow-up. The referring dentist followed the non-attenders, and telephone interview revealed no subjective or objective complications of either the implants or the crowns.

Bone augmentation procedure

In the PRF group, the PRF membranes were prepared before surgery. A venous blood sample (80 ml, distributed in eight 10-ml glass-coated plastic tubes) was collected via puncture of a vein in the cubitial fossa and centrifuged using a Duo Quattro centrifugation device with a 40° rotor angulation with a radius of 88 mm at the clot and 110 mm at the max (A-PRF 12, Process, Nice, France) according to a previously described method [29,30,31]. We followed the manufacturer’s recommendations at the initiation of the study in 2015, producing the membranes using a protocol of 1300 RPM for 14 min (RCF-max = 208 g). The tubes containing the centrifuged blood were placed to rest for approximately 25 min to give the fibrin clot a firmer consistency before collecting it for the final membrane preparation, as previously recommended [31, 32]. The bone at the recipient and donor site were planned to be covered by three membranes, respectively (a total of six PRF membranes), while the last two tubules were held in reserve if for some reason clotting of the PRF matrix in the tubule was inadequate.

Bone graft harvesting

In local anaesthesia (Marcain®adrenalin, 5 mg/ml + 5 μg/ml, AstraZenca, Cambridge, UK) and via an intraoral approach, the lateral aspect of the posterior part of the mandibular corpus was exposed with a mucoperiosteal flap using a standard technique, as previously described [9, 14, 33] (Fig. 1a, b).