The results should not be discounted, but interpreted rather carefully. Previous studies have shown a higher failure rate with single anchors in the maxilla. The reasons cited include: greater off-vertical implant placement angle due to ridge pitch in the maxilla compared to the mandible; this results in greater off-axial loading of the implant and accelerated wear of the resilient insert and the abutment. The second cause of failure is of course the relatively soft bone and typically minimal cortical layer. Next, the maxillary prosthesis is subjected to greater lateral forces than a mandibular as the occlusion and tooth arrangement is typically buccal to the lower.
There are limitations to the study: The study comments on absence of implant mobility or pain associated with them; this is really an inconsequential finding as implants are not mobile unless they are ready to fall out. Tenderness to percussion only seems to occur when the implant is in failure mode. Almost half the implants demonstrated bleeding on probing. is seldom a sequelae unless soft tissue is inflamed. The actual sample size is 12 patients with 2+ years. Bone loss is noted but there isn't a reference to calibration of measurement or baseline. In other words, the information is added to our continuing study of implant therapy but is too limited and somewhat unstructured to greatly influence one's treatment protocol.
With reference to the idea of placing three implants in the maxilla, placing one in each posterior as far back as possible may not be a good choice, as it forms a long triangle, the legs of which are greatly inside the loading area of the canine-second bicuspid chewing zone. This will create a buccal lever arm which can rock the prosthesis. Also, the bone is typically less reliable in the molar area.
I do not treat with solitary, maxillary anchors in edentulous patients, preferring to splint them with a rigid structure to prevent a bending moment and accumulation of strain at the implant site.