101 questionnaires (24%) were returned, consistent with other surgical surveys [², ^8-10]; 67 of these (16% of total sent) contained useful information, which form the basis of the results reported in this paper. The 67 respondents comprised of the following surgical groups: 50 orthopaedic surgeons, 14 neurosurgeons, 1 spinal surgeon and 2 unanswered/are not any of the above.

The majority of respondents (59%) were in favour of a simple device to aid PS placement; 37% were not in favour; 4% did not answer Question 1.

Respondents in favour of a simple device had a relatively high experience of malpositioning. The reduced risk of neurological/vascular complications, reduced incidence of pedicle breach, reduced X-ray exposure, increased confidence, increased accuracy, use as a training tool, use as a supplementary aid and an increased rate of correct screw placement were considered potential benefits. Emphasis was placed on using a simple device to correctly position screws in seriously deformed/degenerated spines. The device was also favoured for revision surgery in the thoracic and the cervical spine.

Respondents not in favour believed that imaging techniques and anatomical knowledge are sufficient aids to PS surgery. Other comments included: the devices not having proper track records; the belief that they already exist (but offer no advantage); the belief that navigation systems are cumbersome and expensive. It was also noted that, in many cases, the pedicle is breached deliberately. Furthermore, respondents explained that blind placement with the aid of a device could lead to the use of PSs by surgeons unfamiliar with the anatomy, and therefore unfamiliar with the clinical indications.

For Question 1, the respondent orthopaedic surgeons answered 61% Yes and 39% No; the respondent neurosurgeons answered 62% Yes and 38% No. No significant difference (p > 0.05) was found between the neurosurgeons and orthopaedic surgeons and whether or not they think there is a need for a simple device to aid PS placement.

The majority of respondents (55%) did not use any systems to aid PS placement; 2% used the systems specified in the question (Fig. 1); 43% used other systems (Fig. 2). The majority using “Other” systems (52%) used C-Arm X-ray image intensifiers.

Most respondents found the specified systems easy to use, safe and accurate (Tables 1). The 3D images provided by some systems, e.g. PRAXIM Medivision navigation system (PRAXIM Medivision SA, Grenoble, France), were considered a major asset. Common limitations were the associated learning time and increased surgery time, as well as high cost and the inability to deal with all spinal conditions.

Respondents either provided general or region-based descriptions/sketches for PS placement, which are described below.

The majority of respondents have inserted PSs in more than one region of the human spine. The proportion of respondents that have inserted PSs in the cervical spine was 14%; 28% have inserted PSs in the thoracic spine; 30% have inserted PSs in the lumbar spine; 28% have inserted PSs in the sacral spine.

Most respondents (78%) had experienced problems listed in Table 2; 16% did not have any problems with PS placement; 6% did not answer Question 6. The respondent orthopaedic surgeons answered 87.5% Yes and 12.5% No; the respondent neurosurgeons answered 71% Yes and 29% No. No significant difference (p > 0.05) was found between the neurosurgeons and orthopaedic surgeons and whether or not they had experienced problems with PS placement.

The percentage of respondents in favour of a simple device for a particular spinal region were: 24% for the cervical spine, 29% for the thoracic spine, 20% for the lumbar spine, 16% for the sacral spine; 3% did not think a device was necessary; 1% did not know; 7% did not answer Question 7 and/or had no view.

This question consisted of 8 sub-questions that provided the answers in Fig. (3). There was no clear consensus in the answers given; however, if the majority opinion is taken for each of the design factors in Question 8, then the overall preference was for a device that is multiple use, one-piece, hand-held, radiolucent, unilateral and uses the line of sight principle in traditional open surgery, with no preference as to whether the PS should be loaded into the device before or after entering the body.

Fig. (4) lists the problems encountered by the respondents whilst undertaking PS surgery in patients with an osteoporotic spine; these problems were already specified in the questionnaire and the respondents were asked to tick as many problems that were appropriate.

The proportion of respondents that answered Yes to Question 10 was 27%; the majority (70%) did not know of any osteoporotic bone screws; 3% did not answer Question 10.

Respondents identified the Biomet® Omega21™ Expandable Screw (Biomet UK Ltd., Bridgend, Wales) as being unsuccessful for fixation in osteoporotic bone and very difficult to revise. Some respondents noted the success of Ulrich Medical® articulated PSs (Ulrich GmbH & Co.KG, Ulm, Germany), the Synthes® Universal Spine System (USS) II anterior screws (Synthes Ltd., Welwyn Garden City, Hertfordshire, UK), and the Ulrich Medical® tango RS™ system (Ulrich GmbH & Co.KG, Ulm, Germany) for use in osteoporotic bone. Other successful screws for osteoporotic bone included hydroxyapatite-coated screws and hollow/cannulated screws to allow cement injection. One respondent noted that another approach “is to use long Schanz screws to engage the anterior cortex.” Some respondents believed that screws designed for osteoporotic bone were very expensive and relatively unsuccessful.

This question provided the opportunity for respondents to give any other opinions on PS surgery, which had not been previously covered in the questionnaire. Several responses consisted of statements that reinforced previous answers; these have been omitted from the following generalised list of final suggestions:

• Any proposed system should secondarily, not primarily, assist stability;
• Distal nail targeting devices should be considered;
• Pss may not have a long-term future.

The posterior-anterior approach for PS fixation is widely used. A large proportion of respondents have experienced problems arising from PS malpositioning (Table 2). The problems specified include breaching of the pedicle and vertebral body walls, pedicle fracture and various neurovascular impairments; these, and similar, problems have been identified previously [¹,²]. Neurologic complications are generally considered the most serious, although some studies have stated that malpositioned PSs do not lead to compromised strength or long-term neuro-vascular impairment [³]. In a review of patients with PS plate fixation to the lumbar spine [¹¹], nerve root impairment was noted in six out of fifty-seven patients, with all but two finally improving to normal. This suggests that only a minority of malpositioned PSs cause severe complications.

The majority of respondents think there is a need for a simple device to aid PS placement, especially in the cervical and thoracic spine. The use of PSs in these regions is becoming more common [¹²]. In a Japanese survey, complications occurred in the cervical (10.3%), thoracic (13.5%) and lumbar (6.7%) spine [¹²]. In a meta-analysis of PS placement, there was no advantage in using the existing navigation options in the thoracic spine [⁵]. It is important to note that surgeons should first, and foremost, rely on their anatomic and technical knowledge with regard to the accurate placement of PSs; any pedicle aid should only be considered as a supplementary tool during surgery.

Part of this questionnaire’s aim was to determine the design criteria for a simple device to aid PS placement. The majority of respondents preferred a multiple use, one-piece, hand-held, radiolucent, unilateral pedicle aid that uses the line of sight principle in traditional open surgery, with no preference as to whether the PS should be loaded into the device before or after entering the body. Only 14% of respondents opted for a device that uses computed navigation. Also, 55% of respondents do not use any systems to aid PS placement. These results are consistent with many surgeons being unhappy with computed navigation systems [^13-15].

The majority of respondents experience multiple problems with PS fixation in the osteoporotic spine, notably poor purchase, loosening and pullout. Several studies have reported on the increased pullout resistance of screws augmented with polymethylmethacrylate and various other bone cements in osteoporotic bone [¹⁶, ¹⁷]. However, 22% of respondents in this study still faced problems with the cement augmentation of PSs, which have been described previously [¹⁸].

Few successful screw designs appear to exist for osteoporotic bone; the majority of respondents were unfamiliar with screws specifically designed for osteoporotic bone. The reason for this could be that screw fixation is not always indicated for use in osteoporotic bone, although it has been performed [¹⁹]. Screw augmentation techniques involving injection of bone cement through a perforated screw may overcome some problems [²⁰].

The 67 surgeons participating in this study were from a list of 422 surgeons provided by Surgicraft Ltd. (Redditch, UK). Whilst this list may introduce a potential selection bias, the authors believe that the list accounted for the majority of surgeons known to perform spinal surgery in the UK and Ireland. However, this list was not able to specify which surgeons actually perform PS surgery, which might explain the 24% response rate and why only 16% of the total questionnaires were completed and returned. The authors acknowledge that a higher response rate might have resulted in a more representative group of surgeons. As with all questionnaire studies, the data from the current study is limited by the response rate; it remains unclear whether these data truly reflect the opinions and preferences of the entire community of UK and Irish spine surgeons performing PS surgery. However, other surgical survey studies have generated responses in the range of 15-49% [², ^8-10] and many of these studies have used a much smaller survey group [², ⁸, ⁹].