Material and method
The instrument : In the H-device two similar parallel intramedullary nails can be fitted. This device enables insertion of one nail within the medullary cavity for the fixation of the fracture. The second free nail outside, parallel and laterally helps for the location and placement of both distal cross screws (Fig. 1.).
The experiment: In eleven sound cadaveric femurs experimental GK nailing was performed. After reaming, nails were inserted from 12 mm to 14 mm in diameter (each nail 1 mm smaller in diameter the reaming was done). The length varied from 38 to 44 cm. The cross screws were inserted using the H device.
The clinical application: After obtaining informed consent, the H device was used during closed intramedullary nailing in 15 patients with fracture of the middle third of the femur. The skeletal traction was exerted via a Steinmann pin inserted through either the femoral condyles or the tibial tubercle. A small lateral incision was made to expose part of the distal femur for the insertion of the cross screws. Using a guide sleeve, a 4 mm Steinmann pin was passed through the free (outer) nail, the lateral cortex, the nail hole and the inner cortex. Consequently a 5 mm drill bit was used to drill through both cortices and a 6 mm drill bit to overdrill the near cortex. Image intensifier of short duration was used to confirm the correct position of the distal screws.

Results
In all experimental cases the insertion of the distal cross screws was possible with the first effort. However in six out of eleven experiments the advancement of the nail towards the distal shaft hit the anterior cortex. In all these six cases the insertion of the distal screws was precise. In the rest five cases the holes of the nails were not absolutely in line, but did not obstruct distal cross screw insertion.
In 9 out of 15 clinical cases the insertion of the distal screws was successful without difficulties. The patients were thin, and the application of the H device was proved easy. In two cases several efforts were needed because the nails initially were not fitted in the H device correctly. In four cases the patients were obese and the depth in the site of the insertion of the nail in the femur did not allow the application of the H device. The distal screws were inserted using the targeting device on image intensifier.

Discussion
The methods up to now successfully used for the insertion of the distal nail cross-screws, need the use of an image intensifier with the known dangers of radiation [5]. At various procedures during nail insertion the radiation by means of an image intensifier used, may be regarded as excessive. Once is used to assess reduction, a second time may be used to confirm correct penetration of the owl at the trochanteric tip, at a third time is again used to confirm the placement of the guide wire within the medullary canal and a possible fourth time to confirm the satisfactory distance at the lower end. Distal cross-screw placement appears to be time consuming procedure.
Nails containing an open section over most of their length frequently undergo torsion on insertion. This changes the relationship between the plane of the distal transfixion screw tunnels and the top of the nail and prohibits correct alignment of the proximal mounted target device when it is reattached. Because these guides often extent for a distance of more than 40 cm between their point of the proximal attachment and the distal holes, the problem of malalignment is further aggravated by the tendency of the guide to sag toward the floor when used in the supine position on the femur or tibia.
The H device is a simple one and was effective for the targeting of the distal screws in all experimental samples. The sagging effect by using a second nail was proved insignificant, and that was obvious in the 6 experimental cases. Reaming by 1 mm more the diameter of the nail to be used reduces the likelihood of increasing the comminution of the fracture during the insertion of the nail. The resistance to torsion supplied by the intramedullary nail is not decreased. Although some deformation of the nail is expected to occur, this is minimal with overreaming and the device was precise.
The ideal device for the targeting of the distal locking screws could be planed if the possible deformation of the nail can be predicted. Despite of this, further investigation to establish the usefulness of the H device and for further improvement of the device may be needed. The effort to place the distal locking screws without radiation must continue.

Acknowledgments. Dr. George Kontakis perceived, invented and designed the H device.

References

1. Grosse A (1981) Manual of osteosynthesis for femoral and tibial shaft fractures. Kiel, West Germany: Howmedica International Inc. 2. Johnson KD (1992) Femoral shaft fractures. Skeletal Trauma, Edited by Browner et al.,Philadelphia: W.B. Saunders Company, pp 1525-1641.
3. Muller M E , Allgower M, Schneider R. Willeneger H (1991). Manual of Internal Fixation. 3rd Ed. Berlin, Heidelberg: Springer-Verlag: pp, 291-365
4. Russell T and Taylor J (1986) Interlocking intramedullary nailing in the femur: Current Concepts. Semin. Orthop 1: 217-231.
5. Sanders R, Koval KJ, DiPasquale T, et al (1993) Exposure of the orthopaedic surgeon to radiation. J Bone and Joint Surg 75A:326-330