Postoperative axial deviations and delayed unions are possible complications after flexible intramedullary nailing (FIN). The goals of this study were to determine if a correlation exists between occurrence of the above complications and the ratio of the diameter between nail and medullary canal [nail diameter (ND)/MCD ratio], to study the interobserver variability in the measurement of MCD, and to define a threshold to be respected to optimize the results.
Eighty-one consecutive diaphyseal fractures treated by means of FIN were evaluated. The ND/MCD ratios were determined by 2 independent observers. Axial deviations were defined as 5-degree angulation or more observed before bone union. Absence of bone union at 3 months was considered as delayed union. Statistical analysis was made for interobserver variability of MCD, dependency between occurrence of complications and ND/MCD ratio and eventual confounding variables (age, weight, sex, and fracture location).
Of the 81 fractures, 14 presented with an axial deviation and 3 with a delayed union. Interobserver variability of MCD diameter was excellent (intraclass correlation: 0.96). Occurrence of the above complications was significantly associated with a low ND/MCD ratio (P=0.0002) but with none of the examined confounding variables. Receiver operating characteristic analysis showed absence of complications with a ND/MCD ratio >35% with a sensitivity of 100% and specificity of 89%. Related with the MCD measurements variability, a safe threshold of 40% can be suggested.
In FIN, ND>40% to the MCD should be chosen to avoid complications, besides respecting the technical principles. Measuring the medullary canal diameter in order to choose correct nail size is reproducible between different observers. In adolescents with a medullary canal diameter of >10 mm in femur or tibia fractures, alternative methods of osteosynthesis than FIN should be considered.
This work statistically confirms that a ND/MCD ratio of >40% must be respected to avoid some complications in FIN.
Level of Evidence: