To verify whether digital three-dimensional model-assisted internal fixation can obtain better effects in the repair of complex long bone fractures compared with conventional internal fixation.

History and current related studies The internal fixator is difficult to completely fit the bone surface at the site of complex long bone fracture, which reduces the firmness of fixation. From a safety point of view, the establishment of a digital three-dimensional measurement model is undoubtedly a very reliable and intuitive way. Digital three-dimensional models are established after CT scan, and these models can greatly reflect the site and degree of fractures, and reduce anatomical parameters, with high degree of simulation and feasibility.

Adverse events

1. The investigators will record adverse events, including screw loosening and falling off, steel plate fracture and bending, joint stiffness, delayed healing or non-healing, peripheral nerve injury and infection.
2. If severe adverse events occur, investigators will report details, including the date of occurrence and measures taken to treat the adverse events, to the principle investigator and the institutional review board within 24 hours.

Data collection, management, analysis and open-access

1. Data collection: Case report forms with demographic data, disease diagnosis, accompanying diseases, drug allergy history, and adverse events will be collected, processed using Epidata software (Epidata Association, Odense, Denmark), collated, and then recorded electronically by data managers using a double-data entry strategy.
2. Data management: The locked electronic database will be accessible and locked only by the project manager, and will not be altered. All data regarding this trial will be preserved by the Affiliated Hospital of Qinghai University, China.
3. Data analysis: The electronic database will be statistically analyzed by a professional statistician who will create an outcome analysis report that will be submitted to the lead researchers. An independent data monitoring committee will supervise and manage the trial data, ensuring a scientific and stringent trial to yield accurate and complete data.
4. Data open access: Anonymized trial data will be published at www.figshare.com.

Statistical analysis

1. Statistical analysis will be performed using SPSS 19.0 software, and will follow the intention-to-treat principle. Normally distributed measurement data will be expressed as mean ± standard deviation, and minimums and maximums. Non-normally distributed measurement data will be expressed as lower quartile (q1), and median and upper quartile (q3).
2. Fisher's exact test will be performed for comparison of total efficacy and the incidence of adverse reactions preoperatively and 6 months postoperatively. Mann Whitney U test will be used to compare Harris score, Visual Analogue Scale, operation time and hospital stay between the two groups. The significance level will be α = 0.05.

Inclusion Criteria:

• Long bone fractures in the four limbs, such as middle segment of humerus, femur, ulna, radius, tibia and fibula; fractures are not related to the articular surface
• Diagnosed with complex long bone fractures via X-ray
• Admission within 6 hours after injury

Exclusion Criteria:

• Pathological fractures with vascular and nerve injuries
• Split fractures of the humeral head
• Neer IV fractures
• Humeral head compression area greater than 40%
• Glenohumeral dislocation
• Conscious disturbance, cerebral infarction, tumor, serious medical complications, such as heart, lung, kidney failure, severe hypertension, diabetes and blood diseases
• History of elbow joint dysfunction or shoulder joint disease
• Unwilling to sign the informed consent