Nonunion age & gender distribution: In this study, nonunion was noticed mainly in the age group 30-39 year (33%), next to this was the age group 20-29 years (28%) while its lowest incidence was in the age group 70-79 (2%), this may be related to the small number of old patients who sustained fracture compared to young people who participate in everyday life activities. The mean age was 36.5 years old, which differs from many studies^2-4. The mean age at the start of treatment was 46.7 years, but goes with Bumbaširević⁵ were the mean age was 30.4 years.

 Gender distribution: It's quite common to find nonunion in males more than females, which was documented strongly in this series by 81% males compared by 19% females out of 84 patients, this may be related to female's life style with all its restricted association as compared to males in our culture. This was supported by the study of Joon-Woo kim³ (18 patients, 13 men and 5 women), ŁukaszSzelerski⁴ in which there was 102 patients (76 men and 26 women), and Bumbaširević⁵ (29 males and 1 female), this is may be explained by the life style in our society as women activities outside home is greatly restricted.

 Fracture pattern: The most common fracture pattern presented in this study was comminuted fracture which was verified in 49 patients (57%), followed by transverse fracture in 25 patients (30%), then oblique fracture in 9 patients (11%) and spiral in one patient (1%), this was related mainly to the severity of the injurious mechanism. These results are supported by the study of Karladani AH⁶ were he recorded 52 comminuted fractures out of 104 tibial shaft fractures and against the study by Cognet JM⁷ were spiral fracture was the most common type presented.

 Mechanism of fracture: The commonest causative mechanism for fracture in this study was road traffic accident in 48 patients (57%), followed by bullet injury in 25 patients (30%) and one patient fall from height. In spite of severity of injuries which caused nonunion, but still closed fracture had high incidence than open fracture, this may be due to changing closed to open fracture by surgical intervention. Closed fracture was recorded to be 57% and open fracture 43% this is supported by the study of Karladani AH⁶ as he recorded 22 open fracture out of 104 patients with tibial shaft fracture, Malik MH⁸ recorded 30 open fractures out of 214 long bone fractures, and Nabil A. Ebraheim⁹ of the 169 nonunion reviewed, 133 (78.7%) were closed fractures and 36 (21.3%) with open fractures.

 Nonunion and bone involved: Tibia was the most common bone involved in nonunion in this study seen in 39 patients (46%), femur 30 patients (36%), and humerus 15 patients (18%), this may be related to the peculiar character of tibia regarding its anatomical position. Results were not similar to the study by Malik MH⁸ who recorded 122 femurs, 75 tibial fractures and 17 humeri and also does not correlate with Arsalan H¹⁰ who estimated 12 femoral fractures and 10 tibial fractures.

 Types of nonunion: In this study, patients with established nonunion from other orthopedic surgeons who were treated by different surgical modalities mainly by changing closed to open fractures, there were 45 patients (54%) developed hypertrophic nonunion while the remaining 39 patients (46 %) had atrophic nonunion, this may be explained according to the type of fixation used for the patients. These results goes with the study by Markus Rupp¹¹ as 162 participants in the study, 99 patients (61.1%) suffered from a hypertrophic nonunion, while 63 patients (38.9%) had an atrophic nonunion.

 Nonunion and chronic disease: Diabetic patients come on the top of the list of those who developed nonunion, 5 patients (6%), they were on oral hypoglycemic drugs with poor dietary control. Most of them had high fasting blood sugar, this contributes in their intolerance to operation and all of them end up with nonunion, this was statically significant. There were; one patient asthmatic, one anemic and one with chronic renal failure, all of them were still set with nonunion due to local and systemic effects of the disease.

 Smoking and nonunion: In this series we recorded 43 patients (51%) as cigarette smoker compared with 41 patients (49%) non-smokers, 57 patients ended with nonunion 31 were smokers, while among those 27 who achieved union, 15 of them were nonsmokers and the remaining 12 patient were smokers, this is reinforced by Jacques Hernigou¹² who suggested solid association between smoking and delayed bone healing, and Dinah¹³ who advice to stop smoking one week before surgery which interfere with process of coagulation and fibrinogen level.

 Causes of nonunion: The causes that stand behind nonunion had been studied systematically to find out the possible correlation between nonunion of the long bone and its causative agents, as we had classified the causes into 3 main groups: Group I:    The cause is one factor (17 patients, 20%). The possible causes were insecure fixation 8/17, then infection 6/17, and open fracture in 3/17 patients (4%). Group II: This is consisting of 2 likely main causes: this was the biggest group represented 55 patients (66%). It is classified into infection and open fracture 11/55 patients, infection and insecure fixation 10/55 patients, insecure fixation and soft tissue interposition 10/55 patients, and infection and comminution 9/55 patients, infection and open fracture 8/55 (10%), and comminution and insecure fixation 7/55 (8%). Group III: Consists of combination of three potential causes represented by 14% (12 patients) here the main collection was infection, open fracture, and comminution 5/12 (6%), also insecure fixation+infection+ soft tissue interposition 5/12 (6%) and the last was combination of infection+ segmental fracture+insufficient time for immobilization 2/12 (2%). So we can identify that the main factors which stand behind nonunion of the long bones in this study which were infection, insecure fixation, comminution, and soft tissue interposition. This does not go with the study by Milind Madhav Chaudhary¹⁴, who blamed recalcitrant infection, complex deformities, sclerotic bone ends, large bone gaps, shortening, and joint stiffness. But goes with study by Mills¹⁵ were a total of 31% of patients had a single attributable cause, 55% had two causes, 14% had three causes and 1% had all the four.

 Treatment of primary fractures: In this series, initial treatment which were done by other orthopedic surgeons (the first surgeon who treated those patients) were:  Conservative group: Those patients were subjected to conservative procedure in form of casting, skin or skeletal traction, this method representing 25% (21 patients).  Operative group: those patients who underwent operative interference like frequent session of wound excision, with essential procedure as external or internal fixation with or without bone graft was estimated to be 39% (33patients). Delayed Operative group: this group consists of patients who had conservative treatments to start with then changed to operative treatment represented 36% (30 patients).

 As we notice from the above figures, the main bulk was surgical interference which may precede the misery of fracture nonunion by changing the closed fracture to open one which reflect the rush decision for treatment of fracture. This is supported by study of Healy WL¹⁶ who exposed his patients to both conservative and operative treatment of 26 humeral shaft fracture, but does not go with studies by Court-Brown-C-M¹⁷ in a review study of 459 patients, tibial fracture nonunion treated operatively (primary reamed nailing and Nabil A. Ebraheim⁹  of the 169 cases reviewed, the initial treatment was ORIF with plating in 108 (63.9%), IMN in 24 (14.2%), external fixation in 13 (7.7%) and unknown in 20 (11.8%). The remaining four initial treatments included one instance of multiple Steinmann pins, one allograft, one plate and cast and one non operative. So our option for choosing both operative and conservative methods based on patient condition, surgical skills, confidence, operative facilities and our general situations.

  Definitive treatment of Nonunion: In this study the ideal target was to achieve union in the non-united fracture of the long bone, this was performed by either operative treatment like internal fixation and autogenous bone graft from ileum or tibia, this was done in thirty patients(36%). External fixation and bone graft was performed in eighteen patients (21%). External fixation alone in fourteen patients (17%). Internal fixation alone in seven patients (8%). Conservative treatment was applied in 9 patients (11%) (Since either refusal for surgical interference or existence of comorbidities which made them inoperable). Delayed operative treatment was performed in 15 patients (18%), so we can distinguish 89% of our patients treated by operative and delayed operative, with just 11% conservative decisions. Here we can touch the big hope of our orthopedic surgeons and there huge desire to get from the sadness of nonunion by achieving this large number of surgical interference while those who treated conservatively either unfit for general anesthesia because of medical precautions or they refuse the operation. This does not go with study by Shih CH¹⁸ who treated 121 cases of femoral neck fracture by Knowles pins, study by Rijnberg WJ¹⁹ were he used fresh autogenous iliac bone graft for 48 tibia.  In Nabil A. Ebraheim⁹  study, out of the 169 nonunions, the specific initial treatments were dynamic condylar screws in 21 (12.4%), anterograde nailing in 18 (10.7%), angled blade plates in 16 (9.47%), retrograde nailing in six (3.55%), and locking compression plates in five (2.96%). In addition, 30 (17.8%) received bone grafts during the initial treatment.

  Follow up time: Group 1: extended from 1-12 weeks (3 months) was included 36% (30 patients). Group 2: from 13-24 week (3-6 months) was included 43% (36 patients). Group 3: those extending from 25-36 week (6-9 months) included 11% (9 patients). Mean time of follow up was 12.4 week, so from these figures above we can recognize that the period of follow up in the group 1 not enough to regard the chance of treatment well established to decide this treatment option not successful (according to the definition of nonunion), so this group needs further follow up time, especially if we look to the follow up time in other studies like Bumbaširević⁵. Bony union was achieved at the tibial docking sites after a mean of 4.5 months in 29 patients (97%), Yin, P²⁰. The mean follow up after removal of the apparatus was 23.12 months (14–46 months), Shi CH¹⁸ 32 months, Rijnberg WJ¹⁹ 6-29 months.

 Conclusion:

 The causes that stand behind nonunion of the long bone in this study were; infection, insecure fixation, comminution, and soft tissue interposition. Internal fixation with autogenous bone graft proved to be the vital step to solve the obstacle of nonunion. The wisdom of clever orthopedic surgeon who recognizes the fracture that should require interference conservatively or operatively from the start is important step to prevent nonunion.

   In order to accomplish union we advise to have healthy viable bone ends, plentiful bone graft, with faultless suitable rigid fixation.

 The main complications which façade in this study were failure to complete union and recurrence of infection in addition to psychosomatic upset of the patients because of dissatisfaction.