• NAVC Brands
Orthopedics, Recovery & Rehab, Rehabilitation

Orthopedic Follow-Up Evaluations: Identifying Complications

Complications can occur during the recovery period following an orthopedic procedure. The author discusses bone healing; delayed unions, nonunions, and malunions; osteomyelitis and infection; bone implant construct failures; and fracture disease in this comprehensive article.

Orthopedic Follow-Up Evaluations: Identifying Complications


Duane A. Robinson, DVM, PhD

Orthopedic procedures, whether performed on an elective or urgent/emergent basis, are common in small animal veterinary patients. In many instances, definitive treatment occurs at a referral center, with postoperative follow-up taking place at the primary care clinic.

During the recovery period, the primary care veterinarian needs to be able to identify complications in order to intervene as soon as possible. This article reviews basic guidelines and provides tools regarding identification of complications that may occur in adult patients recovering from orthopedic procedures.


Stabilization techniques for rupture of the cranial cruciate ligament are undoubtedly one of the most common orthopedic procedures performed on an elective basis; these procedures include:

  • Tibial plateau leveling osteotomy (TPLO)
  • Tibial tuberosity advancement (TTA)
  • Lateral imbrication suture (lateral femoral fabellotibial suture).

The TPLO and TTA procedures require an osteotomy and normal bone healing for a successful outcome.

Fracture repairs are a group of orthopedic procedures that are particularly challenging because normal healing is dependent on a multitude of factors. For example:

  • What was the fracture configuration?
  • What repair method was used?
  • What is the patient’s signalment?
  • Are any comorbidities present and, if so, will they affect fracture healing?


In veterinary medicine, the majority of fractures heal via stabilization of fracture fragments by development of a callus, followed by endochondral ossification, which results in formation of new bone. More specifically, healing of bone can occur via direct (primary)—divided into gap or contact healing— or indirect (secondary) bone healing.1,2

Direct Bone Healing

Clinically, direct (primary) healing occurs via a combination of contact and gap healing,1and requires rigid internal fixation.

Contact direct healing describes situations in which:

  • Fracture/osteotomy surfaces are in direct contact
  • Interfragmentary motion is not present
  • Fragments are usually under compression.

Gap direct bone healing occurs when an interfragmentary gap of < 1 mm is present.

Indirect Bone Healing

Indirect (secondary) bone healing is common in patients with nonreconstructable fracture configurations, in which biologic fixation (biological osteosynthesis) methods are used. These methods minimize the extent to which the fracture site/callus and its blood supply are approached and disturbed (Figure 1). Examples of biologic fixation methods include the use of minimally traumatic surgical approaches (eg, closed alignment using an intramedullary pin), external fixator systems, and cancellous bone grafting.


Figure 1. Radiograph of comminuted radius and ulna fracture at time of injury (A) and approximately 8 weeks later, after removal of failed external fixator (B). This fracture is an example of a delayed union; note extensive callus formation. While fracture is not healed, it also provides an excellent example of indirect (secondary) bone healing. Courtesy UC—Davis VMTH

Under indirect bone-healing conditions, immediately after fracture occurrence, bone union begins by:

  • Accumulation of blood from periosteal, endosteal, and marrow sources, which forms a fracture hematoma
  • Development of a soft tissue envelope, which surrounds the fracture site and delivers the needed blood supply to the healing bone until the endosteal, periosteal, and marrow blood supply sources are reestablished.

The phase described above is the reactive phase, which includes the inflammatory and granulation phases, and it is followed by the reparative and remodeling phases. While the initial phases are relatively short lived, the remodeling phase may continue for years.1

Interfragmentary Strain

Bone healing is also influenced by interfragmentary strain—a measure of the deformation that occurs in the area between fracture ends (Figure 2). The smaller the gap between the fragments, the greater the amount of strain that will occur with a given degree of deformation. Bone formation requires low strain levels.

Orthopedic implants (ie, plate and screw, interlocking nail) are designed to decrease the amount of deformation at the fracture ends. Limiting the amount of deformation decreases interfragmentary strain, which is important because, in order for bone formation to occur via gap direct bone healing, the strain in the fracture gap must be less than 2% (Figure 2).1

Figure 2

Figure 2. Strain is the change in gap length divided by original gap length. This graphical depiction of strain demonstrates how osteoclastic resorption increases the initial fracture gap and, as a result, decreases interfragmentary strain to a point where bone will form.

The clinical manifestation of this concept is somewhat counterintuitive. For example, one method by which the body reduces strain is to increase the original gap length by osteoclastic resorption.1,2 Therefore, during an early (ie, 6—8 weeks after repair) follow-up radiograph, the fracture gap may be wider than it was initially, but this is considered normal (Figure 3). However, if such a gap is noted during subsequent radiographic evaluations, it is considered a complication in healing.

Figure 3

Figure 3. Immediate postoperative radiograph (A) and another radiograph 8 weeks after surgery (B); note that the gap in the ulna (blue arrow) has increased in size, which is normal at this stage of healing.

As it heals, an unstabilized fracture or a fracture that is being addressed using indirect (secondary) bone healing can decrease strain by both formation of a fracture callus (which stabilizes the fracture and, therefore, decreases deformation) and osteoclastic resorption.1,2


Despite veterinarians’ best efforts, complications occur in bone healing, resulting in increased morbidity for patients and increased economic burden for clients.3 Examples of such complications include:

  • Failure to achieve functional clinical union of fragments (Tables 1 and 2)
  • Osteomyelitis
  • Bone-implant construct failure (construct can fail due to problems with the implant or bone, or implant attachment to the bone)
  • Fracture disease (eg, atrophy, stiffness, adhesions).

For most orthopedic cases, follow-up clinical evaluations with radiographs are generally recommended 6 to 8 weeks after surgery; further follow-up depends on the specific needs of the patient. However, in situations with precarious fixations or concern regarding client compliance after surgery, radiographs may be required 4 weeks after surgery.

Screen Shot 2015-05-14 at 12.10.26 PM

  1. For example, if converting from external fixator to plate, remove all fixator components
  2. Surgery involves (1) opening medullary cavity; (2) en bloc removal of affected bone ends, compression of ends using plate and screw fixation (optimal); (3) autologous and/or allogeneic bone graft (essential); (4) if available, consider rhBMP-2

Screen Shot 2015-05-13 at 2.10.46 PM


These complications (Tables 1 and 2) tend to occur when the mechanical and biological environment necessary for bone healing is not optimal.

Biologically, it is essential to minimize disruption of the natural bone healing process by:

  • Minimizing dissection
  • Preserving surrounding soft tissue structures
  • Maintaining the fracture hematoma.

From a mechanical point of view, the aim is to provide:

  • Proper alignment of fracture fragments
  • Adequate stability at the fracture site such that healing (bone formation) can occur (see Interfragmentary Strain).1

A mnemonic has been developed to outline fracture assessment (Table 3).

Screen Shot 2015-05-13 at 2.18.22 PM

Delayed Union

Diagnosis of a delayed union can be challenging; by definition, it is a fracture that has not healed in the typical time frame for a given fracture in a given animal.4,5 Thus, diagnosis is dependent on the knowledge of what is typical for a particular fracture (Table 4).

Screen Shot 2015-05-13 at 2.22.18 PM

Causes. Causes of delayed unions can be classified as mechanical, biologic, or both.

  • Biologic: For osteotomy procedures, such as TPLO or TTA, a delayed union is most often due to biologic causes, such as periosteal damage, infection, and impairment of local blood supply.4,5
  • Mechanical: Mechanical causes relate to excessive fracture gaps when bone is lost during trauma or during surgery, inadequate immobilization or immobilization for an insufficient period of time, or interposition of soft tissue structures between fracture ends (Figures 1 and 4).4,5

Figure 4. Radiographs of an open comminuted segmental fracture of the right tibia (A and B) showing repair with an interlocking nail and screws. Four months postsurgery, fracture had not healed and one screw is bent (*) (C and D). Five months postsurgery the fracture had healed (E and F). Courtesy UC—Davis VMTH

Figure 4A

Figure 4A

Figure 4B

Figure 4B

Figure 4C

Figure 4C

Figure 4D

Figure 4D

Figure 4E

Figure 4E

Figure 4F

Figure 4F

Patient comorbidities are also an important consideration: advanced age, concomitant corticosteroid administration, and metabolic disease (eg, hyperadrenocorticism) can play a role in fracture healing.

Evaluation. If a fracture or osteotomy is not healing in the expected amount of time, careful evaluation of implant construct and thorough patient evaluation are necessary.

The following findings are consistent with instability, and intervention is usually advised:

  • Broken implants (Figures 5 and 6)
  • Radiolucency associated with bone/implant interface (Figure 7)
  • Pain on palpation of the fracture site
  • Increasing lameness.

Figure 5. Open radius/ulna fracture in dog repaired with hybrid external skeletal fixator combining a ring and linear components: 1 month after repair (A) and approximately 8 weeks after repair (B). Note that one of the wires is broken and a portion is missing (*). Courtesy UC—Davis VMTH

Figure 5A

Figure 5A

Figure 5B

Figure 5B

Figure 6

Figure 6. A dog in which a carpal arthrodesis was performed 8 months previously; note the broken screws (*), likely related to excessive mechanical stress associated with this procedure. Courtesy UC—Davis VMTH

Figure 7. Lateral and craniocaudal radiographs of right humerus of a dog that sustained a humeral fracture 4 months previously. The fracture was repaired with an external fixator, threaded intramedullary pin, and single cerclage wire. The cerclage wire appears to have untwisted and is in the fracture site (A and B). Particularly in A, an area/ring of lucency (arrowhead) is apparent around the transcondylar pin, which is consistent with a loose implant. The significant periosteal reaction on the medial aspect of humeral condyle (*) is likely a result of motion but could also result from infection.

Figure 7A

Figure 7A

Figure 7B

Figure 7B

A key radiographic finding that differentiates a delayed union from a nonunion is the absence of sclerotic bone at fracture ends in patients with a delayed union.

Infection. The following can be noted if infection is present: excessive periosteal reaction (Figure 7), radiolucency associated with implant/bone interface, draining tracts (Figure 8), sudden onset lameness, and pain associated with implant or fracture site.

Figure 8. Surgical site in dog that has healed from TPLO, demonstrating drainage that could be associated with implant-associated infection.

Figure 8A Figure 8B

If the construct is otherwise stable, a fracture will heal despite infection. While deep percutaneous aspirates can be valuable in identifying the microbial organism and susceptibility pattern, they must be interpreted carefully because contamination with skin organisms during sampling is common.

While the definitive diagnosis of a delayed union can be difficult, suspicion of an inadequate biologic or mechanical environment warrants prompt intervention.4


A nonunion is characterized by failure of bone healing, cessation of osteogenic activity at the fracture site, and required surgical intervention to achieve a functional outcome.5

Classification. Nonunions are further divided into:

  • Viable: Hypertrophic, moderately hypertrophic (Figure 9), and oligotrophic
  • Nonviable: Dystrophic, necrotic (Figure 10), defect, and atrophic (Figure 11).

Figure 9. Comminuted, with one large fragment, mid diaphyseal fracture of left humerus with proximal, lateral, and cranial displacement of distal fragment (A); repaired with locking plate and intramedullary pin. Radiographs were taken 6 months postsurgery; note the persistent fracture line, moderate nonbridging callus, and 2 broken screws (most distal screws); moderate hypertrophic nonunion demonstrated (B). Original repair was revised by debriding fracture site, placing an autograft, and stabilizing with dynamic compression plate.

Figure 9A-1

Figure 9A-1

Figure 9A-2

Figure 9A-2

Figure 9B-1

Figure 9B-1

Figure 9B-2

Figure 9B-2

Figure 10. Example of necrotic nonviable nonunion: Radiographs from dog that underwent left TPLO procedure 8 months prior to presentation; dog developed a methicillin-resistant Staphylococcus species infection. The TPLO plate was removed 5 months postsurgery. At this point, dog was nonweight-bearing in left pelvic limb. The proximal tibial fragment was palpably unstable, severe muscle atrophy was present in entire hindlimb, and draining tract was noted along medial aspect of stifle.

Figure 10A Figure 10B Figure 10C

Case Example: 

Delayed/Nonunion Complications
Delayed/nonunion complications commonly occur in small/toy breed dogs with fractures in the distal 1/3 to ‘+ of the radial diaphysis (Figure 11).

Morphometric studies have demonstrated a propensity for radial fractures in toy breeds compared with large breed dogs; the radius of toy breed dogs also has a decreased vascular supply compared with that of larger breeds.6These unique mechanical and biologic properties likely contribute to the high rate (83%) of malalignment or nonunion complications when these fractures are treated with external coaptation alone.7,8

This emphasizes the need for adequate apposition and rigid fixation (eg, bone plate or external skeletal fixator) with preservation of the blood supply during fracture repair. In essence, a biological approach to the repair is advocated.

Figure 11

Figure 11. Example of atrophic nonunion: Note that ulna has virtually disappeared and very little radius is present. Despite stable repair with plate, the biological environment was not sufficient to support healing of this fracture.

viable nonunion often has an adequate blood supply and biologic environment but lacks sufficient mechanical stability, while a nonviable union is characterized by its avascular and biologically inactive environment.

Additional Causes. In addition to impaired blood supply, a nonunion can also occur secondary to:4

  • Technical failures during the repair (Figure 7)
  • Bone loss as a result of injury or surgery
  • Devascularization of fragments during surgical approach and dissection
  • Infection (Figure 10)
  • Instability (eg, mismatch of implant to bone stiffness)
  • Poor fracture reduction (eg, inappropriate choice of implants) (Figure 12)
  • Neoplasia.

Figure 12. Example of technical error that led to development of nonunion; the pin and cerclage were not sufficient to stabilize this fracture.

Figure 12B Figure 12A

Clinical Signs. Clinical signs of a nonunion can be variable, but common signs include:

  • New, persistent, or worsening lameness
  • Muscle atrophy and stiffness
  • Palpable instability
  • Pain on palpation or with use of the limb.

Prevention. Prevention of a nonunion is key because treatment can be difficult and nonviable nonunions, in particular, can have a poor to guarded prognosis.

Treatment. Successful treatment of a viable nonunion centers on removal of fibrous tissue in the fracture gap, addition of a graft, and rigid fixation (Figure 9), while a nonviable nonunion must be approached with the focus on preservation of soft tissue structures; as in all other fractures, rigid fixation is of paramount importance.4,5

  1. Remove, reposition, or replace implants.
  2. Open medullary cavity and remove sclerotic/atrophic bone ends.
  3. Lavage area to remove any infection/contamination.
  4. Place a suitable autologous, autogenous, or synthetic graft.

Although potentially cost prohibitive, use of recombinant human bone morphogenetic protein 2 (rhBMP-2) can contribute to a successful outcome in cases of nonviable nonunions, and its use has been documented in veterinary medicine.9


Posttraumatic osteomyelitis is not very common in elective orthopedic procedures and fracture repairs.10 However, open fractures (Figures 4 and 5) are particularly prone to infection, with the risk increasing with severity of injury.11

In the Literature

Despite our best efforts, infection can occur during elective procedures and, therefore, must be a consideration in postoperative monitoring. The infection rate associated with the TPLO procedure (Figure 10) has been reported, at the highest, as 8.4%, but a more recent study identified a rate of 3.8% for superficial or deep surgical site infections (SSI).12 While this information does not apply to orthopedic surgical procedures as a whole, it does provide a current and relevant idea of the impact of posttraumatic osteomyelitis.

Clinical Signs
Clinical signs associated with an infection can be variable and will depend on time since surgery. Signs often include:

  • Inflammation and swelling at the surgery site
  • Pain on palpation over the implant or fracture site
  • Draining tracts (Figure 8)
  • New/worsening or sudden onset lameness.


The following can be noted if infection is present:

  • Excessive periosteal reaction (Figure 7)
  • Radiolucency associated with implant/bone interface.

While deep percutaneous aspirates of the infected area can be valuable in identifying the microbial organism and its susceptibility pattern, they must be interpreted carefully because contamination with skin organisms during sampling is common.

Staphylococcus species are the most common causative organism. That being said, it is crucial that samples of bone, deep tissue, and representative implants are submitted for culture and susceptibility analysis.

Antimicrobial Therapy

Infections associated with the incision or surrounding soft tissue can often be treated with antimicrobial drugs and have minimal impact on bone healing, but those associated with an implant or the bone itself are more problematic. However, if the construct is otherwise stable, a fracture will heal despite infection.

Microbial infections involving orthopedic implants often develop a bacterial biofilm, which confers resistance to systemic antimicrobial drugs. Thus, eradication of the infection necessitates removal of the implant once healing is complete.10 Of similar importance is the removal of any avascular bone and/or sequestra (Figure 13) that may be present.

Antimicrobial therapy should be guided by the results of culture and susceptibility analysis, and should be continued for a minimum of 6 to 8 weeks.

Figure 13

Figure 13. Postoperative TPLO patient that developed an infection and sequestrum (*).

Follow-Up & Prognosis

Serial radiographs every 4 to 6 weeks until complete healing and resolution of radiographic signs of infection are advised.10

Prognosis is generally good for normal function unless there is significant soft tissue loss/involvement or infection is associated with a total joint implant that needs to be removed.10,11

Common Complications: Cranial Cruciate Ligament Rupture Surgical Repair

With elective procedures, such as TPLO or TTA, long-term complications apparent on radiographs can include, but are not limited to:14

  • Fracture of tibial tuberosity
  • Screw breakage or loosening
  • Patellar fracture (Figure 9)
  • Septic arthritis
  • Tibial/fibular fracture (Figure 14). 

The surgical site should be evaluated for draining tracts, swelling and inflammation, and pain associated with palpation of the implant. Postliminary (latent) meniscal tears can occur in either case; therefore, a thorough orthopedic examination should be performed to assess for evidence of meniscal pathology.14

It is important to note that, after TPLO or TTA procedures, dogs continue to exhibit a positive cranial drawer sign but should not have a positive tibial compression (tibial thrust) test.


Bone-implant construct failure can occur:

  • With failure at the implant level, such as a broken screw (Figure 6) or bent plate
  • In association with the bone, such as a tibial tuberosity fracture after TPLO or nonunion (Figure 14).
Figure 14

Figure 14. Catastrophic failure of TPLO: Note that screws in proximal fragment have pulled out and a fracture of proximal tibial fragment is present (*). The fibula is also fractured (+), which eliminates its function as an internal splint. This patient may present with an acute nonweight-bearing lameness after an episode of inappropriate activity (eg, dog escaped from house and was running in yard).

A very small number of these complications are attributed to the implant alone, with the major cause identified as technical errors, including:

  • Inappropriately sized implants
  • Inappropriate implant placements
  • Use of cerclage and intramedullary pin for repair of a transverse long bone fracture (Figure 12)13
  • Poor owner compliance.

Radiographic Evaluation

When evaluating serial follow-up radiographs after fracture fixation, it is generally important to evaluate several criteria:

  • Evidence of implant loosening or breakage (Figures 4, 5, and 7): Breakage may not be obvious and can be obscured on a single radiographic view; thus, orthogonal views are essential. If there is evidence of loosening, assess the position of the implant relative to previous radiographs (Figure 6 and 14).
  • Loss of cortical bone adjacent to the implant(s) or radiolucency: May occur with loosening or infection (Figures 7 and 10)
  • Loss of reduction at fracture site or loss of alignment (Figure 7)
  • Evidence of progression toward normal healing at fracture site: Is there a bridging callus? Do the fracture ends appear more rounded and less distinct?


While some failures require surgical intervention, others can be managed nonsurgically. In general, the greater time since surgery, the less likely bone-implant construct failure will occur. For elective osteotomies, such as TPLO and TTA, once the 6- to 8-week follow-up evaluation is reached, the chance for bone-implant construct failure is quite low.


Fracture disease describes any other postoperative complication associated with the initial injury, fracture, or repair. Some of the more common issues are:

  • Muscle atrophy
  • Joint stiffness
  • Fracture distal to the implant
  • Articular cartilage degeneration
  • Adhesion of muscle to bone/muscle scarring.

Disuse of the affected limb contributes significantly to muscle atrophy, joint stiffness, and osteopenia, which helps emphasize the importance of postoperative physical therapy.

Quadriceps contracture (Figure 15) is the most common and severe form of fracture disease in small animal patients. It is of greatest concern in:

  • Young dogs and cats with femoral fractures
  • Animals managed with prolonged coaptation with the limb in extension.

Figure 15. Dog with quadriceps contracture of right quadriceps muscles; note hyperextended stifle and hock.

Quadriceps Contracture Evaluation

During follow-up evaluations, the quadriceps muscles should be palpated for evidence of persistent firmness (permanent contraction); in addition, stifle range of motion should be assessed. Use of a goniometer to measure and record maximal angle of flexion and extension is important for ongoing comparisons.

For the stifle:15

  • Normal flexion angle is < 45 degrees
  • Normal extension angle is approximately 162 degrees.

In one reported case of quadriceps contracture, a loss of stifle flexion, nonweight-bearing lameness, knuckling, and internal rotation were present 22 days after a second attempt to repair a femur fracture.16 Therefore, young dogs (< 12 months) should be evaluated for these signs at 10 to 14 days after surgery; then at 4- and 8-weeks postoperatively.

Quadriceps Contracture Prevention

Risk for this complication decreases as the fracture heals and improved use of the limb occurs. Prevention is the key because there is no effective treatment; prevention includes:

  • Atraumatic surgery
  • Internal fixation
  • Early mobility
  • Use of affected limb
  • Passive range-of-motion exercises.


Orthopedic procedures are commonly performed in small animal patients, and whether it is an elective procedure or urgent/emergent fracture repair, follow-up evaluations are critical in reaching a desirable outcome.

Due to the increasing frequency of these procedures, primary care veterinarians are often responsible for follow-up visits. In some cases, the primary veterinarian may be comfortable performing the evaluation, taking radiographs, and interpreting progress, while others prefer to perform the evaluation; then consult the surgeon.

In either case, a team approach between the referral surgeon and primary veterinarian is optimal in order to achieve success in managing patients after an orthopedic procedure has been performed.

SSI = surgical site infection; TPLO = tibial plateau leveling osteotomy; TTA = tibial tuberosity advancement

Suggested Reading

Johnson AL, Houlton JEF, Vannini R. AO Principles of Fracture Management in the Dog and Cat. New York City: Thieme Medical Publishers, 2005.

Piermatti DL, Flo GL, DeCamp CE. Brinker, Piermatti, and Flo’s Handbook of Small Animal Orthopedics and Fracture Repair, 4th ed. St. Louis: Saunders Elsevier, 2006.

Tobias KM, Johnston SA. Musculoskeletal system. Veterinary Surgery: Small Animal, Vol 1. St. Louis: Elsevier, 2012.

C09_aDuane Robinson, DVM, PhD, Diplomate ACVS (Small Animal), is an assistant professor of orthopedic surgery at University of California—Davis School of Veterinary Medicine. He received his DVM from University of Guelph Ontario Veterinary College, and his PhD (infectious disease) from University of Minnesota. He completed a rotating internship in medicine and surgery at Ontario Veterinary College, surgical internship at Affiliated Veterinary Specialists (Orange Park, Florida), research fellowship at Iowa State University Veterinary Orthopaedic Research Laboratory, and surgery residency at University of Minnesota.


  1. Cross AR. Fracture biology and biomechanics. In Tobias KM, Jonston SA (eds): Veterinary Surgery: Small Animal, Vol 1. St. Louis: WB Saunders, 2012, pp 565-571.
  2. Griffon DJ. Fracture healing. In Johnson AL, Houlton JEF, Vannini R (eds): AO Principles of Fracture Management in the Dog and the Cat. New York City: Thieme Medical Publishers, 2005, pp 73-97.
  3. Nicoll C, Singh A, Weese JS. Economic impact of tibial plateau leveling osteotomy surgical site infection in dogs. Vet Surg 2014; [ePub ahead of print].
  4. Kraus KH, Bayer BJ. Delayed unions, nonunions and malunions. In Tobias KM, Jonston SA (eds): Veterinary Surgery: Small Animal, Vol 1. St. Louis: WB Saunders, 2012, pp 647-656.
  5. Piermattei DL, Flo GL, DeCamp CE. Delayed union and nonunion. In Brinker, Piermattei, and Flo’s Handbook of Small Animal Orthopedics and Fracture Repair, 4th ed. St. Louis: Saunders Elsevier, 2006, pp 168-176.
  6. Welch JA, Boudrieau RJ, Dejardin LM, et al. The intraosseous blood supply of the canine radius: Implications for healing of distal fractures in small dogs. Vet Surg 1997; 26:57-61.
  7. Lappin MR, Aron DN, Herron HL, et al. Fractures of the radius and ulna in the dog. JAAHA 1983; 19:643-650.
  8. Waters DJ, Breur GJ, Toombs JP. Treatment of common forelimb fractures in miniature-breed and toy-breed dogs. JAAHA 1993; 29:442-448.
  9. Pinel CB, Pluhar GE. Clinical application of recombinant human bone morphogenetic protein in cats and dogs: A review of 13 cases. Can Vet J 2012; 53(7):767-774.
  10. Budsberg SC. Osteomyelitis. In Tobias KM, Jonston SA (eds): Veterinary Surgery: Small Animal, Vol 1. St. Louis: WB Saunders, 2012, pp 669-675.
  11. Millard RP, Towle HA. Open fractures. In Tobias KM, Jonston SA (eds): Veterinary Surgery: Small Animal, Vol 1. St. Louis: WB Saunders, 2012, pp 572-575.
  12. Savicky R, Beale B, Murtaugh R, et al. Outcome following removal of TPLO implants with surgical site infection. Vet Comp Orthop Traumatol 2013; 26:260-265.
  13. Johnston SA, von Pfeil DJF, Dejardin LM, et al. Internal fracture fixation. In Tobias KM, Johnston SA (eds): Veterinary Surgery: Small Animal, Vol 1. St. Louis: WB Saunders, 2012, pp 576-607.
  14. Kowaleski MP, Boudrieau RJ, Pozzi A. Stifle joint. In Tobias KM, Jonston SA (eds): Veterinary Surgery: Small Animal, Vol 1. St. Louis: WB Saunders, 2012, pp 906-998.
  15. Millis DL, Levine D. Appendix 2: Joint motions and ranges. Canine Rehabilitation and Physical Therapy, 2nd ed. Philadelphia: Elsevier, 2014, p 734.
  16. Moores AP, Sutton A. Management of quadriceps contracture in a dog using a static flexion apparatus and physiotherapy. J Small Anim Pract 2009; 50(5):251-254.