Larry Holmes

Outcome of in situ pinning in patients with unstable slipped capital femoral epiphysis: assessment of risk factors associated with avascular necrosis.

Introduction Avascular necrosis (AVN) is a devastating complication following treatment of unstable slipped capital femoral epiphysis (SCFE). The advent of newer methods such as open reduction and surgical dislocation of the hip has increased the debate on the optimal method of treatment. However, the risk or predisposing factors for AVN remain unclear. We aimed to assess the outcome of in situ fixation and the risk factors associated with AVN. Methods We retrospectively reviewed the records of 27 patients (27 hips) with unstable SCFE out of the 280 children managed for SCFE from 1995 to 2006. The mean age in years of the patients at surgery was 12.2±1.58, and our sample comprised 70.4% males, and 29.6% females, with a mean follow-up of 3.1±1.9 years. Univariable and multivariable binomial regression models were used to assess factors predisposing to AVN. Results AVN occurred in 22.2% (6/27) of the children with unstable SCFE. After adjustment for age, race, and time to surgery, sex and preoperative slip angle were the only 2 significant factors related to an increased risk of AVN, risk ratio (RR)=4.15, 95% confidence interval=1.00-17.19, P=0.05 and RR=1.04, 95% confidence interval=1.00-1.07, P=0.03, respectively. Female children constitute a high-risk group for AVN in this subgroup. Conclusions AVN is still prevalent among patients with unstable SCFE who underwent in situ pinning. Female sex and slip magnitude are potential predisposing factors for developing AVN. Level of Evidence Level III (retrospective cohort study).

Operative and clinical markers of deep wound infection after spine fusion in children with cerebral palsy.

Background Infection after spine fusion for neuromuscular scoliosis has been shown to range from 4.2% to 20.0% prevalence. Although there are studies, which have examined deep wound infection and spine fusion surgery as well as risk factors for deep wound infection, there are limited studies evaluating clinical and radiographic factors associated with this complication. We aimed to determine the clinical and operative factors associated with deep wound infection after spine fusion in pediatric patients with cerebral palsy (CP). Methods Medical records of 236 pediatric patients, aged between 5.6 and 21 years (mean=13.8±3.4), with CP who underwent spine fusion from 1995 to 2006 were reviewed. Of these, 22 patients had deep wound infection. To assess the differences in clinical, radiographic, and other predisposing factors, we used &khgr;2 statistic and Fisher exact, and to determine the predisposing factors of deep wound infection, we used binomial regression model. Results The period prevalence of deep wound infection was 9.3%. In the unadjusted model, body weight, residual postoperative Cobb angle, length of hospitalization, packed red blood cells, and skin breakdown were the factors significantly associated with deep wound infection (P<0.05). After controlling for confounding, skin breakdown due to the instrumentation and residual postoperative Cobb angle were the 2 most potent markers of deep wound infection. There was a significant 4% increased risk of deep wound infection for 1-degree increase in the residual Cobb angle from the noncase mean residual Cobb angle of 23.69 degrees (adjusted risk ratio=1.04; 95% confidence interval, 1.01-1.08). Likewise, compared with those without skin breakdown, those with skin breakdown were 12 times as likely to develop deep wound infection (risk ratio=12.92; 95% confidence interval, 1.00-172.00). Conclusions Residual postoperative Cobb angle and skin breakdown due to unit rod instrumentation were the 2 most significant predisposing factors to deep wound infection. Other factors included body weight, packed red blood cells, and length of hospitalization. As the overall prevalence of deep wound infection is relatively high in CP patients after spine fusion, and considering the cost of hospitalization and other related comorbidities, surgeons should recognize these predisposing parameters to prevent deep wound infection in CP patients while correcting curve deformities. Level of Evidence Level III retrospective study.

Factors related to progression of thoracolumbar kyphosis in children with achondroplasia: a retrospective cohort study of forty-eight children treated in a comprehensive orthopaedic center.

Study Design. Retrospective cohort. Objective. To assess the period prevalence and predisposing factors in thoracolumbar kyphosis progression (TLK). Summary of Background Data. TLK is a common presentation in children with achondroplasia. This condition occurs very early in life; it is observed during the first attempts to sit, and has multifactorial etiology. Studies are limited on the radiologic assessment of the angle of wedging of the deformed vertebra, but none to our knowledge on TLK progression in achondroplasia. Methods. The records of 48 of 103 children with achondroplasia seen between 1997 and 2005 were examined. Developmental motor delay (DMD) and other potential predictors of TLK were assessed. The criteria for DMD were a child’s inability to sit without support by 6 months and an inability to walk independently by 15 months. Lateral radiographs were examined for the magnitude of TLK, thoracic kyphosis, lumbar lordosis, percentage of apical vertebral wedging for the height and width of the vertebra, and apical vertebral translation. The association between TLK progression and potential predisposing factors were assessed using &khgr;2 test, Fisher exact, t test, and binomial regression model for predictor identification. Results. The period prevalence of TLK progression in children with achondroplasia was 35.4%. TLK progression was significantly related to DMD, risk ratio = 2.40, 95% confidence interval = 1.10 to 5.58, apical vertebral translation (P = 0.01), and percentage apical vertebral wedging for height (P = 0.01). After adjustment for relevant covariates, the relationship between DMD and TLK progression persisted, risk ratio = 2.65, 95% confidence interval, 1.20 to 5.91. Conclusion. DMD was significantly associated with TLK progression and other radiologic parameters namely, initial TLK of >25°, percentage of apical vertebral wedging for the height of the vertebra, and apical vertebral translation.

Prevalence and risk factors in postoperative pancreatitis after spine fusion in patients with cerebral palsy.

Background: Postoperative pancreatitis is a recognized complication after spine fusion surgery in scoliosis management. There are established risk factors for the development of postoperative pancreatitis for patients with scoliosis in general, but no such factors have been established in patients with cerebral palsy (CP) who have scoliosis. The aim of this study was to assess the association between preoperative comorbidities and other parameters in the development of pancreatitis after spinal fusion surgery in a sample of children with CP. Methods: A retrospective cohort design was used to assess the association between postoperative pancreatitis and preoperative comorbidities and other covariates that might predispose to pancreatitis. The study participants included 355 patients, with a mean age of 13.8 years (range, 5.6-21 years) with CP, who received spine fusion with rod instrumentation. We used the &khgr;2 statistic, mean, and SD to describe study variables that were categorical and continuous, respectively. To test the null hypothesis of no association between our outcome variable (postoperative pancreatitis) and comorbidities and other covariates as predictors, we used a binomial regression model. To simultaneously adjust for confounding effects in the relationship between our outcome and the independent or predictor variables, we used a multivariable binomial regression model. Results: Of the 355 patients who underwent spine fusion, 109 developed postoperative pancreatitis (prevalence, 30.1%). Patients who developed pancreatitis had prolonged number of fasting days (until oral or G-tube feeding initiated; 7.9 vs 5.2 days) and longer hospital stay (23.1 vs 15.6 days). In the univariable binomial regression model, patients with CP who had gastroesophageal reflux disease (GERD) and feeding difficulty were more likely to develop pancreatitis than those without this condition (risk ratio [RR], 1.57; 95% confidence interval [CI], 1.10-2.28, respectively). Likewise, in this model, gastrointestinal tube and reactive airway disease (RAD) were statistically significantly associated with postoperative pancreatitis. Patients with CP who had gastrointestinal tube were 61% more likely to develop postoperative pancreatitis, whereas those with RAD were 54% (RR, 1.61; 95% CI, 1.01-2.55 and RR, 1.54; 95% CI, 1.13-2.10, respectively). However, there was a clinically relevant but nonstatistically significant association between seizure and postoperative pancreatitis (RR, 1.72; 95% CI, 0.96-3.06). After adjustment for the confounding variables in the multivariable model, GERD with feeding difficulties persisted as a single most significant and potent predictor of postoperative pancreatitis (adjusted RR, 1.52; 95% CI, 1.01-2.29). Consequently, patients with CP who had GERD were 52% more likely to develop postoperative pancreatitis. Likewise, there was a statistically significant 49% increase in the risk of postoperative pancreatitis in patients with CP who had RAD (adjusted RR, 1.49; 95% CI, 1.10-2.04). Conclusions: Pancreatitis is a major cause of morbidity after spinal fusion surgery in patients with CP. Patients with preoperative GERD with feeding difficulties and RAD had a higher risk of developing postoperative pancreatitis. Postoperative pancreatitis causes delays in feeding and increases the duration of hospitalization. Clinicians should be aware of the roles of GERD and RAD, as well as seizure and/or antiepileptics in the development of postoperative pancreatitis in patients with CP undergoing spinal fusion. Level of Evidence: Level III.

Comparison of Direct Pars Repair Techniques of Spondylolysis in Pediatric and Adolescent Patients: Pars Compression Screw Versus Pedicle Screw-Rod-Hook.

Study Design:Retrospective clinical cohort study. Objective:To compare the clinical and radiographic outcomes of patients who were treated with intrasegmental pars fixation by either laminar compression screw (LS) or a pedicle screw, rod, and laminar hook (PSRH) construct. Summary of Background Data:Spondylolysis is a nonunion defect of the pars interarticularis. In symptomatic spondylolysis, direct repair of the pars interarticularis defect can preserve motion and prevent abnormal stresses at the adjacent levels. Methods:Sixteen patients who failed nonoperative treatment and underwent direct pars repair by using LS (n=9) or PSRH (n=7) constructs were included in the study. Clinical outcome was assessed by using the MacNab criteria. Radiologic fusion and complications were evaluated using plain radiographs or computed tomography images and patient charts. Results:The healing rate was 100% after 6 months. The healing time was similar in both the groups: LS, 6.5 months; PSRH, 6.2 months. Patients with PSRH (5.9 mo) were more likely to return to sports earlier relative to patients with LS (7.7 mo). There were no complications in the LS group; in the PSRH group, 1 patient had mild sensory deficit and 2 had superficial wound infections. The MacNab criteria for pain assessment showed an excellent or good outcome in 8 of 9 patients in LS group and 6 of 7 patients in PSRH group. Relative to LS patients, there was a significant increase in surgical time and estimated blood loss among PSRH patients. Conclusions:Either of the mentioned 2 techniques appears to produce acceptable results. Biplanar fluoroscopy and navigation systems could minimize the risk of screw misplacement with LS construct. Familiarity with the various fixation techniques will allow the surgeon to select the most appropriate surgical technique.

Labral support shelf arthroplasty for containment in early stages of Legg-Calve-Perthes disease.

Background: Containment of the hip has become a popular strategy for treatment of severe Legg-Calve-Perthes disease; however, the best method remains unclear. This study describes a minimal-incision variation of the labral support shelf arthroplasty, which uses arthroscopic visualization and an allograft buttress on the shelf support. The outcome in a group of hips undergoing this minimal-incision variation is compared by the Stulberg classification with a group treated by the traditional anterolateral open operative approach. In addition, the outcomes of both groups of labral support shelf arthroplasties are compared then with other containment techniques used in our hospital consisting of Petric cast, varus femoral osteotomy, and innominate osteotomy of Salter. Methods: Forty-two patients who fulfilled the following criteria underwent a labral support shelf arthroplasty: (1) unilateral hip disease; (2) age equal to or older than 6 years at diagnosis; (3) Catterall group III or IV; and (4) Herring B or C disease. Twenty patients had a traditional open approach (group I) and 22 had an approach using a minimal-incision arthroscopic visualization and allograft buttress on the shelf support (group II). Outcomes by the Stulberg classification were then compared between these groups (I and II) and with 29 cases of Petri casting, 15 of varus femoral osteotomies, or 15 of innominate osteotomies of Salter. Results: There is no statistical difference in Stulberg classification outcomes between group I and group II, which demonstrates that a minimally invasive technique is satisfactory and prevents substantial tissue dissection as is required in a standard anterior lateral approach to the hip, P>0.05. In addition, there is no statistical difference in Stulberg classification outcomes between the labral support shelf arthroplasties and common containment methods such as Petri casting, varus femoral osteotomies, or the innominate osteotomy of Salter, P>0.05. Conclusions: The labral support shelf arthroplasty by either the traditional method or the minimal-incision variation offers similar Waldenstrom class outcomes when compared with a Petric cast, a varus femoral osteotomy, or an innominate osteotomy of Salter. We believe the labral support shelf arthroplasty technique is simple to perform and does not induce a permanent deformity in the proximal femur or acetubulum.

Treatment Options for Nonunion With Segmental Bone Defects: Systematic Review and Quantitative Evidence Synthesis.

Objectives: To determine which reconstruction treatment of long bones nonunion with segmental bone defects (SBDs) is effective to restore bone length and union with good function. Data Sources: PubMed was used to identify published literature on treatment of SBD caused by fracture nonunion regardless of infection between January 1975 and December 2014. Study Selection: We included retrospective cohort studies with a minimum sample size of 10 consecutive patients with minimum follow-up of 18 months and available data on radiographic and functional outcomes. Data Extraction: Literature review revealed 24 publications with a sample size of 504 patients (395 males, 109 females). Data on bone union and functional outcome and complications were collected and analyzed based on validated classification systems. Data Synthesis: Two outcome groups were categorized for bone union and functional outcome, success, and failure. We then performed heterogeneity test to examine the variability or differences in the methods used by these studies and based on that we determined whether the fixed effect or random effect method is appropriate in examining the summary or pool estimate. Pool estimate was examined for bone union and functional outcome in each surgical modality and in each anatomic location when data were available. Conclusions: Treatment of SBD can be challenging. This quantitative evidence synthesis shows that bone union was achieved by different procedures with variable bone union and functional outcomes. Level of Evidence: Therapeutic Level IV. See Instructions for Authors for a complete description of levels of evidence.

Clinical epidemiologic characterization of orthopaedic and neurological manifestations in children with leukodystrophies.

Background Leukodystrophies (LKDs) are spectra of clinical conditions characterized primarily by brain white matter abnormalities. Although this condition was previously defined around inherited disorders of the white matter of the brain, current application includes acquired and sporadic conditions and some rare conditions that affect gray matter. Over the past 2 decades, information had become available on the clinical subtypes due to neurodiagnostic imaging and improvement in the genetic studies (cytogenetics and molecular genetics) of LKD. However, the epidemiologic profile of LKD remains largely unknown. We aimed in this study to characterize LKD by demographics, family history, orthopaedic and neurological manifestations, and clinical subtypes. Methods Trained medical personnel reviewed medical records of the study population diagnosed with LKD from 1986 to 2008. Using a retrospective review design, we determined the prevalence of the different clinical subtypes of LKD, family history, orthopaedic and neurological manifestations, and the demographics in LKD. The frequency and percentage (proportion, standard error, and 95% confidence interval for proportion) and the &khgr;2 statistic and Fisher exact test for comparison of clinical subtypes were the statistical techniques used in the data analysis. Results Forty-four children were diagnosed with LKD between 1986 and 2008, of whom 25.0% had metachromatic LKD and 20.5% had Pelizaeus-Merzbacher LKD, whereas 40.9% were unspecified LKD. LKDs were more common among boys (63.6%), Whites (77.3%), and more likely to be diagnosed at age <3 years. Scoliosis (70.4%), hamstring contractures (81.8%), acquired hip dysplasia (88.6%), and equinus foot deformity (75.0%) were the most common orthopaedic manifestations. Common neurological manifestations were seizures (45.4%) and spasticity (77.3%). There was a statistically significant difference in sex and family history, seizures, hip dislocation, and hip subluxation, with respect to the clinical subtype of LKD, P<0.05. Conclusions This epidemiologic characterization of LKD validates basic and clinical data on the familial history of LKD and its higher prevalence among boys. The orthopaedic manifestations common in LKD are scoliosis, hamstring contractures, acquired hip dysplasia, and equinus foot deformity, whereas common neurological manifestations are seizures and spasticity. These data are indicative of the need for orthopaedic surgeons to take into consideration this clinical epidemiologic aspect of LKD in the evaluation, treatment planning, and clinical expectations for these patients.

Prevalence of scoliosis in patients with Fontan circulation

Objective To examine the prevalence of scoliosis in patients with Fontan circulation. Design Retrospective cohort (case-only), level IV study. Setting A comprehensive paediatric centre. Methods We examined the radiographs of 194 patients who underwent Fontan completion surgery between 1998 and 2011 at a median age of 1.3 years. Main outcome measure We collected data on the age at the last available radiograph and when scoliosis was diagnosed; we used the first available radiograph with scoliosis. We also measured the magnitude and direction of the spinal curve in patients with scoliosis. Results Median age at radiography was 3.4 years. Nineteen patients (9.8%) developed scoliosis; none of them underwent thoracotomy. Most of the patients with scoliosis were older than 5 years of age at scoliosis diagnosis and the female to male ratio was 5 : 3. The major curve was right thoracic in 12 patients (63.2%), left thoracic in four patients (21.1%) and high thoracic in three patients (15.8%). For every 1 year increase in age there was a 27% increased risk of developing scoliosis. Conclusions There was a high prevalence of scoliosis in patients with Fontan circulation. We recommend interdisciplinary monitoring of these patients to diagnose spinal curve deformities in a timely manner.

Childhood Autism Spectrum Disorder and Epilepsy Co-occurrence: Sub-population Prevalence Variances and Risk Modeling

Autism spectrum disorder (ASD) tends to co-occur with epilepsy, but it remains unclear if ASD predisposes to epilepsy or vice versa. Using the National Survey of Children’s Health (2012), we assessed the relationship between ASD and epilepsy as disease co-occurrence, examined racial/ethnic variability therein and the risk markers. Compared to whites, blacks/African Americans were 54% less likely to be diagnosed with ASD, prevalence odds ratio (POR)=0.46, 95% CI, 0.34-0.63. In contrast, compared to whites, blacks/African Americans were 56% more likely to be diagnosed with epilepsy, POR=1.56, 95% CI, 1.24-1.96. Similarly, racial differences were observed in cooccurrence, with significant 20% higher risk among Blacks relative to Whites, POR=1.20, p<0.0001. Public insurance, poverty, maternal education, and intellectual disabilities were the two most potent predictors of ASDepilepsy co-occurrence. After controlling for age, sex, parental education, insurance coverage, federal poverty level, and intellectual disability, significant racial differences did not persist but Blacks were 74% less likely to report ASD & epilepsy co-occurrence relative to Whites, aPOR=0.26, p=0.23. In summary, racial disparities exist in ASD-epilepsy co-occurrence prevalence, which did not persist after controlling for potential confounders; while public insurance, maternal education, poverty and intellectual disabilities remain potent predictors of the co-occurrence. These findings suggestive of cautious optimism in assessing the causal direction in the co-occurrence of ASD and epilepsy among children.