Declining Rates of Legg-Calvé-Perthes Surgery in the United States: National Trends Using the Kids' Inpatient Database and Pediatric Health Information System.

BACKGROUND: In 2004 and 2008 two large prospective, multicenter studies were published which resulted in improved understanding of operative indications for the treatment of Legg-Calvé-Perthes disease (LCPD) based on patient age, disease severity, and resultant radiographic outcomes. The primary aim of this study is to evaluate the trends in surgical management of LCPD in the United States prior, and subsequent to, the publication of these landmark studies. METHODS: Cross-sectional retrospective analysis of US pediatric hospitalizations for the surgical management of LCPD was conducted using the Kids' Inpatient Database from 2000 to 2016. Patients 12 years of age and younger were included who had a primary admission diagnosis of LCPD and a LCPD-related procedure during the hospitalization. Data was subsequently weighted to produce national-level estimates and variables pertaining to patient age group, procedure, demographics, and hospital characteristics were analyzed. In a post hoc analysis, the results of the Kids' Inpatient Database were also corroborated with the Pediatric Health Information System database. RESULTS: A weighted sample of 2786 LCPD surgical admissions met inclusion and exclusion criteria; 11.2% of surgical admissions were patients below 6 years of age, 35.9% were 6 to 8 years of age, and 52.9% were above 8 years of age. There was a significant decrease in admissions for surgical management of LCPD in all age groups over time, however there was no appreciable change in the proportion of LCPD surgical admissions performed among the above 8 to below 12, above 6 to below 8, or below 6 years age groups. Femoral osteotomy remained the most common surgical procedure, while other osteotomy types, including pelvic and unspecified osteotomies involving the hip, decreased over time ( P <0.001). CONCLUSIONS: There is a decreasing rate of hospital admissions for LCPD surgery since 2000, perhaps indicating a decline in incidence of disease. Furthermore, despite evidence supporting LCPD surgical outcomes related to patient age, there has been no change in the proportion of patients undergoing surgery by age group over time. LEVEL OF EVIDENCE: Level III-retrospective study.

Patients Without Intraoperative Neuromonitoring (IONM) Alerts During VEPTR Implantation Did Not Sustain Neurological Injury During Subsequent Routine Expansions: A Retrospective Multicenter Cohort Study.

BACKGROUND: The purpose of this study was to determine the rate of intraoperative neurological monitoring (IONM) alerts and neurological injury during vertical expandable prosthetic titanium rib (VEPTR) treatment and evaluate the utility of IONM during VEPTR expansion procedures in patients who have not previously had neurological injury or IONM alerts. METHODS: After institutional review board approval, VEPTR procedures and IONM records were reviewed at 17 institutions for patients treated with VEPTR from 2005 to 2011. All consecutive cases in patients with minimum 2-year follow-up were included. Patients with prior history of growing rods or other invasive spine-based surgical treatment were excluded. Surgeries were categorized into implant, revision, expansion, and removal procedures. Cases with IONM alerts or neurological injury had additional detailed review. Descriptive statistics were used for data analysis. RESULTS: In total, 2355 consecutive VEPTR procedures (352 patients) consisting of 299 implant, 377 revision, 1587 expansion, and 92 removal procedures were included. In total, 620 VEPTR procedures had IONM, and 539 of those had IONM records available for review. IONM alerts occurred in 9/539 procedures (1.7%): 3/192 implants (1.6%), 3/58 revisions (5.2%), and 3/258 expansions (1.2%). New neurological injury occurred in 3/2355 procedures (0.1%), 3/352 patients (0.9%). All 3 injuries were in implant procedures, only 1 had an IONM alert. All 3 had upper extremity motor deficits (1 had sensory deficit also). All had full recovery at 17, 30, and 124 days postinjury. One patient without prior neurological injury or IONM alert had an IONM alert during expansion that resolved after an increase in blood pressure. The remaining IONM alerts during expansions were all in children with prior IONM alerts during implant, revision, or exchange procedures. CONCLUSIONS: The highest rate of neurological injury in VEPTR surgery was found for implant procedures. There were no instances of neurological injury during VEPTR expansion, revision, or removal procedures. IONM did not identify new neurological injuries in patients undergoing VEPTR expansion who did not previously have a history of IONM signal change or neurologic injury. LEVEL OF EVIDENCE: Level IV-diagnostic study.

Management of late (>1 y) deep infection after spinal fusion: a retrospective cohort study.

BACKGROUND: The incidence of late infection published in the literature varies from 1% to 12% with varying definition of late infection (range, 3 mo to 1 y). Current evidence suggests implant removal and antibiotic therapy is necessary to clear these infections. A high incidence of late (>1 y) deep infection after instrumented spinal fusion was identified at our institution. We sought to evaluate the efficacy of our management of these patients. METHODS: A total of 1390 patients underwent instrumented spinal fusion from 2000 to 2009. Forty-two patients developed deep infection >1 year after index procedure (3%) and had surgical debridement. Clinical records and microbiology reports were reviewed for details of operative and postoperative management. RESULTS: Advanced imaging was only obtained in 6 patients (5 computed tomography, 1 magnetic resonance imaging). Offending organisms were identified in 39/42 patients, 27 of these grew Propionibacterium acnes. P. acnes grew in culture at a median of 6 days (range, 3 to 10 d), significantly longer than all other organisms, which grew in a median of 1 day (range, 0 to 8 d) (P<0.001). Implants were removed at the index hospitalization in 41 patients. Implant retention was attempted in 1 patient and failed. Primary closure was carried out in 37 patients; the remainder had undergone multiple debridements (4 planned, 1 unplanned). VAC closure was utilized in 2 patients. All patients were treated with organism-specific intravenous antibiotics and transitioned to oral antibiotics on average in 34 days (range, 2 to 186 d). Total length of antibiotic therapy was an average of 141 days (range, 34 to 413 d). CONCLUSIONS: P. acnes was the most common organism identified and took nearly 1 week to grow in culture. Treatment is generally successful with thorough debridement, removal of implants, and antibiotic treatment. LEVEL III: retrospective comparative study.

Are routine postoperative radiographs necessary during the first year after posterior spinal fusion for idiopathic scoliosis? A retrospective cohort analysis of implant failure and surgery revision rates.

BACKGROUND: Radiographs are routinely obtained at postoperative visits during the first year after posterior spinal fusion (PSF) for idiopathic scoliosis (IS). The goal of this study was to determine how often radiographic findings change postoperative care. METHODS: A total of 227 consecutive patients aged 10 to 21 years who underwent surgery for IS at our institution from 2004 to 2010 were identified. Charts were reviewed to determine the frequency of the following clinical symptoms during the first year after surgery: pain greater than expected, implant prominence, and sensory/motor disturbance. Radiographs were reviewed to identify implant failure and curve change. Logistic regression analysis was used to identify clinical symptoms associated with treatment deviation. RESULTS: During the first year after surgery, an average of 6 (range, 2 to 12) radiographs were obtained from patients during an average of 3 (range, 2 to 10) follow-up visits. Pain (14%) was the most common symptom. Neurologic symptoms (13%) and implant prominence (4%) were less common. Implant failure was identified in 4 subjects (2%), of which 3 required revision surgery. The incidence of revision surgery was 2.9/1000 radiographs (95% confidence interval, 0.6-8.3). Curve progression >5 degrees in the uninstrumented curve occurred in 2 patients (0.9%). Curve progression did not result in a change in treatment for any of the patients. Pain was the only clinical symptom associated with implant failure (P=0.0047). 169/227 patients did not have any symptoms and only one of these underwent revision surgery. The sensitivity of a clinical test, which uses the presence of pain to guide the need for radiographic evaluation and rule out implant failure, was 75%, specificity 87%, positive predictive value 10%, and negative predictive value 99.5%. CONCLUSIONS: After obtaining baseline postoperative radiographs, additional radiographs during the first year after surgery for IS may not be required in the absence of clinical symptoms. Reducing the number of radiographs taken during the first year after surgery for IS in patients without symptoms can reduce radiation exposure to patients and health care costs without affecting treatment. LEVEL OF EVIDENCE: Level II, Diagnostic Study.

Predicting Failure of Iliac Fixation in Neuromuscular Spine Deformity.

STUDY DESIGN: Retrospective comparative cohort study. OBJECTIVES: Identify whether there are patient or surgical risk factors to predict the probability of failure of iliac screw fixation after correction of neuromuscular scoliosis. SUMMARY OF BACKGROUND DATA: There are high reported failure rates of pelvic fixation in long posterior spinal fusion (PSF) constructs to the sacrum for neuromuscular scoliosis. METHODS: Patients aged 5 to 25 years, at a single institution, had PSF to the sacrum from 2001 to 2009 with pelvic fixation using iliac screws. Clinical data were retrospectively reviewed to identify patient and surgical variables related to surgery. Failure of iliac fixation was identified strictly as a broken screw, disengagement of the screw from the connector or the connector from the rod, or set plug failure. Lucency around the screws greater than 2 mm was recorded but not considered a failure. Variables were analyzed in a statistical model to identify predictors of failure. RESULTS: A total of 108 patients met inclusion criteria; 100 (38 female and 62 male) had appropriate radiographs and minimum 2-year follow-up (average, 5.5 years). Coronal deformity correction averaged 59%. Most patients (89%) had fill of 6 of 8 possible distal fixation points composed of L4, L5, S1, and ilium bilaterally. Iliac screw failure occurred in 27 patients (27%). The initial single predictor statistical model identified 2 possible predictors of failure (patient: spastic tone; and surgical: absence of distal crosslink). In the multivariable model, spastic tone remained a predictor of failure (p = .0103), whereas absence of distal crosslink bordered on significance (p = .0516). CONCLUSIONS: Iliac screw fixation failure is common in patients with long PSF constructs for neuromuscular scoliosis. Spastic tone is a risk factor for failure of pelvic fixation. A distal crosslink may protect against pelvic fixation failure. Alternative techniques for pelvic fixation should be studied in an attempt to improve the failure rates of iliac screw pelvic fixation.

Bacteriology and Risk Factors for Development of Late (Greater Than One Year) Deep Infection Following Spinal Fusion With Instrumentation.

STUDY DESIGN: Retrospective cohort review. OBJECTIVES: To evaluate patients who underwent instrumented spinal fusion procedures and compare late (more than 1 year) and early infection (less than 1 year). SUMMARY OF BACKGROUND DATA: Centers for Disease Control and Prevention criteria for surgical site infections with implants include infections occurring at less than 1 year postoperatively. The authors observed a high rate of deep infection at more 1 year after instrumented spinal fusion. Retrospective review was conducted to determine whether differences in bacteriology, patient demographics, or surgical factors exist between late and early infection. METHODS: A total of 1,390 patients underwent spinal fusion with instrumentation from 2000 to 2009. Deep infection requiring operative debridement occurred in 112 patients (70 at less than 1 year and 42 at more than 1 year after the index surgery). Clinical, operative, and microbiology reports were reviewed and logistic regression was performed to evaluate the relationship between these factors and time of infection. RESULTS: The most common organisms in the greater than 1 year group were Propionibacterium acnes, compared with Staphylococcus aureus in the less than 1 year group. The odds of late infection in the P acnes group were 15.5 (95% confidence interval [CI], 4.36-54.72) times the odds among all other organisms (p < .0001). Patient demographics and surgical factors were not different between groups. Among all patients, the infection rate was significantly higher in patients who received stainless-steel implants (11.56%) compared with those who received titanium implants (3.53%) (p < .0001). In univariate analysis, the odds of late infection in the stainless-steel group were 6.09 (95% CI, 1.62-39.88) times the odds in the titanium group (p = .0042). However, in multivariate analysis, controlling for organism type reduced the odds ratio to 4.62 (95% CI, .53-40.14), with only a trend to significance (p = .1656). CONCLUSIONS: Propionibacterium acnes was more commonly identified in late infection. Patients with stainless-steel implants had a higher incidence of late infection than those with titanium implants, but this was not significant after controlling for organism type.

Wound complications of vertical expandable prosthetic titanium rib incisions.

STUDY DESIGN: Multicenter retrospective review. OBJECTIVE: To compare the incidence of infection between vertical expandable prosthetic titanium rib (VEPTR) incision locations and determine if the infection risk increases in relation to presence of previous incisions and/or increased number of times incisions are opened. SUMMARY OF BACKGROUND DATA: Patients undergoing treatment for chest and spine deformity with VEPTR require multiple incisions that are opened repeatedly during expansion procedures. METHODS: A prospective database (7 sites) and institutional database (2 sites), were queried to identify their 20 most recent patients with VEPTR with a minimum of 4 expansions for inclusion. A total of 103 patients were identified. Clinical and operative reports were reviewed to determine incision locations, number, and infection complications. RESULTS: Twenty-five of 103 patients (24%) developed an infection during treatment. Six had multiple infections (range, 2-4), providing a total of 34 infection events. Patients averaged 6.4 expansion procedures and 13 total incisions. Infection rate at each incision site was not significantly different, in the range from 1% to 5%: paramedian (6 infections/23 patients with total 185 incisions, 3%), proximal midline (12/39; 224, 5%), thoracotomy (6/61; 455, 1%), iliac (5/37; 143, 4%), and distal midline (5/58; 148, 3%). Infection events occurred after an average of 3 times a particular incision was opened (95% confidence interval: 2.2-3.8). There was a trend toward higher infection rate with increased number of times a particular incision was opened. There was no increased infection rate in patients with surgical incisions prior to VEPTR (26%; 6/23) compared with patients not having prior incisions (24%; 19/80). CONCLUSION: The incidence of infection in patients with 4 or more VEPTR lengthenings was 24% and did not differ across the various incision locations. Presence of prior surgical incisions was not a risk factor for infection. Surgeons should use the most appropriate incision in relation to their patient's pathology when using VEPTR while remaining vigilant for infection. LEVEL OF EVIDENCE: 3.