Can the American College of Surgeons Risk Calculator Predict 30-Day Complications After Spine Surgery?

Michael H. McCarthy, MD, MPH; Partik Singh, BA; Rusheel Nayak, BA; Joseph P. Maslak, MD; Tyler J. Jenkins, MD; Alpesh A. Patel, MD; Wellington K. Hsu, MD

Disclosures

Spine. 2020;45(9):621-628. 

In This Article

Results

Cervical Fusion

A total of 404 patients (207 anterior, 197 posterior) from 2009 to 2015 were identified as having a primary cervical fusion and meeting our inclusion criteria. Baseline demographics, number of levels of fusion and risk estimates are described in Table 1. Our cohort's data were entered into the Risk Calculator and estimates were obtained for each patient in addition to anterior and posterior group estimates as seen in Table 2. Statistical analysis of observed outcomes was calculated and further stratified by anterior versus posterior. Logistical regression models were fit for Risk Calculator estimates and observed outcomes as seen in Table 3. No deaths were noted within our cohorts; thus, no models were fit for mortality. Only "Any complication" and "Skilled Nursing Facility (SNF)/Rehab Admit" met the criteria for c > 0.80 for acceptable concordance and ROC curves for each complication can been seen in Figure 2.

Figure 2.

Cervical fusion receiver-operating characteristic (ROC) curves for the 11 complications measured by the ACS Risk Calculator.

Logistical regression results for anterior and posterior patients demonstrated that Risk Calculator predictions were better in the anterior group for "Any complication," "SNF/Rehab admit," and "Serious complication" than in the posterior group. Although complications occurred at a significantly higher rate in the posterior group, the ability for the Risk Calculator to accurately predict complications within this group was worse when compared to the anterior group. Comparison of single versus multilevel fusions yielded better prediction in the single-level group for "SNF/Rehab admit," although prediction was still acceptable (c > 0.80) in the multilevel group.

Lumbar Fusion

A total of 237 patients from 2009 to 2015 were identified as having a primary lumbar fusion and meeting our inclusion criteria. Baseline demographics, number of levels of fusion, risk estimates, and American Society of Anesthesiologists Classification (ASA) class are described in Table 4. Anterior, oblique lateral, lateral, and posterior fusions were combined within the lumbar cohort; however, fusion length was further specified by number of levels fused. Similar to the cervical cohort, Risk Calculator estimate and observed outcomes were analyzed for the overall sample of lumbar patients. Logistical regression results for the overall sample are represented in Table 5. Of all the calculations, only Pneumonia met the criteria for c > 0.80 for acceptable concordance; however, due to small sample size and low event rate, this was not statistically significant and the confidence interval on the odds ratio was very wide.

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....