A recent study conducted by Peter G Passias (NYU Langone Health, New York Spine Institute, New York, USA) and colleagues showed that male patients and patients with increased comorbidity severity experienced a greater length of hospital stay (LOS) after undergoing a primary spine procedure. Another factor that was found to increase LOS, as well as higher rates of hospital-acquired conditions (HACs), was treatment at large, metropolitan, government teaching hospitals. Overall, the investigators found that hospital characteristics and population density, independent of regional location, appear to play greater roles in predicting longer LOS and HACs compared to patient demographic factors.
At the 19th Annual Conference of the International Society for the Advancement of Spinal Surgery in Anaheim, USA (ISASS; 3 –5 April 2019), co-author Virginie Lafage (Hospital for Special Surgery, New York, USA) presented these results.
Passias and colleagues note that HACs have been the focus of recent policy initiatives by the Centers for Medicare and Medicaid Services in an effort to improve patient safety and outcomes. For example, subsection of the Affordable Care Act’s Pay-for-Performance Program, the Hospital-Acquired Reduction Program lowers patient reimbursement for hospitals with elevated rates of “reasonably preventable” complications.
As elective spine surgery continues to grow in popularity, the investigators note that “It is important to assess the most common postoperative conditions as they relate to invasive spine procedures.” The authors note that the goals of the study were to determine the rates of the three most common HACs that occur within 30 days postoperatively for spine surgeries and compare them to other common surgical procedures, and to investigate the influence of patient and hospital characteristics on length of stay and never events.
Passias and colleagues found that of the 1,348,305 surgeries included in the study, the mean LOS was 4.04 days. The spinal procedures associated with the greatest LOS were any osteotomy (7.57 days), fusion (5.35 days), device insertions (3.4 days), and disc replacement (1.63 days). Additionally, the investigators mention that extended LOS was associated with greater hospital charges (US$143,985.57 vs. US$63,439.5, R: 0.631, p<0.001).
Significant demographic predictors of LOS were male sex, with an odds ratio (OR) of 1.1 (1.1–1.13), Hispanic ethnicity (OR: 1.15 [1.15–1.17]), and CCI (OR: 1.175 [1.174–1.175]). The types of hospitals which predicted the longest LOS were urban teaching hospitals (OR: 1.3 [1.3–1.31]), large hospitals (OR: 1.2 [1.19–1.23]), government-owned hospitals (OR: 1.28 [1.27–1.28]) and hospitals located in metropolitan areas (OR: 1.05 [1.05–1.06]).
The investigators report that 1.4% of the total number of patients included in the study experienced an HCA, or “never event.” Seven point nine per cent of patients who underwent an osteotomy experienced a never event, making it the procedure with the highest rate within this patient cohort. Out of the patients who underwent spinal fusion, 1.2% experienced a never event. The rate for patients undergoing a spinal device insertion was 0.9% and 0.7% of disc replacement patients.
CCI was the most significant predictor of a never event in this patient cohort, with an odds ratio of 1.28 (1.27–1.29). Another important factor was male sex (OR: 1.2 [1.1–1.2]). Hospital factors included mid-Atlantic hospitals (OR: 1.125 [1.068–1.186]), mountain region hospitals (OR: 1.263 [1.189–1.341]), teaching hospitals (OR: 1.752 [1.702–1.805]), government hospitals (OR: 1.335 [1.259–1.416]) and hospitals located in metropolitan areas (OR: 1.122 [1.077–1.17]). Surgical predictors included osteotomy (OR: 5.7 [5.4–6.1]), spinal fusion (OR: 0.5 [0.5–0.6]), spinal device insertion (OR: 0.5 [0.4–0.5]), discectomy (OR: 0.3 [0.3–0.4]) and disc replacement (OR: 0.5 [0.4–0.6]).
Passias and colleagues describe that patients >18 years old who underwent elective spine surgery were identified in the American College of Surgeons’ National Surgical Quality Improvement Program (ACS-NSQIP) database from 2005–2013, and that primary spinal surgeries were isolated via ICD-9-CM codes.
In order to analyse LOS, the investigators used descriptive analysis and Pearson bivariate correlations to determine the mean LOS, LOS associated with specific spinal procedures, and total hospital charges associated with extended LOS (>75th percentile). They used Poisson regression models controlling for CCI, surgical invasiveness, patient and hospital characteristics to identify independent predictors of increased LOS.
For HACs, a descriptive analysis determined the incidence of never events for the overall cohort and specific spinal procedures. Furthermore, binary logistic regression models controlling for CCI, surgical invasiveness, patient and hospital characteristics identified independent predictors of never events.