ABSTRACT
Background: Spinal surgery is associated with an inherently elevated risk profile, and thus far there has been limited discussion about how these outpatient spine patients are benefiting from these same-day procedures against other typical outpatient orthopedic surgeries.
Methods: Orthopedic patients who received either inpatient or outpatient surgery were isolated in the American College of Surgeons National Surgical Quality of Improvement Program (2005–2016). Patients were stratified by type of orthopedic surgery received (spine, knee, ankle, shoulder, or hip). Mean comparisons and chi-squared tests assessed basic demographics. Perioperative complications were analyzed via regression analyses in regard to their principal inpatient or outpatient orthopedic surgery received.
Results: This study included 729 480 surgical patients: 32.5% received spinal surgery, 36.5% knee, 24.1% hip, 4.9% shoulder, and 1.7%ankle. Of those who received a spinal procedure, 74.7% were inpatients (IN), and 25.3% were outpatients (OUT): knee: 96.1% IN, 3.9% OUT; hip:98.9% IN, 1.1% OUT; ankle: 29% IN, 71% OUT; and shoulder: 52.6% IN, 47.6% OUT. Hip patients were the oldest, and knee patients had the highest body mass index out of the orthopedic groups (P < .00). Spine IN patients experienced more complications than the other orthopedic groups and had the lowest OUT complications(both P < .05). This same trend of having higher IN complications than OUT complications was identified for hip, shoulder, and knee. However, ankle procedures had greater OUT procedure complications than IN (P < .05). After controlling for age, body mass index, and Charlson Comorbidity Index, IN procedures, such as knee, hip, spine, and shoulder, were significantly associated with experiencing postoperative complications. From 2006 to 2016, IN and OUT surgeries were significantly different among complications experienced for all of the orthopedic groups (P < .05) with complications decreasing for IN and OUT patients by 2016.
Conclusions: Over the past decade, spine surgery has decreased in complications for IN and OUT procedures along with IN/OUT knee, ankle, hip, and shoulder procedures, reflecting greater tolerance for risk in an outpatient setting.
Level of Evidence: 3.
Clinical Relevance: Despite the increase in riskier spine procedures, complications have decreased over the years. Surgeons should aim to continue to decrease inpatient spine complications to the level of other orthopedic surgeries.
INTRODUCTION
It is estimated that the incidence of spinal pathologies, such as degenerative disc disease, is on the rise in the United States, with a proportional increase in the number of procedures being performed.1,2 In recent decades, the number of lumbar procedures has more than doubled, with a similar surge seen in the number of cervical procedures being performed.3–5 These figures are of concern, as the disparity in healthcare expenditures continues to overwhelm hospital systems. Furthermore, with a shift in reimbursements toward bundle payments in the United States, there is further pressure on hospitals and surgeons to further reduce healthcare expenditures.6
There has been much research into improving patient outcomes with the use of various methodologies, such as preoperative patient optimization, developing risk prediction models, and mitigating anticipated complications postoperatively.7–11 The synergistic effect of these modalities is an attempt to reduce patient morbidity and mortality as well as to reduce healthcare costs. An additional method of controlling costs is to transition from inpatient to outpatient procedures to reduce the high costs of hospital bed per day as well as supplies.12 The development of protocols for transitioning from outpatient to inpatient orthopedic procedures has seen tremendous success for various orthopedic procedures, such as total joint arthroplasty.13–18 These studies have found ambulatory surgery to be safe and effective and may potentially result in a reduction of nosocomial complications when compared to inpatient procedures.
Similarly, there has been a trend toward outpatient spinal procedures, many of which show it to be a feasible option to offset hospital costs without compromising patient safety.19–22 There is generally an inherently higher risk profile in patients undergoing spinal procedures when compared to other divisions of orthopedic surgeries.23 Although there has been much research comparing the safety and efficacy of inpatient versus outpatient spinal procedures, there is very limited (if any) literature comparing the outcomes of this transition to additional orthopedic procedures. Therefore, the purpose of this study was to assess outcomes and improvements in spinal inpatient and outpatient surgery when compared to other orthopedic procedures.
METHODS
Study Design
This is a retrospective cohort study of the American College of Surgeons National Surgical Quality of Improvement Program (ACS-NSQIP). The ACS-NSQIP tracks and audits 30-day perioperative outcomes of surgical patients across more than 400 medical institutions across the United States, and its sampling protocol has previously been described in the literature.7 The database includes demographic, comorbidity, surgical, and preoperative laboratory data as well as complication outcomes. This study includes data from the years 2005–2016. As the ACS-NSQIP data set is deidentified and contains no geographic markers, this study was exempt from local institutional review board review. More information about the ACS-NSQIP program can be accessed at https://www.facs.org/quality-programs/acs-nsqip/about.
Inclusion Criteria
All patients included in this study were adults (age > 18 years) from the ACS-NSQIP database who underwent different types of orthopedic surgeries, such as spine, knee, ankle, shoulder, and hip, as defined via current procedural terminology.
Statistical Analysis
Descriptive statistics provided an overview of patient demographics, comorbidity burden (including a previously published NSQIP-modified Charlson Comorbidity Index,8 surgical factors, and complications. NSQIP is able to track whether a procedure is elective. As the vast majority of outpatient procedures are elective, to more clearly compare the impact of inpatient versus outpatient stay, only elective and nonemergency procedures were included. The NSQIP contains a variable specifically for ambulatory status defined as either inpatient or outpatient. We used this variable to separate inpatient cases from outpatient cases. Individual complications were analyzed among each orthopedic surgery both for its outpatient and for its inpatient components as well as other perioperative complications (return to the operating room, morbidity, and readmission). A subanalysis assessed complication rates when specific patient-specific demographics presented at baseline were controlled for with binary logistic regressions. Descriptive, mean comparison, and binary logistic regression analyses were conducted in SPSS (version 23.0, Armonk, SPSS Inc, New York).
RESULTS
Cohort Overview
A total of 729 480 surgical patients were included: 32.5% received spinal surgery, 36.5% knee, 24.1% hip, 4.9% shoulder, and 1.7% ankle. By basic demographics, hip patients were the oldest, and knee patients had the highest body mass index out of the orthopedic groups (Table 1; P < .001). The top spine procedures were fusions (53.2%), decompressions (54.7%), laminotomies (15.4%), and laminectomies (6.5%).
Breaking Down by Ambulatory Setting
Table 2 displays the breakdown of inpatients (IN) and outpatients (OUT) for these orthopedic surgeries. Spine IN patients experienced more complications than the other orthopedic procedures (14% vs 10.2%; P < .001); however, spine OUT procedures had a lower complication rate (6.3% vs 10.8%; P < .001). A breakdown of type of complications by specific orthopedic procedures is shown in Table 3. Calculated ratios were analyzed for the rate of complication of switching from inpatient to outpatient. Switching spine patients from inpatient to outpatient had an 18.1% decrease in the number of postoperative complications. This switch in ambulatory setting was not identified to be beneficial for other orthopedic surgeries with a 35.5% increase in postoperative complications.
Trends of Orthopedic Surgeries by Setting
From 2006 to 2016, IN spine procedure complications, decreased while other orthopedic surgeries increased. However, during this time, OUT spine complications increased, while other orthopedic complications decreased, as shown in Table 4.
Complication Rate: Spine Versus All Orthopedic Surgeries
Compared to other orthopedic procedures, spine patients had the greatest overall complication rate (12.1% vs 10.3%; P < .001). More specifically, spine patients had greater rates of reoperation (3% vs 2%), sepsis (0.7% vs 0.4%), urinary tract infections (1.3% vs 1.1%), and deep surgical site infections (0.6% vs 0.3%; all P < .001). Table 3 breaks down specific postoperative complications by orthopedic specialties when compared to spine procedures. Multivariate analysis identified postoperative complications associated with various orthopedic surgeries. Table 5 displays significant IN procedures that are independent predictors for developing a complication. After controlling for age and Charlson Comorbidity Index, IN spine patients were significantly associated with postoperative complications (odds ratio [OR]: 0.42; confidence interval [CI]: 0.40–0.43; P < .001) as well as IN ankle patients (OR: 0.54; CI 0.48–0.61; P < .001). OUT knee patients were identified to be significantly associated with this outcome (OR: 1.1; CI: 1.08–1.2; P < .001; Table 5).
Specific Spine Procedures by Complications
Spine procedures, such as fusions and laminectomies, when compared against other orthopedic surgeries, had greater rates of complications, as shown in Table 6. However, decompressions were significantly different only for sepsis complications (0.5% vs 0.4%) and had a lower overall complication rate (10% vs 10.3%; all P < .05). When looking at arthroscopic rotator cuff repair against cervical spine fusions, the latter had greater rates of readmission, reoperation, surgical site infection, sepsis, and overall complications (all P < .05; Table 6). A similar trend was also seen when comparing total hip arthroplasties and lumbar spine procedures.
DISCUSSION
There has been a marked increase in the number of spine procedures being performed in the United States as well as worldwide, which is estimated to have more than doubled in the previous decades.3–5,24–26 Additionally, there has been a concomitant net increase in healthcare costs for these procedures, with estimates of over $35 billion from 2001 to 2010 in the United States.27,28 One method of reducing this expenditure would be to remove the most obviously exorbitant cost of an inpatient hospital bed, which is estimated to be over $5000 per day, and performing procedures on an outpatient basis.12 As a transition takes place toward ambulatory cervical and lumbar procedures, it is important to understand their safety when compared to inpatient procedures as well as in comparison to other orthopedic procedures.29–31 Our study found that spine procedures were associated with higher complications in the inpatient setting and lower complications in outpatient cases when compared to other orthopedic procedures.
Although there are no level I or II studies assessing the safety of inpatient versus outpatient spine procedures, there has been some literature demonstrating it to be a safe medium for surgery.13,15–18,20,21,32,33 A systematic review of 56 articles evaluating ambulatory spine surgery found that a majority of studies reported no differences in peri- and postoperative complications, readmissions, and reoperations when compared to inpatient cases.19,33,34 These findings indicate that outpatient spine cases may be a safe alternative to the traditional inpatient cases. Furthermore, a meta-analysis of studies evaluating cervical and lumbar spine surgeries, with a total of 370,195 patients, demonstrated that outpatient procedures had significantly lower overall costs when compared to inpatient procedures.22 When substratified by age, the study showed that costs for younger patients were markedly lower than procedures for older patients ($555 vs $7290, respectively). This discrepancy in costs is likely attributed to generally higher comorbidity and therefore risk in older patients undergoing surgery.35
Similar trends have been observed in other outpatient orthopedic procedures when compared to inpatient procedures, which showed similar postoperative complications and lower overall costs. A study by Darrith et al16 evaluated 243 consecutive patients undergoing outpatient arthroplasty procedures and compared them to a matched cohort undergoing inpatient arthroplasty and tracked them for 90 days. The authors effectively demonstrated no significant differences when comparing readmission rates, major and minor complications, reoperations, emergency room visits, and unplanned clinic visits between the 2 cohorts.
However, some studies have even showed better overall outcomes of outpatient procedures when compared to inpatient cases. For example, a study by Huntley et al36 reviewed 7672 patients undergoing foot and ankle surgery using the ACS-NSQIP database. When compared to outpatient cases, the authors found that patients undergoing inpatient procedures had significantly higher surgical (8.6% vs 2.0%) and medical complications (16.9% vs 1.7%). These findings appear promising for surgeons considering outpatient procedures as a safe alternative to inpatient cases; however, additional findings in this study highlight the importance of underlying patient demographics and comorbidities as the true deciding components of overall outcomes. Unsurprisingly, the inpatient group represented a higher-risk population when compared to the outpatient group, with a higher mean age, more smokers, poorer functional status, and, most important, a disproportionately higher prevalence of every comorbidity. This is similar to the findings of our study, as patients who had higher comorbidities, which occur often in the spine patient population, had more complications when compared to other orthopedic procedures. This strongly suggests the role and importance of careful patient selection when considering outpatient procedures as well as a potentially existing bias inherent in current studies evaluating these 2 modalities.
There were some limitations in our study, particularly the retrospective nature. Furthermore, there are several inherent known limitations with the use of the NSQIP database, with a reliance on the accuracy of ICD-9 and CPT codes to identify patient outcomes.37 There may have been coding inconsistencies that potentially resulted in an under- or overrepresentation of effects and outcomes. Additionally, as many complications and reoperations can occur for up to 1 year from the index procedure, many of these may be missed since the database allows patient tracking for only 30 days postoperatively. In spite of some of these limitations, we feel that the use of a large database allows us to have adequate power to potentially overcome biases and coding errors to effectively assess the main question of our study.
The pressures to transition from inpatient to outpatient spine procedures appear to be multifactorial, influenced also by changes in reimbursement as bundled payment models.6 Furthermore, as the Centers for Medicare and Medicaid Services transitions from inpatient-only procedures to outpatient, insurance companies are also denying preauthorization for inpatient stays following surgery, which may translate to spine surgeries.32 As outpatient procedures may be riskier due to less postoperative observation, the advent of additional patient optimization strategies and selection is ever important. Additionally, the use of ambulatory procedures has resulted in the development and refinement of minimally invasive techniques in an attempt to further mitigate and reduce complications.29,38
CONCLUSIONS
The results of our study found that inpatient spine surgery patients had higher complications than outpatient when compared to other orthopedic cases. It is possible that inpatient procedures carry with them an inherent sense of safety and comfort with constant medical observation, possibly resulting in less attention to minute patient risk factors and therefore operating on higher-risk patients, which is in contrast to outpatient procedures potentially requiring a more careful evaluation, more aggressive intervention, and improved risk mitigation. Further studies are needed not only to normalize outcomes within a surgical specialty but also to explore variations in subspecialty patient risks as an important means of understanding the multifactorial components contributing to patient complications and hospital costs following any orthopedic procedure, both inpatient and outpatient.
Footnotes
Disclosures and COI: Institutional review board approval was waived, as this study uses a deidentified nationwide database.
- This manuscript is generously published free of charge by ISASS, the International Society for the Advancement of Spine Surgery. Copyright © 2021 ISASS