Anterior cervical diskectomy and fusion (ACDF) is a common option for treating cervical disk disease.1,2 Indeed, the number of ACDF procedures performed each year has increased since 1990.1–3 The increase in the number of procedures has been accompanied by an increase in average patient age and average number of medical comorbidities.1–3 Given the growing prevalence of this procedure, as well as the increasingly complicated patient population, it is more important than ever to understand and minimize the risk of postoperative adverse events.
Surgical duration is an often-studied risk factor for complications following surgical procedures. It has been shown to be independently associated with adverse events in a variety of settings, including cardiac surgery, neurosurgery, bariatric surgery, and urological procedures.4–8 Orthopedic surgeons have also shown interest in this potentially modifiable risk factor, with studies mentioning its association with adverse events involving posterior spinal fusion, knee arthroplasty, calcaneus fracture repairs, and tibial plateau fracture repairs.9–12
Although ACDF studies involving adverse events have noted that increased operative time (as a binary variable) is associated with the total rate of adverse events, they have not attempted to quantify the increased risk associated with greater operative duration across the spectrum of operative times or been specific about which postoperative events are increased.13,14
The purpose of this study was to test for relationships between operative time (as an interval variable) and specific general health postoperative adverse events among patients undergoing ACDF. The analysis was restricted to 1-level procedures and employed appropriate multivariate adjustment to limit the extent that observed associations were due to confounding by baseline or procedural characteristics.
Materials and Methods
This study was granted an exemption from review by the authors' institutional review board. A retrospective cohort study was conducted using the American College of Surgeons National Surgical Quality Improvement Program (NSQIP). The NSQIP is a national surgical safety and data-monitoring program that prospectively identifies patients undergoing major surgical procedures at more than 500 institutions nationwide. The program characterizes patients' baseline characteristics and then follows patients for the development of complications during the 30 days after surgery, regardless of discharge status. The NSQIP data undergo regular and rigorous quality audits and have been shown to have a combined disagreement rate of approximately 2% across all audits.15 The NSQIP has been used extensively in orthopedic research.10,12–14,16,17
Patients undergoing ACDF during 2005 to 2015 as part of the NSQIP were identified using Current Procedural Terminology codes 22551, 22554, and 63075. Multilevel procedures and cervical corpectomies were excluded by eliminating cases that contained the secondary Current Procedural Terminology codes 22552 or 63076 and 63081 or 63082, respectively. Additionally, patients undergoing auto-graft harvest via a separate incision were removed by eliminating cases with the secondary Current Procedural Terminology code 20937. Finally, patients with diagnosis codes for traumatic, oncologic, or infectious indications were also excluded.
In terms of baseline characteristics, the NSQIP includes patient demographics such as age and sex. Height and weight information was used to calculate body mass index. Comorbidities, including diabetes, congestive heart failure, dyspnea on exertion, hypertension, end-stage renal disease, chronic obstructive pulmonary disease, smoking status, and anemia, were identified. Functional status was classified as independent vs dependent. American Society of Anesthesiologists classes 1 to 4 were used. Specific information regarding variable definitions is listed in the NSQIP materials.15
Regarding the independent variable, operative time was defined as time of incision to time of closure. This was available as a continuous variable (in minutes) in the NSQIP data set. Operative time (in minutes) was divided by 15 such that the risk associated with a 15-minute increase in operative time could be directly reported.
Adverse event information collected by the NSQIP and used for this study included mortality, myocardial infarction, anemia requiring transfusion, unplanned intubation, pneumonia, urinary tract infection, surgical site infection (including superficial, deep, and organ/space), sepsis (including systemic sepsis and septic shock), wound dehiscence, and venous thromboembolism (including deep venous thrombosis and pulmonary embolism). These groupings have been used previously in the NSQIP orthopedic literature.18,19 In addition, studied hospital metrics included hospital length of stay and hospital readmission.15 Length of stay was dichotomized as normal or extended, with extended hospital length of stay being defined as greater than the 75th percentile length of stay (>1 day).
Statistical analyses were conducted using Stata version 13.1 software (Stata Corp LLP, College Station, Texas). Multivariate regression was employed so that associations could be reported as the increased risk associated with an increase in surgical duration.
Each regression model had the binary complication or hospital metric as the outcome variable, a primary predictor variable of operative time (in minutes) divided by 15-minute intervals, and control variables, including age, sex, body mass index, and the presence/absence of diabetes, congestive heart failure, dyspnea on exertion, hypertension, end-stage renal disease, chronic obstructive pulmonary disease, smoking status, and anemia. The regression technique used was Poisson regression with robust error variance, an alternative to log-binomial regression that enables direct estimation of relative risk in cohort studies.20
Because multiple statistical comparisons were conducted in this study, a Bonferroni correction was employed to set the level of significance at P<.0036.
A total of 15,241 patients met inclusion criteria. The demographics of this population are presented in Table 1. The average operative time (±SD) was 100.2 minutes (±39.3 minutes). Figure 1 is a histogram of operative times.
Histogram of operative times.
The overall rates of each adverse event and adverse hospital metric are presented in Table 2. Nearly 2% of patients experienced an adverse event. Of note, 21 (31.82%) of the 66 patients experiencing a surgical site infection required a reoperation. Figure 2 shows a trend of rising rates of occurrence of any adverse event (Figure 2A), extended hospital stay (Figure 2B), and readmission (Figure 2C) with 15-minute increases in surgical duration. For all 3 adverse outcomes, increases in rates were observed with nearly all 15-minute intervals across the studied operative times.
Adverse Event and Hospital Metric Rates (N=15,241)
Rates of occurrence of any adverse event (A), extended hospital stay (B), and readmission (C), for 15-minute intervals.
Following adjustment for demographic and comorbidity characteristics, an increase in the operative time by 15-minute intervals was associated with an increase in the risk for the occurrence of “any adverse event” by 10% (99.64% confidence interval, 3%–17%, P<.001; Table 3). More specifically, each 15-minute increase in surgical duration was associated with an increase in the risk for venous thromboembolism by 19% (99.64% confidence interval, 4%–37%, P<.001), sepsis by 44% (99.64% confidence interval, 3%–103%, P=.002), and unplanned intubation by 17% (99.64% confidence interval, 3%–33%, P<.001; Table 3).
Impact of Operative Time on Risk for Adverse Outcomes
On analysis of the hospital metrics, each 15-minute increase in surgical duration was associated with an increase in the risk for extended hospital stay by 14% (99.64% confidence interval, 12%–16%, P<.001) and risk for hospital readmission by 7% (99.64% confidence interval, 1%–14%, P=.001).
There has been increasing emphasis on understanding and minimizing the rate of adverse events after surgical interventions such as ACDF. Prior work has identified overall associations of some complications with longer cases,13,14 with time having been considered a dichotomous variable.
A study by Gruskay et al13 found that ACDF operative time greater than the 75th percentile was associated with an increase in the risk for a “major complication” (odds ratio, 2.1). These findings were echoed by Schoenfeld et al,14 who found that surgical duration 1 SD above the mean for a variety of spinal arthrodesis procedures was associated with an increase in the rate of occurrence of 1 or more complications (odds ratio, 2.2).
As with other research, these 2 studies analyzed operative time using a Boolean approach, which is an appropriate means of testing for an association. However, a Boolean approach relies on 1 arbitrary cutoff point and does not account for the fact that an association may occur over the spectrum of operative times, not just near that cutoff. Moreover, it fails to quantify the increase in risk associated with a specific change in operative time over the spectrum of operative times.
Building on such prior studies, the current study evaluated the effect of incrementally longer operative times on general health outcomes after ACDF. Operative time was found to have a strong and independent association with adverse events and hospital metrics. Specifically, incremental increases in operative time were independently associated with increases in the occurrence of overall general health adverse events, extended hospital stay, and readmission, as well as venous thromboembolism, sepsis, and unplanned intubation. These increased risks represent the average increase in risk for patients, for 15-minute increases in surgical duration, across the spectrum of studied times. These findings are clinically applicable as strategies including defined orthopedic operating rooms and staff, decreased team substitutions, and the use of more efficient surgical techniques can all lead to incremental decreases in operative times.21–23
The association of the individual and overall adverse outcomes identified is consistent with the literature.13,14 The correlation of operative time with venous thromboembolism (deep venous thrombosis and/or pulmonary embolism) may be understood as sequelae of longer immobility during surgery.16,24 The association of operative time with an increased risk of sepsis may be related to patients undergoing longer surgeries being at increased risk for source infections. For example, longer retraction times can increase local tissue trauma and necrosis, leading to a potential infection source.25 Additionally, longer surgeries may be accompanied by longer durations of impaired thermal regulation, which have been associated with complications in elective spine surgeries.26–29
The association of operative time with unplanned intubation is likely multifactorial. Although the occurrence of other complications (eg, sepsis, venous thromboembolism) may contribute to this, it is most likely associated with increased edema related to the anterior cervical approach, which is known to be a greater factor in longer cases.30–33
In addition to adverse events, this study identified associations between increased surgical duration and both extended hospital stay and hospital readmission. The data suggested that an increase in surgical length by 15 minutes led to a 14% increase in risk for extended hospital stay and a 7% increase in risk for hospital readmission. Gruskay et al13 found that operative time in the 75th percentile (>171 minutes) was associated with an increased length of stay by approximately 0.7 days. Again, although aligning the 2 studies is imprecise because of their differing methodologies, the 2 studies are certainly consistent.
To the authors' knowledge, the current study is the first to independently examine and identify the relationship between increased operative time and the risk of hospital readmission. Notably, Lovecchio et al17 examined factors predicting readmission in 2261 patients undergoing ACDF or anterior corpectomy and fusion procedures. They could not identify a statistically significant relationship between operative time and hospital readmission.
The findings of this study are clinically compelling, as the increased risk for adverse events and hospital metrics are all associated with just a 15-minute increase in operative time. Interestingly, a variety of strategies have been studied and shown to improve operative time efficiency in the range of this magnitude. Small et al21 reported that dedicated orthopedic operating rooms and staff decreases operative time by 7 minutes following total joint arthroplasty. Cassera et al22 found that an increase in the number of team members involved in surgery by 1 individual, through mechanisms such as staff substitutions, led to an average lengthening of procedure time by 15 minutes for a variety of laparoscopic procedures. Additionally, technical modifications with a procedure can be associated with decreases in operative time. For example, the use of minimally invasive iliac crest bone harvesting has been shown to decrease operative time by 9 minutes compared with the traditional open window technique.23
Limitations of this study are those of the NSQIP. The program does not capture information on patient symptoms or radiographic findings. Additionally, the NSQIP does not assemble data regarding procedure-specific complications such as postoperative dysphagia, which has additionally been shown to be associated with operative time.34 This limits the analysis to nonspecific, general medical and surgical complications. Certainly not all ACDF procedures are of the same complexity, and greater surgical complexity may have been a confounder for extended operative time that could not be controlled. Nonetheless, restricting the analysis to 1-level procedures and excluding traumatic, oncologic, and infectious indications was done to limit the variability of cases studied. Furthermore, in comparison with nonparticipating hospitals, a larger percentage of NSQIP institutions are academic centers.35 Although impossible to completely account for this, the current study controlled for varying patient populations by including age, sex, body mass index, and the presence/absence of 8 comorbidities in all regressions. Finally, as with all data sets, the accuracy of the database is a potential source of error. However, the NSQIP undergoes routine quality audits and is thought to be of significantly greater data quality than administratively coded databases.15
This study used rigorous methodology to reveal and quantify independent associations between operative duration and several specific postoperative complications. The interval variable approach to analyzing operative time presents major advantages, has not been used previously in the spinal literature, and provides convincing evidence that incremental increases in operative duration independently modulate the risk for several specific adverse events following ACDF. Based on these data, efforts should be made by hospitals and surgeons alike to minimize the duration of surgical interventions.
- Marawar S, Girardi FP, Sama AA, et al. National trends in anterior cervical fusion procedures. Spine (Phila Pa 1976). 2010; 35(15):1454–1459. doi:10.1097/BRS.0b013e3181bef3cb [CrossRef]
- Oglesby M, Fineberg SJ, Patel AA, Pelton MA, Singh K. Epidemiological trends in cervical spine surgery for degenerative diseases between 2002 and 2009. Spine (Phila Pa 1976). 2013; 38(14):1226–1232. doi:10.1097/BRS.0b013e31828be75d [CrossRef]
- Patil PG, Turner DA, Pietrobon R. National trends in surgical procedures for degenerative cervical spine disease: 1990–2000. Neurosurgery. 2005; 57(4):753–758. doi:10.1227/01.NEU.0000175729.79119.1d [CrossRef]
- Totonchi Z, Baazm F, Chitsazan M, Seifi S, Chitsazan M. Predictors of prolonged mechanical ventilation after open heart surgery. J Cardiovasc Thorac Res. 2014; 6(4):211–216. doi:10.15171/jcvtr.2014.014 [CrossRef]
- Ishihara H, Ishihara S, Niimi J, et al. Risk factors for coil protrusion into the parent artery and associated thromboembolic events following unruptured cerebral aneurysm embolization. Interv Neuroradiol. 2015; 21(2):178–183. doi:10.1177/1591019915582375 [CrossRef]
- Abraham CR, Werter CR, Ata A, et al. Predictors of hospital readmission after bariatric surgery. J Am Coll Surg. 2015; 221(1):220–227. doi:10.1016/j.jamcollsurg.2015.02.018 [CrossRef]
- Chen SY, Stem M, Schweitzer MA, Magnuson TH, Lidor AO. Assessment of postdischarge complications after bariatric surgery: a National Surgical Quality Improvement Program analysis. Surgery.2015; 158(3):777–786. doi:10.1016/j.surg.2015.04.028 [CrossRef]
- Matulewicz RS, Sharma V, McGuire BB, Oberlin DT, Perry KT, Nadler RB. The effect of surgical duration of transurethral resection of bladder tumors on postoperative complications: an analysis of ACS NSQIP data. Urol Oncol. 2015; 33(8):338. doi:10.1016/j.urolonc.2015.05.011 [CrossRef]
- Kang J, Jiang X, Wu B. Analysis of risk factors for lower-limb deep venous thrombosis in old patients after knee arthroplasty. Chin Med J (Engl). 2015; 128(10):1358–1362. doi:10.4103/0366-6999.156782 [CrossRef]
- Basques BA, Bohl DD, Golinvaux NS, Smith BG, Grauer JN. Patient factors are associated with poor short-term outcomes after posterior fusion for adolescent idiopathic scoliosis. Clin Orthop Relat Res. 2015; 473(1):286–294. doi:10.1007/s11999-014-3911-4 [CrossRef]
- Wu K, Wang C, Wang Q, Li H. Regression analysis of controllable factors of surgical incision complications in closed calcaneal fractures. J Res Med Sci. 2014; 19(6):495–501.
- Basques BA, Webb ML, Bohl DD, Golinvaux NS, Grauer JN. Adverse events, length of stay, and readmission after surgery for tibial plateau fractures. J Orthop Trauma. 2015; 29(3):e121–e126. doi:10.1097/BOT.0000000000000231 [CrossRef]
- Gruskay JA, Fu M, Basques B, et al. Factors affecting length of stay and complications after elective anterior cervical discectomy and fusion: a study of 2164 patients from the American College of Surgeons National Surgical Quality Improvement Project Database (ACS NSQIP). Clin Spine Surg. 2016; 29(1):E34–E42. doi:10.1097/BSD.0000000000000080 [CrossRef]
- Schoenfeld AJ, Carey PA, Cleveland AW III, Bader JO, Bono CM. Patient factors, comorbidities, and surgical characteristics that increase mortality and complication risk after spinal arthrodesis: a prognostic study based on 5,887 patients. Spine J. 2013; 13(10):1171–1179. doi:10.1016/j.spinee.2013.02.071 [CrossRef]
- American College of Surgeons. User guide for the 2013 ACS NSQIP participant use data file. https://www.facs.org/~/media/files/quality%20programs/nsqip/2013_acs_nsqip_puf_user_guide.ashx. Published November 2014. Accessed June 1, 2017.
- Kim BD, Hsu WK, De Oliveira GS Jr, Saha S, Kim JY. Operative duration as an independent risk factor for postoperative complications in single-level lumbar fusion: an analysis of 4588 surgical cases. Spine (Phila Pa 1976). 2014; 39(6):510–520. doi:10.1097/BRS.0000000000000163 [CrossRef]
- Lovecchio F, Hsu WK, Smith TR, Cybulski G, Kim B, Kim JY. Predictors of thirty-day readmission after anterior cervical fusion. Spine (Phila Pa 1976). 2014; 39(2):127–133. doi:10.1097/BRS.0000000000000051 [CrossRef]
- Webb ML, Nelson SJ, Save AV, et al. Of 20,376 lumbar discectomies, 2.6% of patients readmitted within 30 days: surgical site infection, pain, and thromboembolic events are the most common reasons for readmission. Spine (Phila Pa 1976). 2017; 42(16):1267–1273. doi:10.1097/BRS.0000000000002014 [CrossRef]
- Bohl DD, Sershon RA, Fillingham YA, Della Valle CJ. Incidence, risk factors, and sources of sepsis following total joint arthroplasty. J Arthroplasty. 2016; 31(12):2875–2879. doi:10.1016/j.arth.2016.05.031 [CrossRef]
- Zou G. A modified Poisson regression approach to prospective studies with binary data. Am J Epidemiol. 2004; 159(7):702–706. doi:10.1093/aje/kwh090 [CrossRef]
- Small TJ, Gad BV, Klika AK, Mounir-Soliman LS, Gerritsen RL, Barsoum WK. Dedicated orthopedic operating room unit improves operating room efficiency. J Arthroplasty. 2013; 28(7):1066–1071. doi:10.1016/j.arth.2013.01.033 [CrossRef]
- Cassera MA, Zheng B, Martinec DV, Dunst CM, Swanström LL. Surgical time independently affected by surgical team size. Am J Surg. 2009; 198(2):216–222. doi:10.1016/j.amjsurg.2008.10.016 [CrossRef]
- Burstein FD, Simms C, Cohen SR, Work F, Paschal M. Iliac crest bone graft harvesting techniques: a comparison. Plast Reconstr Surg. 2000; 105(1):34–39. doi:10.1097/00006534-200001000-00006 [CrossRef]
- Sebastian AS, Currier BL, Kakar S, et al. Risk factors for venous thromboembolism following thoracolumbar surgery: analysis of 43,777 patients from the American College of Surgeons National Surgical Quality Improvement Program 2005 to 2012. Global Spine J. 2016; 6(8):738–743. doi:10.1055/s-0036-1579553 [CrossRef]
- Xing D, Ma JX, Ma XL, et al. A methodological, systematic review of evidence-based independent risk factors for surgical site infections after spinal surgery. Eur Spine J.2013; 22(3):605–615. doi:10.1007/s00586-012-2514-6 [CrossRef]
- Kurz A. Physiology of thermoregulation. Best Pract Res Clin Anaesthesiol. 2008; 22(4):627–644. doi:10.1016/j.bpa.2008.06.004 [CrossRef]
- Kurz A, Sessler DI, Lenhardt R. Perioperative normothermia to reduce the incidence of surgical-wound infection and shorten hospitalization: study of Wound Infection and Temperature Group. N Engl J Med. 1996; 334(19):1209–1215. doi:10.1056/NEJM199605093341901 [CrossRef]
- Billeter AT, Rice J, Druen D, et al. Warming to 39°C but not to 37°C ameliorates the effects on the monocyte response by hypothermia. Ann Surg. 2016; 263(3):601–607. doi:10.1097/SLA.0000000000001175 [CrossRef]
- Guest JD, Vanni S, Silbert L. Mild hypothermia, blood loss and complications in elective spinal surgery. Spine J. 2004; 4(2):130–137. doi:10.1016/j.spinee.2003.08.027 [CrossRef]
- Jeyamohan SB, Kenning TJ, Petronis KA, Feustel PJ, Drazin D, DiRisio DJ. Effect of steroid use in anterior cervical discectomy and fusion: a randomized controlled trial. J Neurosurg Spine. 2015; 23(2):137–143. doi:10.3171/2014.12.SPINE14477 [CrossRef]
- Song KJ, Lee SK, Ko JH, Yoo MJ, Kim DY, Lee KB. The clinical efficacy of short-term steroid treatment in multilevel anterior cervical arthrodesis. Spine J. 2014; 14(12):2954–2958. doi:10.1016/j.spinee.2014.06.005 [CrossRef]
- Daniels AH, Riew KD, Yoo JU, et al. Adverse events associated with anterior cervical spine surgery. J Am Acad Orthop Surg. 2008; 16(12):729–738. doi:10.5435/00124635-200812000-00005 [CrossRef]
- Sagi HC, Beutler W, Carroll E, Connolly PJ. Airway complications associated with surgery on the anterior cervical spine. Spine (Phila Pa 1976). 2002; 27(9):949–953. doi:10.1097/00007632-200205010-00013 [CrossRef]
- Cho SK, Lu Y, Lee DH. Dysphagia following anterior cervical spinal surgery: a systematic review. Bone Joint J. 2013; 95-B(7):868–873. doi:10.1302/0301-620X.95B7.31029 [CrossRef]
- Sheils CR, Dahlke AR, Kreutzer L, Bilimoria KY, Yang AD. Evaluation of hospitals participating in the American College of Surgeons National Surgical Quality Improvement Program. Surgery. 2016; 160(5):1182–1188. doi:10.1016/j.surg.2016.04.034 [CrossRef]
| 18–49 y||6593 (43.3)|
| 50–59 y||4717 (30.9)|
| 60–69 y||2699 (17.7)|
| ≥70 y||1232 (8.1)|
| Male||7589 (49.8)|
| Female||7652 (50.2)|
|Body mass index|
| ≤24 kg/m2||3262 (21.4)|
| 25–29 kg/m2||5177 (34.0)|
| ≥30 kg/m2||6802 (44.6)|
| Independent||15,005 (98.5)|
| Dependent||236 (1.5)|
| No diabetes||13,289 (87.2)|
| Non–insulin-dependent diabetes mellitus||1241 (8.1)|
| Insulin-dependent diabetes mellitus||711 (4.7)|
|Dyspnea on exertion||748 (4.9)|
|Chronic obstructive pulmonary disease||520 (3.4)|
|Current smoker||4500 (29.5)|
|American Society of Anesthesiologists class|
| 1||815 (5.3)|
| 2||8980 (58.9)|
| 3||5225 (34.3)|
| 4||221 (1.5)|
Adverse Event and Hospital Metric Rates (N=15,241)
|Adverse Outcome||No. (%)|
|Any adverse event||276 (1.81)|
| Mortality||9 (0.06)|
| Myocardial infarction||12 (0.08)|
| Anemia requiring transfusion||17 (0.11)|
| Unplanned intubation||63 (0.41)|
| Pneumonia||51 (0.33)|
| Urinary tract infection||67 (0.44)|
| Surgical site infection||66 (0.43)|
| Sepsis||5 (0.03)|
| Wound dehiscence||6 (0.04)|
| Venous thromboembolism||42 (0.28)|
|Adverse hospital metric|
| Extended hospital stay (>1 day)||3258 (21.38)|
| Hospital readmission||309 (2.03)|
Impact of Operative Time on Risk for Adverse Outcomes
|Adverse Outcome||Impact of a 15-Minute Increase in Operative Time on Risk for Adverse Outcomes|
|Relative Risk||99.64% Confidence Interval||P|
|Any adverse event||1.10||1.03–1.17||<.001a|
| Myocardial infarction||0.98||0.74–1.29||.798|
| Anemia requiring transfusion||1.06||0.83–1.36||.492|
| Unplanned intubation||1.17||1.03–1.33||<.001a|
| Urinary tract infection||1.03||0.90–1.18||.508|
| Surgical site infection||1.10||0.97–1.25||.032|
| Wound dehiscence||0.95||0.71–1.26||.591|
| Venous thromboembolism||1.19||1.04–1.37||<.001a|
|Adverse hospital metric|
| Extended hospital stay (>1 day)||1.14||1.12–1.16||<.001a|
| Hospital readmission||1.07||1.01–1.14||.001a|