Minimally invasive (MIS) and mini-open spine surgery are becoming more commonplace with technological advances in medicine. Although these procedures are technically more difficult compared to traditional open approaches, they offer many advantages, such as minimizing iatrogenic surrounding soft-tissue damage, atrophy from muscle denervation and retraction, detachment of tendons, and less postoperative pain and complications.1–4 Mini-open procedures are cost effective while providing equivalent improvements in quality adjusted life years through shorter hospital length of stays (LOS), faster decrease in narcotic dependence, and ability to return to work sooner compared to open procedures.5
Mini-open transforaminal lumbar interbody fusion (TLIF) accomplishes solid lumbar interbody fusion through muscle-splitting, tubular retraction, thus protecting the paraspinous muscle complex from injury.6,7 This technically demanding procedure has been successfully used in the treatment of spondylolisthesis, degenerative disc disease, and recurrent disc herniation.8–10 The results of retrospective and prospective cohort studies are promising and suggest advantages such as faster recovery, reduced blood loss, and shorter hospitalization with outcomes comparable to open procedures up to 2 yr postoperatively.5,11–14
Current literature is limited concerning the learning curve associated with mini-open TLIF. The learning curve consists of two phases: the initial improvement phase and the subsequent plateau phase associated with repeated performance of a new task and skill acquisition. Awareness of the learning curve is important for training, patient care, and eventually the widespread acceptance of a new procedure. Numerous studies have validated the LOS15,16 and operative time17–19 in quantifying the learning curve of minimally invasive surgery. Concurrent analysis of outcomes related to the learning curve is equally important. If there is a steep learning curve, postoperative morbidity may be disproportionately high in early cases.
In addition to the effect of a surgeon’s proficiency on LOS, analysis of other factors that correlate to LOS would allow us to combine the cost effectiveness of the novel procedure with proper allocation of limited hospital resources. Identified patients with increased LOS would benefit from preoperative counseling and augmented level of perioperative care.
Our objective in this study was twofold. The first purpose was to evaluate the learning curve and early outcomes associated with mini-open TLIF. The second purpose was to identify factors that affect postoperative LOS.
MATERIALS AND METHODS
Institutional review board approval was obtained for this study. From 2009 to 2014, 100 consecutive patients who underwent a one-level mini-open TLIF by a single surgeon were identified. The surgical indications were degenerative discs or spinal stenosis, with grade 1 or 2 spondylolisthesis presenting with mechanical lower back pain and radicular symptoms.
Retrospective chart review and analysis of operative time, LOS, EBL, Oswestry Disability Index (ODI), visual analog pain scores (VAS), fusion status, age, sex, body mass index (BMI), prior lumbar surgeries, social support, smoking status, grade of spondylolisthesis, preoperative hemoglobin, Charlson comorbidity index, complication rates, and time to physical therapy were performed. Operative time was defined from skin incision to end of closure. LOS was defined as time from leaving the operating room to vacating the hospital bed. Clinical assessments were performed using the ODI and VAS for back and leg, before surgery, and 3 mo postoperatively. Clinical notes after the 3-month follow-up visit were assessed for subjective improvement in pain.
Radiographic assessment of fusion used static anteroposterior (AP) and lateral lumbar views taken at least 1 yr after surgery. Fusion rates were based on the Bridwell classification (Table 1),20 which has been commonly used to assess fusion status in TLIFs.12,21,22 These were reviewed by the primary author and the primary surgeon. When there was disagreement between the two grades, a discussion was had between the two reviewers until an agreement was reached concerning the grading.
By way of comparing operation time and LOS, we determined the proficiency point of the learning curve. Based on the proficiency point, we made two groups: group A (during the learning curve) and group B (after learning curve). We compared the two groups in terms of collected operative (e.g EBL), clinical (ODI, VAS), and radiographic data (fusion status). We also analyzed the factors affecting LOS.
All statistical analyses were performed with SPSS version 20.0 (SPSS, Chicago, IL, USA). First, using data from consecutive patients, operative time and LOS were used to determine whether there was a significant difference (unpaired t-test) in early versus late cases. We started with the first 10 patients and consecutively added one patient at a time until a significant difference was detected in operative time and LOS between the groups. The number of patients required to reach this significant difference was noted as the proficiency point. The proficiency point was then used to compare surgical outcomes by dividing the patients into two groups: group A (patients who had a mini-open TLIF before the proficiency point) and group B (patients having their surgery after the proficiency point). Estimated blood loss, ODI, VAS, and fusion rate were divided by the proficiency point and analyzed for significant differences.
The impact of the different risk factors on LOS was determined by a correlation test for continuous variables and paired t-test for categorical variables. P-values of less than 0.05 were considered statistically significant.
The TLIF was performed on the more symptomatic side. Fluoroscopy was used to determine the operative level. The operative site was infiltrated preoperatively with 10 cc of 1% lidocaine with epinephrine. Two vertical incisions were made over the pedicles, and dissection was carried down to the facet joint (Figure 1A). Blunt dissection was performed to get down to the pedicle. Under both lateral and anteroposterior fluoroscopic guidance, Jamshidi needles were advanced into the pedicles (Figure 1B). Once the needles were confirmed to be in good position, guidewires were inserted through the Jamshidi needles, and the trocars were removed. Overall with four wires, tapping was performed and the screws were inserted. A tubular retractor system was placed to perform a TLIF.
The facet joint of interest was viewed. A total facetectomy was performed with an osteotome and the disc space was identified. The disc was entered with a knife and a complete discectomy was performed. The end plates were decorticated and bone graft (AlloSourceTM Demineralized Cancellous Cube; Centennial, CO and Grafton DBM Putty; Osteotech Inc. Eatontown, NJ) was laid into the disc space with the use of a funnel. A polyetheretherketone (PEEK) oblique cage (CoRoent, NuVasive, Inc, San Diego, CA) was measured, filled with bone graft, and placed into the disc space.
Rods were measured and placed into the tulip heads. Attempts were made to burr the facet joint and place bone graft in the joint on the contralateral side. Final tightening was performed over all the screws and the retractor system was taken out. A deep drain was placed and the wounds were closed in layers. Incision sites were injected with 20 cc of 0.25% marcaine, and prior to extubation 30 mg intravenous ketorolac was administered.
Postoperative pain control was achieved with three doses of ketorolac 15 mg intravenously given every 6 hr along with oral narcotics and breakthrough intravenous morphine as needed.
The mean age for the 100 patients was 59.8 yr (range 32 to 79 yr). There were 73 women and 27 men.
The operative time was significantly less after 21 patients (P=0.028) (Figure 2). Length of stay was significantly shorter after the 22 patients (P=0.002) (Figure 3). For subsequent analysis of perioperative and postoperative outcomes, we defined group A as the first 22 patients (before the proficiency point) and group B as the last 78 patients.
Baseline demographic data are summarized in Table 2. There was a significant difference in EBL between group A and group B (P=0.001). Mean ODI was significantly improved after the procedure (54.6 to 21.1, P=0.001). VAS for leg and back pain at 3 mo postoperatively showed significant improvement compared to preoperation (leg pain 7.80 to 2.20, P=0.001, back pain 7.90 to 2.70, P=0.001). However, the improvement in postoperative ODI and VAS between group A and group B were similar at 3 mo (ODI P=0.77; VAS leg P=0.36; back, P=0.58). Outcome data are summarized in Table 3.
Patient follow-up ranged from 2 mo to over 4 yr, with an average of 17 mo. Seventy-one patients had follow-up for greater than 12 mo. VAS scores were only recorded for 15 patients that averaged 3.88. Four patients in group A and 10 patients in group B reported either persisting back or lower extremity pain at the latest follow-up visit. There was no significant difference in reports of persisting back or leg pain between the two groups (P=0.53) in the latest clinic notes. Eight patients underwent MRI and two patients required epidural injections for pain control.
Group A and B had no significant differences in fusion rates according to the Bridwell classification. Twenty-nine percent of our study population did not have fusion assessment as they were lost to follow-up before their 1-year follow up. At 1 yr after surgery, 92% of group A and 95% of group B patients achieved grade I fusion (P=0.50). Grade III or IV fusions were not seen in either group.
Among the 100 patients, there was one patient who developed foot drop in group A and two patients in group B. Furthermore, one incident of dural tear was seen in group A and two were seen in group B. All patients recovered uneventfully before discharge. Complications, such as hematoma or infection, were not seen.
Factors Affecting LOS
Three factors from the patient’s history significantly correlated with LOS. Higher preoperative BMI was correlated with longer LOS (r=0.35, P=0.005). Patients who had a prior history of lumbar surgery had a shorter LOS (t=2.13, P=0.038). Finally, married patients or those living with a significant other had a shorter LOS (t=2.71, P=0.010).
Two perioperative factors were correlated with LOS (Table 4). Procedure time and LOS were strongly correlated (r=0.44, P=0.001). Average operative time for group B (2.52 hr) was 34 min shorter compared to group A (2.52 hr). Average LOS was 1.28 days shorter for group B compared to group A (Table 4). Patients with increased EBL remained in the hospital for longer (r=0.39, P=0.001).
Factors that did not significantly affect LOS were age, sex, smoking history, grade of spondylolisthesis, preoperative hemoglobin, Charlson comorbidity index, time to physical therapy, and occurrence of complications.
Minimally invasive, mini-open TLIF has emerged as safer and more efficacious alternative to open lumbar fusion surgery.5,8,12,13 These successful outcomes have made acceptance of MIS, mini-open TLIF increasingly widespread. As this technique becomes more prevalent, it is important to document the learning curve of this technically difficult procedure involving more limited surgical field exposure to assist inexperienced surgeons.
In our study, we used surgical time and LOS as a measure of surgeon proficiency and comfort with the mini-open TLIF. Previous literature has cited surgical time as an objective measure to identify the proficiency point on a learning curve. Nowitzke et al.19 reported a decrease in operative time after approximately 30 cases for lumbar microendoscopic discectomy. Similarly, Lonner et al.18 discovered a significant decrease in operative time after the first 28 cases for thoracoscopic spinal instrumentation for the treatment of thoracic adolescent idiopathic scoliosis. Currently, there are no studies investigating the learning curve associated with either MIS or mini-open TLIF. Peng et al.12 found an operating time of 216 min for their first 29 MIS TLIF cases while Lau et al.23 reported an operative time of 389 min for their first 10 cases; however, these likely reflected combining times before and after the proficiency point of the learning curve. Our current study found a proficiency point after 21 cases with OR time of 157 min compared to 187 min before this point.
Additionally, we used the LOS to determine the proficiency point. We felt the LOS was an appropriate measure to judge early successful outcomes in surgery that incorporates multiple factors including pain control, ambulation, and complications. Various studies have used LOS as a measure of a surgeon’s proficiency and competence with a new technique.15,16 The literature suggests a range of LOS for MIS TLIF from 15 hr by Rouben et al.24 to 72 hr.25,26 We compared the early results of mini-open TLIF by LOS and found an average increase in LOS of 31 hr comparing before the proficiency point of 22 patients to after the point (P=0.001), validating both a learning curve and an accurate proficiency point demonstrated by our operative time analysis.
We also assessed other factors besides proficiency of the surgeon that influenced LOS. Insight into the LOS allows us to potentially decrease hospital costs by utilizing fewer resources, decrease risk of nosocomial infection, and facilitate allocation of limited resources. Understanding these factors and the variability in LOS will allow us to predict LOS in future cases.
Previous investigators have identified several risk factors that affected LOS after surgeries. In a retrospective review involving 29 patients undergoing MIS TLIF, Park et al.27 identified BMI as a risk factor for increased LOS. A BMI greater than 25 kg/m2 had a mean LOS of 3.4 days compared to 2 days. Our current study found that an increase in BMI correlated with an increase in LOS regardless of point in the learning curve. Similarly, Vaidya et al.28 found that compared to 12 obese (BMI>30 kg/m2) patients, five morbidly obese (BMI>40 kg/m2) patients had a longer mean LOS (6.1 days vs. 5.4 days) in patients undergoing open TLIF.
There is limited literature relating EBL with LOS in mini-open TLIF. Numerous reports11,12,28 have stated that there is a decreased EBL for MIS TLIF compared to open procedures. Dhall et al.11 found a mean EBL of 194 mL for the mini-open group compared to 505 mL for open TLIF in a 42 patient sample size. In the present study, we observed a lower EBL for cases after the proficiency point (P<0.001) and a shorter LOS (P=0.001). The potential benefit of this is twofold – in addition to a shorter LOS, we avoid the risks and monetary costs associated with transfusions.
Married patients or those living with a significant other had a shorter LOS in this study (P=0.009). However, we did not find an association between LOS and simply living with another person in the house (i.e. sibling, children, or friend; P=0.065) suggesting that the difference in level of care provided at home by these cohabitants translates into variation of hospital discharge. A number of studies29,30 have compared marital status to LOS after surgery but did not find an association. This may be explained by our inclusion of not only marital status but cohabitating with a significant other in our analysis.
Interestingly, we found that a history of prior lumbar surgery, which ranged from laminectomies to microdiscectomies, was associated with a shorter LOS after a mini-open TLIF (P=0.037). This phenomenon is possibly a result of patient familiarity with the postoperative process and ability to facilitate their discharge. A thorough literature review did not yield any prior studies of this relationship. Indications for surgery for patients with a prior history of spine surgery at the location of interest varied across studies. The need for durotomy during revision surgery where scar tissue could possibly encase nerve roots were included in certain studies19 but viewed as a relative contraindication in others.12
Analysis of clinical outcomes postoperatively is equally as important as understanding how to reduce hospital LOS. We saw that all patients had significant improvement in pain and disability at 3 mo postoperatively compared to preoperatively. However, there was no significant difference between the groups operated on before the proficiency point compared to after in VAS and ODI, suggesting that there are no long-term risks of learning to perform mini-open TLIF by surgeons. Similar clinical improvement was demonstrated in clinical notes reviewed at an average follow-up of 17 mo.
Moreover this is demonstrated radiographically in terms of bony fusion. There was no significant difference between the two groups according to the Bridwell classification. Our grade I fusion rate of 92% before the proficiency point and 95% after the point is comparable to the 2-year postoperative fusion outcomes of MIS and open TLIFs by Peng et al.12 (Open 86.7%, MIS 80%) and Lee et al.31 (Open 98.5%, MIS 97%).
Our study was limited in that it was retrospective. Although no differences between LOS and factors such as complications were found, it is possible minor complications were not recorded in the chart but unlikely that major complications such as foot drop or transfusions were missed. We had lost 29% of our patients before their 1-year follow up therefore could not assess their fusion at that time. We had noted that these patients at their latest follow-up did not have any complaints of lower back pain or radicular pain. We could have better assessed fusion rates with CT scans. However, given the excellent clinical results at the patient follow-up, it was difficult to justify the additional radiation of a CT for the sole reason of assessing fusion status. We also assumed all the cases to be identical. We did not take into account the variation in difficulty of surgeries whether they occurred early or late in the learning curve. But, this limitation was somewhat mitigated by the large sample size, thus minimizing confounding factors. In addition, our study is based on the experience of one surgeon. Although other surgeons with unique backgrounds and experiences may differ from the results of this study’s learning curve, it is encouraging to see that it is a practical and safe procedure that can be mastered.
In conclusion, we found the learning curve for mini-open TLIF to be 22 cases when analyzing a surgeon’s proficiency based on LOS and surgical time. The following factors led to a significantly decreased LOS: decreased operative time, lower EBL, later point on the learning curve, lower BMI, being married or living with a significant other, and history of a prior lumbar surgery. Throughout the learning curve, complication rates and outcomes measured by ODI and VAS at 3 mo follow-up and fusion status at 1 yr were comparable. Together, this study supports mini-open TLIF as an acceptable alternative to open procedures allowing patients shorter hospitalization and lasting improvement that decreases pain and disability.
1. Sihvonen T, Herno A, Paljarvi L, et al.. Local denervation atrophy of paraspinal muscles in postoperative failed back syndrome. Spine. 1993; 18:575–581.
2. Kim KT, Lee SH, Suk KS, et al.. The quantitative analysis of tissue injury markers after mini-open lumbar fusion. Spine. 2006; 31:712–716.
3. Bresnahan L, Ogden AT, Natarajan RN, et al.. A biomechanical evaluation of graded posterior element removal for treatment of lumbar stenosis comparison of a minimally invasive approach with two standard laminectomy techniques. Spine. 2009; 34:17–23.
4. Styf JR, Willen J. The effects of external compression by three different retractors on pressure in the erector spine muscles during and after posterior lumbar spine surgery in humans. Spine. 1998; 23:354–358.
5. Parker SL, Adogwa O, Bydon A, et al.. Cost-effectiveness of minimally invasive versus open transforaminal lumbar interbody fusion for degenerative spondylolisthesis associated low-back and leg pain over two years. World Neurosurgery. 2012; 73:178–184.
6. Moskowitz A. Transforaminal lumbar interbody fusion. Orthop Clin North Am. 2002; 33:359–366.
7. Rosenberg WS, Mummaneni PV. Transforaminal lumbar interbody fusion: technique, complications, and early results. Neurosurgery. 2001; 48:569–574. discussion 74-75.
8. Adogwa O, Parker SL, Bydon A, et al.. Comparative effectiveness of minimally invasive versus open transforaminal lumbar interbody fusion: 2-year assessment of narcotic use, return to work, disability, and quality of life. J Spinal Disord Tech. 2011; 24:479–484.
9. Ozgur BM, Hughes SA, Baird LC, et al.. Minimally disruptive decompression and transforaminal lumbar interbody fusion. Spine J. 2006; 6:27–33.
10. Shunwu F, Xing Z, Fengdong Z, et al.. Minimally invasive transforaminal lumbar interbody fusion for the treatment of degenerative lumbar diseases. Spine (Phila Pa 1976). 2010; 35:1615–1620.
11. Dhall SS, Wang MY, Mummaneni PV. Clinical and radiographic comparison of mini-open transforaminal lumbar interbody fusion with open transforaminal lumbar interbody fusion in 42 patients with long-term follow-up. J Neurosurg Spine. 2008; 9:560–565.
12. Peng CWB, Yue WM, Poh SY, et al.. Clinical and radiological outcomes of minimally invasive versus open transforaminal lumbar interbody fusion. Spine. 2009; 34:1385–1389.
13. Schizas C, Tzinieris N, Tsiridis E, et al.. Minimally invasive versus open transforaminal lumbar interbody fusion: evaluating initial experience. Int Orthop. 2009; 33:1683–1688.
14. Park Y, Ha JW. Comparison of one-level posterior lumbar interbody fusion performed with a minimally invasive approach or a traditional open approach. Spine. 2007; 32:537–543.
15. Rege RV, Joehl RJ. A learning curve
for laparoscopic splenectomy at an academic institution. J Surg Res. 1999; 81:27–32.
16. Simons AJ, Anthone GJ, Ortega AE, et al.. Laparoscopic-assisted colectomy learning curve
. Dis Colon Rectum. 1995; 38:600–603.
17. Guttmann D, Graham RD, MacLennan MJ, et al.. Arthroscopic rotator cuff repair: The learning curve
. Arthroscopy. 2005; 21:394–400.
18. Lonner BS, Scharf C, Antonacci D, et al.. The learning curve
associated with thoracoscopic spinal instrumentation. Spine. 2005; 30:2835–2840.
19. Nowitzke AM. Assessment of the learning curve
of lumbar microendoscopic discectomy. Neurosurgery. 2005; 56:755–761.
20. Bridwell KH, Lenke LG, McEnery KW, et al.. Anterior fresh frozen structural allografts in the thoracic and lumbar spine. Do they work if combined with posterior fusion and instrumentation in adult patients with kyphosis or anterior column defects? Spine. 1976; 20:1410–1418.
21. Lee KH, Yue WM, Yeo W, et al.. Clinical and radiological outcomes of open versus minimally invasive transforaminal lumbar interbody fusion. Eur Spine J. 2012; 21:2265–2270.
22. Lee CK, Park JY, Zhang HY. Minimally invasive transforaminal lumbar interbody fusion using a single interbody cage and a tubular retraction system: technical tips, and perioperative, radiologic and clinical outcomes. J Korean Neurosurg Soc. 2010; 48:219–224.
23. Lau D, Lee JG, Han SJ, et al.. Complications and perioperative factors associated with learning the technique of minimally invasive transforaminal lumbar interbody fusion (TLIF). J Clin Neurosci. 2011; 18:624–627.
24. Rouben D, Casnellie M, Ferguson M. Long-term durability of minimal invasive posterior transforaminal lumbar interbody fusion a clinical and radiographic follow-up. J Spinal Dis Tech. 2011; 24:288–296.
25. Schwender JD, Holly LT, Rouben DP, et al.. Minimally invasive transforaminal lumbar interbody fusion (TLIF) – Technical feasibility and initial results. J Spinal Dis Tech. 2005; 18:S1–S6.
26. Isaacs RE, Podichetty VK, Santiago P, et al.. Minimally invasive microendoscopy-assisted transforaminal lumbar interbody fusion with instrumentation. J Neurosurg Spine. 2005; 3:98–105.
27. Park P, Upadhyaya C, Garton HJL, et al.. The impact of minimally invasive spine surgery on perioperative complications in overweight or obese patients. Neurosurgery. 2008; 62:693–698.
28. Vaidya R, Carp J, Bartol S, et al.. Lumbar spine fusion in obese and morbidly obese patients. Spine. 2009; 34:495–500.
29. Forrest G, Fuchs M, Gutierrez A, et al.. Factors affecting length of stay
and need for rehabilitation after hip and knee arthroplasty. J Arthroplasty. 1998; 13:186–190.
30. Zheng FY, Cammisa FP, Sandhu HS, et al.. Factors predicting hospital stay, operative time
, blood loss, and transfusion in patients undergoing revision posterior lumbar spine decompression, fusion, and segmental instrumentation. Spine. 2002; 27:818–824.
31. Lee K, Yue W, Yeo W, et al.. Clinical and radiological outcomes of open versus minimally invasive transforaminal lumbar interbody fusion. Europ Spine J. 2012; 21:2265–2270.
Keywords:Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved
mini-open transforaminal lumbar interbody fusion (mini-open TLIF); learning curve; operative time; length of stay; outcome