UEF//eRepository
DSpace https://erepo.uef.fi
Rinnakkaistallenteet Terveystieteiden tiedekunta
2020
30-Day Recurrence, Readmission Rate and Clinical Outcome After Emergency Lumbar Discectomy
Reito, Aleksi
Ovid Technologies (Wolters Kluwer Health)
Tieteelliset aikakauslehtiartikkelit
© 2020 Wolters Kluwer Health, Inc.
CC BY-NC http://creativecommons.org/licenses/by-nc/4.0/
http://dx.doi.org/10.1097/BRS.0000000000003519
https://erepo.uef.fi/handle/123456789/23692
Downloaded from University of Eastern Finland's eRepository
Downloadedfromhttps://journals.lww.com/spinejournalbyBhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD33nx2+yvo/ef47uwh8SbtgFyIwuJ6UtBj3OsnAfUTfeo=on05/06/2020
Downloadedfrom https://journals.lww.com/spinejournalby BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD33nx2+yvo/ef47uwh8SbtgFyIwuJ6UtBj3OsnAfUTfeo=on
05/06/2020
SPINE An International Journal for the study of the spine, Publish Ahead of Print DOI: 10.1097/BRS.0000000000003519
30-day recurrence, readmission rate and clinical outcome after emergency lumbar discectomy
Aleksi Reito, MD, PhD1,2, Kati Kyrölä, MD1, Liisa Pekkanen, MD, PhD1 Juha Paloneva, MD, PhD, Professor1,3
1 Central Finland Hospital, Department of Orthopedics and traumatology, Keskussairaalantie 19, 40630 Jyväskylä, Finland
2 Tampere University Hospital, Department of Orthopaedics and Traumatology, Teiskontie 35, 33521 Tampere, Finland
3 University of Eastern Finland, School of Medicine, Yliopistonranta 1, 70210 Kuopio, Finland
Correspondence:
Aleksi Reito
Central Finland Hospital, Department of Orthopedics and traumatology Keskussairaalantie 19
40630, Jyväskylä Finland te. +358 3 11 67 605 aleksi@reito.fi
The manuscript submitted does not contain information about medical device(s)/drug(s).
No funds were received in support of this work.
No relevant financial activities outside the submitted work.
Abstract
Study Design: A retrospective cohort study
Objective: To investigate the 30-day recurrence rate after emergency lumbar discectomy.
Secondary aims were to investigate the factors affecting the 30-day recurrence and readmission rates and clinical outcome.
Summary of Background Data: Excluding cauda equine syndrome (CES) due to massive intervertebral disc herniation, emergency surgery for lumbar disc herniation (LDH) is rarely required. The operation may, however, be performed for other reasons such as persistent or progressive motor paresis associated with radiculopathy or uncontrolled pain. Literature on these topics is scarce.
Methods: All patients admitted for inpatient care after a visit to the emergency department (ED) due to acute low back pain and who had subsequently undergone an emergency lumbar
discectomy during the 4-year study period were included in the study. Patients attending the ED who subsequently had a delayed discectomy formed the control group. Recurrence and
readmission rates were analyzed and clinical outcome at 30 days post-surgery was assessed with the Oswestry Disability Index (ODI) and the visual analog pain scale (VAS).
Results: 130 patients were admitted to the hospital after visiting the ED and underwent an emergency discectomy after a median of 1.0 days from admittance. Six patients in the study group [4.6% (95% CI: 2.1 – 9.7)] had recurrent LDH and 9 patients in total [6.9% (95% CI: 3.9 – 12.6)] were readmitted within 30 days. None of the baseline variables clearly predicted
recurrence. Mean ODI difference between the study group and controls was 8.1 (95% CI: -6.7 – 23.2). BMI and surgery by a non-spine surgeon were associated with higher ODI values.
Conclusions: An emergency discectomy is associated with a higher rate than expected of both recurrent LDHs and 30-day readmissions. Surgeon experience and patient-related factors had minor effects on the 30-day clinical outcome.
Key Words: lumbar discectomy; emergency; Radiculopathy; Lumbar disc herniation;
recurrence; Readmission Level of Evidence: 4
Key points:
Emergency surgery for lumbar disc herniation (LDH) is rarely required.
Emergency discectomy is required in cases of cauda equine syndrome and in the presence of persistent or progressive motor paresis and can be considered if uncontrolled pain is associated with acute lumbar disc herniation.
We found a 30-day recurrence rate of 4.8% and a 30-day readmission rate of 6.9% in patients operated for emergency discectomy for acute LDH.
Surgeon experience and patient-related factors had only a small effect on patients’
functional outcome
Introduction
Acute low back pain (LBP) is one of the commonest reasons for visiting an emergency
department (ED) and accounts for approximately 3.2-4.4% of all visits 1,2. Acute LBP is usually categorized as either nonspecific, owing to a serious etiology, or indicative of nerve or spinal cord compression 3. LBP may have a specific cause, such as cauda equine compression, fracture, cancer or infection. The treatment and prognosis of these conditions is disease-specific. It is most important that they are diagnosed or excluded as a likely cause of acute LBP as early as possible in the ED 4.
Literature on the third possibility ‒ low back pain potentially associated with radiculopathy or spinal stenosis ‒ is scarce. Patients presenting with this condition are overrepresented in secondary care 4. Rapid cross-sectional imaging, preferably with MRI, is indicated in these patients if progressive paresis is observed 5. Admission to hospital for imaging is also considered in patients who represent with uncontrolled LBP and radiculopathy without any motor paresis or red flag symptoms.
Emergency surgery for lumbar disc herniation (LDH) is rarely required. Cauda equine syndrome (CES) due to massive intervertebral disc herniation is, however, an indication for an emergency or unscheduled discectomy 6. The operation may also be performed for other reasons such as persistent or progressive motor paresis associated to radiculopathy or uncontrolled pain. Owing to the emergency nature of the procedure, patients cannot always be referred for surgery by a specialist spinal surgeon; instead, the orthopedic surgeon on call must perform the operation.
Clinical outcomes and recurrence rates after lumbar discectomy have almost exclusively been
studied in an elective setting, i.e., among patients undergoing discectomy after unsuccessful nonoperative treatment. To best of our knowledge, the study by Petr et al. is the only one to have systematically reported the results of emergency discectomy 7. The authors only included
patients with motor deficit and concluded that the faster the surgery, the better the recovery of function.
The aim of our study was to investigate the 30-day recurrence rate after an emergency lumbar discectomy. Secondary aims were to investigate the factors affecting the recurrence rate, the 30- day clinical outcome measured using the Oswestry Disability Index (ODI), VAS pain scale and readmission rates in these patients.
Materials and methods
Our institution is the only hospital with an 24-hour emergency department and the only provider of public secondary care in its hospital district, with a catchment area of 250 000 people. Public health care in Finland is available to every citizen. Our hospital is a level II/III trauma center. For tertiary care, patients are referred to either of the two nearest university hospitals. Our hospital operates around-the-clock medical services, including emergency medicine physicians, general surgeons, orthopedic surgery, anesthesiology, radiology and critical care. Patients requiring specialized evaluation or treatment may be referred to our hospital ED by GPs on call either in a local health care center or at our hospital ED or by specialists working in private hospitals.
Patient selection
We identified, from an institutional surgical operation database (Effica Leikkaushoito, Tieto Ltd, Helsinki, Finland), all patients who had been referred to our ED, admitted for in-patient care and undergone an emergency lumbar discectomy during the same treatment period. Patients who had undergone microsurgical or open excision of lumbar intervertebral disc displacement (ABC16/26 in the NOMESCO Classification of Surgical Procedures) were extracted. No endoscopic
surgeries have been performed in our hospital. All patients who had attended our ED between 1 January 2012 and 31 December 2015 were included in the study group. In the database, surgical spinal operations are labeled urgent (class 0, <5 days), semi-urgent (class 1, <30 days) or non- urgent (class 2, >30 days). The control group comprised patients who had attended the ED
during the study period but had been scheduled for semi-urgent lumbar discectomy after discharge home from the ED or from in-patient care after observation.
Patient management
All operations were performed by a consultant orthopedic surgeon. Preferably, operations are carried out by one of the four spine surgeons working at our institution. Operations outside office hours are usually performed by the consultant orthopedic surgeon on call. All patients included in this study were operated by or under direct supervision of a consultant. Electronic anesthesia records (AER), including ASA classification, body mass index (BMI), smoking status, diabetes, name of the operating surgeon and operation time were prospectively collected. Patients
scheduled for discectomy are also given a baseline questionnaire inquiring about their
occupational physical load. Upon discharge from in-patient care, patients are given follow-up questionnaires to be returned 4 weeks after the operation. These include the visual analog scale (VAS) for pain (range 0 to 100 mm, 100=worst) and Oswestry disability index (ODI, range 0 to 100, 100=worst).
Data extraction
In addition to AERs, all the ED discharge summaries were read and evaluated. Patients’ age and sex, urine retention in ml, muscle power on the Medical Research Council (MRC) muscle scale, time from pain onset or symptom exacerbation to operation, previous discectomies, and MRI imaging findings in the ED or during admittance were recorded. If a patient had undergone a discectomy previously, the date of the operation was noted. A grade between 0 and 2 on the MRC muscle scale for lower leg muscle power in either leg was considered paresis. If a prior operation had been performed within the preceding 12 months, the patient was excluded to ensure that only new symptomatic disc herniations were included and not early recurrences. Prior surgery was categorized as performed in either the same or another disc space than the one included in the study. The MRI report was checked to see in which disc space the new herniation was located. The operating surgeon was categorized as a spinal surgeon or non-spinal surgeon, depending on whether the surgeon’s main specialty was the spine or general and some other orthopedic subspecialty. Patients’ occupational physical workload was categorized as follows: 1)
light sedentary, 2) heavy sedentary, 3) light standing or mobile, 4) moderate bodily, 5) heavy bodily and 6) very heavy bodily.
Statistics
Descriptive statistics for normally distributed variables included mean and standard deviation (SD). For skewed and non-normal variables, median, interquartile (IQR) range and total range were used. Normal variables with equal variance were compared using Students t-test and for those with non-equal variance Welchs t-test was used. For each comparison, the mean difference (MD) with 95% confidence intervals (CI) was calculated. Other continuous variables were compared using the Mann-Whitney U-test. In the case of a 2x2 contingency table, categorical variables were compared using Fisher’s exact test. For larger contingency tables, a chi-square test without Yates correction was used. CIs for proportions were calculated using Wilson’s score interval. Logistic regression was used to investigate associations with binary endpoint and linear regression with with continuous outcomes. Directed acyclic graphs (DAGs) were used to assess bias arising from the variable selection in the multivariable analyses. Separate DAGs were drawn for both linear exposures, namely ODI and vas leg pain at 1 month (see Appendix,
http://links.lww.com/BRS/B574). Based on DAGs we omitted age from analyses since including it would have resulted to adjusting for moderator which is a source of bias 8. Univariate analysis was performed using all the baseline variables. Two different multivariable regression analyses were performed. The first included all suitable baseline variables. The second, reduced, model included only the baseline variables that showed the strongest evidence against the null
hypothesis in the Wald test for regression coefficients. Working load was not included in the multivariable models since this information was lacking in 59 patients. The superiority of these two multivariable models was based on the Akaike’s information criteria (AIC) and adjusted R2 value.
Results
During the study period, 130 patients were admitted to hospital immediately after an ED visit and underwent emergency lumbar discectomy. The median and mean number of days from admittance to surgery was 1.0 (IQR: 1-2, range: 0-8) and 1.6 (SD: 1.5). The control group was formed by 16 patients who underwent scheduled discectomy surgery after discharge home from
the ED. Baseline data are shown in the Table 1. Of the 130 patients, 92 had severe pain and 38 was recorded having MRC grade 0-2 paresis. Of these, three did not report having radicular pain in the leg, but only pain in the lower back.
Duration of the current symptomatic episode was longer in the emergency group than delayed group (MD: 13.6 days, 95% CI: 3.6 – 23.6). The operating surgeon was more likely to be a spinal surgeon in the delayed surgery group and operative time was shorter (MD: 20 min, 95%
CI: 5.4 – 34.4).
Six patients in the emergency group had an early recurrent lumbar disc herniation, yielding a 4.6% (95% CI: 2.1 – 9.7) recurrence rate at 30 days. One patient (6.3%, 95% CI: 1.1 – 28.3) in the delayed group had a recurrent lumbar disc herniation (LDH). Data were consistent with the variety of point estimates for the ORs predicting recurrence, and we were unable to exclude the zero effect (OR=1.00) for any predictor variable at a confidence level of 0.05 in the univariable logistic regression.
In addition to the above six patients, another three in the emergency group were readmitted within 30 days of surgery, yielding a 30-day readmission rate of 6.9% (95% CI: 3.9 – 12.6). One patient had a deep infection requiring surgical debridement. Two patients were admitted to ED due to uncontrolled postoperative pain. In both cases, MRI revealed a normal postoperative condition. No secondary surgery was needed in these patients and the pain resolved.
The mean and median Oswestry disability index (ODI) in the emergency group was 23.3 (SD:
18.2) and 19 (IQR: 8 – 36, range: 0 - 72) 4 weeks postoperatively. In the control group the mean and median ODI values were 31.4 (20.0) and 25.9 (IQR: 13.5 – 48.5, range: 4.4 - 76) 4 weeks postoperatively. The MD between groups was 8.1 (95% CI: -6.7 – 23.2). The results of the univariable linear regression analysis are shown in Table 2. Age, diabetes, BMI, operating surgeon and symptom duration were included in the reduced multivariable model. For BMI (β
=0.91, 95% CI: 0.11 to 1.7, p=0.027) and for surgery by a spine surgeon (β =-8.8, 95% CI: -16.3 to -1.2, p=0.024) the CIs of the regression coefficients did not include 0, suggesting linear dependence of the ODI mean scores on these variables. Similar results were seen in the multivariable model including all baseline variables. Reduced model had slightly superior explanatory power based on R2 and AIC.
Mean and median leg pain VAS scores in the emergency group were 21 (SD 23) and 12 (IQR: 2 – 37, range: 0 - 80) 4 weeks postoperatively. In the control group, the mean and median scores were 30 (37) and 13 (IQR: 8 – 51, range: 0 - 84 ) 4 weeks postoperatively. The MD between groups was 9 (95% CI: -13 – 32). The results of the univariable linear regression analysis are shown in Table 3. ASA classification, herniation level and motor deficit were associated with leg pain VAS. Neither of the multivariable models was clearly superior to the other. The presence of motor deficit was most clearly associated with leg pain in both models. The final models
explained 20% to 22% of the overall variance in leg pain VAS.
Discussion
Acute low back pain is a common reason for visiting an ED 1,2. While the majority of patients present with benign nonspecific low back pain, in some patients with radiculopathy the pain is due to acute LDH. The course of acute LDH is usually self-limiting and prognosis excellent.
Occasionally, however, patients may present with progressive motor paresis or uncontrolled pain, in which case an emergency discectomy is performed. Due to the unplanned and unscheduled nature of these cases, patients may not always be referred to a spinal surgeon;
instead, the orthopedic surgeon on call is responsive for the surgery.
In line with our result, Petr et al. reported a 5.5% recurrence rate. They did not report precisely when the recurrences occurred, but as their follow-up was 3 months at most, these can be considered early recurrences. These recurrence rates are considerably higher than those reported after scheduled or elective lumbar discectomies. Häkkinen et al. reported an early recurrence rate of less than 2% in a cohort of patients operated at our institution prior to our study period 9. In the NSQIP database, the 30-day reoperation rate was 2% 10. Since this number also includes cases with deep infection, the true recurrence rate is even lower. According to the German Spine Register, 1.2% of patients presented with an early recurrent LDH “within days of the initial surgery” 11. A precise definition of an early recurrence was not given.
We found no correlations of any variables with early recurrent LDH. An issue of interest was the effect of the expertise of the operating surgeon as patients requiring an emergency discectomy are commonly referred to the orthopedic surgeon on call instead of a specialist spinal surgeon.
Three to four spine surgeons are at work daily in our hospital. Two-thirds of the patients in our
retrospective cohort, however, were operated on by other than spine surgeons. If we assume a recurrence rate of 2% after discectomies performed either in a scheduled setting or by a spine surgeon, we would need over 1 100 patients with an emergency discectomy or surgery performed by a non-spine surgeon per group to obtain a significant difference of 4% in the control group. Larger studies are therefore warranted to further examine the effect of the surgeon experience on the outcome of emergency lumbar discectomy. With the data at hand, we were unable to show a difference in recurrence rates between emergency and delayed discectomy.
Nine patients (6.9%) were readmitted within 30 days of surgery. Hogget reported a readmission rate as low as 0.7% after day case lumbar discectomy 12. In other studies, the 30-day readmission rate has varied between 2.6 and 3.7% 13–15. In previous studies, ASA class, prolonged operative time, age, incidental durotomy, and discharge facility have been associated with risk for
readmission 14,15. We did not observe any predictive factors for early readmission despite including the same baseline variables as in previous studies, namely ASA class, operative time and age.
Increasing BMI and an operation performed by other than a spine surgeon were associated with poorer outcome (i.e., higher ODI). For the latter, the difference was 9.9 points. For BMI, a coefficient β of 1.1 equals a change of 11 points in the ODI when comparing patients with a 10- unit difference in BMI. The CIs for these point estimates suggest a wide range of plausible values, i.e., the effect varied from 2.3 points to 17.6 for the operating surgeon. Therefore, we cannot rule out a possibility that surgeon experience has a meaningful effect on the functional outcome. Similarly can be said about the effect of prior surgery. Larger studies with more precise estimates are needed to clarify their effects. For leg pain, a high LDH level (β =24.8) and the presence of motor deficit (β=13.8) were associated with a higher leg pain VAS score. All these effects, except for the herniation level, may considered minuscule since they are below
established MCIDs.
Our study has its limitations. A major limitation was that we only assessed short-term clinical outcome. This was, however, only a secondary aim of our study, as the main focus was on the factors predicting short-term outcome. Another limitation was the lack of a measurement of preoperative pain and ODI. Selection bias should be also considered. Pain is always a subjective measure and is also a multifaceted issue. It is probable that patients referred to ER and
subsequent unscheduled surgery are different to those who are referred to elective surgery. The major advantage of our study is the representativeness of our study cohort, which is clearly population-based. Public health care is available to everyone in Finland. It is, therefore, extremely unlikely that patients in need of an emergency lumbar discectomy living in the catchment area of our hospital would have undergone this surgery elsewhere.
In conclusion, an emergency discectomy is associated with a higher rate of both recurrent LDHs and 30-day readmissions compared to the rates seen in elective operations. We found no
correlations of any predictors with these outcomes. The surgeon experience and patient-related factors had an effect on the 30-day clinical outcome. Low back pain and pain in general cannot be assessed objectively and they are also highly associated with psychosocial factors. It is likely that patients referred to an ED for an emergency discectomy differ from those undergoing an elective discectomy. Further studies are needed to clarify the characteristics of patients with intolerable pain referred to ED and whether the higher rate of recurrences and readmissions are due to surgery related factors or due to extrinsic factors such as different pathophysiology.
References
1. Edwards J, Hayden J, Asbridge M, et al. Prevalence of low back pain in emergency settings: a systematic review and meta-analysis. BMC Musculoskelet Disord 2017;18:143.
2. Waterman BR, Jr PJB, Schoenfeld AJ. Low back pain in the United States: incidence and risk factors for presentation in the emergency setting. Spine J 2012;12:63–70.
3. Chou R, Qaseem A, Snow V, et al. Diagnosis and treatment of low back pain: a joint clinical practice guideline from the American College of Physicians and the American Pain Society. Ann Intern Med 2007;147:478–91.
4. Reito A, Kyrola K, Pekkanen L, et al. Specific spinal pathologies in adult patients with an acute or subacute atraumatic low back pain in the emergency department. Int Orthop 2018; 42(12):2843-2849.
5. Patel ND, Broderick DF, Burns J, et al. ACR Appropriateness Criteria® low back pain. J Am Coll Radiol. 2016;13(9):1069-78.
6. Todd N V, Dickson RA. Standards of care in cauda equina syndrome. Br J Neurosurg 2016;30:518–22.
7. Petr O, Glodny B, Brawanski K, et al. Immediate versus Delayed Surgical Treatment of Lumbar Disc Herniation for Acute Motor Deficits: The Impact of Surgical Timing on Functional Outcome. Spine (Phila Pa 1976). Epub ahead of print 2017 Jun 27..
8. Ranstam J, Cook JA. Causal relationship and confounding in statistical models. Br J Surg 2016;103:1445–6.
9. Häkkinen A, Kiviranta I, Neva M, et al. Reoperations after first lumbar disc herniation surgery; a special interest on residives during a 5-year follow-up. BMC Musc Dis 2007;8:2
10. Golinvaux NS, Bohl DD, Basques BA, et al. Comparison of the lumbar disc herniation patients randomized in SPORT to 6,846 discectomy patients from NSQIP: demographics, perioperative variables, and complications correlate well. Spine J 2015;15:685–91.
11. Vinas-Rios JM, Sanchez-Aguilar M, Govea FAM, et al. Incidence of early postoperative complications requiring surgical revision for recurrent lumbar disc herniation after spinal surgery: a retrospective observational study of 9,310 patients from the German Spine Register. Patient Saf Surg 2018;12:9–018–0157–1.
12. Hoggett L, Anderton M, Khatri M. 30-day complication rates and patient-reported outcomes following day case primary lumbar microdiscectomy in a regional NHS spinal centre. Ann R Coll Surg Engl 2018;1–5.
13. Webb ML, Nelson SJ, Save A V, 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;42:1267–
73.
14. Kohls MR, Jain N, Khan SN. What are the Rates, Reasons, and Risk Factors of 90-day Hospital Readmission After Lumbar Discectomy?: An Institutional Experience. Clin spine Surg 2018;31:E375–80.
15. Pugely AJ, Martin CT, Gao Y, et al. Causes and Risk Factors for 30-Day Unplanned Readmissions After Lumbar Spine Surgery. Spine (Phila Pa 1976) 2014;39:761–8.
Table 1: Baseline information
Study group Control group p-value
Age Mean (SD) 43.9 (13.8) 41.1 (16.6) 0.48
Gender Male 63 (48.5%) 5 (31.3%) 0.29
Female 67 (51.5%) 11 (68.7%)
Symptom duration Mean (SD) 14.7 (18.3) 28.4 (18.6) 0.007
Median (IQR) 7 (3.3 - 19.5) 26.5 (12.3 - 32.5) 0.010
Smoking Yes 35 (26.9%) 3 (18.8%) 0.76
No 95 (73.1%) 13 (81.2%)
Diabetes Yes 7 (5.7%) 1 (6.3%) 1.0
No 123 (94.3%) 15 (93.7%)
BMI Mean (SD) 27.4 (4.5) 24.8 (3.9) 0.030
Motor deficit MRC 0-2 37 (28.5%) 1 (6.3%) 0.07
MRC 3-5 93 (71.5%) 15 (93.7%)
Urine retention (ml) Median (IQR) 0 (0 – 44) 0 (0-0) 0.20
Operating surgeon Spine surgeon 41 (31.5%) 12 (75%) 0.0014
Other 89 (68.5%) 4 (25%)
Operative time (mins)
Mean (SD) 68 (33) 48 (25) 0.02
Work load 1 21 (29.2%) 4 (30.8%) 0.9
2 3 (4.2%) 0 (0%)
3 10 (13.9%) 2 (15.4%)
4 17 (23.6%) 5 (38.5%)
5 16 (22.2%) 2 (15.4%)
6 5 (6.9%) 0 (0%)
Herniation level L5-S1 53 (40.8%) 9 (56.3%) 0.79
L4-L5 62 (47.7%) 6 (37.5%)
Other 16 (12.3%) 1 (6.3%)
Previous surgery on the same level
Yes 20 (15.4%) 3 (18.7%) 0.71
No 110 (84.5%) 13 (81.3%)
Table 2: Linear regression analysis predicting ODI at 1 month
Simple linear regression
Multivariable model 1
Multivariable model 2 β
coefficient (95% CI)
p-value β coefficient
(95% CI) p- value
β coefficient
(95% CI) p- value
Age Per year 0.35 (0.10 -
0.60)
0.0065
Gender Males -3.0 (-10.6 -
4.7)
0.44 -0.8 (-9.8 - 8.1)
0.85
ASA classification 2 vs 1 2.9 (-5.0 -
10.8)
0.47 2.7 (-6.3 - 11.7)
0.88 3 and 4 vs 1 9.7 (-4.9 -
24.2)
0.19 4.8 (-12.1 - 21.8)
0.79
Diabetes -9.3 (-25.9 -
7.2)
0.27 -13.3 (-30.5 - 3.8)
0.85 -12.2 (-28.1 - 3.8)
0.13
BMI Per unit 0.88 (0.07 -
1.70)
0.034 1.02 (0.11 - 1.9)
0.027 1.09 (0.29 -1.9)
0.008
Smoking 2.23 (-6.5 -
10.9)
0.61 2.9 (-6.1 - 11.8)
0.52
Operating surgeon Spine
surgeon
-8.32 (-18.9 - 0.08)
0.040 -9.4 (-16.5 - 2.4)
0.052 -9.9 (-17.6 - -2.3
0.011
Working status Per category 1.64 (-0.73
- 4.0)
0.17
LDH level L4-5 vs. L5-
S1
0.14 (-7.90 - 8.16)
0.97 -2.2 (-10.9 - 6.5)
0.62 L3-4 vs.
L5-S1
1.10 (-13.7 - 16.0)
0.88 0.09 (-16.1 - 16.3)
0.99 Other vs. L5-
S1
8.52 (-13.5 - 30.3)
0.44 11.4 (-10.3 - 33.1)
0.30
Previous surgery -2.4 (-13.2 -
8.45)
0.67 -0.22 (-11.5 - 11.0)
0.97
Motor deficit 6.7 (-1.5 -
14.8)
0.11 4.1 (-4.6 - 12.7)
0.35
Symptom duration Per day -0.18 (-0.36
- 0.01)
0.057 -0.15 (-0.34 - 0.04)
0.13 -0.15 (-0.32 – 0.03 0.11
Operative time Per minute 0.06 (-0.05
- 0.17)
0.27 -0.01 (-0.13 - 0.12)
0.90
Regression diagnostics Adj. R2 0.061 0.126
AIC 812.6 797.2
Table 3: Linear regression analysis predicting leg pain at 1 month
Simple linear regression Multivariable model 1 Multivariable model 2 β coefficient (95%
CI)
p-value β coefficient (95% CI)
p-value β coefficient (95% CI)
p-value
Age Per year 0.081 (-0.25 -
0.41)
0.63
Gender Males -5.8 (-15.6 - 4.0) 0.24 -1.1 (-12.0 9.5) 0.82 ASA
classification
2 vs 1 10.9 (0.99 - 20.8) 0.032 11.6 (0.98 22.3) 0.033 8.3 (-1.2 - 17.8) 0.088 3 and 4 vs 1 -1.5 (-19.3 - 16.3) 0.86 -0.5 (-20.3 19.2) 0.96 -4.1 (-23.5 -
15.3)
0.67 Diabetes -9.7 (-32.8 - 13.2) 0.40 -14.3 (-36.0 7.3) 0.19
BMI Per unit 0.87 (-0.17 - 1.92) 0.099 1.20 (0.13 2.3) 0.028 1.0 (-0.03 - 2.0) 0.057 Smoking 10.1 (-1.1 - 21.4) 0.077 7.6 (-3.3 18.6) 0.17 9.0 (-1.6 - 19.6) 0.095 Operating
surgeon
Spine surgeon
-2.5 (-12.9 - 7.9) 0.64 -7.3 (-18.6 4.1) 0.20 Working status Per category 0.91 (-1.46 - 3.28) 0.84
LDH level L4-5 vs. L5- S1
-4.0 (-13.9 - 5.8) 0.42 -7.2 (-17.5 3.1) 0.017 -9.9 (-19.4 - -0.4) 0.040 L3-4 vs.
L5-S1
-13.3 (-32.4 - 5.7) 0.17 -10.2 (-30.7 10.4) 0.33 -14.8 (-34.3 - 4.8)
0.14
Other vs. L5-
S1
28.3 (2.32 - 54.3) 0.033 30.2 (5.3 55.2) 0.018 24.8 (0.7 - 48.8) 0.044 Previous surgery 6.1 (-7.8 - 20.0) 0.38 9.0 (-4.3 22.4) 0.18
Motor deficit 11.2 (0.99 - 21.5) 0.032 12.8 (2.8 22.8) 0.013 13.8 (4.0 - 23.7) 0.006 Symptom
duration
Per day -0.050 (-0.30 - 0.21) 0.72 -0.02 (-0.25 0.21) 0.85 Operative time Per minute -0.047 (-0.24 - 0.05) 0.51 -0.10 (-0.26 0.04) 0.19
Regression diagnostics Adj. R2 0.206 0.203
AIC 784.0 779.2
