- Email: [email protected]
Cranioplasty: Is Surgical Education Safe?
Accepted Manuscript Cranioplasty: is surgical education safe? Holger Joswig, Oliver P. Gautschi, Amir el Rahal, Lukas Sveikata, Andrea Bartoli, Gerhard Hildebrandt, Karl Schaller, Martin N. Stienen PII:
S1878-8750(16)30008-0
DOI:
10.1016/j.wneu.2016.03.081
Reference:
WNEU 3904
To appear in:
World Neurosurgery
Received Date: 31 January 2016 Revised Date:
23 March 2016
Accepted Date: 23 March 2016
Please cite this article as: Joswig H, Gautschi OP, el Rahal A, Sveikata L, Bartoli A, Hildebrandt G, Schaller K, Stienen MN, Cranioplasty: is surgical education safe?, World Neurosurgery (2016), doi: 10.1016/j.wneu.2016.03.081. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Cranioplasty: is surgical education safe? 1
2
2
2
2
Holger Joswig , Oliver P. Gautschi , Amir el Rahal , Lukas Sveikata , Andrea Bartoli , Gerhard 1
2
Hildebrandt , Karl Schaller , Martin N. Stienen
2
1
Department of Neurosurgery, Cantonal Hospital St.Gallen, Rorschacher Str. 95, St.Gallen,
RI PT
Switzerland 2
Department of Neurosurgery and Faculty of Medicine, University Hospital Geneva, Rue Gabrielle
SC
Perret-Gentil 4, 1205 Genève Switzerland
The authors declare no conflict of interest.
Correspondence to:
Service de Neurochirurgie
TE D
Martin N. Stienen, M.D.
M AN U
Conflict of interest:
Département des neurosciences cliniques
EP
Hôpitaux Universitaires de Genève Rue Gabrielle Perret-Gentil 4
AC C
1205 Genève, Suisse +41-79-55-33773
Email: [email protected]
Key words: complication; cranioplasty; functional outcome; resident training; risk; safety; surgical education
ACCEPTED MANUSCRIPT Introduction In times of work-hour restrictions in residency, high-quality neurosurgical education in Europe is a 1
concern of paramount importance. While regulating the neurosurgical training program is one 1
solution, surgical education should begin as early as possible without putting patients at risk and should be complemented by surgical training simulators and cadaver courses. In the context of a previously reported structured training program used at the Department of Neurosurgery of the
RI PT
Cantonal Hospital St.Gallen, surgical resident education proved to be safe for several spinal procedures such as lumbar microdiscectomy and decompression as well as anterior cervical 2,8,11,12
discectomy.
Whether these findings also hold true for cranial surgery currently remains
unanswered.
Cranioplasty (CP) is a reconstructive procedure, most commonly performed in patients after 5
SC
decompressive hemicraniectomy for elevated intracranial pressure secondary to ischemic or hemorrhagic stroke or traumatic brain injury (TBI). Less frequently, CP is performed after infectious or wound healing complications following bone flap re-insertion after cranial surgery. On the one hand,
M AN U
CP is a relatively straightforward procedure commonly performed early on in neurosurgery residency training. On the other hand, CP is one of the cranial procedures associated with considerably high 14
rates of overall complications that are reported to be up to 31.32%.
The present work tested the hypothesis that complication rates and postoperative outcome of supervised residents and experienced board-certified faculty neurosurgeons (BCFN) are similar in CP
Material and methods
TE D
surgery.
Study design and patient identification
EP
Retrospective two-center study on consecutive patients scheduled for CP in the time interval 01/2010 to 08/2015 at Cantonal Hospital St.Gallen (KSSG) and between 01/2008 to 08/2013 at the University
AC C
Hospital Geneva (HUG). Patient charts were reviewed and all patients with complete relevant clinical and radiological data were included. All cases were dichotomized into teaching cases (patients operated on by a neurosurgery resident in postgraduate year (PGY)-2 to PGY-6) and non-teaching 2,8,11,12
cases (patients operated on by a BCFN).
At KSSG and HUG, residents fulfill the requirements
to perform their first CP after 18 months of training, at the time when they also begin to perform cranial approaches. Group assignment was mostly influenced by the surgeon’s level of experience at the initial surgery: patients who previously underwent removal of their bone flap by a resident or BCFN were usually re-assigned to the same surgeon in order to pursue continued patient care.
Patients or their next-of-kin were informed on their surgeon’s training level. Every teaching case was supervised by a BCFN, who was usually scrubbed in and intervened when the resident experienced difficulties, e.g. dissection of scar tissue, hemostasis or in case of intraoperative complications such as brain swelling and problems with fitting the flap. The study protocol respected ‘cross-overs’ as 2
ACCEPTED MANUSCRIPT previously handled
2,8,11,12
: Teaching cases were operations that were (almost) completely performed
by the trainee. For the analysis, the operation was declared a ‘non-teaching case’ whenever key parts of the procedure such as the surgical approach (interface dissection), handling complications (brain swelling, bleeding), ventricle tapping (if necessary) or constructing the CP were performed by the BCFN. Informed consent was obtained from all individual participants included in the study. Sample size calculation
RI PT
There was no literature data available on cranial complication rates comparing residents and BCFNs. An estimated complication rate of 20% in teaching cases and 15% in non-teaching cases with a standard deviation of 10% was used for sample size calculation, which revealed that n=64 patients per group were needed to detect a difference with a power of 80% and alpha set at 0.05.
SC
Preoperative factors
Besides patients’ baseline characteristics and details on comorbidities, the etiology and side of the bone defect, time to CP from bone flap removal, number of previous cranial surgeries since the bone
M AN U
flap removal as well as the modified Rankin Scale (mRS) before CP were determined. Computed tomography (CT) measurements of the diameter of the craniectomy site were taken as demonstrated in Figure 1 to compare the size of the bone defect.
Surgical technique and management
Timing of CP was not standardized and depended on the clinical and radiological evolution of the
TE D
patient following craniectomy. Timing for CP in teaching cases was decided by both the resident and his/her supervising BCFN. Surgery was performed under general anesthesia in supine or lateral decubitus position. Usually, cefamandole 1-2g (KSSG) / cefazolin 2g (HUG) (MANDOKEF / KEFZOL, Teva Pharma AG, Basel, Switzerland) was administered for standard perioperative antibiotic
EP
prophylaxis. A lumbar drain was inserted in case of insufficient subsidence of the skin flap with evidence of ventricular enlargement. The scar was re-incised and extended when necessary. The myocutaneous flap was dissected free from the underlying dura or patch. Once the bone edges and were prepared,
a frozen
and autoclaved
autologeous
bone,
a
AC C
central tack-up sutures
polymethylmethacrylat (PMMA) (PALACOS, Heraeus, Yverdon-les-Bains, Switzerland) or polyetherether-ketone (PEEK) (PEEK-SYNTHES, Oberdorf, Switzerland) plastic was inserted and the temporal muscle re-attached. The choice of the material was largely influenced by the etiology of bone flap removal. An extra-cranial drain without suction was inserted at the discretion of the surgeon and wound closure done in usual fashion.
Data collection Operation time in minutes (min), estimated blood loss (EBL) and type of graft material were recorded. 6
Complications up to postoperative day 30 were recorded and classified according to Ibanez et al. : Grade-I complications were any non-life threatening deviations from the normal postoperative course, grade-II complications were those requiring invasive (surgical) treatment (e.g. surgically relevant bone flap necrosis, surgical site infection (SSI), epidural hematoma), grade-III complications refer to life3
ACCEPTED MANUSCRIPT threatening complications requiring management in an intensive care unit (ICU; e.g. large epidural hematoma, intracerebral hemorrhage) and grade-IV complications result in death. SSI and delayed SSI > 30 days postoperative requiring re-do surgery were recorded separately. The modified Rankin Scale (mRS) at the last known follow-up (FU) was used to determine the postoperative outcome.
Statistical methods and study endpoints
RI PT
Comparison of important baseline (Table 1) and surgical parameters (Table 2) between the study 2
groups was done using Pearson Chi tests for categorical and Rank-Sum tests for ordinal variables. The same applies to the comparison of operation time, EBL and complication rates (Table 2 and 3). The primary endpoint was the occurrence of a postoperative complication. The secondary endpoint was patients’ clinical outcome at the last FU. Since we noted a relatively high rate of infectious
SC
complications, a further analysis pertaining to SSI was done. The primary and secondary endpoints 2
were directly analyzed using a Pearson Chi test (Table 2 and 3), uncorrected for group differences at baseline. In order to estimate the effect size of the relationship between teaching case and the primary
M AN U
or secondary endpoints, as well as to correct for existing study group differences, logistic regression analysis was performed using forced-entry methodology. First, a univariate model was created to assess relationships without adjustment. Then, a multivariate model was created to adjust for the above-mentioned parameters. Effect sizes before and after adjustment were analyzed for changes. The software used for statistical analysis was Stata v14 for Mac (StataCorp LP, College Station,
Results
TE D
Texas, USA). P-values <0.05 were considered statistically significant.
A total of n=244 CP cases were identified of which four were excluded due to incomplete FU data. Therefore, n=240 patients were enrolled in the study; 137 (57.1%) of which were teaching cases
EP
(operated on by a resident in training) and 103 (42.9%) were non-teaching cases (operated on by a BCFN). Teaching cases were performed by a total of n=21 residents. Non-teaching cases were performed by a total of n=15 BCFN. No teaching cases were handed over to the BCFN and there was
AC C
no crossover from the non-teaching to the teaching group. There was no significant difference in the distribution of teaching and non-teaching cases between the two teaching hospitals (58.4% versus 2
(vs.) 41.6% teaching cases, Pearson Chi 2.76, p=0.096).
In general, the study groups were well-balanced in terms of important baseline parameters such as age, gender, pre-operative mRS, side and type of CP (Table 1) except etiology. Significantly more decompressive craniectomies for ischemic strokes represent the etiology for a CP in the teaching group. Time to CP from bone flap removal, number of previous cranial surgeries since the craniectomy, as well as the diameter of the bone defect were equal (Figure 1A-B; Table 2). The operation time for residents (129.2 min ± 47.9) was significantly longer than for BCFN (115.8 ± 62.5; p<0.001), while EBL was equal (Table 2). Residents and BCFN used the same materials for CP. The mean FU post-CP for the complete group was about one year and mRS at the last FU was equal 4
ACCEPTED MANUSCRIPT between the groups (Table 2). Overall, the rate of postoperative complications was 25.5% for teaching cases and 29.1% for non-teaching cases with no significant differences when stratified for the Ibanez classification (p=0.106; Table 3). Severe complications resulting in death occurred in n=2 patients (0.8%). SSI requiring CP removal occurred in a total n=27 patients (11.3%; n=19 in the teaching and n=8 in the non-teaching group, Pearson Chi2 = 2.1924, p=0.139).
RI PT
In univariate analysis, supervised residents were as likely as BCFN to have a postoperative complication (Odds ratio (OR) 0.83, 95% Confidence Interval (CI) 0.47-1.48, p=0.537; Table 4). Once corrected for baseline group differences in age, etiology and mRS, this relationship remained stable (OR 0.77, 95% CI 0.42-1.39, p=0.385; Table 4). With respect to infectious complications in particular, no group differences were noted in the uni- and multivariate analysis (Table 5). The postoperative
SC
clinical outcome in terms of mRS is shown in table 2. The likelihood to experience an improvement on the mRS scale after CP was equal among patients operated on by supervised residents or BCFN in
M AN U
the uni- and multivariate analysis (Table 6).
The 137 teaching procedures were performed by residents in the following PGYs: n=3 in PGY1, n=41 in PGY2, n=37 in PGY3, n=25 in PGY4, n=15 in PGY5, and n=16 in PGY6. When dichotomized for junior (PGY 1-3; n=81 (60%)) and senior residents (PGY 4-6; n=54 (40%)), operation time was longer for junior residents (136.7 ± 51.5 min vs. 118.9 ± 40.9; p=0.042) while EBL was similar (242.1 ± 186.6 ml vs. 222.8 ± 197.6 ml; p=0.484). A complication occurred in n=20 (24.7%) patients operated on by a supervised junior resident and in n=15 (27.8%) patients operated on by a senior resident (Pearson 2
TE D
Chi = 0.1607; p=0.688). An infection requiring CP removal occurred in n=11 (13.6%) of patients operated on by a supervised junior resident and in n=8 (14.8%) patients operated on by a senior 2
resident (Pearson Chi = 0.0408; p=0.840). Postoperative neurological improvement on the mRS scale was noted in n=34 (42.0%) of patients operated on by a supervised junior resident and in n=18 2
AC C
Discussion
EP
(33.3%) patients of senior residents (Pearson Chi = 1.0218; p=0.312).
A consecutive cohort of n=240 patients undergoing cranioplasty performed by either supervised residents (teaching cases) or BCFNs (non-teaching cases) at two Swiss teaching hospitals was retrospectively analyzed. Complication rates and postoperative outcomes proved to be equal for teaching and non-teaching cases using uni- and multivariate analysis to account for group imbalances. The present data thus suggest that a relatively simple cranial neurosurgical procedure can be safely performed by residents within a structured training program. To the best of our knowledge, the current study is the first of its kind to analyze in detail the safety of resident involvement in CP. The present results compliment previous reports on patients’ safety in spine surgery performed by supervised residents
2,8,11,12
and provide important information for the current debate on working time restrictions, 1,7,9,10
present and future neurosurgical training.
5
ACCEPTED MANUSCRIPT A recent survey among European neurosurgical residents revealed that only 54.6% were allowed to 9
do a cranial procedure within the first 36 months of training. Although this question referred to craniotomies in general and the timing and operative exposure to CP was not specifically investigated, it can be assumed that European residents are introduced to standard cranial procedures such as CP relatively late in their training. From this survey it became evident that neurosurgical training is by no mean standardized in Europe. Social, economic, legal factors and traditional reasons may be held
RI PT
responsible for this relatively late exposure of European neurosurgical residents to cranial procedures during training. European neurosurgery residency directors need to break with outdated traditions and instead
establish
modern,
goal-orientated
program
guidelines
to
continue
training
junior
neurosurgeons who will be able to compete internationally. The results of the present work suggest that when continuous supervision by a BCFN is guaranteed, CP can safely be handed over to
SC
residents at a relatively early stage of their training. Importantly, except for longer operation times in the junior residents subgroup (PGY 1-3), there were no differences with respect to the rate of complications and the clinical outcome as compared to senior residents (PGY 4-6). Sometimes the 4
M AN U
resident’s role in the operation room remains undisclosed to the patient. With our philosophy of supervised resident care we rarely encounter any patient objection.
In general, the study groups were balanced for most of the important parameters, but etiology. The fact that residents more often performed CP after decompressive hemicraniectomy for ischemic stroke might relate to our management that decompressive hemicraniectomies are often assigned to residents during on-call service. The subsequent CP is thus performed by the same resident in
TE D
training. We did not selectively delete patients and therefore homogenize the study groups to prevent selection bias. Importantly, etiology of CP has previously been shown to not represent a significant 5
predictor for postoperative outcome (Glasgow Outcome Score) or complications. In contrast, in our multivariate analysis, etiology of CP was identified an independent predictor of neurological
EP
improvement post-CP (Table 6). In fact, it appeared that the neurological outcome after CP depended more on the indication for the initial craniectomy than on the skill level of the surgeon at the time of CP. Therefore, in the analysis of the primary endpoint, study group differences in etiology were
AC C
adjusted for. As tendencies for group differences in age and pre-operative mRS were noted, these factors were likewise introduced into the adjusted logistic regression model.
All in all, our complication rate is comparable to what has been reported in the literature. A recent large US-series reported an overall complication rate of 31.32%.
14
With respect to SSI in particular,
our rate of 11.3% is in agreement with the average CP infection rate of 7.9% reported in a systematic 13
review by Yalda et al..
Analysis of putative predictors for SSI was not specifically undertaken in the
current study that focused on resident training. Yalda et al. revealed no associated effect of other factors such as early surgery, material and method of flap preservation.
13
Interestingly, our multivariate
analysis, however, identified mRS before CP to be an independent predictor of SSI (Table 5). Of note, 5.2% of patients with a preoperative mRS 0-2 vs. 15.3% of patients with a mRS 3-5 developed a SSI 2
requiring CP removal (Person Chi 5.84, p=0.016). 6
ACCEPTED MANUSCRIPT 2,8,11,12
Unlike previous single-center reports,
the present analysis combines the data from two Swiss
teaching hospitals with slightly different patient management and educational programs. Thus, the present results might be more applicable to other institutions. While we are advocates of early surgical training, at the same time, we encourage all centers to carefully monitor their local complication rates and outcomes when introducing more resident hands-on training. The presence of a BCFN in the operating room is important even when the resident performs the surgical steps. Having an
RI PT
experienced neurosurgeon confirm the dissection of the correct interface during the approach, decide for ventricle tapping in case of brain swelling, help with controlling hemostasis, or supervise constructing the CP, is very reassuring for the resident and prevents complications. Once again, it 2,8
should be stressed that residents’ operative skills are not equal to those of a BCFN,
as in fact all
teaching cases were supervised and took slightly longer. The aspect of longer operation time should be kept in mind since the infection rate is thought to significantly increase when the total operation
SC
3
time exceeds 2 hours.
M AN U
Limitations
The greatest limitation of the current study lies in its retrospective design with its inherent risk of underreporting adverse events and complications likewise for both study groups. The FU period was only partially standardized as 30-day complications and those occurring later on (late SSI, aseptic necrosis, etc.) were combined in the analysis of the primary endpoint. A more detailed account on patient co-morbidities such as smoking status, diabetes and hypertension that have also been 14
associated with complications,
as well as a detailed account on seizures and shunt-dependent 5
Conclusions
TE D
hydrocephalus that are known to predict postoperative outcome, would have been desirable.
Supervised neurosurgical resident care in the setting of a structured training program is safe for
EP
cranioplasty, a standard cranial procedure. More data on other cranial procedures is warranted to
AC C
assess the safety of resident training in other cranial neurosurgery procedures.
7
ACCEPTED MANUSCRIPT Figures Figure 1: Axial (A) and coronal (B) computed tomography measurements of the bone defect in
AC C
EP
TE D
M AN U
SC
RI PT
patients undergoing cranioplasty.
8
ACCEPTED MANUSCRIPT Acknowledgements We thank Carolin Hock for proofreading the manuscript. We thank all patients that agreed to be operated by residents in training for their trust.
RI PT
Conflict of interest statement All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants; participation in speakers' bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements), or non-financial interest (such as personal or
SC
professional relationships, affiliations, knowledge or beliefs) in the subject matter or materials
Funding
AC C
EP
TE D
No funding was received for this research.
M AN U
discussed in this manuscript.
9
ACCEPTED MANUSCRIPT References
6. 7. 8. 9. 10. 11. 12.
13.
14.
RI PT
SC
5.
M AN U
4.
TE D
3.
EP
2.
Brennum J, van Loon J. Neurosurgical education in Europe. Acta neurochirurgica. Jan 2016;158(1):1-2. Joswig H, Hock C, Hildebrandt G, Schaller K, Stienen MN. Microscopic lumbar spinal stenosis decompression: is surgical education safe? Acta neurochirurgica. Dec 19 2015. Kim H, Sung SO, Kim SJ, Kim SR, Park IS, Jo KW. Analysis of the factors affecting graft infection after cranioplasty. Acta neurochirurgica. Nov 2013;155(11):21712176. Knifed E, Taylor B, Bernstein M. What surgeons tell their patients about the intraoperative role of residents: a qualitative study. American journal of surgery. Nov 2008;196(5):788-794. Krause-Titz UR, Warneke N, Freitag-Wolf S, Barth H, Mehdorn HM. Factors influencing the outcome (GOS) in reconstructive cranioplasty. Neurosurgical review. Dec 1 2015. Landriel Ibanez FA, Hem S, Ajler P, Vecchi E, Ciraolo C, Baccanelli M, Tramontano R, Knezevich F, Carrizo A. A new classification of complications in neurosurgery. World neurosurgery. May-Jun 2011;75(5-6):709-715; discussion 604-711. Schaller K. Neurosurgical training under European law. Acta neurochirurgica. Mar 2013;155(3):547. Stienen MN, Joswig H, Jucker D, Hildebrandt G, Schaller K, Gautschi OP. Anterior cervical discectomy and fusion: is surgical education safe? Acta neurochirurgica. Sep 2015;157(8):1395-1404. Stienen MN, Netuka D, Demetriades AK, Ringel F, Gautschi OP, Gempt J, Kuhlen D, Schaller K. Neurosurgical resident education in Europe-results of a multinational survey. Acta neurochirurgica. Jan 2016;158(1):3-15. Stienen MN, Netuka D, Demetriades AK, Ringel F, Gautschi OP, Gempt J, Kuhlen D, Schaller K. Working time of neurosurgical residents in Europe-results of a multinational survey. Acta neurochirurgica. Jan 2016;158(1):17-25. Stienen MN, Smoll NR, Hildebrandt G, Schaller K, Gautschi OP. Early surgical education of residents is safe for microscopic lumbar disc surgery. Acta neurochirurgica. Jun 2014;156(6):1205-1214. Stienen MN, Smoll NR, Tessitore E, Schaller K, Hildebrandt G, Gautschi OP. Surgical Resident Education in Noninstrumented Lumbar Spine Surgery: A Prospective Observational Study with a 4.5-Year Follow-Up. World neurosurgery. Dec 2015;84(6):1589-1597. Yadla S, Campbell PG, Chitale R, Maltenfort MG, Jabbour P, Sharan AD. Effect of early surgery, material, and method of flap preservation on cranioplasty infections: a systematic review. Neurosurgery. Apr 2011;68(4):1124-1129; discussion 1130. Zanaty M, Chalouhi N, Starke RM, Clark SW, Bovenzi CD, Saigh M, Schwartz E, Kunkel ES, Efthimiadis-Budike AS, Jabbour P, Dalyai R, Rosenwasser RH, Tjoumakaris SI. Complications following cranioplasty: incidence and predictors in 348 cases. J Neurosurg. Jul 2015;123(1):182-188.
AC C
1.
10
ACCEPTED MANUSCRIPT Table 1. Baseline parameters of patients undergoing cranioplasty by either a supervised resident (teaching group) or a board-certified faculty neurosurgeon (non-teaching group). Variables on a ratio scale are presented as mean ± standard deviation. Variables on a nominal scale are presented as count and per cent. ASA = American Society of Anesthesiologists, ICH = intraparenchymal hemorrhage, mRS = modified Rankin scale, SAH
58.4% 41.6%
23.9 ± 4.9
Non-teaching 43.1 ± 18.5 66 37
64.1% 35.9%
23.8 ± 4.6
0.372 0.941
4.4% 95.6%
9.7% 90.3%
30 107
21.9% 78.1%
29 74
28.2% 71.8%
0.265
23 114
16.8% 83.2%
15 88
14.6% 85.4%
0.640
1 59 68 9
0.7% 43.1% 49.6% 6.6%
1 54 41 7
1.0% 52.4% 39.8% 6.8%
58 16 35 5 23
42.3% 11.7% 25.6% 3.7% 16.8%
54 14 7 6 22
52.4% 13.6% 6.8% 5.8% 21.4%
132 2 3
96.3% 1.5% 2.2%
97 2 4
94.2% 1.9% 3.9%
53 79 5
38.7% 57.7% 3.6%
59 38 6
36.9% 57.3% 5.8%
16 57 64 n=137
11.7% 41.6% 46.7% 100%
23 43 37 n=103
22.3% 44.8% 35.9% 100%
TE D
M AN U
6 131
AC C
10 93
p value 0.071
SC
80 57
EP
Age Gender Male Female Body mass index in kg/m2 Diabetes Yes No Arterial hypertension Yes No Smoking Yes No ASA score 1 2 3 4 Etiology TBI SAH Ischemic stroke ICH Other Type of craniectomy Unilateral Bilateral Bifrontal Side Left Right Bilateral mRS 0–1 2–3 4–5
Teaching 46.8 ± 17.1
RI PT
= subarachnoid hemorrhage, TBI = traumatic brain injury.
0.101
0.495
0.006
0.708
0.719
0.057
ACCEPTED MANUSCRIPT Table 2. Surgery-related parameters and follow-up of patients undergoing cranioplasty by either a supervised resident (teaching group) or a board-certified faculty neurosurgeon (nonteaching group). Variables on a ratio scale are presented as mean ± standard deviation. Variables on a nominal scale are presented as count and per cent. EBL = estimated blood loss, min = minutes, mm = millimetres, ml = millilitres, mRS = modified Rankin Scale, PEEK
Teaching
Non-teaching
p value
120.8 ± 107.8
132.8 ± 125.9
0.655
98 39
72 31
0.783
0.278 0.355
129.2 ± 47.9
115.8 ± 62.5
0.001
234.3 ± 189.9
223.1 ± 249.6
0.444
58 66 13 -
52 41 19 1
M AN U
113.5 ± 28.2 94.8 ± 19.4
TE D
42.3% 48.2% 9.5% 0.0%
273.7 ± 398.9
31 62 42 2 n=137
AC C
69.9% 30.1%
SC
71.5% 28.5%
118.1 ± 24.7 97.4 ± 18.1
EP
Time to cranioplasty in days Previous cranial surgeries 1 ≥2 Diameter of craniectomy in mm axial coronal Operation time in min EBL in ml Type of material Autologous bone PMMA PEEK other Time of follow-up in days mRS at follow-up 0–1 2–3 4–5 6
RI PT
= polyetheretherketone, PMMA = polymethyl-methacrylat.
22.6% 45.3% 30.7% 1.4% 100%
50.5% 39.8% 8.7% 1.0%
413.5 ± 639.4 35 40 26 2 n=103
34.0% 38.8% 25.3% 1.9% 100%
0.370
0.841
0.260
ACCEPTED MANUSCRIPT Table 3. Degree of surgical complications after cranioplasty (CP) and the level of experience of the primary surgeon. Surgical complications up to 30 days after CP were classified according to Ibanez et al. (6) Non-teaching 73 70.9% 10 9.7% 8 7.8% 10 9.7% 2 1.9% n=103 100%
p value 0.106
RI PT
Teaching 102 74.4% 6 4.4% 19 13.9% 10 7.3% 0.0% n=137 100%
AC C
EP
TE D
M AN U
SC
Ibanez degree 0 1 2 3 4 Total
ACCEPTED MANUSCRIPT Table 4. Estimation of the effect size of the relationship between the rate of surgical complications after cranioplasty (CP) and the level of experience of the primary surgeon. mRS = modified Rankin Scale.
OR 0.83 0.60 1.00 1.62
p value 0.537 0.085 0.958 0.025
OR 0.77 0.52 1.07 1.85
Multivariate 95% CI 0.42 – 1.39 0.28 – 0.95 0.88 – 1.31 1.18 – 2.90
AC C
EP
TE D
M AN U
SC
Teaching case Age ≥ 50 years Etiology mRS before CP
Univariate 95% CI 0.47 – 1.48 0.34 – 1.07 0.83 – 1.20 1.06 – 2.47
p value 0.385 0.034 0.477 0.007
RI PT
Surgical complication
ACCEPTED MANUSCRIPT Table 5. Estimation of the effect size of the relationship between the rate of postoperative infection after cranioplasty (CP) requiring CP removal and the level of experience of the primary surgeon. mRS = modified Rankin scale.
p value 0.144 0.482 0.142 0.050
OR 1.66 1.04 1.30 2.06
Multivariate 95% CI 0.69 – 4.04 0.45 – 2.39 0.98 – 1.73 1.05 – 4.06
M AN U TE D EP AC C
p value 0.261 0.926 0.062 0.037
RI PT
OR 1.91 1.33 1.20 1.88
Univariate 95% CI 0.80 – 4.56 0.59 – 2.98 0.94 – 1.54 1.00 – 3.54
SC
Infection requiring CP removal Teaching case Age ≥ 50 years Etiology mRS before CP
ACCEPTED MANUSCRIPT Table 6. Estimation of the effect size of the relationship between the rate of patients with improvement on the modified Rankin Scale (mRS) after cranioplasty (CP) and the level of experience of the primary surgeon.
p value 0.526 0.311 <0.001 <0.001
OR 1.11 0.69 1.30 2.49
Multivariate 95% CI 0.62 – 2.00 0.38 – 1.23 0.54 – 0.84 1.59 – 3.90
M AN U TE D EP AC C
p value 0.719 0.210 <0.001 <0.001
RI PT
OR 1.18 0.76 0.66 2.48
Univariate 95% CI 0.69 – 2.02 0.45 – 1.29 0.54 – 0.81 1.63 – 3.77
SC
Improvement on the mRS after CP Teaching case Age ≥ 50 years Etiology mRS before CP
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Abbreviations: BCFN = board certified faculty neurosurgeon CI = confidence interval CP = cranioplasty CT = computed tomography EBL = estimated blood loss
RI PT
FU = follow-up HUG = University Hospital Geneva ICH = intracerebral hemorrhage ICU = intensive care unit KSSG = Cantonal Hospital St.Gallen
SC
min = minutes ml = millilitres mm = millimetres
OR = odds ratio PGY = postgraduate year PEEK = polyether-ether-ketone PMMA = polymethylmethacrylat SAH = subarachnoid hemorrhage SD = standard deviation
AC C
EP
TBI = traumatic brain injury
TE D
SSI = surgical site infections
M AN U
mRS = modified Rankin Scale
ACCEPTED MANUSCRIPT Highlights Safety aspects of cranioplasty by residents or staff surgeons were assessed
-
Operation time was longer in procedures performed by supervised residents
-
The likelihood and severity of postoperative complications were similar
-
Surgical site infections requiring cranioplasty removal were similar
-
The likelihood for neurological improvement after cranioplasty was equal
AC C
EP
TE D
M AN U
SC
RI PT
-
S1878-8750(16)30008-0
DOI:
10.1016/j.wneu.2016.03.081
Reference:
WNEU 3904
To appear in:
World Neurosurgery
Received Date: 31 January 2016 Revised Date:
23 March 2016
Accepted Date: 23 March 2016
Please cite this article as: Joswig H, Gautschi OP, el Rahal A, Sveikata L, Bartoli A, Hildebrandt G, Schaller K, Stienen MN, Cranioplasty: is surgical education safe?, World Neurosurgery (2016), doi: 10.1016/j.wneu.2016.03.081. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Cranioplasty: is surgical education safe? 1
2
2
2
2
Holger Joswig , Oliver P. Gautschi , Amir el Rahal , Lukas Sveikata , Andrea Bartoli , Gerhard 1
2
Hildebrandt , Karl Schaller , Martin N. Stienen
2
1
Department of Neurosurgery, Cantonal Hospital St.Gallen, Rorschacher Str. 95, St.Gallen,
RI PT
Switzerland 2
Department of Neurosurgery and Faculty of Medicine, University Hospital Geneva, Rue Gabrielle
SC
Perret-Gentil 4, 1205 Genève Switzerland
The authors declare no conflict of interest.
Correspondence to:
Service de Neurochirurgie
TE D
Martin N. Stienen, M.D.
M AN U
Conflict of interest:
Département des neurosciences cliniques
EP
Hôpitaux Universitaires de Genève Rue Gabrielle Perret-Gentil 4
AC C
1205 Genève, Suisse +41-79-55-33773
Email: [email protected]
Key words: complication; cranioplasty; functional outcome; resident training; risk; safety; surgical education
ACCEPTED MANUSCRIPT Introduction In times of work-hour restrictions in residency, high-quality neurosurgical education in Europe is a 1
concern of paramount importance. While regulating the neurosurgical training program is one 1
solution, surgical education should begin as early as possible without putting patients at risk and should be complemented by surgical training simulators and cadaver courses. In the context of a previously reported structured training program used at the Department of Neurosurgery of the
RI PT
Cantonal Hospital St.Gallen, surgical resident education proved to be safe for several spinal procedures such as lumbar microdiscectomy and decompression as well as anterior cervical 2,8,11,12
discectomy.
Whether these findings also hold true for cranial surgery currently remains
unanswered.
Cranioplasty (CP) is a reconstructive procedure, most commonly performed in patients after 5
SC
decompressive hemicraniectomy for elevated intracranial pressure secondary to ischemic or hemorrhagic stroke or traumatic brain injury (TBI). Less frequently, CP is performed after infectious or wound healing complications following bone flap re-insertion after cranial surgery. On the one hand,
M AN U
CP is a relatively straightforward procedure commonly performed early on in neurosurgery residency training. On the other hand, CP is one of the cranial procedures associated with considerably high 14
rates of overall complications that are reported to be up to 31.32%.
The present work tested the hypothesis that complication rates and postoperative outcome of supervised residents and experienced board-certified faculty neurosurgeons (BCFN) are similar in CP
Material and methods
TE D
surgery.
Study design and patient identification
EP
Retrospective two-center study on consecutive patients scheduled for CP in the time interval 01/2010 to 08/2015 at Cantonal Hospital St.Gallen (KSSG) and between 01/2008 to 08/2013 at the University
AC C
Hospital Geneva (HUG). Patient charts were reviewed and all patients with complete relevant clinical and radiological data were included. All cases were dichotomized into teaching cases (patients operated on by a neurosurgery resident in postgraduate year (PGY)-2 to PGY-6) and non-teaching 2,8,11,12
cases (patients operated on by a BCFN).
At KSSG and HUG, residents fulfill the requirements
to perform their first CP after 18 months of training, at the time when they also begin to perform cranial approaches. Group assignment was mostly influenced by the surgeon’s level of experience at the initial surgery: patients who previously underwent removal of their bone flap by a resident or BCFN were usually re-assigned to the same surgeon in order to pursue continued patient care.
Patients or their next-of-kin were informed on their surgeon’s training level. Every teaching case was supervised by a BCFN, who was usually scrubbed in and intervened when the resident experienced difficulties, e.g. dissection of scar tissue, hemostasis or in case of intraoperative complications such as brain swelling and problems with fitting the flap. The study protocol respected ‘cross-overs’ as 2
ACCEPTED MANUSCRIPT previously handled
2,8,11,12
: Teaching cases were operations that were (almost) completely performed
by the trainee. For the analysis, the operation was declared a ‘non-teaching case’ whenever key parts of the procedure such as the surgical approach (interface dissection), handling complications (brain swelling, bleeding), ventricle tapping (if necessary) or constructing the CP were performed by the BCFN. Informed consent was obtained from all individual participants included in the study. Sample size calculation
RI PT
There was no literature data available on cranial complication rates comparing residents and BCFNs. An estimated complication rate of 20% in teaching cases and 15% in non-teaching cases with a standard deviation of 10% was used for sample size calculation, which revealed that n=64 patients per group were needed to detect a difference with a power of 80% and alpha set at 0.05.
SC
Preoperative factors
Besides patients’ baseline characteristics and details on comorbidities, the etiology and side of the bone defect, time to CP from bone flap removal, number of previous cranial surgeries since the bone
M AN U
flap removal as well as the modified Rankin Scale (mRS) before CP were determined. Computed tomography (CT) measurements of the diameter of the craniectomy site were taken as demonstrated in Figure 1 to compare the size of the bone defect.
Surgical technique and management
Timing of CP was not standardized and depended on the clinical and radiological evolution of the
TE D
patient following craniectomy. Timing for CP in teaching cases was decided by both the resident and his/her supervising BCFN. Surgery was performed under general anesthesia in supine or lateral decubitus position. Usually, cefamandole 1-2g (KSSG) / cefazolin 2g (HUG) (MANDOKEF / KEFZOL, Teva Pharma AG, Basel, Switzerland) was administered for standard perioperative antibiotic
EP
prophylaxis. A lumbar drain was inserted in case of insufficient subsidence of the skin flap with evidence of ventricular enlargement. The scar was re-incised and extended when necessary. The myocutaneous flap was dissected free from the underlying dura or patch. Once the bone edges and were prepared,
a frozen
and autoclaved
autologeous
bone,
a
AC C
central tack-up sutures
polymethylmethacrylat (PMMA) (PALACOS, Heraeus, Yverdon-les-Bains, Switzerland) or polyetherether-ketone (PEEK) (PEEK-SYNTHES, Oberdorf, Switzerland) plastic was inserted and the temporal muscle re-attached. The choice of the material was largely influenced by the etiology of bone flap removal. An extra-cranial drain without suction was inserted at the discretion of the surgeon and wound closure done in usual fashion.
Data collection Operation time in minutes (min), estimated blood loss (EBL) and type of graft material were recorded. 6
Complications up to postoperative day 30 were recorded and classified according to Ibanez et al. : Grade-I complications were any non-life threatening deviations from the normal postoperative course, grade-II complications were those requiring invasive (surgical) treatment (e.g. surgically relevant bone flap necrosis, surgical site infection (SSI), epidural hematoma), grade-III complications refer to life3
ACCEPTED MANUSCRIPT threatening complications requiring management in an intensive care unit (ICU; e.g. large epidural hematoma, intracerebral hemorrhage) and grade-IV complications result in death. SSI and delayed SSI > 30 days postoperative requiring re-do surgery were recorded separately. The modified Rankin Scale (mRS) at the last known follow-up (FU) was used to determine the postoperative outcome.
Statistical methods and study endpoints
RI PT
Comparison of important baseline (Table 1) and surgical parameters (Table 2) between the study 2
groups was done using Pearson Chi tests for categorical and Rank-Sum tests for ordinal variables. The same applies to the comparison of operation time, EBL and complication rates (Table 2 and 3). The primary endpoint was the occurrence of a postoperative complication. The secondary endpoint was patients’ clinical outcome at the last FU. Since we noted a relatively high rate of infectious
SC
complications, a further analysis pertaining to SSI was done. The primary and secondary endpoints 2
were directly analyzed using a Pearson Chi test (Table 2 and 3), uncorrected for group differences at baseline. In order to estimate the effect size of the relationship between teaching case and the primary
M AN U
or secondary endpoints, as well as to correct for existing study group differences, logistic regression analysis was performed using forced-entry methodology. First, a univariate model was created to assess relationships without adjustment. Then, a multivariate model was created to adjust for the above-mentioned parameters. Effect sizes before and after adjustment were analyzed for changes. The software used for statistical analysis was Stata v14 for Mac (StataCorp LP, College Station,
Results
TE D
Texas, USA). P-values <0.05 were considered statistically significant.
A total of n=244 CP cases were identified of which four were excluded due to incomplete FU data. Therefore, n=240 patients were enrolled in the study; 137 (57.1%) of which were teaching cases
EP
(operated on by a resident in training) and 103 (42.9%) were non-teaching cases (operated on by a BCFN). Teaching cases were performed by a total of n=21 residents. Non-teaching cases were performed by a total of n=15 BCFN. No teaching cases were handed over to the BCFN and there was
AC C
no crossover from the non-teaching to the teaching group. There was no significant difference in the distribution of teaching and non-teaching cases between the two teaching hospitals (58.4% versus 2
(vs.) 41.6% teaching cases, Pearson Chi 2.76, p=0.096).
In general, the study groups were well-balanced in terms of important baseline parameters such as age, gender, pre-operative mRS, side and type of CP (Table 1) except etiology. Significantly more decompressive craniectomies for ischemic strokes represent the etiology for a CP in the teaching group. Time to CP from bone flap removal, number of previous cranial surgeries since the craniectomy, as well as the diameter of the bone defect were equal (Figure 1A-B; Table 2). The operation time for residents (129.2 min ± 47.9) was significantly longer than for BCFN (115.8 ± 62.5; p<0.001), while EBL was equal (Table 2). Residents and BCFN used the same materials for CP. The mean FU post-CP for the complete group was about one year and mRS at the last FU was equal 4
ACCEPTED MANUSCRIPT between the groups (Table 2). Overall, the rate of postoperative complications was 25.5% for teaching cases and 29.1% for non-teaching cases with no significant differences when stratified for the Ibanez classification (p=0.106; Table 3). Severe complications resulting in death occurred in n=2 patients (0.8%). SSI requiring CP removal occurred in a total n=27 patients (11.3%; n=19 in the teaching and n=8 in the non-teaching group, Pearson Chi2 = 2.1924, p=0.139).
RI PT
In univariate analysis, supervised residents were as likely as BCFN to have a postoperative complication (Odds ratio (OR) 0.83, 95% Confidence Interval (CI) 0.47-1.48, p=0.537; Table 4). Once corrected for baseline group differences in age, etiology and mRS, this relationship remained stable (OR 0.77, 95% CI 0.42-1.39, p=0.385; Table 4). With respect to infectious complications in particular, no group differences were noted in the uni- and multivariate analysis (Table 5). The postoperative
SC
clinical outcome in terms of mRS is shown in table 2. The likelihood to experience an improvement on the mRS scale after CP was equal among patients operated on by supervised residents or BCFN in
M AN U
the uni- and multivariate analysis (Table 6).
The 137 teaching procedures were performed by residents in the following PGYs: n=3 in PGY1, n=41 in PGY2, n=37 in PGY3, n=25 in PGY4, n=15 in PGY5, and n=16 in PGY6. When dichotomized for junior (PGY 1-3; n=81 (60%)) and senior residents (PGY 4-6; n=54 (40%)), operation time was longer for junior residents (136.7 ± 51.5 min vs. 118.9 ± 40.9; p=0.042) while EBL was similar (242.1 ± 186.6 ml vs. 222.8 ± 197.6 ml; p=0.484). A complication occurred in n=20 (24.7%) patients operated on by a supervised junior resident and in n=15 (27.8%) patients operated on by a senior resident (Pearson 2
TE D
Chi = 0.1607; p=0.688). An infection requiring CP removal occurred in n=11 (13.6%) of patients operated on by a supervised junior resident and in n=8 (14.8%) patients operated on by a senior 2
resident (Pearson Chi = 0.0408; p=0.840). Postoperative neurological improvement on the mRS scale was noted in n=34 (42.0%) of patients operated on by a supervised junior resident and in n=18 2
AC C
Discussion
EP
(33.3%) patients of senior residents (Pearson Chi = 1.0218; p=0.312).
A consecutive cohort of n=240 patients undergoing cranioplasty performed by either supervised residents (teaching cases) or BCFNs (non-teaching cases) at two Swiss teaching hospitals was retrospectively analyzed. Complication rates and postoperative outcomes proved to be equal for teaching and non-teaching cases using uni- and multivariate analysis to account for group imbalances. The present data thus suggest that a relatively simple cranial neurosurgical procedure can be safely performed by residents within a structured training program. To the best of our knowledge, the current study is the first of its kind to analyze in detail the safety of resident involvement in CP. The present results compliment previous reports on patients’ safety in spine surgery performed by supervised residents
2,8,11,12
and provide important information for the current debate on working time restrictions, 1,7,9,10
present and future neurosurgical training.
5
ACCEPTED MANUSCRIPT A recent survey among European neurosurgical residents revealed that only 54.6% were allowed to 9
do a cranial procedure within the first 36 months of training. Although this question referred to craniotomies in general and the timing and operative exposure to CP was not specifically investigated, it can be assumed that European residents are introduced to standard cranial procedures such as CP relatively late in their training. From this survey it became evident that neurosurgical training is by no mean standardized in Europe. Social, economic, legal factors and traditional reasons may be held
RI PT
responsible for this relatively late exposure of European neurosurgical residents to cranial procedures during training. European neurosurgery residency directors need to break with outdated traditions and instead
establish
modern,
goal-orientated
program
guidelines
to
continue
training
junior
neurosurgeons who will be able to compete internationally. The results of the present work suggest that when continuous supervision by a BCFN is guaranteed, CP can safely be handed over to
SC
residents at a relatively early stage of their training. Importantly, except for longer operation times in the junior residents subgroup (PGY 1-3), there were no differences with respect to the rate of complications and the clinical outcome as compared to senior residents (PGY 4-6). Sometimes the 4
M AN U
resident’s role in the operation room remains undisclosed to the patient. With our philosophy of supervised resident care we rarely encounter any patient objection.
In general, the study groups were balanced for most of the important parameters, but etiology. The fact that residents more often performed CP after decompressive hemicraniectomy for ischemic stroke might relate to our management that decompressive hemicraniectomies are often assigned to residents during on-call service. The subsequent CP is thus performed by the same resident in
TE D
training. We did not selectively delete patients and therefore homogenize the study groups to prevent selection bias. Importantly, etiology of CP has previously been shown to not represent a significant 5
predictor for postoperative outcome (Glasgow Outcome Score) or complications. In contrast, in our multivariate analysis, etiology of CP was identified an independent predictor of neurological
EP
improvement post-CP (Table 6). In fact, it appeared that the neurological outcome after CP depended more on the indication for the initial craniectomy than on the skill level of the surgeon at the time of CP. Therefore, in the analysis of the primary endpoint, study group differences in etiology were
AC C
adjusted for. As tendencies for group differences in age and pre-operative mRS were noted, these factors were likewise introduced into the adjusted logistic regression model.
All in all, our complication rate is comparable to what has been reported in the literature. A recent large US-series reported an overall complication rate of 31.32%.
14
With respect to SSI in particular,
our rate of 11.3% is in agreement with the average CP infection rate of 7.9% reported in a systematic 13
review by Yalda et al..
Analysis of putative predictors for SSI was not specifically undertaken in the
current study that focused on resident training. Yalda et al. revealed no associated effect of other factors such as early surgery, material and method of flap preservation.
13
Interestingly, our multivariate
analysis, however, identified mRS before CP to be an independent predictor of SSI (Table 5). Of note, 5.2% of patients with a preoperative mRS 0-2 vs. 15.3% of patients with a mRS 3-5 developed a SSI 2
requiring CP removal (Person Chi 5.84, p=0.016). 6
ACCEPTED MANUSCRIPT 2,8,11,12
Unlike previous single-center reports,
the present analysis combines the data from two Swiss
teaching hospitals with slightly different patient management and educational programs. Thus, the present results might be more applicable to other institutions. While we are advocates of early surgical training, at the same time, we encourage all centers to carefully monitor their local complication rates and outcomes when introducing more resident hands-on training. The presence of a BCFN in the operating room is important even when the resident performs the surgical steps. Having an
RI PT
experienced neurosurgeon confirm the dissection of the correct interface during the approach, decide for ventricle tapping in case of brain swelling, help with controlling hemostasis, or supervise constructing the CP, is very reassuring for the resident and prevents complications. Once again, it 2,8
should be stressed that residents’ operative skills are not equal to those of a BCFN,
as in fact all
teaching cases were supervised and took slightly longer. The aspect of longer operation time should be kept in mind since the infection rate is thought to significantly increase when the total operation
SC
3
time exceeds 2 hours.
M AN U
Limitations
The greatest limitation of the current study lies in its retrospective design with its inherent risk of underreporting adverse events and complications likewise for both study groups. The FU period was only partially standardized as 30-day complications and those occurring later on (late SSI, aseptic necrosis, etc.) were combined in the analysis of the primary endpoint. A more detailed account on patient co-morbidities such as smoking status, diabetes and hypertension that have also been 14
associated with complications,
as well as a detailed account on seizures and shunt-dependent 5
Conclusions
TE D
hydrocephalus that are known to predict postoperative outcome, would have been desirable.
Supervised neurosurgical resident care in the setting of a structured training program is safe for
EP
cranioplasty, a standard cranial procedure. More data on other cranial procedures is warranted to
AC C
assess the safety of resident training in other cranial neurosurgery procedures.
7
ACCEPTED MANUSCRIPT Figures Figure 1: Axial (A) and coronal (B) computed tomography measurements of the bone defect in
AC C
EP
TE D
M AN U
SC
RI PT
patients undergoing cranioplasty.
8
ACCEPTED MANUSCRIPT Acknowledgements We thank Carolin Hock for proofreading the manuscript. We thank all patients that agreed to be operated by residents in training for their trust.
RI PT
Conflict of interest statement All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants; participation in speakers' bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements), or non-financial interest (such as personal or
SC
professional relationships, affiliations, knowledge or beliefs) in the subject matter or materials
Funding
AC C
EP
TE D
No funding was received for this research.
M AN U
discussed in this manuscript.
9
ACCEPTED MANUSCRIPT References
6. 7. 8. 9. 10. 11. 12.
13.
14.
RI PT
SC
5.
M AN U
4.
TE D
3.
EP
2.
Brennum J, van Loon J. Neurosurgical education in Europe. Acta neurochirurgica. Jan 2016;158(1):1-2. Joswig H, Hock C, Hildebrandt G, Schaller K, Stienen MN. Microscopic lumbar spinal stenosis decompression: is surgical education safe? Acta neurochirurgica. Dec 19 2015. Kim H, Sung SO, Kim SJ, Kim SR, Park IS, Jo KW. Analysis of the factors affecting graft infection after cranioplasty. Acta neurochirurgica. Nov 2013;155(11):21712176. Knifed E, Taylor B, Bernstein M. What surgeons tell their patients about the intraoperative role of residents: a qualitative study. American journal of surgery. Nov 2008;196(5):788-794. Krause-Titz UR, Warneke N, Freitag-Wolf S, Barth H, Mehdorn HM. Factors influencing the outcome (GOS) in reconstructive cranioplasty. Neurosurgical review. Dec 1 2015. Landriel Ibanez FA, Hem S, Ajler P, Vecchi E, Ciraolo C, Baccanelli M, Tramontano R, Knezevich F, Carrizo A. A new classification of complications in neurosurgery. World neurosurgery. May-Jun 2011;75(5-6):709-715; discussion 604-711. Schaller K. Neurosurgical training under European law. Acta neurochirurgica. Mar 2013;155(3):547. Stienen MN, Joswig H, Jucker D, Hildebrandt G, Schaller K, Gautschi OP. Anterior cervical discectomy and fusion: is surgical education safe? Acta neurochirurgica. Sep 2015;157(8):1395-1404. Stienen MN, Netuka D, Demetriades AK, Ringel F, Gautschi OP, Gempt J, Kuhlen D, Schaller K. Neurosurgical resident education in Europe-results of a multinational survey. Acta neurochirurgica. Jan 2016;158(1):3-15. Stienen MN, Netuka D, Demetriades AK, Ringel F, Gautschi OP, Gempt J, Kuhlen D, Schaller K. Working time of neurosurgical residents in Europe-results of a multinational survey. Acta neurochirurgica. Jan 2016;158(1):17-25. Stienen MN, Smoll NR, Hildebrandt G, Schaller K, Gautschi OP. Early surgical education of residents is safe for microscopic lumbar disc surgery. Acta neurochirurgica. Jun 2014;156(6):1205-1214. Stienen MN, Smoll NR, Tessitore E, Schaller K, Hildebrandt G, Gautschi OP. Surgical Resident Education in Noninstrumented Lumbar Spine Surgery: A Prospective Observational Study with a 4.5-Year Follow-Up. World neurosurgery. Dec 2015;84(6):1589-1597. Yadla S, Campbell PG, Chitale R, Maltenfort MG, Jabbour P, Sharan AD. Effect of early surgery, material, and method of flap preservation on cranioplasty infections: a systematic review. Neurosurgery. Apr 2011;68(4):1124-1129; discussion 1130. Zanaty M, Chalouhi N, Starke RM, Clark SW, Bovenzi CD, Saigh M, Schwartz E, Kunkel ES, Efthimiadis-Budike AS, Jabbour P, Dalyai R, Rosenwasser RH, Tjoumakaris SI. Complications following cranioplasty: incidence and predictors in 348 cases. J Neurosurg. Jul 2015;123(1):182-188.
AC C
1.
10
ACCEPTED MANUSCRIPT Table 1. Baseline parameters of patients undergoing cranioplasty by either a supervised resident (teaching group) or a board-certified faculty neurosurgeon (non-teaching group). Variables on a ratio scale are presented as mean ± standard deviation. Variables on a nominal scale are presented as count and per cent. ASA = American Society of Anesthesiologists, ICH = intraparenchymal hemorrhage, mRS = modified Rankin scale, SAH
58.4% 41.6%
23.9 ± 4.9
Non-teaching 43.1 ± 18.5 66 37
64.1% 35.9%
23.8 ± 4.6
0.372 0.941
4.4% 95.6%
9.7% 90.3%
30 107
21.9% 78.1%
29 74
28.2% 71.8%
0.265
23 114
16.8% 83.2%
15 88
14.6% 85.4%
0.640
1 59 68 9
0.7% 43.1% 49.6% 6.6%
1 54 41 7
1.0% 52.4% 39.8% 6.8%
58 16 35 5 23
42.3% 11.7% 25.6% 3.7% 16.8%
54 14 7 6 22
52.4% 13.6% 6.8% 5.8% 21.4%
132 2 3
96.3% 1.5% 2.2%
97 2 4
94.2% 1.9% 3.9%
53 79 5
38.7% 57.7% 3.6%
59 38 6
36.9% 57.3% 5.8%
16 57 64 n=137
11.7% 41.6% 46.7% 100%
23 43 37 n=103
22.3% 44.8% 35.9% 100%
TE D
M AN U
6 131
AC C
10 93
p value 0.071
SC
80 57
EP
Age Gender Male Female Body mass index in kg/m2 Diabetes Yes No Arterial hypertension Yes No Smoking Yes No ASA score 1 2 3 4 Etiology TBI SAH Ischemic stroke ICH Other Type of craniectomy Unilateral Bilateral Bifrontal Side Left Right Bilateral mRS 0–1 2–3 4–5
Teaching 46.8 ± 17.1
RI PT
= subarachnoid hemorrhage, TBI = traumatic brain injury.
0.101
0.495
0.006
0.708
0.719
0.057
ACCEPTED MANUSCRIPT Table 2. Surgery-related parameters and follow-up of patients undergoing cranioplasty by either a supervised resident (teaching group) or a board-certified faculty neurosurgeon (nonteaching group). Variables on a ratio scale are presented as mean ± standard deviation. Variables on a nominal scale are presented as count and per cent. EBL = estimated blood loss, min = minutes, mm = millimetres, ml = millilitres, mRS = modified Rankin Scale, PEEK
Teaching
Non-teaching
p value
120.8 ± 107.8
132.8 ± 125.9
0.655
98 39
72 31
0.783
0.278 0.355
129.2 ± 47.9
115.8 ± 62.5
0.001
234.3 ± 189.9
223.1 ± 249.6
0.444
58 66 13 -
52 41 19 1
M AN U
113.5 ± 28.2 94.8 ± 19.4
TE D
42.3% 48.2% 9.5% 0.0%
273.7 ± 398.9
31 62 42 2 n=137
AC C
69.9% 30.1%
SC
71.5% 28.5%
118.1 ± 24.7 97.4 ± 18.1
EP
Time to cranioplasty in days Previous cranial surgeries 1 ≥2 Diameter of craniectomy in mm axial coronal Operation time in min EBL in ml Type of material Autologous bone PMMA PEEK other Time of follow-up in days mRS at follow-up 0–1 2–3 4–5 6
RI PT
= polyetheretherketone, PMMA = polymethyl-methacrylat.
22.6% 45.3% 30.7% 1.4% 100%
50.5% 39.8% 8.7% 1.0%
413.5 ± 639.4 35 40 26 2 n=103
34.0% 38.8% 25.3% 1.9% 100%
0.370
0.841
0.260
ACCEPTED MANUSCRIPT Table 3. Degree of surgical complications after cranioplasty (CP) and the level of experience of the primary surgeon. Surgical complications up to 30 days after CP were classified according to Ibanez et al. (6) Non-teaching 73 70.9% 10 9.7% 8 7.8% 10 9.7% 2 1.9% n=103 100%
p value 0.106
RI PT
Teaching 102 74.4% 6 4.4% 19 13.9% 10 7.3% 0.0% n=137 100%
AC C
EP
TE D
M AN U
SC
Ibanez degree 0 1 2 3 4 Total
ACCEPTED MANUSCRIPT Table 4. Estimation of the effect size of the relationship between the rate of surgical complications after cranioplasty (CP) and the level of experience of the primary surgeon. mRS = modified Rankin Scale.
OR 0.83 0.60 1.00 1.62
p value 0.537 0.085 0.958 0.025
OR 0.77 0.52 1.07 1.85
Multivariate 95% CI 0.42 – 1.39 0.28 – 0.95 0.88 – 1.31 1.18 – 2.90
AC C
EP
TE D
M AN U
SC
Teaching case Age ≥ 50 years Etiology mRS before CP
Univariate 95% CI 0.47 – 1.48 0.34 – 1.07 0.83 – 1.20 1.06 – 2.47
p value 0.385 0.034 0.477 0.007
RI PT
Surgical complication
ACCEPTED MANUSCRIPT Table 5. Estimation of the effect size of the relationship between the rate of postoperative infection after cranioplasty (CP) requiring CP removal and the level of experience of the primary surgeon. mRS = modified Rankin scale.
p value 0.144 0.482 0.142 0.050
OR 1.66 1.04 1.30 2.06
Multivariate 95% CI 0.69 – 4.04 0.45 – 2.39 0.98 – 1.73 1.05 – 4.06
M AN U TE D EP AC C
p value 0.261 0.926 0.062 0.037
RI PT
OR 1.91 1.33 1.20 1.88
Univariate 95% CI 0.80 – 4.56 0.59 – 2.98 0.94 – 1.54 1.00 – 3.54
SC
Infection requiring CP removal Teaching case Age ≥ 50 years Etiology mRS before CP
ACCEPTED MANUSCRIPT Table 6. Estimation of the effect size of the relationship between the rate of patients with improvement on the modified Rankin Scale (mRS) after cranioplasty (CP) and the level of experience of the primary surgeon.
p value 0.526 0.311 <0.001 <0.001
OR 1.11 0.69 1.30 2.49
Multivariate 95% CI 0.62 – 2.00 0.38 – 1.23 0.54 – 0.84 1.59 – 3.90
M AN U TE D EP AC C
p value 0.719 0.210 <0.001 <0.001
RI PT
OR 1.18 0.76 0.66 2.48
Univariate 95% CI 0.69 – 2.02 0.45 – 1.29 0.54 – 0.81 1.63 – 3.77
SC
Improvement on the mRS after CP Teaching case Age ≥ 50 years Etiology mRS before CP
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Abbreviations: BCFN = board certified faculty neurosurgeon CI = confidence interval CP = cranioplasty CT = computed tomography EBL = estimated blood loss
RI PT
FU = follow-up HUG = University Hospital Geneva ICH = intracerebral hemorrhage ICU = intensive care unit KSSG = Cantonal Hospital St.Gallen
SC
min = minutes ml = millilitres mm = millimetres
OR = odds ratio PGY = postgraduate year PEEK = polyether-ether-ketone PMMA = polymethylmethacrylat SAH = subarachnoid hemorrhage SD = standard deviation
AC C
EP
TBI = traumatic brain injury
TE D
SSI = surgical site infections
M AN U
mRS = modified Rankin Scale
ACCEPTED MANUSCRIPT Highlights Safety aspects of cranioplasty by residents or staff surgeons were assessed
-
Operation time was longer in procedures performed by supervised residents
-
The likelihood and severity of postoperative complications were similar
-
Surgical site infections requiring cranioplasty removal were similar
-
The likelihood for neurological improvement after cranioplasty was equal
AC C
EP
TE D
M AN U
SC
RI PT
-