Cardiovasc Intervent Radiol (2018) 41:1067–1073 https://doi.org/10.1007/s00270-018-1921-4 CLINICAL INVESTIGATI O N N ON- V ASCULAR I NT ERVENTIONS High-Frequency Jet Ventilation During Cryoablation of Small Renal Tumours 1 1 2 2 • • • • Thea Buchan Miles Walkden Kathryn Jenkins Pervez Sultan Steve Bandula Received: 25 October 2017 / Accepted: 24 February 2018 / Published online: 7 March 2018 The Author(s) 2018. This article is an open access publication Abstract Conclusion By reducing target tumour motion during CT- Aim To evaluate the effect of high-frequency jet ventila- guided renal cryoablation, HFJV can reduce procedure tion (HFJV) in place of standard intermittent positive- times and exposure to ionising radiation. HFJV provides an pressure ventilation (IPPV) on procedure duration, patient important adjunct to complex image-guided interventions, radiation dose, complication rates, and outcomes during with potential to improve safety and treatment outcomes. CT-guided cryoablation of small renal tumours. Materials and Methods One hundred consecutive CT- Keywords Kidney Cancer Cryoablation guided cryoablation procedures to treat small renal tumours Ventilation Computed Tomography under general anaesthesia were evaluated—50 with stan- dard IPPV and 50 after the introduction of HFJV as stan- dard practice. Anaesthesia and procedural times, ionising radiation dose, complications, and 1-month post-treatment outcomes were collected. Introduction Results HFJV was feasible and safe in all cases. Mean procedure time and total anaesthetic time were shorter with Modern management of small renal masses has moved HFJV (p = \0.0001). The number of required CT acqui- towards nephron-sparing techniques—that include partial sitions (p = 0.0002) and total procedure patient radiation nephrectomy (PN) and image-guided ablation (IGA). IGA dose (p = 0.0027) were also lower in the HFJV group allows minimally invasive tumour destruction with signif- compared with the IPPV group. There were a total of four icantly reduced patient impact , preserved renal function complications of Clavien–Dindo classiﬁcation 3 or , and oncological outcomes that are equivalent to sur- above—three in the IPPV group and one in the HFJV gical resection . group. At 1-month follow-up, two cases (both in the IPPV Minimising rates of local recurrence following IGA group) demonstrated subtotal treatment. Both cases were requires diligent technique to ensure ablative energy is subsequently successfully retreated with cryoablation. delivered precisely to the tumour with an appropriate margin. Cryoablation (CYA) has become the technique of choice, with multiple probes carefully positioned simulta- neously within the tumour to produce a conformal treat- & Steve Bandula ment zone or ‘iceball’ that can be visually conﬁrmed on firstname.lastname@example.org intra-operative CT or MRI. Imaging Department, University College London Hospital, Despite immobilisation with general anaesthesia, tradi- 235 Euston Road, London NW1 2BU, UK tional intermittent positive-pressure ventilation (IPPV) Department of Anaesthesia, University College London causes signiﬁcant abdominal organ motion making tumour Hospital, 235 Euston Road, London NW1 2BU, UK targeting more difﬁcult . Probes must be inserted Centre for Medical Imaging, University College London, 250 between respiratory cycles or during short periods of Euston Road, London NW1 2PG, UK 123 1068 T. Buchan et al.: High-Frequency Jet Ventilation During Cryoablation of... apnoea—often requiring additional image acquisitions and Anaesthetic Protocol extending procedure length. High-frequency jet ventilation (HFJV) delivers contin- All patients were subject to anaesthetic pre-assessment and uous small tidal volume ventilation at high frequency were classiﬁed per the physical status classiﬁcation system of the American Society of Anaesthesiologists (ASA). resulting in a near stationary diaphragm and minimal target organ motion [5–8]. Several studies have reported a General anaesthesia was induced by one of two anaes- thetists with 1–2 mg midazolam, followed by a total reduction in technical difﬁculty associated with probe placement, procedure time, and patient radiation dose intravenous anaesthesia target controlled intravenous infusion of propofol (maintenance range 1.8–6 mcg/ml) during image-guided interventions and single-probe ther- mal ablation of metastatic tumours in the liver and lung and remifentanil (0.2–8 ng/ml). After muscle relaxation, [9–11]. HFJV has the potential to simplify renal tumour patients were intubated orally and a reinforced endotra- cryoablation, where multiple simultaneous probe place- cheal tube (ETT) inserted. The patients were then posi- ments are required, and where treatment margins must be tioned in the CT scanner to allow optimal tumour access in adequate but not excessive to preserve renal function . either a prone or lateral position. The aim of this study was to evaluate the effect of HFJV All patients were initially ventilated with conventional volume- or pressure-controlled intermittent positive-pres- in place of standard IPPV on procedure duration, patient radiation dose, and complication rates during CT-guided sure ventilation, via a standard circle anaesthetic breathing system using a Datex Aestiva anaesthetic machine. For cryoablation of small renal tumours. those patients in the HFJV group, a jet swivel adaptor (Acutronic Medical Systems) was introduced into the cir- Methods cuit between the patient’s cuffed oral endotracheal tube (ETT) and the anaesthetic breathing circuit. Study Design HFJV was commenced at 100–120 breaths per minute, with a driving pressure (DP) of 1.0 bar. DP was adjusted to maintain the end-tidal carbon dioxide between 4.5 and This retrospective review was undertaken at a single ter- tiary interventional oncology (IO) centre with prior expe- 5.5 kPa using intermittent checks measured through IPPV via the anaesthetic machine capnography device (GE rience of over 150 renal tumour cryoablation cases. The IO team had also completed eight CT-guided ablation proce- Healthcare gas module E-CAi0-00 (Helsinki, Finland)). Once CYA treatment was complete, the patients were dures (ﬁve lung and three liver) using HFJV before this study. One hundred consecutive patients undergoing CT- ventilated conventionally until awake. guided cryoablation of a small renal tumour under general anaesthesia between December 2015 and January 2017 CT Protocol were evaluated—50 with standard IPPV and 50 after the introduction of HFJV as standard practice at this centre. All renal cryoablation procedures were performed in a The study was reviewed by the University College London dedicated interventional CT suite (Toshiba Prime, Toshiba and University College London Hospitals joint research Medical Systems, Japan, 2014) following the standard CT- and development ofﬁce and considered to be service guided renal cryoablation protocol at our institution. Patients were positioned in either the prone or lateral evaluation exempt from research ethics committee approval. position to allow optimal target tumour access—conﬁrmed using a low-dose unenhanced positioning CT scan. A dual The decision to treat with CYA was made at the local renal cancer network multidisciplinary team meeting. All (arterial and venous)-phase CT scan of the kidneys was patients gave fully informed consent, and in all cases, performed to plan probe placement and adjunctive hydro- treatment was performed with curative intent. The primary or pneumo-dissection. outcome for this study was procedure duration. Secondary The CT ﬂuoroscopic package in intermittent biopsy outcomes included: probe placement time, patient radiation mode was used for guidance of cryoprobe placement (Galil dose (represented by dose length product (DLP) generated Medical System, Minneapolis, USA). Each acquisition comprised a volume (100 kV, 30 mA with 0.5 s gantry automatically at the scanner console), number of CT acquisitions (ﬂuoroscopic acquisitions required for needle rotation) reconstructed into three congruent 4 mm slices. Final cryoprobe placement was conﬁrmed with a low-dose insertion), and complication rates (anaesthetic, medical or surgical score according to the Clavien–Dindo scale) fol- scan volume of the target organ before commencement of treatment with a double freeze (10 min) and thaw (6 min lowing CT-guided cryoablation of small renal tumours. passive and 2 min active) protocol. During each cycle, low-dose scans of the target organ were obtained at 5 and 123 T. Buchan et al.: High-Frequency Jet Ventilation During Cryoablation of... 1069 10 min to monitor the ice ball formation. A ﬁnal low-dose were carried out using PRISM (GraphPad Software, San post-procedure scan was performed following cryoprobe Diego, Calif). removal as a safe check to look for immediate complica- tions such as haemorrhage or pneumo-thorax. Hydrodis- section was administered when required using a 21-gauge Results spinal needle to infuse sterile saline containing Omnipaque at a ratio of 20 per 1000 ml of saline. Hydrodissection During the review period, 100 renal tumours were treated needles were left in situ during treatment cycles. by CT-guided cryoablation in 100 separate treatment ses- All helical scans used the following standard parame- sions involving 95 patients. Fifty treatments were per- ters: 100 kV tube potential, automatic tube current modu- formed using IPPV and 50 using HFJV. No patients were lation with a range of 40–180 mA, 0.5 s gantry rotation, excluded from the analysis, and patient records and data 0.5 detector collimation, a pitch factor of 0.813, a helical were available for all cases. None of the patients intended pitch of 65, and iterative reconstruction with 1 mm slice for HFJV required a switch in ventilation technique to thickness. IPPV. Treatment was successfully completed (technical For each procedure, data related to procedure duration, success) in all cases at the time of procedure. However, at patient-ionising radiation dose, technical difﬁculty, and 1-month follow-up, subtotal treatment was found in two complications were collected. Technical difﬁculty was patients in the IPPV group, requiring retreatment of assessed through the surrogate measures of probe place- cryoablation. None of the patients treated using HFJV ment time and total number of CT acquisitions for probe required repeat treatment. placement. Total ionising radiation dose was recorded as Patient demographics and tumour characteristics are dose length product (DLP) generated automatically at the summarised in Tables 1 and 2, with comparable charac- scanner console. teristics and ASA classiﬁcation between groups. Table 3 Anaesthetic induction time was deﬁned as the period summarises details of treatment sessions including patient from initiating induction of anaesthesia to commencement position, anaesthetic induction time, number of cryoprobes of the positioning scan, and total procedure time was placed, and the use of hydro- and pneumo-dissec- deﬁned as time from positioning scan to the ﬁnal treatment tion. Comparison between groups showed no statistically conﬁrmation scan. The number of cryoprobes, hydro- and signiﬁcant difference in these variables. Mean total pro- pneumo-dissection needles used, number of biopsy-mode cedure time was 144 min (SD 37.32) for IPPV versus CT volume acquisitions, and total time required to place 106 min (SD 29.69) for HFJV (p = \0.0001) (Fig. 1), and the probes and needles were also recorded. mean needle placement time was 100 min (SD 38.24) for IPPV versus 66 min (SD 27.83) for HFJV Patient Follow-Up (p = \0.0001) (Fig. 2). The number of CT acquisitions required for needle placement was 78.8 (SD 29.02) for All patients were admitted overnight for observation and if IPPV versus 59.2 (SD 21.98) for HFJV standard hospital discharge criteria met, discharged the (p = 0.0002) (Table 4), and total procedure patient radia- following morning. Patients were reviewed in an outpatient tion dose 1975 mGycm2 (SD 1055) for IPPV versus 1449 clinic at 1 month post-treatment with renal contrast-en- mGycm2 (SD 597.4) for HFJV (p = 0.0027) (Fig. 3). hanced CT. Treatment was considered successful if at 1 month post-procedure, the ablation zone encompassed the target tumour with a margin greater than 5 mm in all Table 1 Patient demographic details—mean (range) planes, and with no tumour enhancement . All com- IPPV group HFJV group plications within 30 days of treatment were recorded and scored using the Clavien–Dindo classiﬁcation system. Number of procedures 50 50 Sex Statistical Analysis Male 35 39 Female 15 11 Data between IPPV and HFJV groups were compared Age (years) 64 (28–86) 69 (41–85) using unpaired t tests. A p value of 0.05 or lower was Weight (Kg) 91 (73–138) 85 (65–128) considered statistically signiﬁcant. IPPV and HFJV groups Height (metres) 1.70 (1.56–1.91) 169 (1.5–1.87) were assessed for normal distribution using Kolmogorov– Body mass index (kg/m2) 31.4 (22.2–43) 29.5 (21.1–42.2) Smirnov tests. Nonparametric data between groups were ASA 2.54 2.68 compared using Mann–Whitney U test. All statistical tests 123 1070 T. Buchan et al.: High-Frequency Jet Ventilation During Cryoablation of... Table 2 Baseline tumour characteristics delivery of treatment , with improved tumour targeting a crucial factor in achieving these aims. IPPV HFJV p value By reducing target organ motion, HFJV has potential to PADUA score 7.2 7.41 reduce technical difﬁculty associated with treatment probe Size (mm) 25.63 28.74 0.1322 placement in solid organs such as the liver and kidney, Histology thereby improving tumour targeting and accuracy. Clear cell 31 35 Chromophobe 3 5 Papillary 12 8 P < 0.0001 Spindle cell 1 0 Oncocytoma 1 0 150 Not biopsied 2 2 There were no complications or delays to discharge attributable to ventilation technique. There were no deaths within the 30-day follow-up period. There were 17 com- plications, four of which with Clavien–Dindo (CD) score of 3 or above. These were one case of transient brachial plexus injury (CD3) which resolved without intervention; one urine leak (CD3b), managed conservatively with insertion of a ureteric stent; one case of haematuria Fig. 1 Box-and-whisker plot showing total procedure time (mean (CD3a), requiring bladder irrigation via a three-way and standard deviation) in minutes for the two ventilation groups catheter; and one patient developed a right-sided pul- monary embolus (CD4a) requiring an ITU admission and delaying discharge by 15 days. P < 0.0001 At 1-month follow-up imaging, two cases (both in the IPPV group) demonstrated subtotal treatment. Both cases were subsequently successfully retreated with cryoablation. Discussion Image-guided ablation has established itself within the toolbox of modern oncology. For small renal tumours, advances in technique, image guidance, and ablation technology allow image-guided ablation (IGA) with equivalent oncological outcomes to surgery , and IGA has moved away from being used only in high-risk surgical groups to an effective primary therapy for localised cancer Fig. 2 Box-and-whisker plot showing needle placement time (mean [13, 14]. Technique development is now focused on and standard deviation) in minutes for the two ventilation groups achieving more robust, consistent, and cost-effective Table 3 Comparison of IPPV HFJV p value procedural details Patient position Prone 38 31 – Left lateral 5 9 – Right lateral 7 10 – Induction time (total anaesthesia time minus procedure time) 30 min 4 s 31 min 0.2788 Number of probes 3.922 4.24 0.3427 Hydro or pneumo-dissection 25 23 – IPPV HFJV IPPV HFJV Needle placement time (mins) Total procedure time (mins) T. Buchan et al.: High-Frequency Jet Ventilation During Cryoablation of... 1071 Table 4 Comparison of IPPV HFJV p value time/dose/difﬁculty Total procedure time (minutes) 142.4 105.8 \ 0.0001 Needle placement time (time to achieve probe positions—minutes) 100.2 66.2 \ 0.0001 Number of acquisitions for probe placement 78.8 59.2 0.0002 Patient dose collected as dose length product (mGycm2) 1975 1449 0.0027 4000 acquisitions required to place the probes (24%), HFJV has P = 0.0027 a direct effect on the ionising radiation dose to the patient. We report a 27% reduction in patient radiation dose (DLP—mGycm ) for procedures performed under HFJV compared with IPPV. The patient dose reduction achieved is signiﬁcant (an increasing number of whom are \ 50 years), and although not measured here, would likely lead to a reduced scatter dose to the operator. This dose reduction is, however, less than the 54% reduction reported by Chung et al., achieved with HFJV during thermal ablation of lung tumours, which may reﬂect the higher number of set helical acquisitions (e.g., position checking and treatment monitoring helical scans) within the renal tumour treatment protocol. Fig. 3 Box-and-whisker plot showing total procedure dose (mean We used needle placement time and number of CT and standard deviation) as dose length product for the two ventilation acquisitions to evaluate the ability of HFJV to reduce the groups technical difﬁculty of CT-guided renal cryoablation by reducing ventilatory motion. Needle placement time fell Technical difﬁculty is problematic to quantify, but logical from 100 to 66 min between groups (p = \0.0001), and surrogate measures related to procedure duration and number of CT acquisitions from 78.8 to 59.2 between required level of image guidance can be evaluated. groups (p = \0.0002)—giving a clear indication of the In this retrospective review, we demonstrate that HFJV effect of HFJV. Factors other than motion may affect can be safely implemented and produces a signiﬁcant needle placement time, including depth of tumour from the reduction in procedure time, number of required CT skin, tumour endophycity, and patient position. Patient acquisitions, and ionising radiation dose for patients BMI, position, and tumour PADUA score were well mat- undergoing cryoablation of small renal tumours, with no ched between groups (Table 2), minimising this potential effect on complication rate or length of inpatient stay. bias. Although we included placement of dissection nee- CT-guided cryoablation procedures frequently place dles in the evaluation, we did not correct for variation in signiﬁcant pressure on CT capacity due in part to the length the time and CT acquisitions required to achieve satisfac- of procedure. The 26% reduction in total procedure time tory hydro- or pneumo-dissection. demonstrated here will have cost and resource beneﬁts to Importantly, we demonstrate that in a representative patient interventional oncology departments, justifying the group, HFJV can be applied ubiquitously without the need for investment in additional equipment and training that may conversion to IPPV and that establishing anaesthesia with be required to perform HFJV. A signiﬁcant reduction in HFJV was safe and took no longer. The requirement for the procedure and anaesthetic time will reduce anaesthetic risk anaesthetic team to undergo additional training to be familiar and allow more rapid recovery [16, 17]. Reduced anaes- and conﬁdent with the use of HFJV is recognised , and thetic risk will also widen access to patients who may have special arrangements were made during the evaluation period comorbidities and are unable to undergo other treatments. to ensure speciﬁc anaesthetist staff were available. However, Complex interventions require multiple CT acquisitions. despite the beneﬁts described, limitations in training and staff Modern scanners implement a range of dose sparing availability may limit use with IGA. techniques such as dose modulation and iterative recon- The use of HFJV in the prone position may be unfa- struction; however, dose to the patient and operator during miliar to many anaesthetic teams. Prone ventilation is used these procedures is signiﬁcant. With the current trend of widely in the intensive care environment to optimise PaO2: increasing numbers of IGA treatments being performed, FiO2 ratio, with no increase in risk of barotrauma, CO methods that reduce dose exposure to patient and operator clearance, or ability to deliver tidal volumes . These are particularly important. By reducing the number of IPPV HFJV DLP (mGycm2) 1072 T. Buchan et al.: High-Frequency Jet Ventilation During Cryoablation of... Compliance with Ethical Standards risks are not thought to be altered by the use of HFJV in the prone position, but this has not been speciﬁcally investi- Conﬂict of interest The authors declare that they have no conﬂict of gated. Concerns associated with a prone position are dis- interest. lodging tubes and lines; abdominal discomfort and compartment syndrome in obese patients; and nerve dam- Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http:// age associated with arm positioning and pressure point creativecommons.org/licenses/by/4.0/), which permits unrestricted damage rather than issues associated with the mode of use, distribution, and reproduction in any medium, provided you give ventilation. appropriate credit to the original author(s) and the source, provide a Patient age was also not a limitation to HFJV. Out of 95 link to the Creative Commons license, and indicate if changes were made. patients in this study, 72 were aged over 60 years. We experienced no issues with hypercapnoea, acidosis, or tube placement that required intervention or deviation from our References predetermined protocol. Severe COPD has been considered to be a relative contraindication to HFJV, due to the 1. Katsanos K, Mailli L, Krokidis M, McGrath A, Sabharwal T, potential risk of gas trapping and barotrauma in this pop- Adam A. Systematic review and meta-analysis of thermal abla- ulation; however, in the range of lung function normally tion versus surgical nephrectomy for small renal tumours. Car- diovasc Interv Radiol. 2014;37(2):427–37. acceptable for ablation under general anaesthesia at our 2. Yang Q, Meng F, Li K, Wang T, Nie Q, Che Z, et al. Safety and institution (i.e., an FEV1 [ 1L), this has not thus far been efﬁcacy of thermal ablation for small renal masses in solitary a signiﬁcant concern. Further studies are however required kidney: evidence from meta-analysis of comparative studies. to speciﬁcally explore the acceptability of this technique in PLoS ONE. 2015;10(6):1–12. 3. Thompson RH, Atwell T, Schmit G, Lohse CM, Kurup AN, patients with severe cardiorespiratory disease to determine Weisbrod A, et al. Comparison of partial nephrectomy and per- whether any patient cohorts should be precluded from cutaneous ablation for cT1 renal masses. Eur Urol. Eur Assoc HFJV in the prone and lateral positions. Urol. 2015;67(2):252–9. There were several further limitations to this study. The 4. Wu QJ, Thongphiew D, Wang Z, Chankong V, Yin F-F. The impact of respiratory motion and treatment technique on stereo- review was conducted retrospectively and without ran- tactic body radiation therapy for liver cancer. Med Phys. domisation and therefore potentially subject to several 2008;35(4):1440–51. sources of bias. However, the compared groups did consist 5. Raiten J, Elkassabany N, Mandel JE. The use of high-frequency of consecutively treated patients, with good matching of jet ventilation for out of operating room anesthesia. Curr Opin Anaesthesiol. 2012;25(4):482–5. key patient and tumour characteristics and no missing 6. Olive DJ, Cormack JR. Kidney stone movement during litho- data—thus allowing meaningful comparison between the tripsy under general anaesthesia: high frequency jet ventilation HFJV and the IPPV control group. Although sufﬁcient to versus spontaneous ventilation. Anaesth Intensive Care. demonstrate change in the endpoints assessed here, the case 2006;34(6):832–3. 7. Brandner ED, Wu A, Chen H, Heron D, Kalnicki S, Komanduri numbers were relatively small, and not sufﬁcient to probe K, et al. Abdominal organ motion measured using 4D CT. Int J differences in technical success or complication rate. Radiat Oncol Biol Phys. 2006;65(2):554–60. The post-treatment follow-up period was also short 8. Fritz P, Kraus H-J, Mu ¨ hlnickel W, Sassmann V, Hering W, (1 month) and insufﬁcient to demonstrate any effect on Strauch K. High-frequency jet ventilation for complete target immobilization and reduction of planning target volume in treatment efﬁcacy. We hypothesise that reduced tumour stereotactic high single-dose irradiation of stage I non-small cell motion and more accurate targeting would lead to higher lung cancer and lung metastases. Int J Radiat Oncol Biol Phys. treatment efﬁcacy and improved oncological outcomes, but 2010;78(1):136–42. this requires future investigation. 9. Chung DYF, Tse DML, Boardman P, Gleeson FV, Little MW, Scott SH, et al. High-frequency jet ventilation under general anesthesia facilitates CT-guided lung tumor thermal ablation compared with normal respiration under conscious analgesic Conclusion sedation. J Vasc Interv Radiol. 2014;25(9):1463–9. 10. Perkins PE. High-frequency jet ventilation during radiofrequency ablation: a case report. AANA J. 2008;76(3):209–12. By reducing target tumour motion during CT-guided renal 11. Denys A, Lachenal Y, Duran R, Chollet-Rivier M, Bize P. Use of cryoablation, HFJV can reduce the technical difﬁculty high-frequency jet ventilation for percutaneous tumor ablation. associated with probe insertion—leading to reduced pro- Cardiovasc Interv Radiol. 2014;37(1):140–6. cedure times and exposure to ionising radiation. HFJV 12. Breen DJ, Bryant TJ, Abbas A, Shepherd B, McGill N, Anderson JA, et al. Percutaneous cryoablation of renal tumours: outcomes provides an important adjunct to complex image-guided from 171 tumours in 147 patients. BJU Int. 2013;112(6):758–65. interventions, with potential to improve safety, treatment 13. Georgiades CS, Rodriguez R. Efﬁcacy and safety of percutaneous outcomes and extend the scope of IGA therapies; however, cryoablation for stage 1A/B renal cell carcinoma: results of a further studies are required to evaluate these effects. prospective, single-arm, 5-year study. Cardiovasc Interv Radiol. 2014;37(6):1494–9. 123 T. Buchan et al.: High-Frequency Jet Ventilation During Cryoablation of... 1073 14. Hayes MC, Breen DJ. Excision versus ablation in renal cancer: 17. Fernandez-Bustamante A, Frendl G, Sprung J, Kor DJ, Subra- optimising outcome and minimising risk. Eur Assoc Urol. maniam B, Martinez Ruiz R, et al. Postoperative pulmonary 2016;69:2015–6. complications, early mortality, and hospital stay following non- 15. Chehab M, Friedlander JA, Handel J, Vartanian S, Krishnan A, cardiothoracic surgery: a multicenter study by the perioperative Wong CO, et al. Percutaneous cryoablation versus partial research network investigators. JAMA Surg. nephrectomy: cost comparison of T1a tumors. J Endourol. 2017;152(2):157–66. 2016;30(2):170–6. 18. Messerole E, Peine P, Wittkopp S, Marini JJ, Albert RK. The 16. Nazemi AK, Gowd AK, Carmouche JJ, Kates SL, Albert TJ, pragmatics of prone positioning. Am J Respir Crit Care Med. Behrend CJ. Prevention and management of postoperative delir- 2002;165(10):1359–63. https://doi.org/10.1164/rccm.2107005. ium in elderly patients following elective spinal surgery. Clin Spine Surg. 2017;30(3):112–9.
CardioVascular and Interventional Radiology
– Springer Journals
Published: Mar 7, 2018