Downloaded from http://bjo.bmj.com/ on November 2, 2015 - Published by group.bmj.com
BJO Online First, published on July 3, 2015 as 10.1136/bjophthalmol-2014-306467 Clinical science
The Royal College of Ophthalmologists’ National Ophthalmology Database Study of vitreoretinal surgery: report 5, anaesthetic techniques Ahmed A B Sallam,1,2 Paul H J Donachie,1,3 Tom H Williamson,4 John M Sparrow,5 Robert L Johnston1,3 1
Gloucestershire Hospitals NHS Foundation Trust, Gloucestershire, UK 2 Ain Shams University, Cairo, Egypt 3 The Royal College of Ophthalmologists’ National Ophthalmology Database, London, UK 4 Guy’s and St. Thomas’ NHS Foundation Trust, London, UK 5 Bristol Eye Hospital, University of Bristol, Bristol, UK Correspondence to Dr Ahmed A B Sallam, Gloucestershire Hospitals NHS Foundation Trust, Cheltenham, Gloucestershire GL53 7AN, UK; [email protected]/ [email protected] Received 17 December 2014 Revised 11 March 2015 Accepted 11 June 2015
ABSTRACT Aims To explore trends over time and variation in the use of anaesthetic techniques for vitreoretinal (VR) surgery in the UK. Methods Prospectively collected data from 13 centres contributing >50 VR operations, including either pars plana vitrectomy (PPV) or scleral buckle (SB), between May 2000 and November 2010 were retrospectively analysed. Anaesthesia was categorised as general anaesthesia (GA) or local anaesthesia (LA) and results were reported by year, centre, grade of surgeon and type of operation. Results 160 surgeons performed 12 124 operations on 10 405 eyes (9935 patients); 6054 (49.9%) under GA and 6070 (50.1%) under LA. The percentage performed under GA decreased from 95.3% in 2001 to 40.9% in 2010. Within LA techniques, peribulbar or retrobulbar injection was used in 2783 (45.8%) operations and sub-Tenon’s cannula in 3287 (54.2%). The proportions of operations performed under GA or LA were similar for consultants and trainees. Primary SB, primary combined PPV and SB for retinal detachment (RD), repeat RD surgery and complex vitrectomy surgery were more commonly performed under GA (85.8%, 67.0%, 63.5% and 69.4%, respectively), while primary PPV for RD, simple vitrectomy surgery and macular surgery were more commonly performed under LA (58.1%, 53.7% and 58.2%, respectively). Marked intercentre variation existed with the extremes being one centre with 100% of operations performed under GA and one centre with 98.3% under LA. Conclusions LA for VR surgery has steadily increased over the last decade in the UK but marked intercentre variation exists.
INTRODUCTION
To cite: Sallam AAB, Donachie PHJ, Williamson TH, et al. Br J Ophthalmol Published Online First: [ please include Day Month Year] doi:10.1136/bjophthalmol2014-306467
Vitreoretinal (VR) surgery has rapidly expanded in the last four decades in the UK National Health Service (NHS) with the number of operations increasing more than sixfold from 1968 to 2004.1 The choice between general anaesthesia (GA) and local anaesthesia (LA) for VR surgery depends on a number of factors, including surgeon’s preference, patient’s age and preference as well as the anticipated difficulty and duration of the operation.2 3 Marked variation in anaesthetic techniques used for VR surgery exists between countries and may be related to the number of cases per surgical list, fee schedule for physician and institution reimbursement and whether surgery is delivered as an inpatient or in an ambulatory day case centre. In the USA, 85–90% of VR surgery is carried out under
LA,2 4 while in contrast, the majority of VR surgery in the UK has historically been performed under GA.3 5 The results of a national questionnaire of British and Eire Association of Vitreoretinal Surgeons (BEAVRS) members in 2001 showed that 77% of surgeons preferred GA for most VR procedures, including macular surgery, diabetic vitrectomy and scleral buckles (SBs).3 However, more recently published data from single centres highlight an increasing usage of LA in the UK.6–8 The UK Royal College of Ophthalmologists (RCOphth) initiated the National Ophthalmology Database (NOD) project to collate anonymised ophthalmic data that are collected during routine clinical care, using electronic medical records (EMRs). By providing representative and generalisable data, the RCOphth NOD aims to determine national practice patterns, establish benchmark standards and enable ophthalmologists to compare their surgical processes and outcomes with those of their anonymised peers as a quality improvement tool. Additionally, it will be possible to compare ophthalmology practice in the UK with healthcare systems in other countries. The aims of this report are to analyse the trends in anaesthesia used for VR surgery between 2000 and 2010 in the UK and to describe variation between centres, grade of surgeon and type of VR surgery.
MATERIALS AND METHODS The methods used to establish the RCOphth NOD have been described in detail in previous papers.9–11 Prospectively collected data were automatically extracted from 31 centres that use the same EMR system (Medisoft Ophthalmology, Medisoft Limited, Leeds, UK) up to 30 November 2010. In October 2013, data were also extracted from an in-house non-commercial database (VITREOR database, Guy’s and St. Thomas’ NHS Foundation Trust, London, UK) used at three centres and added to the RCOphth NOD. In this analysis, the data from the three centres that use the non-commercial database are combined because it is the same surgical team operating at these three sites. The lead clinician and Caldicott Guardian (who oversees data protection) at each centre gave written approval for the data extraction. This study was conducted in accordance with the Declaration of Helsinki and the UK’s Data Protection Act. In this report, data analysis was restricted to centres that had contributed >50 VR operations performed before 1 December 2010 and operations
Sallam AAB, et al. Br J Ophthalmol 2015;0:1–7. doi:10.1136/bjophthalmol-2014-306467
Copyright Article author (or their employer) 2015. Produced by BMJ Publishing Group Ltd under licence.
1
Downloaded from http://bjo.bmj.com/ on November 2, 2015 - Published by group.bmj.com
Clinical science that included a pars plana vitrectomy (PPV) and/or SB. Other procedures such as retinal laser or cryopexy, pneumatic retinopexy, intravitreal injections of gas as well as removal of silicone oil or removal of SB were excluded unless they were part of PPV and/or SB surgery.
Statistical analysis For the purpose of this study, all operations were allocated to one of four complexity groups: 1. Retinal detachment (RD) surgery, which was further subclassified into primary PPV, primary SB, primary combined PPV and SB and repeat RD surgery (repeat PPV and/or SB). 2. Macular surgery, which included: macular hole, epiretinal membrane peel and surgery for vitreomacular traction. 3. Simple vitrectomy defined as PPV for vitreous haemorrhage for any indication except when combined with diabetic delamination/segmentation, surgery for ectopia lentis, dropped nucleus or vitreous opacity/floaters. 4. Complex vitrectomy when PPV was undertaken for diabetic retinopathy with advanced traction, posterior segment trauma, intraocular foreign body, endophthalmitis and when PPV was combined with glaucoma surgery. The grade of surgeon was categorised as consultant, independent non-consultant (staff grade, associate specialists and Trust doctors) or trainee (fellows, specialist registrars and specialty trainees). Anaesthesia was categorised as GA (even if it included supplementary LA) or LA. Information on anaesthetic techniques, anaesthetic and sedative drugs and local anaesthetic complications was recorded. One source of data used in this study, the non-commercial database (VITREOR), did not have an anaesthetic complication field and it was therefore not possible to record complications of LA for cases that were operated in centres using this database. Pearson’s χ2 test and Fisher’s exact test were used to test for differences between the anaesthetic techniques for the number of anaesthetic complications and for each individual anaesthetic complication. Potential differences in patient’s age at the time of a first VR surgery between GA and LA were investigated by a Student’s t test using the Welch approximation. Yearly results follow the NHS year, which runs from 1 April to 31 March and all analyses were conducted using STATA V.11.
66 trainee surgeons and 8 surgeons had data recorded at more than one grade. The number of operations performed by each surgeon ranged from 1 to 1553, where 54 surgeons had data for 1 operation, 46 surgeons had data for 2–10 operations, 26 surgeons had data for 11–99 operations and 34 surgeons had data for ≥100 operations. The most common surgical category was RD comprising 5755 operations (3677 for primary PPV, 639 for primary SB, 356 for primary PPV+SB and 1083 were for repeat RD surgery). Of the remaining eyes, 2951 were for macular surgery, 2476 for simple vitrectomy and 942 for complex vitrectomy.
Anaesthetic technique GA was used for 6054 (49.9%) operations and LA for 6070 (50.1%). Of those operations performed under GA, 761 operations also had additional LA (31 (4%) by peribulbar or retrobulbar injection and 730 (96%) by sub-Tenon’s). In all patients (n=9935), GA tended to be more used for younger patients than LA (the mean age at the time of first VR operation; 57.3 years for GA vs 67.7 years for LA; p≤0.001). The tendency for GA use in younger patients was also evident when restricting the analysis to patients >25 and 70% of these operations were performed under GA. However, wide intercentre variations were present for other types of surgery and in four centres, >70% of operations for each type of VR surgery were performed under GA, table 1.
LA complications Of the 6070 operations performed under LA, 5225 (86.1%) were performed in centres that recorded anaesthetic complications data and one or more complications were recorded in 184 (3.5%) of these operations. The complication rate was lower for sub-Tenon’s anaesthesia compared with sharp needle injections (2.9% vs 4.2%, respectively). A slightly higher percentage of patients having sub-Tenon’s anaesthesia had one of the following structured complications of anaesthesia: mild, moderate or severe ‘discomfort/pain’, recorded in the EMR, compared with sharp needle administered anaesthesia ( p=0.048). Statistically significantly higher rates of conjunctival chemosis and eyelid haemorrhage/bruising were found with sharp needle injections compared with sub-Tenon’s anaesthesia ( p≤0.001 and 0.003, respectively). Potentially sightthreatening or life-threatening complications of LA were very uncommon with both techniques: serious systemic complications were documented in 0.1% of operations (six cases), with no retrobulbar haemorrhages or ocular perforations being recorded, table 2.
DISCUSSION This study of 12 124 VR operations performed in 13 geographically dispersed centres in the UK surveyed trends in anaesthesia for VR surgery over a 10-year period. Our data demonstrate that from 2000 to 2010, the use of GA for VR surgery has 4
decreased, while the use of LA increased substantially from 5% in 2001 to almost 60% in 2010. In eyes operated under LA, the use of sub-Tenon’s cannula techniques has exceeded sharp needle techniques in recent years and complications of both techniques were uncommon. While there were marked differences between retinal surgeons and centres in our study, GA is still preferred for SBs, complex vitrectomy operations and younger patients. LA using sub-Tenon’s cannula12–16 and sharp needle peribulbar and retrobulbar injections6–8 15 have both been documented as effective for VR surgery. There is evidence that LA is associated with comparable rates of per-operative surgical complications to GA and does not influence the anatomical or functional outcomes that are achieved.17 LA has several advantages over GA: it is associated with less cardiopulmonary risks to patients, it represents a viable option for emergency operations in patients who are not fasting and theatre turnover time is significantly shorter making it more cost effective.17 In the UK, GA has traditionally been the preferred method for delivering anaesthesia for VR surgery.3 5 This multicentre study adds weight to the findings of single-centre studies6–8 17 that have demonstrated an increased uptake of LA for VR surgery. It is likely that the increased use of LA is related to the trend for VR centres to perform surgery as day case without an overnight stay, to reduce healthcare costs and increase efficiency. Additionally, we speculate that the rising popularity of LA in the UK is linked to the increased use of smaller gauge transconjuctival vitrectomy instruments, which have been shown to be associated with less ocular discomfort for patients18 as well as shorter duration of surgery compared with 20-gauge instruments.19 Intravenous sedation in addition to LA was used in a very small number of operations (2%) in this study. This is in accord with other studies from the UK where the use of adjunctive Sallam AAB, et al. Br J Ophthalmol 2015;0:1–7. doi:10.1136/bjophthalmol-2014-306467
Downloaded from http://bjo.bmj.com/ on November 2, 2015 - Published by group.bmj.com
Clinical science Table 1 The percentage of operations performed under general anaesthesia (GA) in each centre by grade of surgeon and complexity of operation Centres Operative details
A
B
C
D
E
F
G
H
I
J
K
L
M
Total
Number of operations Number of surgeons* Grade of surgeon Consultant surgeons; n surgeons Number of operations Percentage of operations under GA Independent non-consultant surgeons; n surgeons Number of operations Percentage of operations under GA Trainee surgeons; n surgeons Number of operations Percentage of operations under GA Type of surgery Primary PPV for retinal detachment (RD); n Percentage of operations under GA Primary SB RD; n Percentage of operations under GA Primary PPV+SB for RD; n Percentage of operations under GA Repeat RD; n Percentage of operations under GA Macular surgery; n Percentage of operations under GA Simple PPV; n Percentage of operations under GA Complex PPV; n Percentage of operations under GA
3615 23
3265 38
1694 32
1409 24
653 1
424 8
238 6
224 6
167 4
152 2
147 8
74 3
62 5
12 124 160
1 1542 85.5 0
11 1361 36.5 11
11 1587 41.9 20
11 850 23.6 3
1 653 78.9 0
3 279 96.1 4
2 182 2.2 1
3 58 22.4 3
2 156 55.1 2
1 151 30.5 0
3 88 28.4 2
2 72 100.0 0
3 46 69.6 2
54 7025 53.3 48
0 N/A 23 2073 70.0
748 25.0 19 1156 28.5
102 29.4 2 5 0.0
4 0.0 11 555 16.2
0 N/A 0 0 N/A
16 87.5 2 129 98.4
2 0.0 3 54 0.0
120 28.3 1 46 34.8
11 36.4 0 0 N/A
0 N/A 1 1 0.0
5 0.0 3 54 27.8
0 N/A 1 2 100.0
16 75.0 0 0 N/A
1024 27.4 66 4075 49.8
1042 71.3 247 96.4 76 78.9 381 79.8 689 72.9 761 71.4 419 90.7
1147 19.3 220 73.6 136 58.1 360 49.4 682 20.1 536 28.9 184 44.0
488 30.3 82 91.5 56 57.1 123 65.0 414 35.7 411 33.6 120 61.7
439 17.8 43 88.4 28 53.6 85 34.1 456 16.2 279 11.5 79 31.6
162 90.1 1 100.0 19 94.7 48 85.4 247 71.3 150 72.0 26 96.2
111 97.3 4 100.0 5 100.0 38 100.0 117 97.4 126 93.7 23 95.7
62 0.0 11 9.1 1 0.0 21 4.8 91 0.0 38 5.3 14 0.0
51 37.3 12 91.7 3 33.3 11 45.5 59 16.9 73 16.4 15 33.3
46 45.7 5 80.0 6 100.0 4 100.0 56 50 35 40.0 15 86.7
38 21.1 5 100.0 3 100.0 6 83.3 56 14.3 21 23.8 23 52.2
54 24.1 4 100.0 0 N/A 3 0.0 56 21.4 15 13.3 15 60.0
17 100.0 4 100.0 21 100.0 3 100.0 14 100 10 100.0 5 100.0
20 95.0 1 100.0 2 100.0 0 N/A 14 78.6 21 38.1 4 75.0
3677 41.9 639 85.8 356 68.0 1083 63.5 2951 41.8 2476 46.3 942 69.4
The centres are ranked by the total number of operations they contributed to the analysis. *Eight surgeons had data recorded at multiple grades, reflecting progress through training. PPV, pars plana vitrectomy; SB, scleral buckle.
sedation was uncommon and ranged from 0% to 20.2%.6 7 16 17 20 In contrast, intravenous sedation in conjunction with LA is standard practice in the USA.2 14 15 GA is still more often used for complex surgery or when a SB is included. These procedures are expected to be either prolonged and may involve traction on extraocular muscles as well as cryopexy with increased level of patients’ discomfort compared with more straightforward vitrectomy surgery.5 17 Several studies have shown that LA can still be used for SBs and/or prolonged VR surgery; however, there was often a need for additional LA during surgery and an increased requirement for intravenous sedation.6 7 12 20 Young patients have been previously identified as an important factor for considering GA.2 8 However, LA is still an option in young adults (
BJO Online First, published on July 3, 2015 as 10.1136/bjophthalmol-2014-306467 Clinical science
The Royal College of Ophthalmologists’ National Ophthalmology Database Study of vitreoretinal surgery: report 5, anaesthetic techniques Ahmed A B Sallam,1,2 Paul H J Donachie,1,3 Tom H Williamson,4 John M Sparrow,5 Robert L Johnston1,3 1
Gloucestershire Hospitals NHS Foundation Trust, Gloucestershire, UK 2 Ain Shams University, Cairo, Egypt 3 The Royal College of Ophthalmologists’ National Ophthalmology Database, London, UK 4 Guy’s and St. Thomas’ NHS Foundation Trust, London, UK 5 Bristol Eye Hospital, University of Bristol, Bristol, UK Correspondence to Dr Ahmed A B Sallam, Gloucestershire Hospitals NHS Foundation Trust, Cheltenham, Gloucestershire GL53 7AN, UK; [email protected]/ [email protected] Received 17 December 2014 Revised 11 March 2015 Accepted 11 June 2015
ABSTRACT Aims To explore trends over time and variation in the use of anaesthetic techniques for vitreoretinal (VR) surgery in the UK. Methods Prospectively collected data from 13 centres contributing >50 VR operations, including either pars plana vitrectomy (PPV) or scleral buckle (SB), between May 2000 and November 2010 were retrospectively analysed. Anaesthesia was categorised as general anaesthesia (GA) or local anaesthesia (LA) and results were reported by year, centre, grade of surgeon and type of operation. Results 160 surgeons performed 12 124 operations on 10 405 eyes (9935 patients); 6054 (49.9%) under GA and 6070 (50.1%) under LA. The percentage performed under GA decreased from 95.3% in 2001 to 40.9% in 2010. Within LA techniques, peribulbar or retrobulbar injection was used in 2783 (45.8%) operations and sub-Tenon’s cannula in 3287 (54.2%). The proportions of operations performed under GA or LA were similar for consultants and trainees. Primary SB, primary combined PPV and SB for retinal detachment (RD), repeat RD surgery and complex vitrectomy surgery were more commonly performed under GA (85.8%, 67.0%, 63.5% and 69.4%, respectively), while primary PPV for RD, simple vitrectomy surgery and macular surgery were more commonly performed under LA (58.1%, 53.7% and 58.2%, respectively). Marked intercentre variation existed with the extremes being one centre with 100% of operations performed under GA and one centre with 98.3% under LA. Conclusions LA for VR surgery has steadily increased over the last decade in the UK but marked intercentre variation exists.
INTRODUCTION
To cite: Sallam AAB, Donachie PHJ, Williamson TH, et al. Br J Ophthalmol Published Online First: [ please include Day Month Year] doi:10.1136/bjophthalmol2014-306467
Vitreoretinal (VR) surgery has rapidly expanded in the last four decades in the UK National Health Service (NHS) with the number of operations increasing more than sixfold from 1968 to 2004.1 The choice between general anaesthesia (GA) and local anaesthesia (LA) for VR surgery depends on a number of factors, including surgeon’s preference, patient’s age and preference as well as the anticipated difficulty and duration of the operation.2 3 Marked variation in anaesthetic techniques used for VR surgery exists between countries and may be related to the number of cases per surgical list, fee schedule for physician and institution reimbursement and whether surgery is delivered as an inpatient or in an ambulatory day case centre. In the USA, 85–90% of VR surgery is carried out under
LA,2 4 while in contrast, the majority of VR surgery in the UK has historically been performed under GA.3 5 The results of a national questionnaire of British and Eire Association of Vitreoretinal Surgeons (BEAVRS) members in 2001 showed that 77% of surgeons preferred GA for most VR procedures, including macular surgery, diabetic vitrectomy and scleral buckles (SBs).3 However, more recently published data from single centres highlight an increasing usage of LA in the UK.6–8 The UK Royal College of Ophthalmologists (RCOphth) initiated the National Ophthalmology Database (NOD) project to collate anonymised ophthalmic data that are collected during routine clinical care, using electronic medical records (EMRs). By providing representative and generalisable data, the RCOphth NOD aims to determine national practice patterns, establish benchmark standards and enable ophthalmologists to compare their surgical processes and outcomes with those of their anonymised peers as a quality improvement tool. Additionally, it will be possible to compare ophthalmology practice in the UK with healthcare systems in other countries. The aims of this report are to analyse the trends in anaesthesia used for VR surgery between 2000 and 2010 in the UK and to describe variation between centres, grade of surgeon and type of VR surgery.
MATERIALS AND METHODS The methods used to establish the RCOphth NOD have been described in detail in previous papers.9–11 Prospectively collected data were automatically extracted from 31 centres that use the same EMR system (Medisoft Ophthalmology, Medisoft Limited, Leeds, UK) up to 30 November 2010. In October 2013, data were also extracted from an in-house non-commercial database (VITREOR database, Guy’s and St. Thomas’ NHS Foundation Trust, London, UK) used at three centres and added to the RCOphth NOD. In this analysis, the data from the three centres that use the non-commercial database are combined because it is the same surgical team operating at these three sites. The lead clinician and Caldicott Guardian (who oversees data protection) at each centre gave written approval for the data extraction. This study was conducted in accordance with the Declaration of Helsinki and the UK’s Data Protection Act. In this report, data analysis was restricted to centres that had contributed >50 VR operations performed before 1 December 2010 and operations
Sallam AAB, et al. Br J Ophthalmol 2015;0:1–7. doi:10.1136/bjophthalmol-2014-306467
Copyright Article author (or their employer) 2015. Produced by BMJ Publishing Group Ltd under licence.
1
Downloaded from http://bjo.bmj.com/ on November 2, 2015 - Published by group.bmj.com
Clinical science that included a pars plana vitrectomy (PPV) and/or SB. Other procedures such as retinal laser or cryopexy, pneumatic retinopexy, intravitreal injections of gas as well as removal of silicone oil or removal of SB were excluded unless they were part of PPV and/or SB surgery.
Statistical analysis For the purpose of this study, all operations were allocated to one of four complexity groups: 1. Retinal detachment (RD) surgery, which was further subclassified into primary PPV, primary SB, primary combined PPV and SB and repeat RD surgery (repeat PPV and/or SB). 2. Macular surgery, which included: macular hole, epiretinal membrane peel and surgery for vitreomacular traction. 3. Simple vitrectomy defined as PPV for vitreous haemorrhage for any indication except when combined with diabetic delamination/segmentation, surgery for ectopia lentis, dropped nucleus or vitreous opacity/floaters. 4. Complex vitrectomy when PPV was undertaken for diabetic retinopathy with advanced traction, posterior segment trauma, intraocular foreign body, endophthalmitis and when PPV was combined with glaucoma surgery. The grade of surgeon was categorised as consultant, independent non-consultant (staff grade, associate specialists and Trust doctors) or trainee (fellows, specialist registrars and specialty trainees). Anaesthesia was categorised as GA (even if it included supplementary LA) or LA. Information on anaesthetic techniques, anaesthetic and sedative drugs and local anaesthetic complications was recorded. One source of data used in this study, the non-commercial database (VITREOR), did not have an anaesthetic complication field and it was therefore not possible to record complications of LA for cases that were operated in centres using this database. Pearson’s χ2 test and Fisher’s exact test were used to test for differences between the anaesthetic techniques for the number of anaesthetic complications and for each individual anaesthetic complication. Potential differences in patient’s age at the time of a first VR surgery between GA and LA were investigated by a Student’s t test using the Welch approximation. Yearly results follow the NHS year, which runs from 1 April to 31 March and all analyses were conducted using STATA V.11.
66 trainee surgeons and 8 surgeons had data recorded at more than one grade. The number of operations performed by each surgeon ranged from 1 to 1553, where 54 surgeons had data for 1 operation, 46 surgeons had data for 2–10 operations, 26 surgeons had data for 11–99 operations and 34 surgeons had data for ≥100 operations. The most common surgical category was RD comprising 5755 operations (3677 for primary PPV, 639 for primary SB, 356 for primary PPV+SB and 1083 were for repeat RD surgery). Of the remaining eyes, 2951 were for macular surgery, 2476 for simple vitrectomy and 942 for complex vitrectomy.
Anaesthetic technique GA was used for 6054 (49.9%) operations and LA for 6070 (50.1%). Of those operations performed under GA, 761 operations also had additional LA (31 (4%) by peribulbar or retrobulbar injection and 730 (96%) by sub-Tenon’s). In all patients (n=9935), GA tended to be more used for younger patients than LA (the mean age at the time of first VR operation; 57.3 years for GA vs 67.7 years for LA; p≤0.001). The tendency for GA use in younger patients was also evident when restricting the analysis to patients >25 and 70% of these operations were performed under GA. However, wide intercentre variations were present for other types of surgery and in four centres, >70% of operations for each type of VR surgery were performed under GA, table 1.
LA complications Of the 6070 operations performed under LA, 5225 (86.1%) were performed in centres that recorded anaesthetic complications data and one or more complications were recorded in 184 (3.5%) of these operations. The complication rate was lower for sub-Tenon’s anaesthesia compared with sharp needle injections (2.9% vs 4.2%, respectively). A slightly higher percentage of patients having sub-Tenon’s anaesthesia had one of the following structured complications of anaesthesia: mild, moderate or severe ‘discomfort/pain’, recorded in the EMR, compared with sharp needle administered anaesthesia ( p=0.048). Statistically significantly higher rates of conjunctival chemosis and eyelid haemorrhage/bruising were found with sharp needle injections compared with sub-Tenon’s anaesthesia ( p≤0.001 and 0.003, respectively). Potentially sightthreatening or life-threatening complications of LA were very uncommon with both techniques: serious systemic complications were documented in 0.1% of operations (six cases), with no retrobulbar haemorrhages or ocular perforations being recorded, table 2.
DISCUSSION This study of 12 124 VR operations performed in 13 geographically dispersed centres in the UK surveyed trends in anaesthesia for VR surgery over a 10-year period. Our data demonstrate that from 2000 to 2010, the use of GA for VR surgery has 4
decreased, while the use of LA increased substantially from 5% in 2001 to almost 60% in 2010. In eyes operated under LA, the use of sub-Tenon’s cannula techniques has exceeded sharp needle techniques in recent years and complications of both techniques were uncommon. While there were marked differences between retinal surgeons and centres in our study, GA is still preferred for SBs, complex vitrectomy operations and younger patients. LA using sub-Tenon’s cannula12–16 and sharp needle peribulbar and retrobulbar injections6–8 15 have both been documented as effective for VR surgery. There is evidence that LA is associated with comparable rates of per-operative surgical complications to GA and does not influence the anatomical or functional outcomes that are achieved.17 LA has several advantages over GA: it is associated with less cardiopulmonary risks to patients, it represents a viable option for emergency operations in patients who are not fasting and theatre turnover time is significantly shorter making it more cost effective.17 In the UK, GA has traditionally been the preferred method for delivering anaesthesia for VR surgery.3 5 This multicentre study adds weight to the findings of single-centre studies6–8 17 that have demonstrated an increased uptake of LA for VR surgery. It is likely that the increased use of LA is related to the trend for VR centres to perform surgery as day case without an overnight stay, to reduce healthcare costs and increase efficiency. Additionally, we speculate that the rising popularity of LA in the UK is linked to the increased use of smaller gauge transconjuctival vitrectomy instruments, which have been shown to be associated with less ocular discomfort for patients18 as well as shorter duration of surgery compared with 20-gauge instruments.19 Intravenous sedation in addition to LA was used in a very small number of operations (2%) in this study. This is in accord with other studies from the UK where the use of adjunctive Sallam AAB, et al. Br J Ophthalmol 2015;0:1–7. doi:10.1136/bjophthalmol-2014-306467
Downloaded from http://bjo.bmj.com/ on November 2, 2015 - Published by group.bmj.com
Clinical science Table 1 The percentage of operations performed under general anaesthesia (GA) in each centre by grade of surgeon and complexity of operation Centres Operative details
A
B
C
D
E
F
G
H
I
J
K
L
M
Total
Number of operations Number of surgeons* Grade of surgeon Consultant surgeons; n surgeons Number of operations Percentage of operations under GA Independent non-consultant surgeons; n surgeons Number of operations Percentage of operations under GA Trainee surgeons; n surgeons Number of operations Percentage of operations under GA Type of surgery Primary PPV for retinal detachment (RD); n Percentage of operations under GA Primary SB RD; n Percentage of operations under GA Primary PPV+SB for RD; n Percentage of operations under GA Repeat RD; n Percentage of operations under GA Macular surgery; n Percentage of operations under GA Simple PPV; n Percentage of operations under GA Complex PPV; n Percentage of operations under GA
3615 23
3265 38
1694 32
1409 24
653 1
424 8
238 6
224 6
167 4
152 2
147 8
74 3
62 5
12 124 160
1 1542 85.5 0
11 1361 36.5 11
11 1587 41.9 20
11 850 23.6 3
1 653 78.9 0
3 279 96.1 4
2 182 2.2 1
3 58 22.4 3
2 156 55.1 2
1 151 30.5 0
3 88 28.4 2
2 72 100.0 0
3 46 69.6 2
54 7025 53.3 48
0 N/A 23 2073 70.0
748 25.0 19 1156 28.5
102 29.4 2 5 0.0
4 0.0 11 555 16.2
0 N/A 0 0 N/A
16 87.5 2 129 98.4
2 0.0 3 54 0.0
120 28.3 1 46 34.8
11 36.4 0 0 N/A
0 N/A 1 1 0.0
5 0.0 3 54 27.8
0 N/A 1 2 100.0
16 75.0 0 0 N/A
1024 27.4 66 4075 49.8
1042 71.3 247 96.4 76 78.9 381 79.8 689 72.9 761 71.4 419 90.7
1147 19.3 220 73.6 136 58.1 360 49.4 682 20.1 536 28.9 184 44.0
488 30.3 82 91.5 56 57.1 123 65.0 414 35.7 411 33.6 120 61.7
439 17.8 43 88.4 28 53.6 85 34.1 456 16.2 279 11.5 79 31.6
162 90.1 1 100.0 19 94.7 48 85.4 247 71.3 150 72.0 26 96.2
111 97.3 4 100.0 5 100.0 38 100.0 117 97.4 126 93.7 23 95.7
62 0.0 11 9.1 1 0.0 21 4.8 91 0.0 38 5.3 14 0.0
51 37.3 12 91.7 3 33.3 11 45.5 59 16.9 73 16.4 15 33.3
46 45.7 5 80.0 6 100.0 4 100.0 56 50 35 40.0 15 86.7
38 21.1 5 100.0 3 100.0 6 83.3 56 14.3 21 23.8 23 52.2
54 24.1 4 100.0 0 N/A 3 0.0 56 21.4 15 13.3 15 60.0
17 100.0 4 100.0 21 100.0 3 100.0 14 100 10 100.0 5 100.0
20 95.0 1 100.0 2 100.0 0 N/A 14 78.6 21 38.1 4 75.0
3677 41.9 639 85.8 356 68.0 1083 63.5 2951 41.8 2476 46.3 942 69.4
The centres are ranked by the total number of operations they contributed to the analysis. *Eight surgeons had data recorded at multiple grades, reflecting progress through training. PPV, pars plana vitrectomy; SB, scleral buckle.
sedation was uncommon and ranged from 0% to 20.2%.6 7 16 17 20 In contrast, intravenous sedation in conjunction with LA is standard practice in the USA.2 14 15 GA is still more often used for complex surgery or when a SB is included. These procedures are expected to be either prolonged and may involve traction on extraocular muscles as well as cryopexy with increased level of patients’ discomfort compared with more straightforward vitrectomy surgery.5 17 Several studies have shown that LA can still be used for SBs and/or prolonged VR surgery; however, there was often a need for additional LA during surgery and an increased requirement for intravenous sedation.6 7 12 20 Young patients have been previously identified as an important factor for considering GA.2 8 However, LA is still an option in young adults (
