Visual Abstract
Abstract
Background Regional anesthesia improves short-term blood flow through arteriovenous fistulas (AVFs). We previously demonstrated that, compared with local anesthesia, regional anesthesia improves primary AVF patency at 3 months.
Methods To study the effects of regional versus local anesthesia on longer-term AVF patency, we performed an observer-blinded randomized controlled trial at three university hospitals in Glasgow, United Kingdom. We randomly assigned 126 patients undergoing primary radiocephalic or brachiocephalic AVF creation to receive regional anesthesia (brachial plexus block; 0.5% L-bupivacaine and 1.5% lidocaine with epinephrine) or local anesthesia (0.5% L-bupivacaine and 1% lidocaine). This report includes findings on primary, functional, and secondary patency at 12 months; reinterventions; and additional access procedures (primary outcome measures were previously reported). We analyzed data by intention to treat, and also performed cost-effectiveness analyses.
Results At 12 months, we found higher primary patency among patients receiving regional versus local anesthesia (50 of 63 [79%] versus 37 of 63 [59%] patients; odds ratio [OR], 2.7; 95% confidence interval [95% CI], 1.6 to 3.8; P=0.02) as well as higher functional patency (43 of 63 [68%] versus 31 of 63 [49%] patients; OR, 2.1; 95% CI, 1.5 to 2.7; P=0.008). In 12 months, 21 revisional procedures, 53 new AVFs, and 50 temporary dialysis catheters were required. Regional anesthesia resulted in net savings of £195.10 (US$237.36) per patient at 1 year, and an incremental cost-effectiveness ratio of approximately £12,900 (US$15,694.20) per quality-adjusted life years over a 5-year time horizon. Results were robust after extensive sensitivity and scenario analyses.
Conclusions Compared with local anesthesia, regional anesthesia significantly improved both primary and functional AVF patency at 1 year and is cost-effective.
Clinical Trial registry name and registration number Local Anaesthesia versus Regional Block for Arteriovenous Fistulae, NCT01706354
Arteriovenous fistulas (AVFs) are the hemodialysis (HD) access modality of choice for patients with ESKD1 and are associated with lower rates of systemic sepsis and mortality compared with other vascular access options.2,3 However, the early failure rate is approximately 30%4⇓⇓⇓–8 and is influenced by preoperative vessel size, arterial inflow, and early postoperative blood flow, all of which can be affected by anesthetic technique.9,10
Regional anesthesia (RA), such as brachial plexus block, involves targeted injection of local anesthetic (LA) to block motor and sensory nerves supplying the operative site. Unlike LA infiltration for AVF creation, RA also blocks sympathetic nerves resulting in vasodilation, improved blood flow, and reduced vasospasm both perioperatively and in the early postoperative period.10⇓⇓–13 Until recently, there was no evidence that short-term perioperative hemodynamic changes secondary to anesthesia could improve longer-term fistula patency.11,14
We previously demonstrated that medium-term (3 month) primary AVF patency rates were higher in patients randomized to RA compared with LA infiltration at the time of radiocephalic (RCF) or brachiocephalic (BCF) fistula creation.11 However, functional patency rates at this 3-month time point were lower than anticipated in both cohorts. The effect of anesthetic technique on longer-term AVF functional patency remains unknown.
Long-term functional patency is the ultimate goal of vascular access surgery, reducing both the need for further vascular access procedures and complications associated with tunneled dialysis catheters (TDCs). If using RA improved long-term outcomes it could therefore result in cost savings, although this would need to be offset against the financial costs of a dedicated anesthetist, anesthetic equipment, and additional time required.
The purpose of this study was to analyze the clinically relevant outcome of functional AVF patency at 1 year using follow-up data from our original randomized controlled trial (RCT) comparing RA with LA for AVF creation, and to evaluate the cost-effectiveness of each anesthetic modality.
Methods
Study Design and Participants
An observer-blinded RCT was performed at three university hospitals in Glasgow, United Kingdom (Stobhill Ambulatory Care Hospital, Western Infirmary, and Queen Elizabeth University Hospital). The full trial design, methodology, and initial (3 month) outcomes have been published previously.11,15 The trial protocol was approved by West of Scotland Research Ethics Committee 5 (12/WS/0199) and is available at https://trialsjournal.biomedcentral.com/articles/10.1186/1745-6215-14-263.
A research team member approached eligible patients preoperatively and assessed them for eligibility. Patients were provided with study information before informed, written consent was obtained. The research was undertaken in accordance with the Declaration of Helsinki and was in keeping with the standards set by the International Conference on Harmonization of Good Clinical Practice.
Adults (aged 18 or older) undergoing primary RCF or BCF fistula creation for the purposes of HD were eligible for inclusion. Patients were excluded if they were unable or unwilling to provide informed consent, had previous ipsilateral attempts at AVF creation, the radial or brachial artery was <1.8 mm or cephalic vein was <2 mm at the wrist or <3 mm at the elbow on preoperative ultrasound (without tourniquet), had an allergy to LA, significant peripheral neuropathy or neurologic disorder affecting the upper limb, infection at the anesthetic or surgical site, coagulopathy, or known ipsilateral central vein stenosis (even if treated).
Randomization and Masking
Patients were randomly assigned (1:1, in blocks of eight) using a computer-generated allocation system to receive either RA or LA. Study allocation was by opaque, sealed envelopes as produced by a member of staff independent of the research team. After obtaining consent, each patient was assigned a study number and corresponding sealed envelope containing their study allocation. This was opened by the anesthetist allocated to the theater list. Because of the nature of the study intervention, neither the anesthetist, surgeon, nor patient were blinded. The vascular access nurse performing the assessment of study outcomes was blinded to study allocation.
Procedures
A detailed description of surgical and anesthetic techniques has previously been published11,15 and is available at: https://trialsjournal.biomedcentral.com/articles/10.1186/1745-6215-14-263. In brief, a standard end-to-side RCF or BCF was created. Patients in the RA cohort all received an ultrasound-guided brachial plexus block performed by one of two experienced consultant anesthetists. The supraclavicular approach was chosen unless there was a contraindication, in which case an axillary block was undertaken. A 1:1 mixture of 0.5% L-bupivacaine and 1.5% lidocaine with epinephrine (1 in 200,000) was injected up to a maximum volume of 40 ml. Patients in the LA infiltration group received infiltration of LA into the surgical site by the operating surgeon under sterile conditions using a combination of 0.5% L-bupivacaine and 1% lidocaine injected subcutaneously immediately before the commencement of surgery. Maximum dose limits of 2 mg/kg for bupivacaine and 3 mg/kg for lidocaine (7 mg/kg with epinephrine) were observed throughout, recognizing that these effects are additive.
Clinical Outcomes
The primary outcome has been reported previously.11 Secondary end points reported here include primary, functional, and secondary patency at 1 year. Functional patency was assessed both clinically as an AVF suitable for dialysis and by ultrasound (>6 mm diameter, <6 mm from skin surface, and flow rate >600 ml/min16). Additional interventions (angioplasty, stenting, and surgical revision) and alternative vascular access formation and adverse events, including access-related complications (infection, stenosis, thrombosis), were also recorded.
Definitions of patency are derived from Sidawy et al.17 and are outlined in the Supplemental Appendix.
Statistical Analysis
Results were analyzed using SPSS Statistics version 22 (Armonk, NY). Data were tested for normality. Assuming normal distribution, t test (two tailed) was used to compare continuous data, and chi-squared test was used to compare categoric data. Mann–Whitney U tests were used for non-normally distributed data. Logistic regression analysis was carried out to examine the interaction between AVF site (RCF/BCF) and anesthesia on primary and functional patency. P<0.05 was considered significant. Results are presented as mean (95% CI), median (IQR), or as a percentage of the total population and odds ratio (OR). Missing data were limited and assumed to be missing at random. If a data point was missing, this case was removed from analysis of the specific variable of interest. A “last-forward” approach was taken toward fistula patency in patients who died. Data were analyzed on an intention-to-treat basis. The study was powered according to the primary outcome11 and no formal power calculation was performed for the secondary end points reported in this paper.
Cost-Comparison and Cost-Effectiveness Analysis
A probabilistic state-transition (Markov) model was developed, tracking the progression of patients with ESKD across various health states representing alternative vascular access modalities. Nested decision trees captured the pathways associated with the creation and maturation of new AVFs (Figure 1). Transition probabilities were derived from clinical data observed in the RCT across the 1-year follow-up (Supplemental Table 1).
(A) State-transition Markov model and (B and C) nested decision trees outlining potential treatment pathways. Patients enter the decision tree in state A (new AVF creation). This state is structured as a nested decision tree depicted (A) which captures the two operative processes being compared. The AVF may successfully mature and become functional (state B); or may not reach full maturation, being in a state of primary (but nonfunctional) patency (state C) or fail completely (state D), necessitating an alternative access (state E). The patient may also die (state F). The outcome at 3 months determines the state to which the patient transitions within the Markov model. The model works in discrete time cycles of 3 months. Transition probabilities for the decision tree in state A were derived based on the previously reported 3-month primary and functional patency rates.11 These determined the distribution of patients across states B, C, D, and F at 3 months after the creation of a new AVF. Past this point, transition probabilities were derived from the longer-term patency rates observed at the 1-year follow-up of the original trial cohort and were assumed to apply throughout the time horizon of the model. For patients with a functional AVF (state B), there is a probability that their fistula remains functional or fails (transition to state D) that applies at each cycle as the time goes forward. For patients with a primary (but nonfunctional) AVF (state C), there is a probability that their fistula matures spontaneously and becomes functional, or matures as a result of a surgical or radiologic revision. State C is also structured as a nested tree as detailed (C). State D is an atemporal health state. Patients do not spend time in this state within the model. Instead, they are referred for a new AVF creation when the previous one fails (transition to state A), which occurs in the same cycle. This is assumed to occur up to four times in the model, i.e., a patient can have up to four new AVFs created. If the fourth AVF also fails, the patient is switched permanently to an alternative vascular access modality (TDC) until the end of the time horizon in the model. A patient can die within any given cycle in the model and get transferred to the absorption state F. For patients on dialysis, it is assumed they are dialyzing via AVF while their fistula is functional and via TDC otherwise. A proportion of the starting cohort is predialysis. This subgroup is assumed to be starting dialysis by 1 year, via AVF if a functional one is present or via TDC otherwise. While dialyzing, patients incur a risk of developing sepsis, which is dependent on the dialysis modality, and require further treatment. The incidence of infection was derived from the literature: 0.2 and 1.4 cases per 1000 dialysis days was applied for the time spent dialyzing via AVF and TDC, respectively.35
All costs were estimated from the perspective of the National Health Service (NHS) and were converted to 2019/2020 United Kingdom pound sterling (£). A “bottom-up” approach was used to estimate costs associated with each anesthetic technique including medication, equipment, and staff time. Costs were derived from the British National Formulary,18 Personal Social Services Research Unit Costs of Health and Social Care 2018,19 national procurement data,20 or market prices (Supplemental Tables 2–5) and were validated by clinical experts.
Methodology described by Shechter et al.21 was applied to derive baseline health utility scores for HD via different access modalities using previously published data.22,23 As such, baseline utility values of 0.767 and 0.677 were assigned to health states “HD via AVF” and “HD via TDC,” respectively (Supplemental Table 6).
Two sets of results were estimated. Firstly, resource utilization data derived directly from the 1-year follow-up of patients in the RCT were used for cost-comparison analysis. A one-way sensitivity analysis was undertaken (±20%) to evaluate the influence of key factors. Secondly, the decision-analytic model described above was used to estimate the relative cost-effectiveness of RA compared with LA across a 5-year time horizon. Costs and health benefits were tracked and aggregated, and incremental cost-effectiveness ratios (ICERs) were estimated (expressed in £ per life-year and per quality-adjusted life year [QALY]). One-way sensitivity analysis was conducted and structural uncertainty in the model was investigated through a series of scenario analyses. Finally, the joint parameter uncertainty was explored by second-order Monte Carlo simulation in which a probabilistic distribution was fitted around each model parameter and 1000 simulations run, repeatedly sampling different point estimates from these distributions and estimating alternative probabilistic model results. The parameters for the distributions were estimated from available or assumed sample statistics using the method of moments (Supplemental Table 6).
Results
Between February 6, 2013 and December 4, 2015, 163 patients were assessed for eligibility and 126 patients were randomly assigned to LA (n=63) or RA (n=63) (Figure 2). One patient breached protocol having been randomly assigned before vein mapping ultrasound (no suitable vessels identified). Six patients (four in the RA cohort and two in the LA cohort) died between 3 and 12 months follow-up. Otherwise all patients completed 12-month follow-up on an intention-to-treat basis.
Trial profile and Consolidated Standards of Reporting Trials diagram demonstrating flow of participants through the trial. Follow-up and analysis figures are reported at 1 year.
Patient demographics have been described previously.11 The groups were similar in terms of age, sex, comorbidities, renal replacement modality, and other baseline variables (Table 1). A total of 51 of 126 (40%) patients randomly assigned had an RCF created whereas the remaining 75 (60%) had a BCF creation.
Baseline characteristics
Primary patency at 12 months was higher in the RA group compared with the LA group (50 [79%] versus 37 [59%] patients; OR, 2.7; 95% CI, 1.6 to 3.8; P=0.02; fragility index=2). Similarly, functional patency at 1 year was higher in the RA cohort (43 [68%] versus 31 [49%] patients; OR, 2.1; 95% CI, 1.5 to 2.7; P<0.01; fragility index=2) (Table 2).
Patency rates of AVF (primary, secondary, and functional patencies at 3 and 12 mo)
The observed benefits of RA were more marked in RCF than BCF, although formal interaction tests did not find significance (Tables 3 and 4). Primary and functional patency rates at 12 months in RCF were 77% versus 48% (P=0.02) and 78% versus 48%, respectively (P=0.02). No statistically significant difference in functional patency at 1 year was observed in BCF (76% versus 63%; P=0.25). Although not significantly different between RA and LA, overall functional patency of BCF was better at 12 months than 3 months (53 [69%] versus 15 patients [20%]; P<0.01).
Logistic regression analysis examining the interaction between anesthetic type and AVF site on primary patency at 12 mo
Logistic regression analysis examining the interaction between anesthetic type and AVF site on functional patency at 12 mo
A total of 21 revisional procedures were performed in 14 patients (Table 5); 13 patients (93%) successfully achieved functional patency as a result of these interventions. All revisional procedures were performed in BCF. No RCF required intervention. Of the 21 revisional procedures, 15 (71.4%) were performed in the RA cohort. Conversely, more new AVFs and temporary dialysis catheters were required in the LA cohort. Additionally, a further 24 AVFs (13 RA, 11 LA), which had not achieved functional patency by 3 months, subsequently achieved functional patency by 12 months without additional intervention.
Number of additional procedures first year
Seven patients (four RA, three LA) died during the 12-month follow-up period. No death was associated with vascular access complications. There were no complications of anesthetic administration. One patient (LA) experienced a superficial wound infection. There was one case of line sepsis in the LA cohort (treated with 14 days of intravenous antibiotics). Mean duration of hospital stay for vascular access issues was 2.5 days (range, 1–17) in 12 months.
High initial staff time costs in the RA cohort were offset by cost savings of improved AVF maturation (fewer additional new AVF procedures and reduced TDC complications), delivering an overall net cost saving of £195.10 per patient at 1 year (Table 6). Full details of costs associated with anesthesia, initial AVF surgery, revisions, additional accesses, and sensitivity analyses are outlined in the Supplemental Table 3. The incremental cost was most sensitive to the number of alternative accesses (both AVFs and TDCs) created but RA remained a cost saving in most cases (Supplemental Figure 1).
Breakdown of costs per patient in the RA and LA arms at 1-yr follow-up
Base case results demonstrating relative cost-effectiveness of RA compared with LA over a 5-year time horizon are summarized in Table 7. Over 5 years, RA resulted in a cost saving of approximately £2100 per patient. In the RA cohort, patients spent relatively more time dialyzing via AVF and derived an incremental survival and QALY benefit. The ICERs realized in the RA arm compared with the LA arm were approximately £10,300 per life year gained and £12,900 per QALY gained, respectively. The robustness of base case ICER (£/QALY) to variations in a wide range of model parameters and under various scenarios was investigated by one-way sensitivity analysis as outlined in the Supplemental Appendix (Figure 3, Supplemental Table 7). RA was cost-effective at a £30,000/QALY threshold in all but one case: when the cost of dialyzing via AVF was increased by 20%, the ICER increased to approximately £33,000/QALY. The ICER was not higher than £30,000/QALY in any other case and in several cases RA dominated LA, offering a higher health benefit for a lower cost.
Base case results cost-effectiveness analysis (per patient)
Base case ICER (£/QALY) was robust to variations in a wide range of model parameters. BPB, brachial plexus block; TCVC, tunneled central venous catheter.
The ongoing costs of HD are a big driver of cost-effectiveness results, with the NHS England cost of “HD via AVF” being paradoxically higher than “HD via TDC.” Excluding these costs resulted in RA dominating LA. The same effect is observed when the time horizon of the analysis is reduced. The proportion of the cohort that are predialysis at the start of the model also has a substantial effect on results, with RA being particularly cost-effective in patients already on dialysis at the time of initial AVF creation.
Probabilistic sensitivity analysis results show a high probability of RA being cost-effective (89.6% at £30,000/QALY threshold and 76.3% at £20,000/QALY threshold) compared with LA (Figure 4). Moreover, the probability of RA being dominated by LA is substantial (39.1%), whereas the probability of RA being dominated by LA is virtually zero (0.0%) (Figure 5).
Probabilistic sensitivity analysis. Cost-effectiveness acceptability curve. BPB, brachial plexus block.
Probabilistic sensitivity analysis demonstrating a high probability of RA being cost-effective compared to LA. BPB, brachial plexus block; CE, cost-effectiveness; PSA, probabilistic sensitivity analysis.
Discussion
These results confirm an enduring superiority of RA over LA in achieving primary and functional AVF patency at 12 months and are, to our knowledge, the first randomized data to demonstrate an anesthetic technique improving any long-term surgical outcome. The results also prove RA is cost-effective in AVF creation. Our findings serve to address criticisms ascribed to our original RCT,11 namely poor functional patency rates, particularly in BCF.
The 12-month functional patency rates described here are comparable to those observed in other large vascular access RCTs8,23,24 and are reflective of the contemporaneous Scottish population.25 We hypothesized that absence of assisted maturation techniques, comorbidity, and obesity (difficulties cannulating deep AVFs, or failing to meet our functional patency criteria of <6 mm from skin surface) may explain the relatively poor 3-month functional patency rates previously observed. These follow-up data suggest that, although revisional procedures were required in 14 patients, a further 24 AVFs developed functional patency between 3 and 12 months without intervention, simply requiring additional time to mature. This supports the assertion that the early maturation period is key to long-term functional patency, especially among RCFs where the functional patency rates at 12 months closely mirror primary patency at 3 months. Our data suggest that, if early patency (assisted by increased blood flow and vasodilation secondary to RA) is established, it will ultimately be possible to achieve functional patency. This should be considered when determining end points for future clinical trials.
As with our previous study,11 the beneficial effect of RA was more marked in the small vessels of RCF. Multiple interventions were needed, mainly in the RA cohort, to assist in the maturation of poorly developed fistulas. However, the cost of these interventions was offset against a need for more alternative accesses (de novo AVFs and TDCs) in the LA cohort. Due to the relatively small sample size, cost-effectiveness analysis was not performed for RCF and BCF as independent subgroups. It follows, however, that most of the cost savings are likely to be observed in patients undergoing distal fistula creation.
Three smaller randomized trials all demonstrated the beneficial effects of RA in short-term AVF maturation, although these have been considered to have a high level of bias with outcomes often limited to surrogate data.26⇓–28 Two recent meta-analyses also favored RA,14,29 but no trial to date has studied the long-term effects of RA. Nevertheless, the recent European Vascular Access and European Renal Best Practice guidelines recommend the use of RA for AVF creation.1,30 This guidance also states that RA may increase costs or delay the access procedure, which we have now demonstrated not to be the case. The recently commissioned National Institute for Health Research systematic review demonstrated a reduction of AVF failure by 72% with RA, concluding that future studies should also include a cost analysis.31
The study population is largely reflective of practice in the United Kingdom, however, it is acknowledged that demographic differences exist internationally. For example in the United States there is a larger proportion of patients who are obese and diabetic (factors known to be associated with adverse AVF outcomes32). Similarly, in the United States, assisted maturation techniques are commonly used early in the fistula life span, in part to address targets imposed by Centers for Medicare & Medicaid Services which aim to achieve freedom from TDC by 90 days. Recent United States analysis suggests rates of secondary AVF interventions to be as high as 44%.33 Therefore, it is difficult to extrapolate our results to this population because only 11% of patients in this study had an intervention to assist maturation, and none were within the first 90 days.
The study is limited by the lack of original quality of life (QoL) data. Baseline utility scores were extrapolated from other studies of health-related QoL outcomes in patients on HD.22,23 These studies used Kidney Disease Quality of Life Short Form and the 36-Item Short Form Survey to measure QoL in patients on dialysis. However, to date, there is no validated QoL tool evaluating the effect of either dialysis modality or vascular access specifically. Such a tool is urgently needed for performing future cost-effectiveness analyses of vascular access.
With these results we demonstrate, for the first time, net cost savings and long-term cost-effectiveness of RA. The cost-analysis presented reflects practice in the United Kingdom with health economic analysis performed from the perspective of the NHS and United Kingdom Personal and Social Services. It is acknowledged that costs in other healthcare systems will vary with differences in both perioperative care pathways (e.g., availability of block rooms, different anesthetic agents, availability of trained anesthetists) and subsequent interventions to provide a functional access for dialysis (e.g., assisted maturation techniques, resource implications of TDC) such that extrapolation of the absolute cost savings demonstrated within a United Kingdom setting may be limited. However, results appear robust throughout extensive sensitivity and scenario analyses and our model could be adapted to reflect variations in healthcare in an international setting.
Even within our model, the distribution of cost savings is complex and controversial. The national tariff in England and Wales incentivizes AVF use by providing cost savings to individual dialysis centers for “HD via AVF.” However, the overall net costs to the health service, society at large, and therefore to cost-effectiveness analyses, are higher than “HD via TDC.” The “costs” of care are not actually higher with AVF, rather the commissioner-to-institution reimbursement is. The decision whether to include ongoing HD costs in cost-effectiveness analysis is therefore a matter of debate,34 given that the more clinically effective an intervention is, the more it will disproportionately affect the costs in favor of the less effective comparator. Similarly, the higher overall survival in the RA arm translates to the accruing of relatively higher ongoing costs of HD. These idiosyncrasies lead to underestimating the true cost-effectiveness benefit derived from improving access outcomes. The absence of dialysis-associated costs (such as the need for alternative accesses) limits cost savings in patients who are predialysis. These patients should be considered as a separate cohort for future studies of cost-effectiveness. Future studies of cost-effectiveness in vascular access should focus on measurement of long-term overall healthcare costs rather than basic maintenance HD costs and reimbursement fees.
In conclusion, this is the first randomized study of any perioperative intervention to demonstrate enduring improvement in AVF patency at 12 months. We have presented mechanistic explanation,11 clinical benefit, and evidence of cost-effectiveness. Moreover, this trial demonstrates the value of a multidisciplinary approach to vascular access with motivated surgeons, anesthetists, nephrologists, access nurses, and health economists all contributing. On this basis, we reaffirm our assertion that RA should be used for all de novo AVF creation.
Disclosures
E. Aitken’s salary was funded by Darlinda’s Charity for Renal Research. A. Macfarlane received funding from Regional Anaesthesia UK. Neither funder played any role in the design, conduct, or reporting of this study. All remaining authors have nothing to disclose.
Funding
None.
Acknowledgments
The funder of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the writing of the report. The corresponding author had full access to all of the data in the study and had final responsibility for the decision to submit for publication.
Miss Emma Aitken was principally responsible for recruitment, collated and analyzed the data, and wrote the final manuscript; Dr. Marc Clancy conceived the study and was the principal surgeon; Dr. Lucian Gaianu performed the cost-effectiveness analysis; Dr. Andrew Jackson assisted with recruitment and data collection; Dr. Rachel Kearns and Dr. Alan Macfarlane designed the study, wrote the protocol, and contributed to the manuscript; Dr. John Kinsella reviewed the design of the study and revision of the final manuscript; Dr. Alan Macfarlane and Dr. Mark Steven were anesthetists for the trial; and Dr. Alan Macfarlane had principal responsibility for the study.
Data Sharing Statement
Data collected for the study will be made available to others on request. The study protocol and informed consent forms will also be made available. Data will be available to researchers who provide a methodologically sound proposal to achieve the aims in the approved proposal and reviewers of the manuscript at time of submission. To gain access, proposals should be directed to the corresponding author. Data will be available for 5 years from the date of the original study.
Supplemental Material
This article contains the following supplemental material online at http://jasn.asnjournals.org/lookup/suppl/doi:10.1681/ASN.2019111209/-/DCSupplemental.
Supplemental Table 1. Adjusted transition probabilities used in the model.
Supplemental Table 2. Corresponding dosages and list prices for the medicines utilised in preparing the anaesthetic solutions and other auxiliary medicines required per treatment arm.
Supplemental Table 3. Medical equipment and consumables used for anaesthesia.
Supplemental Table 4. Break-down of anaesthesia staff and afferent costs involved in the two anaesthetic procedures.
Supplemental Table 5. Additional costs relating to new AVF creation, AVF revision, TDC insertion, and on-going dialysis.
Supplemental Table 6. Parameters included in the probabilistic sensitivity analysis and their associated distributions.
Supplemental Table 7. Scenario sensitivity analysis.
Supplemental Figure 1. One-way sensitivity of RA versus LA incremental cost at 1-year to key analysis parameters.
Footnotes
Published online ahead of print. Publication date available at www.jasn.org.
- Copyright © 2020 by the American Society of Nephrology
References
In this issue
Jump to section
More in this TOC Section
Cited By...
- No citing articles found.